A while ago I wrote a short article on a rumour about Apple potentially using a POWER4 derived CPU. Since then there has been a lot of talk of Apple switching to an x86 based solution and now a desktop POWER4 derivative has been announced. Some people suggest this will be too expensive and in it’s cut down form slower than a x86. What is Apple more likely to go for?
The speculation about Apple going x86 is based mainly around a comment by Apple CEO Steve Jobs in Apple’s Q3 Financial Analyst Meeting Q&A here.
“Then we’ll have options and we like to have options”
What seems to be missing is the rest of his comment and this puts things into much clearer perspective:
“The roadmap on the PowerPC actually looks pretty good and there are some advantages to it. As an example, the PowerPC has something in it called AltiVec, we call the Velocity Engine — it’s a vector engine —
it dramatically accelerates media, much better than, as an example, the Intel processors or the AMD processors… so we actually eke out a fair amount of performance from these things when all is said and done. And the roadmap looks pretty good. Now, as you point out, once our transition to Mac OS 10 is complete, which I expect will be around the end of this year or sometime early next year and we get the top 20% of
our installed base running 10, and I think the next 20 will come very rapidly after that. Then we’ll have options, then we’ll have options and we like to have options. But right now, between Motorola and IBM,
the roadmap looks pretty decent.”
Thats the full quote and to me it doesn’t sound very much like there’s going to be an Apple with x86 flavouring anytime soon.
We know Motorola are working on a G5 PPC processor and the most recent rumours peg it at 2.5GHz. Motorola’s roadmap states it as being 64 bit and having a 13 stage pipeline. The jump from 1GHz to 2.5GHz
may seem unduly vast however the longer pipeline and 0.13um manufacturing technology will allow such clock speeds so 2.5GHz does not sound unreasonable. We can expect to see similar clock rates from AMDs hammer when it arrives using as it has a similar length pipeline and the silicon process is the same (Motorola and AMD are process partners) – The current rumour is Hammer will start shipping around 2.6GHz (or in sales speak “3400+”). Apart from that Motorolas roadmap doesn’t give much away although we can expect higher bandwidth busses and backward compatibility.
However IBM has not been inactive and have not only announced a “cut down” version of the POWER4 but will be giving out details at Octobers Microprocessor
Forum, an industry leading event where new microprocessors and technologies are often announced.
Notably IBM are pegging the new CPU as a “desktop” processor and I don’t know of many other companies in the desktop PowerPC space other than Apple. I think however we can safely assume that IBM has not produced this CPU for the soon to be resurrected Amiga platform (but thats not to say they couldn’t utilize it).
Another notable feature of the new processor is the addition of vector instructions. I pointed in the previous article out one of the problems of the existing POWER4 CPU was the lack of support for Altivec
instructions. It would appear that this problem has already been taken care of although we don’t at this point know it they are the same instructions as used in Altivec but I for one would be more than surprised if they were not.
It will be interesting to see if dual precision floating point instructions are featured in the vector unit, these are featured in Intel’s SSE2 but not in Altivec, Altivec added 162 instructions whereas IBM have mentioned 160 so this may not be the case. That said the POWER4 FPU is a mighty beast already capable of issuing 4 floating point operations per cycle so it’s questionable if dual precision vector instructions would actually be necessary.
Performance and price.
Some people get the idea POWER4 is a horribly expensive CPU and oddly enough they’d be right. However they also assume that it will remain expensive and even if it were to be “cut down” it’s performance will
plummet to levels below that of the x86 processors.
I disagree on both counts:
Firstly POWER4 in it’s current form is horribly expensive yes, but thats probably because it’s shipped as an 8 way SMP system. Yes thats right, the smallest POWER4 has 8 CPUs. Cutting that down in price is going to be easy, dump 6 of those CPUs and the large Multi chip packaging and you’ve just cut a great chunk off the price.
It also has 128MB cache, 32MB per CPU pair so you can dump 96MB of this as well. This will impact performance but probably by not a significant margin. Like everything the law of diminishing returns applies so going from 32 to 128MB cache isn’t likely to have a massive impact although this very much depends on the application in use.
There will be a big change in the memory interface however. The POWER4 having 8 CPUs needs a large fast memory system to keep those processors fed with data and this is implemented in a special memory controller attached to the L3 Cache chips. This controller has multiple 400MHz ports going out to main memory.
This isn’t going to be necessary for a single processor and removing it will also cut costs further. IBM however have not skimped on the memory system however as the new PowerPC has a bandwidth 6.4GB per second. 50% higher then the current top Pentium 4, and 15% higher than the forthcoming AMD Opteron. Apples professional target markets (audio/visual) need bandwidth by the bucket load and this is sure a great improvement over the current 1GB per Second (although this figure looks like it might go upwards rapidly in a couple of weeks).
Performance.
Cutting the Cache and memory bandwidth will have an effect on performance but as above the law of diminishing returns applies, I don’t think the performance drop will be very great and certainly nothing a die shrink to 0.13um wont fix. Also current SPEC marks for the POWER4 are for a single processor, the single chip version may keep the dual cores of the full POWER4. A single core may be behind the Itanium 2 in SPEC marks but once the second core kicks in I can see it showing Itanium 2 a clean pair of heels.
A G5 will bring performance levels back up to those delivered by the x86 CPUs, a POWER4 CPU will go straight past. Yes they will still not be cheap but who says Apple hardware is cheap? I can see a POWER4
based CPU going into the PowerMacs at least at the top end (or possibly a new, more expensive Super-PowerMac). You wont however see them appearing in your iMac anytime soon and they’ll have problems fitting them into a laptop unless they cut the clock speed whilst on batteries.
As to the question of which CPU Apple will use next, read the CEOs comments:
“But right now, between Motorola and IBM, the roadmap looks pretty
decent.”
Note he mentions Motorola and IBM, he never mentioned AMD or Intel producing CPUs. Some people have mentioned x86 as a future CPU for Macs, Others have suggested Itanuim. Switching the CPU to POWER4 –
even just at the top end – will not only be a great deal easier but compared to x86 IBMs new POWER4 derivative will outgun them, How? The current SPEC marks for POWER4 are for a 0.18um chip, whereas Intels are for 0.13um expect them to leap upwards when POWER4 transitions to 0.13um.
About the Author:
Nicholas Blachford is a Software Engineer / Architect currently living in Amsterdam. He has numerous geeky interests including CPUs. He does not own a Mac but wants one.
Apple’s path of least risk is to go with IBM’s new POWER RISC desktop chip (POWER4i I am calling it). Apple is the POWER/PowerPC volume platform.
Apple will get good performance and have a good reason to raise prices. So, Apple marketing and Apple bean counting are both happy campers.
Although it would be a huge fun adventure to go with AMD or Intel, Apple does not have the guts to do this. PowerPC is a core comptency, x86 is not.
What strikes me as intriguing, though is the coincidence between Microsoft’s Longhorn announcement and the availability of the cheaper 64 bit POWER4i. As you may recall, Microsoft has said that they will be supporting “3 to 5” 64 bit architectures.
Itanium – already done – 1
AMD Hammer – already done – 2
IBM POWER4i – ??? – 3
Intel x86-64 – ??? – 4
??? – ??? – 5
Perhaps IBM will be doing something very interesting and creating a new 64 bit PC. With the huge architecture changes afoot in the PC world, this is the time to do it.
Selling commodity hardware isn’t working for IBM. But creating a new PC based on POWER4i and Longhorn?? With .NET this wild speculation becomes more possible.
As for Apple, I hope they get new machines out soon. They are far behind when it comes to processor speed and platform innovation.
#m
Step 1: Pushing Linux
Step 2: Making desktop PowerPC chips
Step 3: ???
I still don’t know what are they gonna do with them PowerPC chips. Make pin compatible ones with P4 chips? IBM is dumping their PC computer division. I am still baffled by this. I still think analyst are off base. TO WHOM ARE THEY GONNA SELL THESE DESKTOP CHIPS?
>>I still don’t know what are they gonna do with them PowerPC chips. Make pin compatible ones with P4 chips? IBM is dumping their PC computer division. I am still baffled by this. I still think analyst are off base. TO WHOM ARE THEY GONNA SELL THESE DESKTOP CHIPS?
Apple, according to the article.
Anybody who can figure any reason whatsoever for Apple to raise prices even MORE is out of this world… Hello, Apple has 2,5% market share – and why is that? Have a guess…
Told Ya!
That’s according to the article. No press release, no proof. I want to know. Is Mac OS X even ready for a 64-bit PowerPC chip?
Darwin is. In the begenning of 2001 (If I recall correctly) some discussion on the darwin mailing list took care of 64 bit cleaneness.
—
http://islande.hirlimann.net
> That’s according to the article. No press release, no
> proof. I want to know. Is Mac OS X even ready for a 64-
> bit PowerPC chip?
BSD runs 64 bit. There are plenty of Machs that run 64 bit. As mentioned by Ludovic the darwin group has checked for 64 bit. gcc is used all the time for 64 bit. OpenGL host platform is currently on 64 bit. Porting window managers isn’t hard. I’d assume though I don’t know that Carbon and Cocoa are 64 bit clean. Getting OSX (with the possible exception of the classic component) on 64 bit is not a problem.
Well IBM has a .1 micron fab up and running now. There is no reason why they won’t use it here. Frankly this appears to be a full replacement for the power4, possibly making the power5 a “Xeon” type derivitive of the new chip. Creating cheaper and more scalable servers, that make the whole range of power servers much more marketable and ween costumers off Intel. If they pull all the stops 3ghz is certainly possible. I don’t see the consumer as a tarket market, but the rest of the nix vendors and the Itanium, which IBM doesn’t like very much. We can only hope that IBM provides an open achitecture and sells the chips to third parties. (which is what I think they are referencing when the say it will be a “desktop” chip) Perhaps all the recent powerpc projects like the Amigaone have resarted the intitial idea that IBM had of an open powerpc platform. I am not sure weather Apple will bite. Apple has made some terrible hardware decisions lately. My bet as for the price, it will probly have a better Price/performace than Itanium, and perhaps even the AMD opteron (as much of the development cost on the chip, have been spent on the years of Power(PC) development. IBM may push for Linux on it and use it to anniliate the professional workstation market. One think is for sure, there is no way IBM will ever try to compete in the general consumer market.
To conclude this chip appears to be everything the Alpha should have been, but will actually succeed as it has a solid codebase and a strong company backing it.
go with x86
i’m sure apple’s engineers are as disgusted with this option as many seasoned mac users are. it would lower prices and appear to raise performance at the cost of massive increases in heat dissapation and decreases in battery life.
stick with ppc
ibm and motorola are going to produce these new chips and then apple will decide. with OS X it is possible to produce “fat” binaries which can support many architectures in the same application, so moving to a 64 bit processor will be easier than people think. primarily apple want to reclaim that graphics niche that they have been losing out on and the best way is to put something (expensive) on the desktop which beats the pants off x86 and something (light, long battery life) for the road.
i think we’ll see smaller G3/4s for some time to come living in iBooks and powerbooks and something heavy on the desktop in the coming years.
I recall reading somewhere that back in 91/92 that the Apple engineers were considering using Alpha as their next CPU after m68k, anyway to cut a long story short the then CEO of Apple had dinner with the CEO of DEC where the matter of Apple using Alpha’s was dicussed. What i heard happened was that the DEC CEO basically said no, cause Alpha was a new platform and he didn’t want DEC to be dependent on it sales (he was supposedly a big VAX man himself)
Moral of the story, if he hadn’t being so short sighted no doubt DEC would never have being bought by Compaq and Mac would have being a 64bit desktop for years now. Imagine an apple based on a Alpha EV8
BeOS running on an 8 processor xeon.. oh wait.. we don’t have to. It happened like 5 years ago.
Close your Windows and Eat your Apple. Go OpenBeOS!!
Well IBM has a .1 micron fab up and running now.
You should probably write it like 0.10 so as not to confuse people with 0.01 (It confused 95% of slashdotters).
Intel x86-64 – ??? – 4
Vapourware. Intel said there is no such thing as Yamhill, nor is there any plans to make a X86-64 processor.
Selling commodity hardware isn’t working for IBM. But creating a new PC based on POWER4i and Longhorn?? With .NET this wild speculation becomes more possible.
Microsoft said it would support 3-4 64-bit architectures. It didn’t say with its desktop offering. Most likely, it is with their server offering.
That’s according to the article. No press release, no proof. I want to know. Is Mac OS X even ready for a 64-bit PowerPC chip?
That’s exactly what I though of when I read the article about IBM’s thingy. Sure, even CNet has the report, but I’m still spectical about Apple not moving to x86. Sure, IBM could perhaps pump out 2GHz, but if history is any indicator, Intel and AMD would outbeat it in a year ot maybe two.
that make the whole range of power servers much more marketable and ween costumers off Intel.
IBM SELLS Intel-based servers for the low end.
it would lower prices and appear to raise performance at the cost of massive increases in heat dissapation and decreases in battery life.
God, this is so old. By time Apple finishes its migration to Mac OS X, at the rate things are going, Intel’s and if the rumours are correct, AMD’s processor would have heat dissapation as low as current G4 offerings. As for battery life, that isn’t a problem. My laptop has a max of 6 hours (but only under Windows XP), and that could be higher…
All the low-end machines have been moved to G4 except for iBook.
The Power4i would slot in really well, providing a reason for Apple thralls to fork over the big bucks for a new PowerMac. If the new PowerMac were to ship with a 64 bit version of OS X, it would provide the cachet that Apple has been missing for a long time. Early next year would be plenty soon as AMD is not shipping any volume of Hammers until 1H03.
Apple could become a world leader if they went x86, but as someone insightfully remarked before, Steve Jobs doesn’t care about market share. He wants to be the all powerful monarch of a small scuderia that makes luxury computers and entertainment devices.
#m
Many of you will remember that IBM bite the dust with their OS/2. If you ever have the chance to talk to any IBM engineer, you will see that the pain is still there and it’s deep, very deep.
Microsoft hurted IBM on more than one occasion, and IBM remembers it. There’s more than one employee/engineer/executive at IBM that would like Microsoft to suffer for their deed.
I can picture IBM deciding to be bold and take a sweep at an impressive part of the market. Mac OS X is ready for the desktop, unlike Linux/X11 atm. An IBM backed Mac OS X could literally shatter the current desktop market trends.
IBM powering the desktop computer? I can see it, I drool for it.
Microsoft hurted IBM on more than one occasion, and IBM remembers it. There’s more than one employee/engineer/executive at IBM that would like Microsoft to suffer for their deed.
Microsoft only “shattered” IBM’s dream dirrectly only once, with Windows 3.1, which ended in the divorce between IBM and Microsoft. But really, the divorce was caused by IBM, ultimately. Microsoft had spent more money promoting OS/2 than Windows 3, but IBM’s attitude towards clones [read: pricing] ultimately caused Windows to win.
So I would say, most of the fault lies at IBM. Sure, you could bring up SmartSuite, database etc. but that’s what competitors ultimately do to each other. If IBM was really planning to commit revenge, why aren’t they planning to do so to Oracle and Sun before Microsoft?
Why is it nearly everyone is assuming IBM is building this chip for Apple and OS X? With all the moves and money IBM has put towards Linux, I think they’re creating a new desktop computer based on Linux. Think about it – if IBM goes with Apple the future market for their chip is based on the future of Apple. But if they go with Linux they have much better control over the marketing and programming for their chip. IBM likes control.
I’m predicting a new Linux-based workstation – not OS X. That to me makes much more sense.
Why not the two? Doesn’t it make much much more sense?
Seriously, the sucess of Linux on the desktop is not guaranteed. The sales to Apple are. And for servers too.
I could see it working with OSX, but I still doubt that will be the primary OS run on it.
Linux on the every man’s desktop is not guaranteed. However, Linux as a workstation is pretty much guaranteed. You can see converts to it all over the place. The chip IBM is coming out with screams WORKSTATION.
Oh, and sales to Apple are not guaranteed. They’ve been losing marketshare for a long, long time. Could more power change that? Possibly, but I really doubt it. Apple’s problems (IMHO) are more based on pricing than power.
Not to be mean or anything, but will this rumoured future super chip really have the same kind of performance that future AMD/iNTEL offerings will have? Could it battle it out with a 3.0Ghz P4? What can it do against an Itanium 2 or even 3? The new Athlon XPs and Opterons?
Why would IBM want to compete directly with AMD/iNTEL, which own the desktop computer market by now? There are so many ifs and whys surrounding the successor CPU for Apple, and it changes all the time. Rumours are all fine an interesting, but it is not until the hardware hits the street that the company behind it can make money, or claim that it is faster than the competitors.
With the POWER chips Mhz isn’t an issue….It crushes and stomps everything in it’s path
Couple of totally ridiculous assumptions:
1) The desktop processor will have dual cores: In desktop space, there are a limited number of applications that can take advantage of multiprocessing, thus it is a very likely possibility that the Power4-jr will be a single core CPU. Besides, if it takes dual cores to push Power4 over the x86 performance edge, IBM is screwed. The dual core CPU pushes up transistor counts, which lowers yeild, which pushes up cost. AMD’s Hammer will have glueless SMP, which will make 2-way machines quite cheap. Thus, if the Power4 comes in somewhere around $1000, a dual Hammer setup would be very price-competitive, and beat it soundly in performance.
2) The cut-down version will have 32MB of cache: No way. Do you realize how many transistors that takes up? If you look at die pictures, the cache is the single largest component of a modern CPU, and we’re talking the 512K things on P4 processors. Plus, Intel has better fab capability at the small-die, huge volume level than IBM does, so if they can’t get much more than 512K on a CPU at a decent price, IBM’s Power4-jr will certainly come in at less than 2MB.
3) IBM can keep the clock-speed heat on: If you take a look at the benchmarks, a Power4 is only about 50% faster in floating point. Even if the Power4-jr can keep that kind of performance (the cut-downs offset by increased clockspeed from a smaller process) how long can it hold the lead? Intel and AMD are clock-speed demons. Their business model is built around doubling CPU clock very often. Can IBM keep up in that kind of race? All the other RISC manufacturers have fallen behind x86 in performance due to its sheer clockspeed. The only thing saving them is the inferiority of the hardware surrounding most x86 CPUs.
first of all, it’s a scaled-down power4 so it’s gonna be slower than a power4…NOT necessarily faster than x86s…second it’s probably gonna cost a good chunk b/c it can’t and won’t have the economies of scale that Intel and AMD can have or even Motorolla (Motorolla sells most of their chips for embedded development, Apple’s not really a crucial Motorolla customer and it shows).
i seriously doubt that this chip will have a better price/performance ratio than a P4…this chip is probably not being made for Apple…if Apple signs on to it…great for IBM…but Big Blue is probably making a new line of AIX/Linux servers and workerstations…NOT commodity computers…IBM making commodity computers would be insane!
it’s a nice dream, but i think that if you think this is gonna be Apple’s saviour you’ll find yourself very dissapointed
-bytes256
That’s what it’s going to be.
Something similar to Sun’s Blade stations.
IBM already put enough compatibility with Linux in AIX 5L.
New release of AIX will be even more compatible with Linux,
may be it will have OpenOffice ported to it.
Apple may be going with Power4 but it’s not IBM’s decision.
IBM cannot say “we have developed CPU for desktop” for no reason – and so far Apple didn’t give them that reason.
Apple was saying “may be”. So I conclude that IBM has another desktop system in mind – and it’s not OS/2.
Will IBM make their own Linux built for 64-bit Power4 ? May be – they have invested enough R&D in Linux and not all of it is going for Linux on OS/390.
>>>Apple is the POWER/PowerPC volume platform.
Cisco and Lucent sell more PowerPC systems than Apple.
>>>>Microsoft hurted IBM on more than one occasion, and IBM remembers it. There’s more than one employee/engineer/executive at IBM that would like Microsoft to suffer for their deed.
>>>>If IBM was really planning to commit revenge, why aren’t they planning to do so to Oracle and Sun before Microsoft?
Where have you people been? IBM and Microsoft has been best buddies and outplaying Oracle and SUN on the web services standards: SOAP, WSDL, WS-I, WDSL, UDDI, …..
>>>>With all the moves and money IBM has put towards Linux, I think they’re creating a new desktop computer based on Linux.
IBM’s ultimate goal is to have ONE single hardware (POWER6 running from workstations to mainframes) and ONE single OS (AIX running from workstation to mainframes). IBM has been incorporating all the mainframe tricks, partitioning and self-healing things into AIX, not linux.
Well…scratching head…there are many pros and cons. However, it is clear that IBM has not been idle. And, Motorola is continuing work on the G5 (or whatever they end up calling it). So, it appears that there are at least some options out there. Very interesting options. I try to look long term too, but I also look short term, especially when Apple is in a pickle about their current processor speeds. I still want to see what the next incremental upgrade is. Apple has to do something. Right now my Microtel PC has faster RAM than the high end Power Mac. Well anyway, I do believe the next short term upgrade will tell us alot. If it’s puny, Apple must have something else in the works. They cannot continue as they have been.
If Mac OS X is going to be able to run on the Power4 why wouldn’t IBM and Apple team up to put it on the larger chips?
AIX is nice but for some segment of the potential IBM marketplace, they’re going to want more MS compatibility with that IBM hardware than AIX seems to be going for. Mac OS X on those systems might be an interesting offering that would give Apple a small but nice extra revenue hit but would certainly give them huge credibility in a lot of markets they’ve either not been in for awhile or *never* have been in.
While i do think that moving Mac OSX to a POWER derivative would be neat and a better idea than x86, i think the exciting part should be the fact that IBM wants to make the a workstation type chip. This will make some competition in the 64bit market for workstations and lowerend servers. Force x86-64 and IA-64 and all the other competitors to be better. I love competition, its better for everyone if you ask me. I wonder though if if MS is going to try again at the workstation market (anyone remember NT Workstation for Alpha?).
If apple and Ibm hope to have a chance to compete with the x86 crowd then they need volumes.
that means that it makes sense, from IBm’s perspective, to offer that chip to as many OS’s and vendors as possible, including linux. That would help apple too by driving costs down but it could also injure apple by competing with OS X and apple hardware.
It will be interesting to see this take shape. The two could just go after different segments of the market and try to avoid head to head competition. Ibm might lean towards the enterprise while apple sticks to the creative and consumer.
I don’t doubt that IBM is gunning for MS and Intel so we’ll see what happens.
Moral of the story, if he hadn’t being so short sighted no doubt DEC would never have being bought by Compaq
DEC died because Ken Olsen was an engineer and wasn’t cut out to run a company. He did some brilliant stuff with the PDP-1, but unfortunately had his head stuck up his ass on the whole Unix issue, continuing to push VMS when it was obvious Unix had become the commodity platform. Having completely lost touch with reality, he blindly ran the whole company into the ground, where its shattered remains were bought up by Compaq.
Revolutions devour their offspring.
IBM is:
“Step 1: Pushing Linux
Step 2: Making desktop PowerPC chips
Step 3: ???
I still don’t know what [IBM is] gonna do with them PowerPC chips. Make pin compatible ones with P4 chips? IBM is dumping their PC computer division.”
Well, their PC division has been a money loser for a while. Why keep it? Let Hell, I mean Dell, sucker the gray suit market.
“I’m predicting a new Linux-based workstation – not OS X. That to me makes much more sense.”
A winning strategy: IBM= big to medium metal; LINUX. Apple= medium to small metal, plus desktop; same chips; OS X. Motorola= cell phones and Furbys.
” first of all, it’s a scaled-down power4 so it’s gonna be slower than a power4…NOT necessarily faster than x86s…”
I wouldn’t bet that way, particularly since Intel’s 64 bit foray has been pretty anemic so far.
“second it’s probably gonna cost a good chunk b/c it can’t and won’t have the economies of scale that Intel and AMD can have or even Motorolla (Motorolla sells most of their chips for embedded development, Apple’s not really a crucial Motorolla customer and it shows). ”
The economy comes out of the greatly reduced number of transistors you use with RISC vs. CISC– cheaper fabrication.
I hope this IBM rumor is for real….
Seriously, the sucess of Linux on the desktop is not guaranteed.
In emerging markets (China, India etc.), success is almost a fact. How well would IBM tap into the emerging markets?
With the POWER chips Mhz isn’t an issue….It crushes and stomps everything in it’s path
But with something like, say, 120mhz, even with a high IPC, I doubt it could crush and stomp everything in its path.
MSFT pulled out, fearing that people who could run Mac and NT on the same box would soon become pretty disenchanted with NT.
A clear disregard for history. Both IBM and Microsoft pulled out because of Apple. Remember Apple killed Copland? Plus, Apple adopted the CHRP so late that Microsoft already gave up on the PPC port of NT because of lack of sales and Apple refusing the pay Microsoft for the loss of profits. (Be also left the platform partly because of Apple – or it is just one of the many excuses, and the reason why we are using x86 mostly is because of Apple).
The prices of the chips will inevitably fall with time as they get “older” (read: out for more than 2 months) and are produced in larger quantities.
The main reason why I think x86 is a better choice.
Some of the quotes above came from here, and the rest came from http://www.osnews.com/comment.php?news_id=1506&offset=0&rows=15 (should go to bed soon).
IBM will not make this new processor exclusively for Linux. IBM already ships Linux on servers and workstations why would they reduplicate the effort on PowerPC.
PPC Linux and derivatives do not even use “altivec like” acceleration in the OS with the exception of MacOS X, there is no way this processor is exclusive for Linux.
You wonder why IBM would make this exclusive to Linux? Control. If they make it exclusive to OSX they have no control over which direction the OS takes, how well it’s marketed, etc.
As to why they would ship this new processor with Linux when they already have workstations with Linux is simple: processor power. If this chip blows x86 out of the water then they have a killer platform. Altivec-like acceleration works perfectly with programs like CAD, engineering, 3D, video and imaging. Guess what most high-powered workstations are being used for?
IBM uses PPC604s in it’s low-end severs and workstations. Altivec is designed for graphic & multimedia applications in workstations. IBM is creating the POWER4i for their workstation market; since no one is going to by an 8-way workstations which is what the POWER4 would be. It looks like IBM has came around and decided to adopt Mot’s Altivec design (they had access to it through the Apple-IBM-Mot aggreement anyways).
Apple will use the new G5 when it comes out. It would be nice if they would also extend their server line to include systems based on the POWER4i system.
As for price and chip size, the current G4 is 1/3rd the size of the P4. The G4 also has a 1MB to 2MB backside cache; this cache is on the die with the CPU; so, why is a 32MB cache on the CPU so hard to understand.
As for speed, the P4 has a 20 stage pipe line compared to the normal 12 and the G4’s 7. It’s easy to add stages and incress the clock speed; however, it’s hard to make each stage count.
As for compatability, the new 64bit Intel chips aren’t compatable with the x86 software. AMD is creating the x86-64 chips that will be backwards compatible. Thus, the windows market is in for a big problem soon. They will have 3 choices:
1) Keep the current P4 (no planed upgrades from Intel)
2) Move to Intel’s 64 bit system & replace all software
3) Move to AMD’s x86-64.
MS has already given their support to Intel.
As for compatability, the new 64bit Intel chips aren’t compatable with the x86 software.
They are, just not very well
AMD is creating the x86-64 chips that will be backwards compatible. Thus, the windows market is in for a big problem soon. They will have 3 choices:
1) Keep the current P4 (no planed upgrades from Intel)
2) Move to Intel’s 64 bit system & replace all software
3) Move to AMD’s x86-64.
MS has already given their support to Intel.
Microsoft is also backing x86-64, and seems overwhelmingly to support it over Itanium, for the obvious reason that a move to an incompatible architecture could possibly cause enough unrest in the software world to allow one of their competators to gain a better foothold. x86-64 makes for a nice, smooth transition to a better platform.
As is, Intel is making an x86-64 processor of their own. Right now they’re poised to take the server market with Itanium while continuing to distribute lower end x86-64 processors.
The big difference is AMD will finally have a leg up on them, thanks to the spectacular failure that was the Pentium 4.
2.5 percent? try 5 and CLIMBING. I work for apple at retail, and we have 20-30 people EVERY DAY coming in and saying ‘my peecee keeps crashing, can you help?’ and we convert about 80 percent of them. you can only hide the fact your ‘mercedes’ runs on gasohol until the first few backfires’, then the secret is out, and people want something that WORKS.
secondly, 5percent is the number because THOSE ARE THE PEOPLE WHO CAN’T AFFORD TO CRASH EVERY DAY!.
someday you too may have an important job, where they will give you a GOOD computer (read OS X or Solaris) that actually works, and you can get work done.
Yeah, OSX works great, not many crashes, but I still can’t get my work done. OSX just doesn’t have what I want/need. People complain about PC’s crashing and “not working” because they just don’t know how to properly maintain it. Nobody is forcing them to use a high maintaince computer, I just think it’s the wrong idea to label Macs as the one and only decent platform.
Apple has 2,5% share worldwide – if you happen to have 5% around your place – good for you, therefore, others will have less. If you really sell Mac, then this is known to you of course, as well as the FACT that they are still declinig, not climbing.
Google is your friend, you need not sell Macs for such wisdom…
People complain about PC’s crashing and “not working” because they just don’t know how to properly maintain it.
Fair enough, but why is it that people expect to have a rock stable linux system without any type of maintenance whatsoever? “Linux is stable my ass” is what people say. Oh if only people could be consistent.
I agree, OS X isn’t the only decent platform but for it’s age, it’s doing very well. Apple hasn’t pushed OS X because it won’t be pollished enough until 10.2 is relaesed. I think then you’ll see Apple step up the effort.
It also depends on how you use it. Hell, I replaced my Gateway laptop for work with a new Powerbook running VPC and Win2000. I run SQL Server with all my company’s proprietary apps, outlook, office 2000 etc and I connect to our remote office thru a VPN. The bottom line for OS X is that it’s only going to get better and faster and have more support.
As far as this new chip from IBM, I doubt that it is new news to Apple, this has probably been a nicely kept secret between the two for some time. I think IBM is playing it off as “Wooing Apple” and is probably already set to produce it in the very short term. My guess is that we’ll see this chip in new PowerMacs and servers in early 2003.
actually Drone…OS X is quite old…based on NeXTStep which dates back almost a decade which is based on Mach which dates back even further which is based on BSD which goes back to the 70s which is based on UNIX…
you get the picture…so OS X damn well better be stable…actually, Windows is the young whippersnapper even if you include DOS in the lineage, Windows is only 20 years old…do be quite honest, age has nothing to do with stability…i could write an OS tomorrow that’s more stable than anything we’ve got…what does that prove?
This is basicly the way I see Apple withing the next year or two with recard to processors:
G3s – If these are still around, they will be push very cheap, low entry Macs like the old iMac, and iBooks, but these will probably die as things advance.
G4s – Entry level Macs, The eMac, and new iMac, and probably soon a G4 iBook.
G5s- Midrange Macs, For people who need a little more power, maybe lower PowerMacs, and Power Books.
POWER4s – Top or the line Macs, with the Power Macs for multimedia designers, and Apple’s real work stations and server.
G4s – Entry level Macs, The eMac, and new iMac, and probably soon a G4 iBook.
I thought this was laid to rest when the ibook’s headguy said they would stay with g3 for awhile longer.
The Power4 has 32MB off chip L3 cache. Those who questioned 32MB on chip cache were right to do so, that is insane. The L1 (on chip not shared) , and shared L2 caches (on chip) are much smaller.
Go read for yourself at:
http://www-1.ibm.com/servers/eserver/pseries/hardware/whitepapers/p…
And nobody really knows how much they cost because you can’t actually buy one without buying the whole machine.
http://commerce.www.ibm.com/content/home/shop_ShopIBM/en_US/eServer…
Something like this will be your new desktop CAD/CAM workstation.Named Innellistaion Power4 or about it.
Nicely priced at $12K+ for single CPU version and $28K+ for more CPUs.
As for price and chip size, the current G4 is 1/3rd the size of the P4. The G4 also has a 1MB to 2MB backside cache; this cache is on the die with the CPU; so, why is a 32MB cache on the CPU so hard to understand.
>>>>>>>>…
No, the G4’s backside cache is not on-die. It is a backside L3 cache running at 1/2 the CPU frequency. (BTW, the Power4 L3 cache is also off-die). The reason I find 32MB of cache on the CPU hard to understand is this: the P4’s L2 cache consists of about 20 million transistors. That’s 40 million per meg. That’s over a billion transistors. Feasible for an off-die cache module in a machine where the sales tax will run you as much as many PCs, but not on a desktop or workstation.
>>>>You wonder why IBM would make this exclusive to Linux? Control. If they make it exclusive to OSX they have no control over which direction the OS takes, how well it’s marketed, etc.
For the Nth time, 75% of all PowerPC chips manufactured by both IBM and Motorola are used in network/communications systems. Therefore the first OS’es supporting these 64 bit PowerPC chips will be the vxworks and the qnx’s. PAYING customers always come first, even before their own AIX porting work for this chip.
Then IBM will get AIX running on PowerPC chips — since AIX runs on the POWER4 chip already, that’s not going to be much of a problem.
Then IBM will try to convince Apple that this chip is not that expensive to manufacture by going over IBM’s work on cheap manufacturing techniques used in the Nintendo’s chip and PlayStation3’s cell chip. Then they will help apple to port MacOS to the 64 bit chip.
Then, after all that, linux comes to play.
Thats not a typo. Probably its for a telecomm G3 type variant. still a PowerPC core. But still guys >>> 100 GHz in 2003.
IBM made big headlines with this.
If people are coming in cauce there PC is crashing to much a mac won’t change things over them getting a new Windows PC. If you bought an Apple OSX PC today or Windows XP PC both would crash as often (probably never). The only reason their computer crashes is A) they screw with it to much in the wrong ways, which will have the same effect in windows or mac B) they are using something like win98, but if they have that and need a old computer you could compair them to a person running OS9 which crashes more then win98.
If they are going new ether a new apple or windows computer will be just as reliable from the factory. Also stop with the “PeeCee” stuff, it’s childish and dumb, Apple makes PC’s same as Dell, PC stands for personal computer not Intel CPU running MS Windows.
I’m curious why those who didn’t buy a mac from you didn’t buy a mac from you.
Microsoft only “shattered” IBM’s dream dirrectly only once, with Windows 3.1, which ended in the divorce between IBM and Microsoft. But really, the divorce was caused by IBM, ultimately. Microsoft had spent more money promoting OS/2 than Windows 3, but IBM’s attitude towards clones [read: pricing] ultimately caused Windows to win.
So I would say, most of the fault lies at IBM. Sure, you could bring up SmartSuite, database etc. but that’s what competitors ultimately do to each other. If IBM was really planning to commit revenge, why aren’t they planning to do so to Oracle and Sun before Microsoft?
I’d agree with you that IBM was mainly at fault. I disagree the problem was clones, they didn’t push OS/2 for their Microchannel machines either and never gave their customers a price break. I actually bought an IBM Ambra during the OS/2 2.x days and they wouldn’t sell it to me with OS/2. Quite simply as a company the couldn’t decide on whether to kill or committ to OS/2; and so they spent money to develop the software but then made it difficult to buy the software and didn’t support. The IBM OS/2 group was betrayed by other groups within IBM far more than Microsoft.
So it certainly fair to say that IBM was the one who dropped the ball on OS/2. Its also fair to say that Microsoft’s refusal to cooperate with IBM on OS/2 from the time when 2.0 went into production was what turned dropping the ball into being pushed down the feed into their endzone. OS/2 was vastly superior and if it had won PC software would be years of where it is today.
recall reading somewhere that back in 91/92 that the Apple engineers were considering using Alpha as their next CPU after m68k, anyway to cut a long story short the then CEO of Apple had dinner with the CEO of DEC where the matter of Apple using Alpha’s was dicussed. What i heard happened was that the DEC CEO basically said no, cause Alpha was a new platform and he didn’t want DEC to be dependent on it sales (he was supposedly a big VAX man himself)
I tend to doubt this story. In ’91 DEC wanted Alpha to replace x86 line for NT; I’m not even sure they made an alpha VMS box that year. My guess is the problem was Apple. Motorolla at that made a better chip than Intel and they supported Mac strongly. Further the PPC move made more sense in terms of natural progression. Finally the Alpha chip cost over a $1 and used very expensive memory. Even generic Alpha boxes in ’92-93 cost around $5k with an Apple mark up (which was larger then) it might have been $7k which was out of their customer’s price range.
Why is it nearly everyone is assuming IBM is building this chip for Apple and OS X? With all the moves and money IBM has put towards Linux, I think they’re creating a new desktop computer based on Linux. Think about it – if IBM goes with Apple the future market for their chip is based on the future of Apple. But if they go with Linux they have much better control over the marketing and programming for their chip. IBM likes control.
I’m predicting a new Linux-based workstation – not OS X. That to me makes much more sense.
Why either/or they can easily do both? For that matter toss in AIX. Linux is very hard for IBM to control since too much of the core is GPL and a great deal of the development is non commercial and international. Conversely Apple and the RS/6000 line complement one another nicely. Photoshop for Mac-64 could very easily mean Photoshop for RS/6000; heck Apple might give IBM Aqua for RS/6000 in exchange for a price break on the chips. The partnership works like this: Apple makes the inexpensive easy to use desktops and low end servers while IBM RS/6000 gets makes the high end workstations, and mid range servers; with S390 and OS/400 for the very high end server market. Neither side has interest in the other’s core markets. Its going to be a cold day in hell before the print system for OSX is PFS or the print system for MVs is CUPS.
The partnership with Linux allows IBM to sell hardware, and also puts IBM in place for designing custom embedded apps inexpensively. Further it gets all sorts of goodies into AIX easily. If they need to they can drop AIX and focus on hardware / services. But at this point IBM is not looking to become VALinux.
1) Keep the current P4 (no planed upgrades from Intel)
You are probably living under a rock.
Intel Pentium 4 3.0Ghz (Northwood A) CPU is expected to be released in Q4.
Intel Pentium 4 3.06Ghz (Northwood B) CPU is expected to be released in Q4.
Intel Xeon (Prestonia B) 2.8Ghz is expected to be released in Q4. This CPU will have a 533Mhz FSB.
Intel Xeon MP (Gallatin) 1.5, 1.9 & 2Ghz are expected to be released in Q4. Gallatin is the 0.13micron successor to Foster MP, and is expected to be available with either 1Mb or 2Mb of L3 cache and feature Intel’s HyperThreading SMT technology.
Intel Xeon 3.0 & 3.06Ghz (Prestonia) are expected to be released in Q1 2003.
Intel Pentium 4 3.2Ghz is expected to be released in Q2 2003.
Intel Pentium 5 (Prescott) IA32 CPU is the 0.09micron successor to the Northwood core of the Pentium 4 due between Q2 and late Q3 or 2003, depending on the state of the CPU industry. Prescott is expected to launch at 3.2Ghz and higher and will introduce Intel’s HyperThreading technology to the desktop (first introduced in Prestonia). Prescott will also feature a number of architectural improvements including a larger 1Mb L2 cache and potentially an updated instruction set. Prescott is expected to feature an 667Mhz (166Mhz Quad Pumped) Front Side Bus.
A lot of upgrades for something with no planned upgrades, don’t you think?
As is, Intel is making an x86-64 processor of their own.
Actually, they aren’t. Intel had denied the existance of Yamhill nor any x86-64-like processors in the making.
The big difference is AMD will finally have a leg up on them, thanks to the spectacular failure that was the Pentium 4.
The “spectacular failure” caused Intel’s market share to go up. Plus, the latest variation of P4 hold the speed crown in the x86 market (and the 32-bit market), a position AMD held for so long.
2.5 percent? try 5 and CLIMBING. I work for apple at retail, and we have 20-30 people EVERY DAY coming in and saying ‘my peecee keeps crashing, can you help?
Of world wide market share, Apple had 2.4%. In the US, it has somewhere around 4%. The US market is getting more and more irrevelant.
As for crashing, that may be true for PCs running Windows 98/Me back then, but I don’t see mcuh complain now with Windows XP. It is practically the same as OS 9, crashing all the time. So in other words, if you were selling computers with Windows XP, you would probably have as much success as selling Macs.
secondly, 5percent is the number because THOSE ARE THE PEOPLE WHO CAN’T AFFORD TO CRASH EVERY DAY!.
25% of my time I spend on WIndows XP, another 25% of my time on Win2k, and 50% on Linux. None of these OS had ever crashed on my. You are comparing with PCs with an old OS, why not compare with Macs with OS 9 (which is used by most new Mac buyers)
Fair enough, but why is it that people expect to have a rock stable linux system without any type of maintenance whatsoever? “Linux is stable my ass” is what people say. Oh if only people could be consistent.
Most people tied Linux stablity with KDE’s stablity, and KDE 2.0 wasn’t the best example of KDE’s robustness. It also happens to be release at the peak of the Linux hype… oh well 🙂
G5s- Midrange Macs, For people who need a little more power, maybe lower PowerMacs, and Power Books.
Oh I love this… there have been no indicator that the G5 was even made for the desktop.
This is basicly the way I see Apple withing the next year or two with recard to processors:
It’s very clear that Apple is going with Motorola for embedded-type devices, i.e. laptops and Digital Hub devices (iPod), and with IBM for desktops. Both of those devices, of course, require very different processors.
The POWER4 is poised to replace the G4 as Apple’s fast processor. With IBM’s new .10-micron plant, Macs based on POWER4s with single or dual cores should become barely more expensive than their PC counterparts, and more powerful. I think it’s safe to say that the POWER4 is the G5; whatever Apple decides to call it, if anything, there will be no chip in between it and the present-day G4.
Laptops and iPods, on the other hand, require very low-power, low-heat chips; those are fast enough now that speed is no longer quite so important. People don’t expect a notebook to do Photoshop rendering work at high speed; they expect it to let them do simple work, like word processing or Internet, for as long as possible in a mobile setting. Therefore, Apple will go with Motorola processors for all its mobile devices. This means the G4 or its successor; processors that are increasingly specially designed and optimized for mobility, not power.
The eMac and new iMac are left in the middle. But that’s okay; I think it’s fair to say that these are both specialty machines. Both computers, really the same computer with a different screen, appeal to schools, students, and other people who want cheap, sturdy computers; people who will do little more on their computer, and require as little power, as they would get from a notebook, but value price over portability. There’s a complicating factor, too; consumers are increasingly opting for laptops/notebooks (some of the new ones can’t rightly be called notebooks, they’re too big) over desktops, and even schools, led by the state of Maine, are doing the same. I predict that in five years, the eMac/new iMac will have been all but obsoleted by new, cheaper notebooks, with sales continuing only to schools that want the absolute cheapest computer they can buy that are compatible with their existing ones.
The new iMac and eMac require AltiVec for OS X, but other than that don’t need anything special in the way of speed. I expect those computers to use the fastest Motorola processors that exist; say, a 1.1GHz G4. Motorola has no wish to continue improving the high-end G4 line, so instead Apple simply tries to make their desktops as cheap as possible, positioning their notebooks as a replacement.
What does that leave in the way of processors? The G3 is dead and gone, replaced by AltiVec-enabled, low-power, low-heat processors from Motorola; even for the iBook, as soon as price makes that a reality. The G4, or future Motorola procs, becomes Apple’s dominant processor, being present in all but two of Apple’s product lines. And the fabled G5 never exists, being replaced instead by IBM’s POWER4 and successors, being present in future high-end workstations and rack servers, once again making Macs the most powerful computer a consumer can realistically buy.
continuing to push VMS when it was obvious Unix had become the commodity platform.
While I’m a Unix guy I’m very impressed with the features that VMS offered that Unix doesn’t. I think choice is good and VMS offered genuine choice, as well as technology. IMHO many companies that have gone to Unix for their enterprise systems instead of VMS or MVS are going to be shocked when they discover some of the downsides in terms of security and long term maintance that will come with moving enterprise apps and data over to enterprise platforms. While many of these issues are more cultural than technical I’m not sure in the end that’s going to make a great deal of difference.
Anyway here’s a link about VMS vs. Unix written by a VMS guy with a good sense of humor: http://www.mindspring.com/~blackhart/requium.html
> Apple has 2,5% share worldwide – if you happen to have 5%
> around your place – good for you, therefore, others will
> have less. If you really sell Mac, then this is known to
> you of course, as well as the FACT that they are still
> declinig, not climbing.
Just as a quick point here, macs stay in service about 2x as long as PCs. So macs have about 5% of the desktop market in terms of usage and 2.5% in terms of sales.
http://www.silicon.com/public/door?6004REQEVENT=&REQINT1=55033&REQS…
you heard it here folks!
I got too excited… it’s the same darn thing as what Cnet wrote ARGH!!!
Just as a quick point here, macs stay in service about 2x as long as PCs. So macs have about 5% of the desktop market in terms of usage and 2.5% in terms of sales.
I have seen this one thrown around so much. What prove is there that PCs have this built in self-destruct mechanism that goes out before Mac’s built in one? Nothing. All Apple could prove is a Gartner report on ONE university spending less on upgrades on their Mac lab. That’s all.
Most of PC users are using old computers (below 400mhz) running Windows 98. Sure, PCs get obsolete faster than Macs (that in many cases good, not bad), but doesn’t mean it becomes useless faster than Macs.
Don´t get to excited about ibm’s power4 cpu, because apple designed OSX to take advantage of altivec. As you all might know altivec is motorola’s property. I guess the next generation cpu’s will come from motorola. If not apple would just have lost time and money developing osx to take advantage of altivec.
Some of you might be thinking this is no big deal because g3 doesn’t have altivec, so it could run on power4. This is not the point. The power4 has a similar altivec unit, ok…
but runing osx, which is optimized for altivec, on a new cpu(ibm) is stupid.
You might think why don´t they port osx to the vector unit of power4?
Again i think switchs like this are a waste of money. If you invest in something it should be kept a few years. See mmx and sse, these are around for a while and being upgraded now to sse2. One might say what about the 3dnow, again i say games are developed, usually and i say usually, for one of these technologies.
So why change, if apple wants more market share in digital video editing, the have an optimized os, they only need better hardware. This better hardware does not necessarily include a faster cpu but a better buses to all system.
The next few months will reveal apple’s path…
quote “OS/2 was vastly superior and if it had won PC software would be years of where it is today.”
is os/2 vastly superior to windows NT? i think windows xp is pretty good.
as far as software being years of where it is today, today;s software is pretty fast and stable. what would years ahead of where improve upon what?
Quite a few few people seem to be under the impression Altivec is “owned” by Motorola. It was part designed by Apple and I wouldn’t be the least bit surprised if IBM had a hand in the design also.
IBM did talk about SIMD a few years ago and from what I remember it was said to be pretty much identical. They choose however not to use it in their G3s as it would have limited the clock frequency.
IBM and Motorola both use an identical instruction set for their CPUs (Book E) so why should they differ in vector instructions?
Even if IBM didn’t have rights over Altivec and couldn’t license it they could certianly clone it, they own god only knows how many patents and are most likely to have a cross-patent agreement with Motorola.
I suspect the only thing 100% owned by Motorola is the name
Altivec and I doubt IBM really care about that.
PenguinPPC.org has a new piece of info : (i quote)
The new vector unit _is_ altivec compatible and it is called VMX. Feel free to draw your own conclusions
interesting isn’t it ? 😉
Just as a quick point here, macs stay in service about 2x as long as PCs. So macs have about 5% of the desktop market in terms of usage and 2.5% in terms of sales.
Gartner actually did not allow Apple to use the results of their Mac study publically. Yes, some information leaked out of Apple but Gartner clamped down quickly as their report was not fit for public comsumption.
As it turns out, the Gartner study, which Apple paid for, was totally cooked. There is no real data showing Macs are used longer than PCs.
PC users upgrade more frequently as they are able to afford it. With the price of Macs so high (this is what really used to be 2X vs. a PC), many Apple owners cannot afford to purchase a new Mac very often.
It is a very common theme in Mac world “want to upgrade but don’t have the money”. With a PC, it is easy and affordable to upgrade your computer in stages.
#m
VMX… a proposed but never implemented SIMD archetecture for the POWER family
heavy rah-rah PowerPC is god site
http://www.mackido.com/Hardware/MMX_VMX.html
something a bit more balanced
http://slashdot.org/articles/99/09/15/1141230.shtml
#m
It’s quite surprising what you find if you go digging in Google groups. One post about a year old from someone obviously very informed had all sorts of details about the G4 and explained that once they hit 0.13um the version on G4 (G4+ actually) should perform pretty much in line with AMDs Hammer. One of the reasons being a memory speed upgrade – which was cited as the main reason for Apples current poor performance (which I agree with in part) but another reason being to do with the silicon process itself which is rather different from AMDs 0.18um. AMD and Motorola will use the same 0.13um process. Also interesting was the assertion that G4 was not an embedded chip – it was pretty much designed for Apple.
I for one wouldn’t be surprised if the G4 suddenly shoots up performance wise but I guess we will soon see if this is correct – the current rumours are dual 1.4GHz with a 333MHz bus – This Tuesday. That maens a > 50% improvement in bandwidth per processor and this could seriously effect benchmarks along with compiler and OS improvements.
Another thing I read was about Altivec, it was very much a joint effort with the cheif Architect being an Apple guy (Now employed by MDR – who host the Microprosser Forum) with design by Motorola and testing equipment from IBM.
quote “OS/2 was vastly superior and if it had won PC software would be years of where it is today.”
is os/2 vastly superior to windows NT? i think windows xp is pretty good.
You have to remember the high point of OS/2 was the early 90’s. Windows XP is more advanced than OS/2, though there are still features I miss which Microsoft could add to XP, but the real comparison in terms of when released are things like:
OS/2 2.0 vs Windows 3.0
OS/2 2.1 vs. Windows 3.1
OS/2 3.0 vs. WinNT 3.51
(I don’t remember exact dates so corrections from others welcome).
I think its fair to say that OS/2 2.1 was clearly better than NT 4.0 and NT 4.0 better than any of the systems in my list.
With Dos and 286’s you could task swap between several apps. The big “test” was getting a modem download (remember you are talking 2400 baud modems at the time downloading 1 meg files) to work at the same time you ran a productivity ap. OS/2 1.3 could run a modem, 4 apps or even multi task networking with applications. It brought to the 286 the kind of multi tasking Unix users had on expensive workstation setups. I think Windows 95A was the first version of Windows where you could reliable use a modem and run apps at the same time (with 3.1 it was possible if you were careful, with 3.0 you needed to run Windows inside of Desqview). By the time of 2.0 you had applications docs. Individual customizable of folders (if memory serves introduced to Windows around the time of Internet explorer 5 but really not present until Windows 2000). The ability to completely specify the environment for each application (example: how much ram it could have at most, how much would be the least it would get, percentage of CPU, whether it was high low or medium priority relative to other apps) still doesn’t exist in windows (though it does exist in Server starting around Windows NT 4.0 Server). Attaching resources to files (like in MacOS) still doesn’t exist in Windows and the extensions system is still a terrible problem.
I could go on and on but just do a web search and you’ll see how far advanced OS/2 was. Windows has caught and passed OS/2 ten years later, but those ten years were lost.
as far as software being years of where it is today, today’s software is pretty fast and stable. what would years ahead of where improve upon what?
I can’t go to the universe where OS/2 won and Windows lost and see what 2002 looks like there. My guess is that 2002 in that universe probably look similar to the PC apps that are created in 2007 that not terribly demanding on hardware (i.e. hardware advances could not have gone much faster IMHO even if OS/2 had won).
I’ll speculate though in a few areas:
1) Networking took off 5 years earlier since PCs were able to network at the same time they running CPU intensive apps. As a result PC users got used to networking and the home / corporate internet revolution happened in 1993. However routing technology wasn’t advanced enough to handle huge loads and so the corporate intranets by the mid 90’s are where we are today though the global internet is playing catch up. However the main advantage they have is that technology and interoperability difficulties have already been worked out on the intranet side (where routing was not as much of a problem) so when the routing technology came we didn’t have these problems too.
a) So today on the internet search engines we can use all kinds of complex queries “I’m looking for the name of the famous movie with James Cagney where he is in a blue rain coat near a lamppost” and get an answer back from the computer with the name of the movie.
b) Applications are completely distributed in the sense that there is no distinction between local apps and networked apps no distinction between local data and network data. Your data is availability to you on your cell phone or PDA, and can be sent to anyone and they will be able to use it. (Sort of like a Unix shop LAN setup but global)
c) XML and Soap and other things are actually finished so data can be migrated from app to app easily (no proprietary lock ins). So for example a standard design document can import from word into line items in project. More importantly this can between different application companies so it works just as well with a conceptual design document to a source forge project plan.
2) Database based applications came much earlier to PCs (while they’ve been around for decades they were never the bulk of the applications like they are on good multitasking systems like VMS and MVS). Because 5 years ago people became genuinely familiar with Access in they way they are Excel today we have associative data models and not relational data models in all our applications including the OS. Commands like “copy into email all the files used for the McKinney presentation” are possible. You are able to make strange connections between data because the end user has a way to record information that the designers hadn’t considered: the NIH health databases are associative and so strange connections between drugs are noticed, nuanced pubic opinion surveys can be taken instead of simple polling, etc…
3) Because home/office PCs could multitask running complex voice recognitions while using a mainstream office app (what’s starting slowly today) happened in the early 90’s. Today we have voice control completely worked out and so where ever there is space to put a computer but a keyboard is impractical we have fully computerized systems. That means HDTV + Tivo + high end sound (i.e televisions can use compressed signals). That means fully computerized cars: so the “neverlost” features are standard as well as standardized complex commands to other subsystems. “look for a talk station on the AM band coming in clearly” (BMW has recently started this but poor voice recognition plus lack of experience show obviously in their latest offerings).
Again these are guesses, no one knows what would have happened. My point is that having technology sooner means that issues we are working on today would be solved by now.
Taking a look at what Apple is trying to do with OS X and extend itself beyond the current domain of Consumer O/S Platform and Low to Mid Tier Creative Platform and drive into the high-end multimedia creative market. With it’s new marketing campaign to drive into the traditional Professional workstation and low to mid tier Unix server business. Apple will need to move beyond the G4 now.
Apple very future existence to hold any market share in these market mean it must be looking higher end platform that pull over the likes of SGI traditional workstation and low to mid tier server business. Otherwise it will continue to loose this business to ever growing advancement of Intel based products (X86 and Itanium).
Market areas that would accept the additional cost for existing Power4 Platform are.
Digital Video (HDTV)
High end Post Production
Broadcast Graphics EDITING and Serving
Mac OS X server based Web Farms
Mac OS X server based Rendering Farms
Enterprise class database server (Oracle/Sybase)
Streaming Media Servers
Visualization
Bioinformatics
Molecular Modeling
Medical Imaging
For Apple to be serious contender in the rendering space they would need this to effectively win at account like ILM, Dreamworks, Cinesite, etc.
Also the cost for the current Power4 Platform today very much looks in line as natural product extension to Xserver family line today..
$2999
PowerPC G4 1 GHz 1-way L3 cache 2MB HD 60GB System Memory 512MB
$3999
PowerPC G4 1 GHz 2-way L3 cache 2MB HD 60GB System Memory 512MB
$7999
PowerPC G4 1 GHz 2-way L3 cache 2MB HD 4x120GB System Memory 2.0GB
$12,495.00 IBM Web price* 7028-6C4A
POWER4: 1GHz 1-way L3 32MB HD 18.2GB System Memory 1GB
$16,977.00 IBM Web price* 7028-6C4B
POWER4: 1GHz 1-way L3 32MB HD 72.8GB System Memory 2GB
$28,043.00 IBM Web price* 7028-6C4C
POWER4: 1GHz 2-way L3 32MB HD 109.2GB System Memory 4GB
$50,591.00 IBM Web price* 7028-6C4D
POWER4: 1GHz 4-way L3 64MB HD 220.2GB System Memory 8GB
Unlike perhaps a lot of u i actually remember os2 when it started. Some ppl claim IBM wrote rewrote os2 or something ive never heard of this.
This is what i remember of it all..
OS2 was made by microsoft, (i have os2 specs in my MSDOS bible cause it was based on an enhanced MSDOS).
MS then decided it was going to make OS2 with IBM and they formed this partnership (this was just before os2 was released if i remember correctly hey it was a long time ago).
At the time i thuoght it was strange and what the hell was IBM doing there in the partnership.
Soon after ms anounced that it was going to make windows and that it would of course still support OS2.
MS announcs it dosent want to do any more on os2 and gives it IBM who declare its going to replace dos and make windows redundant.
Its worth noting WHY these things happened… microsoft saw that all the developers could see benefits in os2.. but the whole industrty was running dos and ppl didnt want to make this big break. To be honest itwasnt a big break as ppl thought the BIOS ints and file formats are all based on DOS. (Perhaps this would have limited it later????)
Anyway so yeah at this point everyone but idiots realised os2 was going to die (ibm didnt seem to realise) because the thing that was different about pcs to all the other systems .. was it was open.. and so MANY ppl were developing for it. THIS WAS WHEN MICROSOFT LEARNED A LESSON APPLE SUN AND EVERYBODY ELSE DOSENT UNDERSTSAND- COMPATABILITY COUNTS! ..(ppl want to leverage their investments not through it away like apple or sun). This whole culture of geek and programming was created BY microsoft with MSDOS and its amusing to hear everyone criticise MS now and say ppl should be using linux or some bull. The whole mindset of free software was made from DOS.. wanna see a catalog 15 years old with 20,000+ free programs for dos just in australia? its here somewhere…
So yeah IBM went on and made OS2.. it was slightly faster and slightly more reliable but also had sever flaws. These arent talked about now of course .. but ill mention 1 that percisted for about 10 years after os2 started.
The internal timming os OS2 is so bad that nobody could make a midi drivre for os2 .. let alone a low latency audio driver.
Windows 3.1 with its renound pathetic midi timming.. was used and still is because it worked.. If u imagine the crappies os withthe crappiest midi timming was considered windows.. YET os2 was considered a complete joke because it couldnt (despite 10 years of programming) be made to work with midi without so much latency it made windows look good.
There are other performance and thread issues with OS2 i believe which is why it was also crappy for multimedia. If u think if the world was os2 it would be better thats pretty nieve asumption. Imagine windows XP with audio and multimedia crippled by design (say 10 years behind).
x86 are the fastest processors
windows is the fastest opperating system
These prove that incremental design and a competative industry allways beat redesigning the project every time.
Glenn –
Early versions of OS/2 were joint Microsoft IBM products. During the 1.3.1 -> 2.0 period was when Microsoft dropped out entirely, so OS/2 2.0 was the first version relesed exclusively by IBM and the 2.0 -> 2.1 changes was the first update that was fully IBM’s.
As for the break with Dos, IBM’s claim about “a better Dos than Dos and better Windows than Windows” (version 2.0 slogan) was certainly true by 2.1. Among office apps everything worked at least as well in OS/2 as either Dos or Windows. Most single tasking games actually ran in OS/2 and dual booting was easy for people who were heavy game players. OS/2 did not fail due to lack of compatability. I know of no OS/2 userswhich had serious compatability problems
As for multimedia I can’t answer that. Multimedia wasn’t a big deal then. Hardware issues on PCs were so serious regarding multimedia that OS related issues paled in comparison. If you wanted good multimedia you bought an Amiga or an SGI (depending on your budget). Mac offered pretty good multimedia in a more generalized package.
>>x86 are the fastest processors<<
Wrong that would be either the Alpha or POWER4 (amongst other incarnations)
>>windows is the fastest opperating system<<
Wrong that would be BeOS (now OBOS)… and I don’t know where you get that Windows is fast, that’s a load of crap because I work with that operating system daily!!!
In guessing which processor Apple might go with (and that is all we can do, none of us has as complete a picture as the insiders at Apple, IBM, etc. have) you need to look at the future of all the candidate chips not just the new Power4 chip.
The Intel 32 bit Pentium chips would be a popular choice and if done today would probably yield Apple some performance gains. But what is the future of this architecture? Will Intel soon start de-emphasizing this architecture in favor of their new 64 bit architecture? Apple would probably want to wait past that phase and go straight for the 64 bit chips. But switching to Intel’s 64 bit chips isn’t practical right now.
AMD’s x86 compatible 64 bit architecture looks intriguing. But is a sure thing? Will it have the legs to survive in the market? Can the architecture continue to compete for many years?
As you can see arguments can be made that these other chips have risks associated with them as well.
Don