This article started life when I was asked to write a comparison of x86 and PowerPC CPUs for work. We produce PowerPC based systems and are often asked why we use PowerPC CPUs instead of x86 so a comparison is rather useful. While I have had an interest in CPUs for quite some time but I have never explored this issue in any detail so writing the document proved an interesting exercise. I thought my conclusions would be of interest to OSNews readers so I’ve done more research and written this new, rather more detailed article. This article is concerned with the technical differences between the families not the market differences.
History and Architectural Differences
The x86 family of CPUs began life in 1978 as the 8086, an extension to the 8 bit 8080 CPU.
It was a 16bit CISC (Complex instruction Set Computing) processor.
In the following year the 8088 was introduced which was used in the original IBM PC. It is this computer which lead to todays PCs which are still compatible with the 8086 instruction set from 1978.
The PowerPC family began life with the PowerPC 601 in 1993, the result of a collaboration started in 1991 between Apple, IBM and Motorola.
The family was designed to be a low cost RISC (Reduced Instruction Set Computing) CPU, it was based on the existing IBM POWER CPU used in the RS/6000 workstations so it would have an existing software base.
RISC Vs CISC
When Microprocessors such as x86 were first developed during the 1970s memories were very low capacity and highly expensive. Consequently keeping the size of software down was important and the instruction sets in CPUs at the time reflected this.
The x86 instruction set is highly complex with many instructions and addressing modes. Additionally it also shows it’s age by the small number and complex nature of registers (internal stores) available to the programmer. The x86 only has 8 registers and some of these are special purpose, PowerPC has 32 general purpose registers.
RISC was originally developed at IBM by John Cocke in 1974 [1]. Commercial RISC microprocessors appeared in the mid 80s first in workstations later moving to the desktop in the Acorn Archimedes.
These use a simplified instruction set which allow the CPUs to be simpler and thus faster. They also included a number of architectural improvements such as pipelining, super scalar execution and out of order execution which enabled the CPUs to perform significantly better than any CISC CPUs.
CISC CPUs such as the 68040 and the Intel 80486 onwards picked up and used many of these architectural improvements.
In the mid 1990s a company called NextGen produced an x86 CPU which used a translator to convert x86 instructions to run within a RISC core. Pretty much all x86 CPUs have since used this Technique. Even some RISC CPUs such as the POWER4 / PowerPC 970 use this technique for some instructions.
The high level internal architecture of the vast majority of modern desktop CPUs is now glaringly similar be they RISC or CISC.
Current State Of x86 And PowerPC CPUs
The current desktop PowerPC and x86 CPUs are the following:
x86
AMD Athlon XP
Intel Pentium 4
PowerPC
IBM 750xx (G3)
Motorola 74xx (G4)
IBM 970 (G5)
The current G4 CPUs run at significantly lower speeds compared with the x86 CPUs which are now above 2GHz (P4 > 3GHz). The recently announced PowerPC 970 currently runs up to 2GHz and delivers performance in line with the x86 CPUs.
CPUs break down all operations into stages and these are performed in a pipeline, these stages can be big or small and the number of stages depends on what’s done in each stage, the more an individual stage does the less stages you need to complete the operation. However if the stages are simple you will need more of them but each stage can complete quicker. The clock speed of the CPU is limited by the time an individual stage needs to complete. A CPU with simpler but greater number of stages will operate at a higher frequency.
Both the Athlon and Pentium 4 use longer pipelines (long and thin) with simple stages whereas the PowerPC G4s use shorter pipelines with more complex stages (short and fat). This is the essence of the so called “megahertz myth”. A CPU with a very high clock speed may not be any faster than a CPU with a lower clock speed. The Pentium 4 is now at 3.2 GHz yet a 1.25 GHz Alpha can easily outgun it on floating point operations.
The longer pipelines allow the x86 CPUs to attain these very high frequencies whereas the PowerPCs G4s are somewhat restricted because they use a smaller number of pipeline stages and this limits the clock frequency.
The amount of voltage the CPU can use restricts the power available and this effects the speed the clock can run at, x86 CPUs use relatively high voltages to allow higher clock rates, to boost clock speeds further, power hungry high speed transistors are used. A long thin pipeline is very fast but also very inefficient power wise. All these things add up so a 3GHz CPU may be fast but are also very power hungry with maximum power consumption rates now approaching or even exceeding 100 Watts. Intel in fact have taken to using a much lower frequency part for laptop computers than the top end Pentium 4. Yet, despite the fact it is only 1.6GHz, the Pentium M performs just as well as the 2.2GHz Pentium 4.
The Law Of Diminishing Returns (Aka Amdahl’s Law)
The Law of diminishing returns is not exactly a new phenomenon, it was originally noticed in parallel computers by IBM engineer Gene Amdahl, one of creators of the IBM System 360 Architecture. The original describes the problem in parallel computing terms however this simplified version pretty much describes the problem in terms of any modern computer system:
“Each component of a computer system contributes delay to the system
If you make a single component of the system infinitely fast…
…system throughput will still exhibit the combined delays of the other components.” [3]
As the clock speeds goes upwards the actual performance of the CPU does not scale exactly with the clock speed. A 2GHz CPU is unlikely to be twice the speed of a 1GHz CPU, indeed on everyday tasks people seem to have some difficulty telling the difference between these speeds.
The reason for the lack of scaling is the fact that memory performance has not scaled with the CPU so the CPU is sitting doing nothing for much of it’s time (HP estimate this at 70% for server CPUs). Additionally the latency of memory has barely improved at all so any program which requires the CPU to access memory a lot will be effected badly by memory latency and the CPU will not reach anything near it’s true potential. The CPU memory cache can alleviate this sort of problem to a degree but it’s effectiveness depends very much on the type of cache and software algorithm used.
Many of the techniques used within x86 CPUs may only boost performance by a small amount but they are used because of the need for AMD and Intel to outdo one another. As the clock speed increases ever higher the scaling problem increases further meaning that the additional effort has less and less effect on overall performance. Recent SPEC marks for two Dell workstations show that a greater than 50% increase in CPU speed and the addition of hyper-threading results in only a 26% increase in SPEC marks [2]. Yet when the Itanium 2 CPU got an 11% clock speed boost and double the cache the SPEC mark increased by around 50%
Of course there are other factors which effect the performance of CPUs such as the cache size and design, the memory interface, compiler & settings, the language it’s programmed in and the programmer who wrote it. Changing the language can in fact be shown to have a much greater effect than changing the CPU [4]. Changing the programmer can also have a very large effect [5].
Performance Differences Between The PowerPC And x86
Since AMD began competing effectively with Intel in the late 1990s both Intel and AMD have been aggressively developing new faster x86 CPUs. This has lead them to becoming competitive with and sometimes even exceeding the performance of RISC CPUs (If you believe the benchmarks, see below). However RISC vendors are now becoming aware of this threat and are responding by making faster CPUs. Ironically however if you were to make all CPUs at the same geometry the Alpha 21364 is the fastest CPU going – yet it uses a 7 year old core design.
PowerPCs although initially designed as desktop processors are primarily used in embedded applications where power usage concerns outweigh raw processing power. Additionally, current G4 CPUs use a relatively slow single data rate bus system which cannot match the faster double or quad data rate busses found on x86 CPUs.
The current (non G5) PowerPC CPUs do not match up to the level of the top x86 CPUs however due to the effects of the law of diminishing returns they are not massively behind in terms of CPU power. The x86 CPUs are faster but not by as much as you might expect [6]. (Again, see below section on benchmarks).
Vector Processing Differences
Vector processing is also known as SIMD (Single Instruction Multiple Data) and it is used in some types of processing. When used it speeds up operations many times over the normal processing core.
Both x86 and PowerPC have added extensions to support Vector instructions. x86 started with MMX, MMX2 then SSE and SSE2. These have 8 128 bit registers but operations cannot generally be executed at the same time as floating point instructions. However the x86 floating point unit is notoriously weak and SSE is now used for floating point operations. Intel has also invested in compiler technology which automatically uses the SSE2 unit even if the programmer hasn’t specified it boosting performance.
The PowerPC gained vector processing in one go when Apple, IBM and Motorola revised the powerPC instruction set and added the Altivec unit which has 32 128 bit registers. This was added in the G4 CPUs but not to the G3s but these are now expected to get Altivec in a later revision. Altivec is also present in the 970.
Currently the bus interface of the G4 slows down Altivec as it is very demanding of memory. However the Altivec has more registers than SSE so it can operate without going to memory too much which boosts performance over SSE. The Altivec unit can also operate independently from and simultaneously to the floating point unit.
Power Consumption Differences
One very big difference between PowerPC and x86 is in the area of power consumption. Because PowerPCs are designed for and used in the embedded sector their power consumption is deliberately low. The x86 CPUs on the other hand have very high power consumption due to the old, inefficient architecture as well as all the techniques used to raise the performance and clock speed. The difference in power consumption is greater than 10X for a 1GHz G4 (7447) compared with the 3GHz Pentium 4. The maximum rating for a G4 is less than 10 Watts whereas Intel do not appear to give out figures for power consumption rather referring to a “thermal design rating” which is around 30 Watts lower than the maximum figure. The Figure given for the design rating of a P4 3GHz is 81.9 Watts so the maximum is closer to and may even exceed 100 Watts.
A single 3GHz Pentium 4 CPU alone consumes more than 4 times power than a Pegasos PowerPC motherboard including a 1GHz G4.
Low Power x86s
There are a number of low power x86 designs from Intel, AMD, VIA and Transmeta.
It seems however that cutting power consumption in the x86 also means cutting performance – sometimes drastically. Intel still sell low power Pentiium III CPUs right down at 650MHz. The Pentium 4 M can reduce it’s power consumption but only by scaling down it’s clock frequency.
Transmeta use a completely different architecture and “code morphing” software to translate the x86 instructions but their CPUs have never exactly broken speed records.
VIA have managed to get power usage down even at 1GHz levels but they too use a different architecture. The VIA C3 series is a very simple CPU based on an architecture which forgoes the advanced features like instruction re-ordering and multiple execution units. The nearest equivalent is the 486 launched way back in 1989. This simplified approach produces something of a compromise however, at 800MHz it still requires a fan and even at 1GHz the performance is abysmal – a 1.3GHz Celeron completely destroys it in multiple benchmarks [7].
Why The Difference?
PowerPCs seem to have no difficulty reaching 1GHz without compromising their performance or generating much heat – how?
CISC and RISC CPUs may use the same techniques and look the same at a high level but at a lower level things are very different. RISC CPUs are a great deal more efficient.
No need to convert CISC -> RISC ISA
x86 CPUs are still compatible with the large complex x86 Instruction set which started with the 8080 and has been growing ever since. In a modern x86 CPU this has to be decoded into simpler instructions which can be executed faster. The POWER4 and PPC 970 also do this with some instructions but this is a relatively simple process compared with the multi-length instructions or the complex addressing modes found in the x86 instruction set.
Decoding the x86 instruction set is not going to be a simple operation, especially if you want to do it fast.
How for instance does a CPU know where the next instruction is if the instructions are different lengths? It could be found by decoding the first instruction and getting it’s length but this takes time and imposes a performance bottleneck. It could of course be done in parallel, guess where the instructions might be and get all possibilities, once the first is decoded you pick the right one and drop the incorrect ones. This of course takes up silicon and consumes power. RISC CPUs on the other hand do not have multi-length instructions so instruction decoding is vastly simpler.
Related to the above is addressing modes, an x86 has to figure out what addressing mode is used so it can figure out what the instruction is. A similar parallel process like the above could be used. RISC CPUs on the other hand again have a much simpler job as they usually only have one or two addressing modes at most.
To RISC Or Not To RISC
Once you have the instructions in simpler “RISC like” format they should run just as fast – or should they?
Remember that the x86 only has 8 registers, this makes life complicated for the execution core in an x86 CPU. x86 execution cores use the same techniques as RISC CPUs but the limited number of registers will prove problematic. Consider an loop which uses 10 variables in an iteration. An x86 will need hardware assist just to perform a single iteration.
Now consider a RISC CPU which generally have in the order of 32 registers. It can work across multiple iterations simultaneously, the compiler can handle this without any hardware assist.
The Hardware assist in question is Out-Of-Order execution and the tools of this trade are called rename registers. Essentially the hardware fools the executing program into thinking there are more registers than there really are and in the example this will allow for instance an iteration to be completed without the CPU needing to go the cache for data, the data needed will be in a rename register.
OOO execution is mainly used to increase the performance of a CPU by executing multiple instructions simultaneously. If so the instructions per cycle increases and the CPU gets it’s work done faster.
However when the x86 includes this kind of hardware the 8 registers becomes a problem. In order to perform OOO execution, program flow has to be tracked ahead to find instructions which can be executed differently from their normal order without messing up the logic of the program. In x86 this means the 8 registers may need to be renamed many times and this requires complex tracking logic.
RISC wins out here again because of it’s larger number of registers. Less renaming will be necessary because of the larger number of registers so less hardware is required to do register usage tracking. The Pentium 4 has 128 rename registers, the 970 has less than half at 48 and the G4 has just 16.
Because of the sheer complexity of the x86 ISA and it’s limited number of architectural registers a RISC processor requires less hardware to do the same work.
Despite not using the highly aggressive methodologies used in the x86 CPUs, IBM have managed to match and even exceed the computing power of x86 CPUs with the PowerPC 970 – at lower power consumption. They were able to do this because of the efficiency of RISC and the inefficiency of x86 CPUs. IBM have already managed to get this processor to run at 2.5GHz and this should perform better than any x86 (with the possible exception of the Opteron).
The idea that x86 have RISC-like cores is a myth. They use the same techniques but the cores of x86 CPUs require a great deal more hardware to deal with the complexities of the original instruction set and architecture.
PowerPC And x86 Get More Bits
Both families are in the process of transitioning to 64 bit.
AMD
Opteron
Athlon 64 (due September)
IBM
PowerPC 970
The AMD Opteron adds 64 bit addressing and 64 bit registers to the x86 line. There is already some support for this CPU in linux and the BSDs, a 64 bit version of Windows is also due.
The Opteron is designed as a server CPU and as such both the CPU and motherboards cost more than for normal desktop x86 CPUs. The Athlon 64 can be expected to arrive at rather lower prices.
Despite performing better than the best existing 32 bit Athlon, the Opteron has a slower clock speed (1.8GHz Vs 2.2GHz).
AMDs x86-64 instruction set extensions give the architecture additional registers and an additional addressing mode but at the same time remove some of the older modes and instructions. This should simplify things a bit and increase performance but the compatibility with the x86 instruction set will still hold back it’s potential performance.
The PowerPC 970 is as predicted on OSNews [8] is a 64 bit PowerPC CPU based on the IBM POWER 4 design but with a smaller cache and the addition of the Altivec unit as found in the G4. It supports 32 bit software with little or no changes although some changes to the original 64bit PowerPC architecture have been made in the form of a “64 bit bridge” to ease the porting of 32 bit Operating Systems [9]. This bridge shall be removed in subsequent processors.
The hardware architecture of the 970 is similar to that of any advanced CPU however it does not have the aggressive hardware design of the x86 chips. IBM use automated design tools to do layout whereas Intel does it by hand to boost performance.
The 970 has a long pipeline however it is not run at a very high clock rate, unusually the CPU does more per clock than other long pipeline designs so the 970 is expected to perform very well.
In addition to the new architecture the 970 includes dual floating point units and a very high bandwidth bus which matches or exceeds anything in the x86 world, this will boost performance and especially boost the Altivec unit’s capabilities.
The IBM PPC 970 closes the performance difference between the PowerPC and x86 CPU without consuming x86 levels of power (estimated 20 Watts at 1.4GHz, 40W at 1.8GHz). It has been announced in Apple Power Macintosh computers for August 2003, with the pent up demand I think we can expect Mac sales to increase significantly.
Benchmarks
There has been a great deal of controversy over the benchmarks that Apple has published when it announced the new PPC 970 based G5 [10].
The figures Apple gave for the Dell PC were a great deal lower than the figures presented on the SPEC website. Many have criticised Apple for this but all they did is use a different compiler (GCC) and this gave the lower x86 results. GCC may not be the best x86 compiler but it contains a scheduler for neither the P4 or PPC 970 however it is considerably more mature on x86 than PowerPC. In fact only very recently has the PowerPC code generation began to approach the quality of x86 code generation. GCC 3.2 for instance produced incorrect code for some PowerPC applications.
However, this does lead to the question of why the SPEC scores produced by GCC are so different from those produced by Intel’s ICC compiler which it uses when submitting SPEC results. Is ICC really that much better than GCC? In a recent test [11] of x86 compilers most results turned out glaringly similar but when SEE2 is activated ICC completely floors the competition. ICC is picking up the code and auto-vectorising it for the x86 SSE2 unit, the other compilers do not have this feature so don’t get it’s benefit. I think it’s fairly safe to assume this at least in part is the reason for the difference between the SPEC scores produced by Apple and Intel.
This was a set of artificial benchmarks but does this translate into real life speed improvements? According to this comment [12] by an ICC user the auto-vectorising for the most part doesn’t make any difference as most code cannot be auto-vectorised.
In the description of the SPEC CPU2000 benchmarks the following is stated:
“These benchmarks measure the performance of the processor, memory and compiler on the tested system.”
SPEC marks are generally used to compare the performance of CPUs however the above states explicitly this is not what they are designed for, SPEC marks also also test the compiler. There are no doubt real life areas where the auto-vectorisation works but if these are only a small minority of applications, benchmarks that are effected by it become rather meaningless since they do show reliably how most applications are likely to perform.
Auto-vetorisation also work the other way, The PowerPCs Altivec unit is very powerful and benchmarks which are vectorised for it can show a G4 outperforming a P4 by up to 3 1/2.
By using GCC Apple removed the compiler from the factors effecting system speed and gave a more direct CPU to CPU comparison. This is a better comparison if you just want to compare CPUs and prevents the CPU vendor from getting inflated results due to the compiler.
x86 CPUs may use all the tricks in the book to improve performance but for the reasons I explained above they remain inefficient and are not as fast as you may think or as benchmarks appear to indicate. I’m not the only one to hold such an opinion:
“Intel’s chips perform disproportionately well on SPEC’s tests because Intel has optimised its compiler for such tests”[13]* – Peter Glaskowsky, editor-in-chief of Microprocessor Report.
I note that the term “chips” is used, I wonder does the same apply to the Itanium? This architecture is also highly sensitive to the compiler and this author has read (on more than one occasion) from Itanium users that it’s performance is not what the benchmarks suggest.
If SPEC marks are to a useful measure of CPU performance they should use the same compiler, an open source compiler is ideal for this as any optimisations added for one CPU will be in the source code and can thus be added to the other CPUs also keeping things rather more balanced.
People accuse Apple of fudging their benchmarks, but everybody in the industry does it – and SPEC marks are certainly not immune, it’s called marketing.
Personally I liked the following comment from Slashdot which pretty much sums the situation up:
“The only benchmarks that matter is my impression of the system while using the apps I use. Everything else is opinion.” – FooGoo
The Future
x86 has the advantage of a massive market place and the domination of Microsoft. There is plenty of low cost hardware and tons of software to run on it, the same cannot be said for any other CPU architecture.
RISC may be technically better but it is held in a niche by market forces which prefer the lower cost and plentiful software for x86. Market forces do not work on technical grounds and rarely chose the best solution.
Could that be about to change? There are changes afoot and these could have an unpredictable effect on the market:
1) Corporate adoption of Linux
Microsoft is now facing competition from Linux and unlike Windows it is not locked into x86. Linux runs across many different architectures if you need more power or low heat / noise you can run Linux on systems which have those features. If you are adopting Linux you are no longer locked into x86.
2) Market saturation
The computer age as we know it is at an end. The massive growth of the computer market is ending as the market is reaching saturation. Companies wishing to sell more computers will need to find reasons for people to upgrade, unfortunately these reasons are beginning to run out.
3) No more need for speed
Computers are now so fast it’s getting difficult to tell the difference between CPUs even if their clock speeds are a GHz apart. What’s the point of upgrading your computer if you’re not going to notice any difference?
How many people really need a computer that’s even over 1GHz? If your computer feels slow at that speed it’s because the OS has not been optimised for responsiveness, it’s not the fault of the CPU – just ask anyone using BeOS or MorphOS.
There have of course always been people who can use as much power as they can get their hands on but their numbers are small and getting smaller. Notably Apple’s software division has invested in exactly these sorts of applications.
4) Heat problems
What is going to be a hurdle for x86 systems is heat. x86 CPUs already get hot and require considerable cooling but this is getting worse and eventually it will hit a wall. A report by the publishers of Microprocessor Report indicated that Intel is expected to start hitting the heat wall in 2004.
x86 CPUs generate a great deal of heat because they are pushed to give maximum performance but because of their inefficient instruction set this takes a lot of energy.
In order to compete with one another AMD and Intel will need to keep upping their clock rates and running their chips at the limit, their chips are going to get hotter and hotter.
You may not think heat is important but once you put a number of computers together heat becomes a real problem as does the cost of electricity. The x86’s cost advantage becomes irrelevant when the cooling system costs many times the cost of the computers.
RISC CPUs like the 970 are at a distinct advantage here as they give competitive performance at significantly lower power consumption, they don’t need to be pushed to their limit to perform. Once they get a die shrink into the next process generation power consumption for the existing performance will go down. This strategy looks set to continue in the next generation POWER5.
The POWER5 (of which there will be a “consumer version”) will include Simultaneous Multi-Threading which effectively doubles the performance of the processor unlike Intel’s Hyper Threading which only boosted the performance by 20% (although this looks set to improve). IBM are also adding hardware acceleration of common functions such as communications and virtual memory acceleration onto the CPU. Despite these the number of transistors is not expected to grow by any significant measure so both manufacturing cost and heat dissipation will go down.
Conclusion
x86 is not what it’s sold as. x86 benchmarks very well but benchmarks can and are twisted to the advantage of the manufacturer. RISC still has an advantage as the RISC cores present in x86 CPUs are only a marketing myth. An instruction converter cannot remove the inherent complexity present in the x86 instruction set and consequently x86 is large and inefficient and is going to remain so. x86 is still outgunned at the high end and perhaps surprisingly also at the low end – you can’t make an x86 fast and run cool. There is a lot of marketing goes into x86 and the market -technical people included- just lap it up.
x86 has the desktop market and there are many large companies who depend on it. Indeed it has been speculated that inefficient or not, the market momentum of x86 is such that even Intel, it’s creator may not be able to drag us away from it [14]. The volume of x86 production makes them very low cost and the amount of software available goes without saying. Microsoft and Intel’s domination of the PC world has meant no RISC CPU has ever had success in this market aside from the PowerPCs in Apple systems and their market share is hardly huge.
In the high end markets, RISC CPUs from HP, SGI, IBM and Sun still dominate. x86 has never been able to reach these performance levels even though they are sometimes a process generation or two ahead.
RISC vendors will always be able to make a faster, smaller CPUs. Intel however can make many more CPUs for less.
x86 CPUs have been getting faster and faster for the last few years, threatening even the server vendors. HP and SGI may have given up but IBM has POWER5 and POWER6 on the way and Sun is set to launch CPUs which handle up to 32 threads. Looks like the server vendors are fighting back.
Things are changing, Linux and other Operating Systems are becoming increasingly popular and these are not locked into x86 or any other platform. x86 is running into problems and PowerPC looks like it is going to increasingly become a real, valid alternative to x86 CPUs both matching and exceeding the performance without the increasingly important power consumption or heat issues.
Notes:
Both Amdahl’s Law (of diminishing returns) and Moore’s Law date from around the same time but notably we hear a great deal more about Moore’s law. Moore’s Law describes how things are getting better, Amdahl’s Law says why it’s not. There is a difference however: Moore’s Law was an observation, Amdahl’s Law is a Law.
References:
[1] John Cocke, inventor of RISC (obituary)
http://www.guardian.co.uk/Print/0,3858,4469781,00.html
[2] SPEC benchmark results
http://www.spec.org/cpu2000/results/
[3] Amdahl’s Law Simplified – Richard Wiggins
http://www.ucalgary.ca/library/access97/wiggins/tsld027.htm
[4] Speed differences in different languages
http://www.kuro5hin.org/story/2002/6/25/122237/078
[5] Coding competition shows humans are better than compilers
http://www.realworldtech.com/page.cfm?ArticleID=RWT041603000942
[6] Combined CPU Benchmarks
http://www.cpuscorecard.com/
[7] C3 V’s Celeron benchmarks
http://www.digit-life.com/articles2/roundupmobo/via-c3-nehemiah.html
[8] Speculation on the PowerPC G5
http://www.osnews.com/story.php?news_id=1357
[9] Details of the 64bit bridge can be found in the Software Reference Manual.
http://www-3.ibm.com/chips/techlib/techlib.nsf/products/PowerPC_970_Microprocessor
[10] Apples G5 benchmarks
http://www.apple.com/powermac/performance/
[11] ICCs optimisations can greatly effect performance
http://www.aceshardware.com/read_news.jsp?id=75000387
[12] But [11] does not appear to continue into real life code
http://www.osnews.com/comment.php?news_id=3931&limit=no#117135
[13]* Article on G5 benchmarks
http://news.zdnet.co.uk/story/0,,t271-s2136537,00.html
*I do not know if this is an exact quote.
[14] Escape from planet x86 – Paul DeMone
http://www.realworldtech.com/page.cfm?ArticleID=RWT060503232439
Further Reading
Article covering the differences between RISC and CISC
http://www.realworldtech.com/page.cfm?articleid=RWT021300000000
Article on PowerPC 970
http://arstechnica.com/cpu/03q1/ppc970/ppc970-0.html
About the Author:
Nicholas Blachford has been interested in CPUs for many years and has written on the subject for OSNews before. He works for Genesi who produce the Pegasos G3 / G4 PowerPC based motherboard and the MorphOS Operating System.
“The x86 can never be designed for a mission critical task. It started as a toy and should remain that way. It has very basic design flaws.”
Actually, the space shuttle computers use old x86 technology. We all know how that story ends. Maybe its time they update.
Great article BTW. Thats the kind of stuff I like to read at OSNews. The horrid power consumption of x86 processors is really starting to make more people aware of some of the shortcomings.
Anyone know what the i960 was? Wasn’t that an updated processor design?
What is up with these lame rollover ads? Why is the word “computer” linked to a microsoft wireless ad?? These aren’t even real links! What were you thinking?!
What is up with these lame rollover ads? Why is the word “computer” linked to a microsoft wireless ad?? These aren’t even real links! What were you thinking?!
What ads? I don’t see anything.. Oh, it’s because I paid OSNews.com for no ads. 😛
Here’s more info about be an OSNews member:
http://www.osnews.com/story.php?news_id=3878 😉
if i am remembering correctly, the i960 is either intel’s attempt at risc OR their early attempt at a somewhat EPIC like cpu (ie, a cpu that depended extremely heavily on the quality of the compiler and not its own ability to compensate)
as far as the article goes, the bias is horrible. yes, arstechnica is somewhat more biased on the pc side, but i give credit for it not screaming through in their cpu arch articles (one also has to consider that ars had some of the most in depth and extensive osx articles of any site i have seen, and IIRC none of the article staff of ars blatantly dislikes macs nor pcs) i also give kudos to realworldtech. both of these sites are really good places for technical comparasons and reviews.
Any article written by DKE (david every) should be reguarded as marketing propoganda. He isnt able to hold his own in an actual debate on any computer architecture discussion and is shown to be extremely biased, and alot of times blatantly wrong. I have had personal dealings with him in which he equated having drive letters to having “dos underpinnings”. on 2k and xp, drive letters are superficial(sp?) for the purpose of software compatability. When he finally backed down from his position (in which he was obviously wrong) the backpeddling was astounding. by his logic, mandrake 9 has dos underpinnings as well because wine provides drive letters to applications in the name of compatability. this is obviously not the case.
this article, OTOH, doesnt really say anything, nor does it really tell you anything. at one point it looks like he is equating heat dissipation measurements power consumption measurements as though 1w for one means 1w for the other … which is certainly not the case (although the actual relation is fairly linear). if you want to read an article with merit, go to realworldtech or arstechnica. skip the fluff that was this article.
as far as benchmarks go, the number of compilers used should have been increased as well as the oses used.
windows – gcc
linux x86 – gcc
windows – icc
linux x86 – icc
windows – vs
linux ppc – gcc
osx ppc – gcc
osx ppc – codewarrior
spec is cpu AND compiler dependent. they should have also noted the x86 config that is expected to be used more (windows – vs) and the ppc config expected (likely osx and gcc, but maybe cw, not 100% sure, dont know any mac developers who do it for more than a hobby)
for the record im not biased towards any particular cpu arch, im only anti via. i have always liked ppc architecture, if not the macos 9.x and below (and imacs, i really detest the imac look, how friggin gaudy can a computer get?)
> True he may havg gotten it wrong – the x86 architecture
> actually goes all the way back to the 4004. The truth is
> when the PC choose the 8088, it was already somewhat
> handicapped by it’s ties to the past.
Guys, do you even have the slightest idea of what “x86” means? look, 8086, 286, 386, 486, 586, 686… Do you get what “x86” means now? Now tell me, do 8080 and 4004 fit in that scheme? Some people…
Actually, the space shuttle computers use old x86 technology. We all know how that story ends. Maybe its time they update.
Maybe you should tell NASA about your theory the last shuttle crash was because of x86 processors and not a chunk of foam smashing a hole in the wing…
Great article BTW. Thats the kind of stuff I like to read at OSNews. The horrid power consumption of x86 processors is really starting to make more people aware of some of the shortcomings.
The power consumption argument is bogus.
In environments where power consumption is a critical issue, neither the PPC 970 _or_ a plain P4 are going to be used.
In average environments, the additional electricity costs will be dwarfed by the higher hardware costs.
Added to that, there are intel CPUs that have low power consumption, for applications that demand it (like laptops). A Pentium M @ 1.6Ghz requires less than 25W. For comparison, IBM says a 1.8Ghz 970 requires about 42W (1Ghz G4 = 30W, 2.8Ghz P4 = 68W).
” (and imacs, i really detest the imac look, how friggin gaudy can a computer get?)”
“gaudy”???
Why? Because it’s white? Because it’s screen is able to move in all directions? Because it has a round base?
I’m curious, because if not for it’s design, you would not have:
a.) an all-in-one design (and perhaps you are just against an all-in-one design, fair enough, although very subjective).
b.) the only reason the monitor CAN move in every direction, is because the screen sits via a flexible arm onto a round body. Other designs would sacrifice the screen movement.
c.) I can’t come up with a c.). Does a computer have to look like a Dell or HP to please you?
Needs to look more like a COMPUTER, right?
Sheesh
Actually, the G4 running at 1GHz is rated at 23 Watts per CPU maximum.
Compare that to AMD’s Athlon XP 2000+ (which runs at a 1667MHz clock speed despite its name) that also uses a 0.18-micron copper process, but not SOI, and it consumes up to 70 Watts.
The PPC’s are efficient……is that so bad?
A second grader writes with better grammar.
Josh Goldberg, do you sit or stand in front of your computer? I’d guess you stand, because it must be difficult to sit with that huge stick shoved up your….
Great job, Nicholas! This type of material, explained at a level most of us can follow, is just what sites like OSNews need. Thanks to everyone involved!
http://www.scimag.com/scripts/ShowPR.asp?PUBCODE=030&ACCT=300003960…
Several here have complained about the age of x86, saying that this comparison isn’t fair. It most certainly IS fair, because these architectures are what the respective companies are currently offering for sale. Now, if there were a P5 around the corner with an all new architecture, perhaps these people would have apoint. But there isn’t. The 970 and the P4 are both going to be on lots of desktops running head-to-head…there really is no other comparison (right now).
“Actually, the space shuttle computers use old x86 technology. We all know how that story ends. Maybe its time they update.
Maybe you should tell NASA about your theory the last shuttle crash was because of x86 processors and not a chunk of foam smashing a hole in the wing…
”
Duh… I realize that the crash wasn’t because of the the processors. I only made a comment that x86 are, in fact used for mission critical applications. ANd that even the 8086 is still being used and sought after by NASA:
http://www.geek.com/news/geeknews/2002may/chi20020514011664.htm
Actually, the G4 running at 1GHz is rated at 23 Watts per CPU maximum.
Ars Technica disagrees:
http://www.arstechnica.com/cpu/02q2/ppc970/ppc970-1.html
The PPC’s are efficient……is that so bad?
No, but they sacrifice performance to use less power. That’s hardly surprising.
I’m not going to argue that the old Pentium chips that NASA uses in its shuttles are the very best for them, but they purposefully do not use the latest technologies. The reason for this is that before any processor can be trusted to run these systems, they must undergo about two years’ worth of rugged testing to see if they will be reliable under such stressful conditions. Just thought I’d throw that in.
Hey, there’s no accounting for taste.
I personally loved the G4 Cube design. I guess I was in the minority there. Never much cared for the iMac design because of included monitor. I prefer them separate, but I wasn’t the target market for iMacs anyway. The newer flat-panel iMacs are definitley cool looking though (just my opinion).
The G3/G4 PowerMac designs were pretty, but not really earth shattering. I kinda like the new look of the G5 PowerMacs. I’m especially impressed with their “silent” cooling.
While not exactly competing with Apple on the design front, the new Shuttle XPCs are at least a step forward in the Intel world. Personally, I’m a bit more utilitarian and am interested in some of the “quiet” cases (Noise Control Stealth Tower and Antec Sonata).
Why do you use arstechnica.com to prove your point? Try looking at http://e-www.motorola.com/webapp/sps/site/taxonomy.jsp?nodeId=03C1T…
Notice that the MPC7455 @ 1Ghz is 18.5W typical and 28W max.
The MPC74xx family is the series that Apple calls the G4.
I agee, the Cube was great looking, just overpriced for what it was.
As for the older iMacs, I though they were extremely ugly, but the new ones are not only cool, but functional as hell.
Silence is golden, and way overlooked, thus I love my 17in iMac, it is almost totally silent. If the Shuttle XPC is a winner in this regard, I will definitely take a look at it.
My Athlon spanks my iMac, and is faster than my dual G4, but for most things, the iMac is just perfect, it gets used more than the other two.
Apple delivers a very efficient, logical and Easy to implement CPU, yet they cost to much. They need to get their cost down and they have all the tools in their hand to do it.
Great article! I think a G5 Cube would be rather nice. Don’t know if thats possible, maybe a G6 Cube.
which model of imac looks gaudy? is it the ilamp? no
is it the translucent cased candy coloured all in one model? yes
geez, some ppl have to have it spelled out completely
i would be glad to own an ilamp, just not the original version of the imac
Just want to say that I found this article very
interesting. It really helps clarify many of the
differences in design and some of the possible
advantages of the PPC platform…
“which model of imac looks gaudy? is it the ilamp? no
is it the translucent cased candy coloured all in one model? yes
geez, some ppl have to have it spelled out completely”
Geez, I must not have transported back to 1999 when the iMac was “candy coloured”.
Maybe some of us assume you would be talking about the only iMAC BEING SOLD RIGHT NOW.
Why the hell would I think it was an older model?
Most people will assume that it is one THAT IS NOT DISCONTINUED.
LOL!!!!!
Why do you use arstechnica.com to prove your point?
Because it was the first decent link my search returned.
That figure is also for a 1Ghz G4, which is significantly slower than both a 2.8Ghz P4 and a 1.6Ghz Pentium M (except for a few corner cases).
Incidentally, the page you link to is rather unclear as to which figures apply to which processor.
There are no power consumption figures for the (most likely overclocked) >1Ghz G4s Apple are currently using, so it’s somewhat more difficult to give comparable performance/power usage figures. Undoubtedly the P4 is still going to be higher, but a Pentium M won’t (the 25W figure is a maximum).
As I said elsewhere, the power consumption argument is largely bogus, as the tiny amount of money saved in power will be vastly overshadowed by the extra expense involved in purchasing Apple hardware (and that’s not even counting the costs of a platform migration). Electricity is cheap, Macs are not.
I must say, I am impressed at the way that you so quickly shifted the focus from heat and energy to the vast expense involved in Mac ownership!!
A beautiful shift from one straw-man to the other.
(although I think California might disagree with you on the “Electricity is cheap” argument)
Again, a quite skillful sleight of hand, congrats!!!!
@bobby
>x86 is bigger requires twice the clock speed, generates 4 times the heat do do the same amount of work as the PPC.
Regarding “X86 requires twice the clock speed” and the amount of work. Your generalised statement is generally not true with AMD’s Opteron @1.8Ghz and nForce3**. Careful with any generalisation.
**Engineering release of ASUS SK8N. Reference:
http://www.amdzone.com/articleview.cfm?articleid=1304
I must say, I am impressed at the way that you so quickly shifted the focus from heat and energy to the vast expense involved in Mac ownership!!
That would be because 90% of the post was about power consumption while half of the last sentence pointing out why it’s not really an issue happened to mention Apple, right ?
A beautiful shift from one straw-man to the other.
None of my points (that there are low power x86 chips, that x86 chips using more power are also faster (as are ones using less power), that the figures reported on the motorola website are difficult to interpret and that there are no power consumption figures for the fast G4s Apple are using) could be described as strawmen.
Neither is the Macintosh cost issue, given pretty much the only PPC systems out there are Macs. And this discussion is about an article comparing PPCs to x86s.
(although I think California might disagree with you on the “Electricity is cheap” argument)
Compared to the price difference between a Mac and a PC, it is.
Again, a quite skillful sleight of hand, congrats!!!!
Hardly. I wasn’t trying to deceive anyone, so how could I have done it sucessfully ?
Interesting read, but there are a few oddball statements in it, like “RISC vendors will always be able to make a faster, smaller CPUs”. Huh?
We are dumping our Sun workstations left and right (Blade 2000s, no less) and replacing them with MUCH faster x86 Linux boxes at 1/3 the price. I don’t simply mean faster clocks, I mean better performance. Simulation runs that take days on Blades run in LESS than a day on one of our 2.8G P4 boxes. Sure, Sun may one day offer a CPU that is faster than a P4 or Opteron, but they will want $10K for it and it will only be 2% faster. As far as I am concerned, Sun is dead, at least as far as 99.99% of the market is concerned. Hope the other RISC vendors can figure out a way to compete with x86, because at this point what they have to offer is both slower and more expensive.
I for one want a processor that doesn’t suck up the earth resources. Just think about the quantity of PC out there and that will grow with the population.
Since I read in one of Eugenia’s article that MS will break some backward compatability in software, I hope they do this with hardware too. Hopefully .NET has been designed for this purpose.
I would love to have Apple hardware that can run both OS’s extremely fast.
Nice article, but why do people complain about OSX unresponsiveness, although I have run a couple of high quality quicktime streams and multitasking was beautiful.
Those x86’s were a very special run of a few thousand and the simple computer was not a mission critical computer nor was it an IBM compatable design. I worked at IBM when they did that just for the fact that they could say they used those. You will never see one of those processors on the market. Every one was tested until only a few working ones were left.
I’m not going to argue that the old Pentium chips that NASA uses in its shuttles are the very best for them, but they purposefully do not use the latest technologies.
I doubt NASA uses Pentiums in the shuttle. I’d be amazed if it was anything more modern than a 286.
smithey try to stay on topic. We already know you can’t afford a Mac and it doesn’t apply to everyone and its not really pertinent to this topic.
The 1GHZ G4 is not overclocked, a litte bit of not too hard research will prove this and its a full desktop processor that works well in a laptop case less than 1 inch thick.
http://e-www.motorola.com/webapp/sps/site/prod_summary.jsp?code=MPC…
Back to the topic, its a good article that is a bit biased but a good read nontheless. I prefer the way information was presented in the arstechnica articles concerning the G4 and G5.
s
The x86 is bigger requires twice the clock speed, generates 4 times the heat do do the same amount of work as the PPC.
This isn’t true for (at least) P3s, Athlons, Pentium Ms or P4s, except in a few *very* specific cases, so which x86 CPUs are you thinking of ?
They may be about the same speed, but the PPC has a lot more room to grow.
I remember this being said about PPC and x86 once before, about a decade ago. Since then, PPC processors had the better performance of the two for less than a year, around the release of the first G4s,
Intel have a much better track record for delivering than any of the members of the AIM alliance.
486 on the Space Shuttle
http://www.spaceref.com/directory/exploration_and_missions/human_mi…
486 in the AirPort
http://freebase.sourceforge.net/hardware.html
I remember that these processors can run without a heatsink!
I for one want a processor that doesn’t suck up the earth resources. Just think about the quantity of PC out there and that will grow with the population.
This is another bogus issue, too. In terms of overall environmental impact by the average computer, the amount of electricity the CPU uses is barely even measurable.
Since I read in one of Eugenia’s article that MS will break some backward compatability in software, I hope they do this with hardware too. Hopefully .NET has been designed for this purpose.
Microsoft has had a portable operating system running on multiple architectures since 1993. They’ve just never had a reason to consider anything but x86 since the early ’90s (back then it really did look like x86 had hit the wall – then intel released the Pentium). NT *was* designed for “this purpose”, fifteen years ago.
I would love to have Apple hardware that can run both OS’s extremely fast.
Not going to happen.
Nice article, but why do people complain about OSX unresponsiveness, although I have run a couple of high quality quicktime streams and multitasking was beautiful.
Because compared to – well, just about anything else – it is unresponsive on anything short of the fastest machines available. Your example is not going to demonstrate unresponsveness. Try resizing some windows or switching between applications.
smithey try to stay on topic. We already know you can’t afford a Mac and it doesn’t apply to everyone and its not really pertinent to this topic.
Heh. You’re posting about how much money you think I have and apparently accusing me of “hijacking” an article. The irony.
Incidentally, nothing in that link you provided suggest Motorola endorse a G4 at >1Ghz. Not to mention most of it emphasises using the G4 in embedded scenarios…
486 on the Space Shuttle
I don’t think that’s quite what people are thinking when they talk about the computers that run the space shuttle – but still, useful info.
I enjoy reading the mac zealots poking fun at the x86 because it’s a frankenstein of a processor (tons of additions to make up for previous generations of shortfalls, you know).
Yet, we have Quartz Extreme, whose sole purpose in life was to make up for the shortfall in the OS X rendering system. Apparently megs of PDF are hard to draw in real time, so lets make the GPU do it! Look at that window slide down!
I’m not a fan of pc’s in general, but anyone who thinks that QE was a credit to Apple’s ingenuity needs to realize that 4% of the entire computer market was going to throw their macs out the window if they couldn’t move a window without the mouse leading it by 3 seconds.
“Why the hell would I think it was an older model?”
maybe because i also made reference to an older version of macos? maybe because between the different models carrying the name “imac” only 1 style was considered by MANY to be extremely gaudy?
forgot this box doesnt have my name cookie set :
Well Beta was better than VHS – but which won ? and what’s around today ?
Beta is/was used in TV studios (probably more digital these days).
VHS in the home.
“That figure is also for a 1Ghz G4, which is significantly slower than both a 2.8Ghz P4 and a 1.6Ghz Pentium M (except for a few corner cases).
”
I don’t think so!!!!, the G4 at 1.0 ghz is very effective as effective and even more than the pentium M 1.6 ghz (a pentium without very high frequency is nothing, and here the more advanced arcitecture of the G4 makes the difference ), and can beat in a a lot of applications the P4 2.8 ghz.
Anyway!!!!
Interested, I checked out the website of MorphOS, in a paper about MorphOS “in Detail” it said the below. I think this would have been a big point in the article but it was not mentioned. Is it true and how does it work that it is 10x faster? And, more importantly, is that fast enough to provide a speedy OS?!
Thanks for the good article!
From the PDF:
Microkernel Vs Macro Kernel
A common problem encountered in the development of microkernel Operating Systems is speed. This is due to the CPU having to context switch back and forth between the kernel and user processes, context switching is expensive in terms of computing power. The consequence of this has been that many Operating Systems have switched from their original microkernel roots and become closer to a macrokernel by moving functionality into the kernel, i.e. Microsoft moved graphics into the Windows NT kernel, Be moved networking inside, Linux began as a macrokernel so includes everything. This technique provides a speed boost but at the cost of stability and security since different kernel tasks can potentially overwrite one another’s memory.
Given the above, one might wonder why Q can be based on a microkernel (strictly speaking it’s only “microkernel like”) and still expected to perform well. The answer to this lies in the fact that MorphOS runs on PowerPC and not x86 CPUs. It is a problem with the x86 architecture that causes context switches to be computationally expensive. Context switching on the PowerPC is in the region of 10 times faster, similar in speed to a subroutine call. This means PowerPC Operating Systems can use a microkernel architecture with all it’s advantages yet without the cost of slow context switches. There are no plans for an x86 version of MorphOS, if this changes there will no doubt be internal changes to accommodate the different processor architecture.
Good document Nicholas!
Come see Nicholas and the Pegasos at Linuxtag. There will be a Pegasos at the Linuxtag in Karlsruhe this weekend. Linuxtag is Europe’s largest GNU/Linux exhibition. The Linuxtag is at the Karlsruhe Convention Center from 10-13 July. Nicolas will be joined by Genesi’s newest team member Sven Luther. Sven is a developer known to both the MorphOS and Linux development communities. Sven Luther’s LinuxPPC kernel was recently announced on MorphOS-News here:
http://www.morphos-news.de/index.php?lg=en&nid=371&si=1
At Linuxtag, the Genesi Team and the Pegasos will be found in the Debian Booth.
The following weekend (18-20 July 2003) Sven will also be attending the Debian Developers Conference, Oslo, Norway with the Pegasos. Details about that conference can be found here:
http://debconf.org/debconf3/
Sven has joined Genesi to manage the LinuxPPC development effort for the Pegasos. Sven is a PhD candidate at the University of Strausbourg and an Associate Professor in the Computer Sciences Department. The University of Strasburg was founded in 1621, with a long tradition of academic excellence. Louis Pasteur, John Calvin, Marc Bloch and four Nobel Prize winners have studied or taught there. We can now add Sven to the list…;-) Welcome Sven!
Of course, Nicholas and Sven will be able to demo MorphOS too!
Have a great day!
Raquel and Bill
[email protected]
“Well Beta was better than VHS – but which won ? and what’s around today ?
Beta is/was used in TV studios (probably more digital these days). “
Beta hasnt been used in most major tv stations in a really really long time. It was never super popular. For lowend use SVHS was used (Which came out a while after beta, and was better quality) but Betacam (no relation to beta) was used almost exculusivly in all major stations untill recently where we have professional DV vtrs, digital betacam, and HDCAM (digital high definition)…
A) There is no diff between RISC and CISC now.
B) The law of diminishing returns actually means that because there are no new instructions added the amount of space required to decode the SAME instructions gets relativly smaller and less important every year. (The artical is illogical .. either cause the writer is or because his trying to manipulate the reader.)
C) Altivec makes G4 a slow RISC computer. Thats why it cant scale well because of the COMPLEX Alitvec instruction set.
D) x86 benchmarks are invalid.. gawd what a load of bull….
This artical is a transparent piece of tripe .. trying to sell PPC chips and whatever else he is interseted in.
I am wondering, would IBM ever consider making their new chips the beguining of a PPC open platform where people can build their systems and other hardware vendors other than Apple can pump out these systems for the desktop?
This coupled with Linux or something would really give the consumer some choice and add more competition into the pot for Apple, Intel and AMD.
Just a thought.
Well Beta was better than VHS – but which won ? and what’s around today ?
Did you ever watch anything on Beta? Do you remember having to switch tapes in the middle of a movie because each Beta tape could only hold an hour of video?
The difference in picture quality was virtually undiscernable. Meanwhile, any movies over an hour in length had to be provided to the viewers on multiple tapes.
So please tell me, how is Beta better than VHS again?
Ok, we’re getting WAY off topic here, but…beta certinly, was, by all accounts (including my own) superior to VHS, in both video and audio. However, it was shorter in length (which might have been corrected had the format lived longer) and slightly more expensive because of the more complex cassette architecture. Had Beta survived longer, R&D would have come up with improvements, as is always the case with a financially successful technology. Don’t compare old Beta wares with current VHS…compare them with old VHS.
Beta was a home video format as was VHS. I forget what format most TV stations used but it was these mammoth 1 or 2 inch tapes. Beta & VHS are 1.5 inch.
A) There is no diff between RISC and CISC now.
B) The law of diminishing returns actually means that because there are no new instructions added the amount of space required to decode the SAME instructions gets relativly smaller and less important every year. (The artical is illogical .. either cause the writer is or because his trying to manipulate the reader.)
C) Altivec makes G4 a slow RISC computer. Thats why it cant scale well because of the COMPLEX Alitvec instruction set.
D) x86 benchmarks are invalid.. gawd what a load of bull….
This artical is a transparent piece of tripe .. trying to sell PPC chips and whatever else he is interseted in.
Did you even read the article?
The entire point is that A) is not true.
It is true to say that CISC uses the same techniques as RISC but the inefficiency cannot be hidden by an instruction decoder and consequently CISC has to expend a great deal more energy getting up to high performance rates.
B) What are you on about?
“LAW OF DIMINISHING RETURNS
An economic principle asserting that the application of additional units of any one input (labor, land, capital) to fixed amounts of the other inputs yields successively smaller increments in the output of a system of production. (Krippendorff)”
C) Perhaps you should tell IBM that.
D) If Apple produces a benchmark it automatically assumed to be fake – which I guess is only to be expected given their record.
However If Intel produced a benchmark it’s “gosh look how fast they are”. They should both be treated for what they are: benchmarks produced by a manufacturer of a product – no matter their intention – are marketing.
—
Heat:
Yes there are low power x86s but the latest G4s (7447) go down to 7.5 Watts at 1GHz. There’s nothing in the x86 world that even comes close to that level of performance at such a low power consumption.
VCRs
It’s quite correct that the best or most advanced technology does not always win the market, if it did we would all be Alpha users running BeOS with an Amiga AAAAA chipset…
Just hop to: http://www.aceshardware.com/list.jsp?id=4
Till then, why don’t you stop spamming the boards with irrelevancies ?
Interested, I checked out the website of MorphOS, in a paper about MorphOS “in Detail” it said the below. I think this would have been a big point in the article but it was not mentioned. Is it true and how does it work that it is 10x faster? And, more importantly, is that fast enough to provide a speedy OS?!
Thats what I was told yes.
However…
I was later told that actually deciding which contect switch to make takes a lot longer so it doesn’t have that much effect in the grand scheme of things.
However if you were running a lot of tasks and were switching rapidly between them it could then make an impact.
Also PowerPC is mainly used in embedded / real-time applications so this could be beneficial there.
Nice reading! Perhaps not always as objective as it could’ve been, but overall a very nice read.
“C) Perhaps you should tell IBM that. ”
I agree the guy is wrong but perhaps you should go look and see how many altivec equipped G4s IBM uses in its scaled up offerings. 🙂
Thanks for writing the most decent article around here in ages. About all this complaining though, you have to wonder if they could do any better. Im putting bets on no….
So please tell me, how is Beta better than VHS again?
They didn’t suffer from the Y2K problem
A Joke okay!!!
With All Due respect Nicholas Blachford, you article was full of half truths (selectively using out of date information!), and some really big errors.
I don’t understand it!! When will you learn that you don’t have to lie to show how good the PPC is.
Let the damn thing stand on it’s own two feet, I’m sure it will run…
A very good article and thoroughly enjoyable. It’s just a real shame that a minority of people on OSNews seem to have massive insecurities and are nit picking tiny little holes because they want to be the cleverest.
Well get over it, grow up and I await reading your absolutely perfect in every way articles that you manage to explain the entire concepts and history of CPU in full detail, never making one mistake (not even a typo) and fit it into a 5 page document. Go on – I dare you.
Oh, and don’t forget to reference all of your sources in a comprehensive bibliography making sure that none of the websites you referenced are ever going to be down!
Just want to bring up the point that Intel/Dell when producing there SPEC scores,
used the Microquill SMP Heap management library. A $1200 US Dollars piece of software, and,
ah, Dell isn’t packing that Microquill library with every P4 it ships.
So, don’t expect to get those SPEC scores on your machine.
Which, in my definition of a benchmark is Cheating.
Have you been Duped today?
There are always people who will concentrate on the most trivial points of a subject, and distort it to their own individual certainty- as depicted by a majority of the comments on this topic.
Most comments I have read are merely personal justifications of insecurities, combined with some factual information….
To the author; a fantastic read, thank you
“Incidentally, nothing in that link you provided suggest Motorola endorse a G4 at >1Ghz. Not to mention most of it emphasises using the G4 in embedded scenarios…”
http://e-www.motorola.com/webapp/sps/site/taxonomy.jsp?nodeId=03C1T…
Smithey look at the chart a little closer. It shows righ there that the G4 is running at 1GHZ. Apple uses the MPC7455 and MPC7457. Why is it so hard to admit you are wrong and in turn learn something? Now if you were to mention that G4s at 1.24 and 1.42 GHZ are overclocked then there might be some truth to your statement.
Motorola has a long history of being months behind in their web site doc. What Moto tells the world, and what they tell their real customers, are two different things.
Only Moto can tell us why.
Cherry picking: Intel also has a problem supplying the P4 3.2 ghz processor, i.e. cherry picking. So, what’s your point?
With California (and now Ontario) facing rolling blackouts due to power shortages you have to wonder how much power hungry PCs are contributing to the problem, power management may help, but when todays systems gobble up 10x the power that a system from 5 years ago did combined with more and more PCs being bought you know we’re headed for trouble.
My prediction: much like the auto industry is regulated for emissions, the PC industry will become regulated for power consumption.
The article is just not very good. How can you take an article seriously when it says something like “the x86 floating point unit is very weak”? Sir, it’s quite possible to rig up an x86 CPU with an extremely strong fpu. No one’s done it because the consumer market space doesn’t need it, and high clockspeeds can disguise it in consumer apps. Don’t confuse instruction set with good CPU design.
Another good one: “you can’t make a low-power consumption, low-end x86 CPU”. Try the C3.
Sorry, but the article is riddled with inaccuracies like those. The ArsTechnica articles BLOW IT AWAY.
-Erwos
“Heat:
Yes there are low power x86s but the latest G4s (7447) go down to 7.5 Watts at 1GHz. There’s nothing in the x86 world that even comes close to that level of performance at such a low power consumption.”
Thermal Design Power of a ultra-low voltage PentiumM at 1GHz is 7 Watts. While TDP is not equal to power usage, it does scale fairly linearly, so I doubt its power usage is above 10 Watts (could be wrong). The PentiumM performs better per clock than a P3, so it shouldn’t be too far off from the G4 (though certainly not quite as fast per clock). While not quite as fast and possibly not quite as energy efficient, this is certainly in the same ballpark as the G4.
RE: G5 Cube
I too would like to see a G5 Cube. It would probably need a fan (no convection cooling), but I think it could be kept fairly silent. BTW, I’ve heard that the Shuttle XPCs are NOT terribly silent. 🙁 You could probably mod it to be silent, but it might require lowering the clock and possibly the voltage of the CPU. Apple pretty much has the market cornered on cool looking AND quiet.
Actually, the general x86 FPU (referred to as the x87 because it used to be a separate coprocessor) pretty much sucks. The programming interface to the FPU is a REAL problem. This issue is largely fixed by SSE(2). With the P4, Intel designers pretty much decided to abandon the x87 FPU. It still works, but is REALLY slow (significantly slower than an Athlon). This is one reason that the P4 specifically is VERY dependent on SSE(2) usage.
Oh, and the C3 is able to be low power, but it sacrifices a LOT of performance to do this. I agree with your point about it being possible. The PentiumM is really the best example of this that I know of.
The instruction set DOES have an impact on the design of a processor, but that impact has been reduced through tricks suck as out-of-order execution and register renaming. Also, other factors such as memory bandwidth multiprocessor scalability have become more relevant over the years (these are more dependent on system rather than processor architecture). Designing a fast x86 processor is “harder” than designing a fast PowerPC processor. Intel has largely been able to stay ahead of Apple in the last few years because their fabrication and design capabilies were so far ahead of Motorolla. They still have an advantage in fabrication over IBM due to volume (higher volume = lower per unit cost), but technology wise, IBM is MUCH more competetive than Motorolla.
And again, yes, the ArsTechnica article rocks. THAT writer really knows what he is talking about.
Roy stated: “Apple pretty much has the market cornered on cool looking AND quiet”.
What? I have a dual-proc G4 here that rivals Larry Ellisons GulfStream V jet in noise.
It is cool looking, but I don’t agree that they have the market cornered.
a very good read, and full of actual information. shockingly so for some wintel trolls i notice, but that is their trouble, not mine.
again, thanks for the great read.
“This was added in the G4 CPUs but not to the G3s but these are now expected to get Altivec in a later revision.”
The G4 is simply a G3 with Altivec. Currently the G3 is the “consumer” processor while the G4 is the “Professional” processor. This is similar to the P4 vs Celeron.
Now that the G5 is around the G4 will become the “Consumer” processor.
I’m really not sure what role the G3 will have after the G5 ships but i wouldnt expect it to disapear untill apple gets a portable G5 on the market.
>>Dude — an apostrophe does not mean “watch out, here comes an ‘s’ !!” Posessive pronouns, “its, hers, yours,” do not have apostrophes. Use apostrophes when you are using a contraction, for instance “it’s” means “it is” and the apostrophe stands for the (space and) vowell. <<
You obviously do not understand English grammar then.
Jenny’s house. The house belonging to Jenny.
An apostrophe also represents posession. I could bore you with the reasons why, but will leave it at “English was formerly an inflected language; the >’s< is a hangover from this.” Suffice to say, some other languages of a Germanic origin do not use the apostrophe in this case (Swedish comes to mind), but propper English does.
I only wish that some of the posters worried more about the difference between ‘Your’, ‘You’re’ and such like rather than picking on an otherwise readable article.
You are correct that Nicholas has a few cases where he meant to write ‘Its’ rather than ‘it’s’, but on the whole his usage is good.
>> [I had to stop reading the article ] because it was making my head hurt. Can you say proofreading? spellcheck? A second grader writes with better grammar. Perhaps you need to put the pipe down a bit sooner before writing your next article.
Let me suggest looking two words up in the dictionary: effect and affect. <<
This is simply being picky. In British English (as I beleive Nicholas is originally from the UK) the difference between the words ‘Affect’ and ‘Effect’ is only in the written language. We don’t prescribe to the greater English speaking worlds tendancy to say ‘Ah-fect’ and ‘Ee-fect’. Both sound like the former to my ears (actually, technically we use the Schwa sound for the initial syllanble’s vowel sound, ‘@-fekt’ with the stress on the second syllable.) This would be like asking a middle wenterner to acknowledge the difference between Marry, Mary and Merry. Believe me all three sound different to my ears 😉
If you look at the root of the word ‘effect’ you’ll find that it actually is directly related to ‘affect’ and the fact that we choose to use both words is an anomily. Much like the use of both Dispatch and Despatch… both of which have exactly the same meaning. Whilst I realise that affect and effect have divergent meaning, it’s on a similar path. This happens a lot in liguistics 😉
Also, baring in mind British spelling and grammar were mostly in use, please realise that we do not speak in the same way over here.
I have gotten a card from Mary – I got a card from Mary.
I could use a cold one – I could do with a pint.
July 10, 2003 (july 10 [th]*) – 10th July 2003 (tenth of july)
* some US speakers seem to insist on dropping the ordinality of the date.
Ah well… tom-ah-to tom-ay-to.
The article starts off declaring PPC’s larger number of registers and then fails to say *why*. I find this disturbingly biased. RISC architectures are supposed to simplify code, and one of the ways it does this is by using register to register operations. With a CISC architecture, you can add two numbers together and put them in memory in one instruction. For a RISC architecture, you add two numbers, put them in a register, then do a second instruction to move the register contents to memory. This important fact was wholly omitted. Nor is it mentioned that the reason micro-ops are used is for pipelining purposes! Pipelining is an important feature of RISC. Of course, it isn’t mentioned that one of the reasons RISC and pipelining are associated is because the instructions need to all be of the same length, and how micro-ops gets around the fact that long CISC instructions impede pipelining. What’s even more is that RISC is there to help people write compilers, but it isn’t mentioned that because Intel is so large, they can throw plenty of money at the compilers. Microsoft also has no shortage of people to put to work on it. And then SMT is mentioned, but not explained in basic terms how it works.
This article is nothing more than biased junk. And I even like the new G5s. They are good processors, speedy, I’m sure.
“D) If Apple produces a benchmark it automatically assumed to be fake – which I guess is only to be expected given their record.
However If Intel produced a benchmark it’s “gosh look how fast they are”. They should both be treated for what they are: benchmarks produced by a manufacturer of a product – no matter their intention – are marketing. ”
Exactly why you should talk about benchmarks from real-world applications instead of SPEC. But then again, Adobe is inching away from PPC, and a number of benchmarkers have found x86 to be giving higher numbers in applications, including traditionally Mac ones. But don’t take my word for it.
http://www.macnn.com/news/18887
The sad thing is, I was eagerly looking at the tech paper for the G5, and Apple did a music software comparison, using Logic on the Mac and Cubase on the PC. Logic 5.1 works on the PC; in fact I use it on this very PC all the time. Why wouldn’t they compare results using the *same* software, since comparing results from two different programs is useless? My only conclusion can be that it didn’t give the results they wanted.
I thought the article was great. It summed up a lot of facts that obviously offend MANY people that are at the butt end of the article. But even more entertaining was reading the comments left by those people.
I bet the president of Intel himself could publicly declare that RISC is superior to CISC and the wintellies will still be unsatisfied. But for no real reason other than the fact that they laid down hard earned cash for their machines, and are almost obligated to defend their decisions, in whichever way they can. So its understandable.
And maybe they DO have real reasons. I know a lot of people do. They’re required to run software that runs only on intel machines, or RISC machines (we’ll name none specific) are too cost prohibitive. That too is understandable.
But after 12 years of working with Intel powerd machines (and a few AMD) I had nothing really tying me to them, other than years of experience. So I moved on.
“My prediction: much like the auto industry is regulated for emissions, the PC industry will become regulated for power consumption.”
You ever wonder how much juice those hybrid or electric cars suck up? Are you aware that the lead acid from those batteries is sure to destroy the environment?
The article is decent as far as an overview goes. However, it seems like a rehash of common computer architecture knowledge. In other words, it doesn’t give any special insight or even the depth one might expect. My intermediate level computer systems and architecture class went into much greater depth that this. This is not such a problem if not for the fact that the author is trying to compare two architectures and settle a very important (and heated) question. The article seem biased in many ways (one has to wonder if the author’s profession have something to do with this). The author undermines the credibility of the x86 claims by campanioning the PPC’s abilities. It is not so much a comparison as “Why the PPC is better than x86”. Lastly, the conclusions at the end, even if there is a source or two to support them, are very bold. The lack of substantial evidence does not warrant such strong claims. The author would do well to consider that DEC no longer exists, Sun is having trouble in the server market, and the Linux + x86 combination is gaining market shares. With these considerations in mind, one can hardly justify the claims made at the end. In conclusion, this article seem nothing more than another opinion piece inspired by Jobs’ recent claims disguised as a comparison of the two architecture.
That is the point, the CISC instructions are fat and require the CPU to break them down further in micro-code. In addition CISC instructions are not uniform in size which means that scheduling, out of order processing, decoding, etc. are a lot more expensive. The RISC vs CISC debate is long dead. RISC won a long time ago and x86 has been adopting as much RISC type design over the years as possible. The problem as the Author and even Intel has pointed out, is that the current generation of processors require compatibility with the original x86 CISC platform. Intel’s Itanium design finally walks away from that, but it has been a failure, last year selling only 700 units. It is going to be a herculean effort for Intel to break free from its legacy processors. IBM does not have these kinds of handcuffs on them, the PowerPC Architecture had the foresight to be well designed as a 64-bit RISC instruction set since its inception.
That is why the G5 is such a clean and elegant design.
It is important to note that PowerPC Specification does not dictate the actual design of a PowerPC processor. So Apple, IBM and Motorola are free to develop their own processor designs while being confident that binary compatibility will be maintained. Thus the G5 can have a different caching mechanism, different pipelining, scheduling, branch prediction algorithms, staging and pipelining. A high degree of parallelism (almost 20 to 1), significantly greater functional units, hardwired instructions such as square rooting which motorola performs in code. Yet, although designs are radically different, they are binary compatible. That is an incredible engineering feat, and illustrates why the PowerPC processors will continue to be advanced at a faster clip than legacy processors.
“You ever wonder how much juice those hybrid or electric cars suck up? Are you aware that the lead acid from those batteries is sure to destroy the environment?”
You’re right we’re screwed either way. So why bother doing anything.
The real problem is that there’s too many friggin people. We need a return to the good ‘ol days when entire generations would get wiped out by war or disease. These modern conflicts just ain’t cutting it.
Intel has largely been able to stay ahead of Apple in the last few years because their fabrication and design capabilies were so far ahead of Motorolla. They still have an advantage in fabrication over IBM due to volume (higher volume = lower per unit cost), but technology wise, IBM is MUCH more competetive than Motorolla.
Umm… earth to Roy…
IBM is the #1 manufacturer of integrated circuits in the world
IBM just opened the largest, most advanced chip fabrication facility in the world:
http://www.internetnews.com/infra/article.php/1437171
IBM is already manufacturing ICs using a 90nm process:
http://albany.bizjournals.com/albany/stories/2002/12/16/daily6.html
Intel’s Itanium design finally walks away from that, but it has been a failure, last year selling only 700 units.
I’m glad others are aware of what a dismal failure Itanium was. I always get a laugh out of people saying that the lingering big iron RISC vendors (i.e. Sun, SGI) should drop their designs and standardize on IA64.
Except SPARC and MIPS together outsold Itanium last year by two orders of magnitude. Way to go, Intel!
The fabrication disadvantage that I was referring to was ONLY volume, which affects cost per CPU. Are you suggesting that IBM will produce as many PPC970s as Intel produces P3s? As I said, technology wise, IBM and Intel are competetive. The G5 is significantly smaller on an equivalent process though, so this may make up for the volume difference. Also, some of pooling of resources from Intel competitors (IBM and AMD) may help alleviate this issue.
Itanium certainly got off to a bad start. Merced (Itanium 1) sucked big time. McKinley (Itanium 2 – largely an HP design, I think) is much more promising though it certainly isn’t taking the world by storm. Deerfield (very low cost version of McKinley – ~$800) can either be seen as interesting value or as a last ditch effor for IA64 to gain some marketshare, depending on whether you think Itanium has a future. I personally don’t know.
“as Intel produces P3s”
oops. Meant P4s.
…until you’ve designed your own processor!!! 8-bit, all the way!
Go CS curriculum, GO!
Nice article, by the way.
There are many good facts in this article but they are all in support of the opinion that the PowerPC architecture is better than x86. I have a couple of nits to pick with the implications of this article.
The author goes into great detail about power consumption and heat. The author neglects to point out why power consumption and waste heat are important. Any modern power supply can provide for even the most power hungry processor. So the only problem with high power consumption is the generated heat. If a cooling solution exists to dissipate the waste heat of the processor then there is no problem. That is currently the case for even the highest performing x86 processors.
I don’t think PPC has much of an edge in the heat problem; the new G5 case is optimized for high airflow. It is an excellent example of how engineering can work around a heat problem.
The other problem I have is that the author attempts to justify the Apple benchmarks of a Dell system. He states that the optimizations achieved by the ICC are unrealistic for normal software, and that GCC is at least as suboptimal for PPC as it is for x86. The author neglects to point out that the ICC is commonly available to software developers, who would be crazy not to take advantage of the provided optimizations.
If the author wants to argue that SPEC benchmarks are not relevant because they don’t compare to modern applications or that current PPC compilers aren’t very good then be my guest. These are good points, which is why SPEC is looking for new benchmarks right now. But there is no defense for how Apple crippled that Dell system to get benchmark figures they liked.
“That is the point, the CISC instructions are fat and require the CPU to break them down further in micro-code. In addition CISC instructions are not uniform in size which means that scheduling, out of order processing, decoding, etc. are a lot more expensive.”
No, not in addition. You’ve got it backwards. Micro-ops exist to get around these issues. The way micro-ops are done in hardware, the translation process is not the bottleneck slowing down the CPU.
Maybe I’m missing the point of your comment, but I don’t understand how this is different from the x86 world. The Pentium, Pentium Pro/2/3, Pentium4, PentiumM, Althon/AthlonXP, and the Via C3 all run compatible code, yet have all the same differences. This is true of ANY ISA with multiple generations of products. No doubt, the PowerPC ISA is superior to the x86 ISA, but I don’t see what that has to do with most of your comment.
IBM has a definite advantage in its ISA in that is doesn’t have to worry about binary compatibility with ancient procs (like the i386). Yet, this is the very reason why the x86 is king: backwards compatiblity sells chips. End users are eager to see better performance from existing applications and developers prefer it because it doesn’t disrupt the tool chain (compilers, profilers, debuggers, linkers, etc). Developers can gradually improve the apps as the user base migrates to the new iteration of chips. The P4 is heavily reliant on compiler vectorization and instruction scheduling and as a result initial P4 performance was poor due to the immaturity of the compilers. There is an obvious tradeoff in the form of the compiling for the most common arhitecture because releasing apps for specific x86 architectures is a pain (granted Red Hat and other Linux vendors do this with RPMs).
I’m a big fan of the PowerPC architecture but it’s not the second coming. It will give Apple a new lease on life in the PC market and provide Intel with its most serious non-x86 competition in years. My biggest question about the Apple benchmarks is why Apple failed to compare its top-end workstation (and they are workstations, people) with the top-end AMD x86-64 workstations? After all, Athlon MP workstations were demolishing the very best that Apple had in G4 workstations. I suspect the same would apply to x86-64 workstations. Apple’s reluctance might be attributable to AMD’s relationship with IBM. I like Apple but its highly selective and highly suspect benchmarking techniques are discouraging (not to mention all of the macaddicts who would do well to read some books by Patterson and Hennesey).
Finally, for a good article on the RISC/CISC debate take a look at: http://www.embedded.com/story/OEG20030205S0025
Smithey look at the chart a little closer. It shows righ there that the G4 is running at 1GHZ. Apple uses the MPC7455 and MPC7457. Why is it so hard to admit you are wrong and in turn learn something? Now if you were to mention that G4s at 1.24 and 1.42 GHZ are overclocked then there might be some truth to your statement.
Which part of the “>” in “[…] a G4 at >1Ghz” are you having trouble with ?
For those defending the article, I’d like you to read these two paragraphs:
“By using GCC Apple removed the compiler from the factors effecting system speed and gave a more direct CPU to CPU comparison. This is a better comparison if you just want to compare CPUs and prevents the CPU vendor from getting inflated results due to the compiler.
x86 CPUs may use all the tricks in the book to improve performance but for the reasons I explained above they remain inefficient and are not as fast as you may think or as benchmarks appear to indicate.”
Excuse me while I remove the bias from the factors effecting (sic) this post. okay, I’m done, wow that was easy. glad that’s over with.
I found that article by Nicholas Blachford a very informative read.
IBM’s web site confirms your power dissapation figures for the various 970 chip frequencies. Impressive work from Big Blue.
My only problem with the article is not the answers it provides in comparing PPC 970 v x86 (IA-32) but why make such an odd comparison in the first place?
Intel knows that IA-32 is nearing the end of its useful life, that is why they spent so much time and money together with HP to create the VLIW IA-64 arcitecture.
The correct comparison is Itanium 2 (and especially the upcoming “Deerfield” version) versus PPC 970. The reason is that these two have more similar 64-bit addressing, memory bandwidth, SMP scalability and floating point math capabilities, and both are manufactured in 130 nanometer copper processes.
Deerfield will have a much more similar cache size to PPC 970, but we will have to wait until the Intel chip comes out before the complete system testing could be done.
Why don’t Apple compare the Dual G5 (970) Mac against the allready shipping HP ZX-6000 Dual Itanium 2 Workstation??
I also think your implication about the IA-32 software base not making an eventual migration to IA-64 does not take into account the “IA-32 execution layer” (code name: btrans) feature coming out for linux-64 and Windows XP 64-bit.
Intel claims that IA-32 software run on an IA-64 system via the “IA-32 execution layer” will provide a GHz equivalent, ie. 1.5 GHz Itanium 2 will run IA-32 software at the same speed as a 1.5 GHz Pentium 4. Thats good enough IMHO.
When memory technology improves to the point when an 8GB memory module is the minimium size commonly produced, then that is the time for the 64-bit chips to take over in new desktop computer systems sold.
I think that by the time that happens then the current IA-64 chip together with the “IA-32 execution layer” will be very, very fast indeed. Then IA-32 software will migate with little or no resistance.
If necessary the Intel cash reserve will be partly used to use differential pricing to boost IA-64 and kill off IA-32 in the same way it killed off the 486 with aggressive Pentium pricing. Since its a battle Intel cannot afford to loose, they WILL go all out.
Then IMHO we have a “Final Unity” in which Intel can drop IA-32, and go all out on IA-64 production in an attempt to crush the PPC with sheer economy of scale.
At the end of the day, software availability is the key factor. PPC is a great technology, but without plentiful native software its just another Betamax.
The Itanium 2 should be compared to the Power4 not the 970. The Power4 and the Itanium 2 are the natural competitors. And they have been compared and the Power4 whips the Itanium in every test. Nothing emotional here, those are just the facts. In order for the Itanium to match the Power4’s performance, they require a 128-way Server against IBM’s 32-way server.
Really AMD has not released a competitor for the 970 either. The Opteron competes also against the Power 4 and Itanium families. Athlon 64 is the competing product and that will not be out till mid-September, and I hope it does provide healthy competition for the 970,
Software availability is a key factor and since there is now Windows XP 64 yet and Microsoft has no plans to develop a 64-bit desktop OS, with longhorn still being 32-bit, it looks like Athlon 64 will be running some form of Linux. Even if it is capable of running a modified version of Windows XP, it will still be crippled to the IA-32 instruction set. The 970 contains a special 64-bit instruction bridge that allows OS X to access the entire 64-bit address space and all of the 970 instruction set, even the 64-bit instructions. In addition 64-bit Apps will run fine over Apple’s current OS X version (10.2.7) and 10.3 (Panther).
When it comes to software OS X has an abundance of native software, and all the Unix programs you can stick a fork in, plus is compatible with OS 9 software and with a software emulator runs most Windows software. And, if the software emulator bothers you, that is exactly what Intel is doing with Itanium to run Windows 32-bit apps!.
OS X also has some of the most innovative software around today in almost every space. Apple’s development frameworks have stimulated one of the most exciting development environments, popping out software faster than you can blink. Stay up with the times.
This is perhaps the most thoughtful, knowledgable, fair and balanced article that I have come across on this subject. This is an excellent article and a must read.
Other fanatical and totally biased and propagantised articles don’t help me as a consumer, but just lead me astray. This objective and incisive article is what I need.
Thank you for a great job on this article. I hope there will be more objective pieces like this in the future.
I’m not even here to defend the Intel and MS world but this article is ridiculous. All he did was hit up corporate websites and quoted them, and even worse all he did was hit up the websites of people who agreed with him in the first place (that RISC is great). “The IBM site says”, my god man, thats called marketing, theres nothing factual about it, writing an entire paper based on marketing is the most absurd thing I’ve ever seen. And the author REALLY CLEARLY has a bias going into this article, its glaring from the beginning.
Here’s a tip. Get out your stopwatch and perform common tasks between the most powerful Mac and the most powerful PC. Load a web browser, load a web page, load a document, etc etc and you’ll notice something. They will perform very similarly, yet one costs a good lord lot more than the other, and one runs a good lord lot more software than the other. Its just economics, the tasks are performed similarly, but one is a LOT less expensive.
And finally, those “small groups” who really need alot of CPU power aren’t small at all. Office/Internet software has run just fine since the 1ghz days. Games are driving the x86 world for the most part, and they’re not a small group at all (outsold the movie industry JUST considering software).
In conclusion, I can’t believe you wasted your time writing this piece of propaganda instead of sticking to the technical +’s and -‘s of each one. If you had stuck to the technical differences this could have been a mildly interesting read, even if its just a rehash of what other people have already written. Throwing in your blatantly biased opinion turned this into a grade school report, where the child copies quotes from the internet and pastes them as fact. Shame on you.
Whaterver is better, whether RISC or CISC, why can’t I look at some hardware shop and buy a RISC cpu + mobo ?
If i’d like to buy a RISC arch, i’ll have to look at SGI (mips) or apple.com (ibm/motorola), wich ARE extremely expensive!.
Might happen that Motorola or IBM sell their CPUs like an AMD/INTeL one? What ’bout mobos?
“Software availability is a key factor and since there is now Windows XP 64 yet and Microsoft has no plans to develop a 64-bit desktop OS, with longhorn still being 32-bit, it looks like Athlon 64 will be running some form of Linux.”
So which 64 bit Windows OS is it that’s not in the making? This one for x86-64:
http://www.amd.com/us-en/Processors/ProductInformation/0,,30_118_88…
Or this one for IA-64:
http://www.microsoft.com/windowsserver2003/64bit/overview/default.m…
And the x86-64 Windows has been spotted running on a demo machine. How functional and stable it is, who knows, but it does exist.
Please stop making statements about things without doing even basic research.
IBM was pushing CHRP for the PPC line. The bios is openstandard and the board can be openstandard. Only thing needed is a company other than IBM or Apple to build it. Osnews posted a photo of the booting of a PPC a few weeks ago and the bios said chrp enabled. Hummmm.
Actually an Alpha vs G5/970 would be better to compair in terms of real numbers. I think we all suspect the PPC would trounce the pants off the Alpha. Shame, I liked Alpha’s but with no native OS the design couldn’t win.
unless you are a ee developing processors, you really shouldn’t care about the architecture. to argue one way or the other is pointless. want to know why? because NONE of you have access to the proprietary multi-billion dollar research and development that has been invested by corporations in order to determine the most optimal way to process information. the fact remains that different companies took different development paths, each architecture does have its strengths and weakness when compared. people drum up the relative importance of the strengths and weaknesses while discounting others to make one seem better than the other. the fact remains that they are both fairly equal. if one was in real world situations vastly superior the industry would move to it, but that is not the case. that is also why apples and x86 both perform about the same in the real world – with differences due mainly to development decisions and not architecture i.e. heat vs power vs performance vs die size vs intended market etc. you need to decide what to buy based on real world performance and price. because of economies of scale and because apple is 100% proprietary the x86 will remain the cost/performance leader.
Really needed comparisons on things like cache efficiency, branch prediction etc (some of the major areas of performance on modern CPUs). Also, it used the old model of the RISC vs CISC argument, which has long been doubtful (CISC instructions are more expensive, but RISC often has a higher instruction count). Didn’t go into things like register renaming etc (if you’re going to compare register count, going into register renaming on the x86 is very important as it gets around the register limits).
Also, MMX uses 64 bit registers (it re-uses the floating point registers), but SSE and SSE2 are 128bit and don’t reuse.
Finally, I would’ve liked to have seen a few words about VLIW.
Otherwise good and fairly well done – especially the power and heat consumption bits.