“The dual Athlon is still the fastest PC we’ve tested, but the single Intel P4 2.53 GHz machine runs a close second, and even beats the dual Athlon on some of the tests. And, as expected, the Mac dual 1GHz G4 could not even come close to keeping up with these two PCs. Even though the P4 machine has only a single processor, it was easy for it to leave the dual-processor Mac far behind.” Read the benchmarks at DigitalVideoEditing.
Our Take: Some readers were discussing two days ago in our recent story of Apple’s possible jump to another architecture, that speed does not matter for an Apple user. I completely disagree with this statement (MacOSX is still very slow on my G4 Cube 450 Mhz 1 MB L3 Cache, 448 MB SDRAM, while WindowsXP literally flies on my dual Celeron 533, 128 KB cache, with 256 MB SDRAM), as Apple is leaning and betting a lot towards digital/multimedia professional software with the buyout of several key multimedia companies the last few months. It is clear that they want to become the “professional” image/digital OS, something similar to the market SGI used to have a few years back. For that plan to succeed, they need more processing power, power that do not have with the current G4s. And even simple users could use more speed, as the OS itself or the browsing experience is a far cry, speed-wise, from other OS alternatives.
Both G4 vs x86 benchmarks (first benchmark here) show that the G4s are lagging behind in speed a lot (even on Photoshop tests, where Apple used to have an edge), while rumors only want Jobs to introduce a 1.2 Ghz G4 next month, which is still very far away from what a PC can do these days speed-wise (and Intel will release a 3 Ghz Pentium4 before the end of the year).
On a related note, please read our recent editorial “Apple To Consider x86 After MacOSX Transition Done“.
) and a pIII 550 running 
Keep in mind though that good multi-processing is largely dependent on the operating system you are running. Apple has NOT a reputation with regard to this. So a major aspect you are testing is MacOS X vs WindowsXP, not just Mac vs PC.
I have also used a single G4 933 Mhz 1 MB L3, 512 MB SDRAM on my local shop. OSX does not work much faster on it either. And please do not forget that my G4 Cube has 1 MB of L3 while the Celerons have only 128 KB and 66 Mhz memory (as opposed to Cube’s 100 Mhz). That compensates the fact that my Celeron is a dual. And please do not forget that only multithreaded applications will make use of the extra CPU on the celeron machine (and not many apps are actually properly multithreaded on Windows). Based on Apple’s statements about how faster the G4 is clock per clock with an x86 CPU, supposedly the G4 Cube is faster than my dual celerons. Even if this might be true, the OS that runs on top of the Cube does not “show” that fact. WindowsXP flies on the Celerons and OSX is literally a painful experience everytime I boot into it. Please open your OSX’s IE on this page: http://www.sport.gr/track.asp and now try to resize that page…
That page has caracters I can’t read. Probably is evil.
across architectures, use the Ars Technica testbench software suite. It runs a variety of standard CPU and cache tests – the same ones on both PowerPC and x86-based CPUs. The URL is
http://www.arsware.org
Get it and run it – it’ll put these silly CPU contention threads to rest with real, hard data. Maybe we could post our
own results, do an OSNews benchmark using a CPU agnostic tool instead of consumer tools.
Doing benchmarks to measure performance without measuring the performance of the system is stupid. It’s like saying one person is healthier because they both got virii with similar symptoms. One recovered faster. The virii aren’t the same, however, so any comparison is moot. Using an agnostic tool that heavily optimizes for both platforms approaches the question from a technical standpoint. I know that most folks are interested in “real-world” performance. I’m banking on the fact that the crowd here is also interested in true performance.
And for the record, the PPC-based machines are going to lose.
Go down fighting, though – and the results will be pretty good.
Peace,
‘Rithm
PS
And remember Captain Sisko’s ghostly voice:
“It’s the SOFTWARE!”
I will in no way advocate for MacOS X. IMO the OS is dog slow.
I never ever thought Apple’s OSes, ever were really that good. 68k AmigaOS runs circles beyond imagination around 68k MacOSes on equivalently powered hardware. Strangely 68k MacOSes running emulated on 68k Amigas are even faster than real equivalently powered Macintoshes, and you could even multi-task between several emulated Macintoshes. In a strange way Amigas were the first real multi-tasking Macintoshes.
MacOS X has some nice eye-candy. But overall I am not that impressed. IMO WindowsXP is better, but also bloated and extremely inefficient.
>it’ll put these silly CPU contention threads to rest with real, hard data
I do not find the DigitalVideoEditing’s benchmark’s silly. The guys are professionals, they use digital video/3D/imaging applications and they do care to be fast in these apps. And that is what they test!
As for my blurb regarding my Celerons Vs the Cube, I *never* said that my Celerons are faster than the Cube G4 on clock per clock (I believe that the G4 1 MB L3 cache *is faster* than a Celeron). What I said, is that WindowsXP is faster than OSX on these two machines. My point was that OSX’s overa;; speed is bad, therefore it is even yet one more reason to be in need of faster PPC CPUs to run OSX.
You don’t find those benchmarks silly? Personal preference, I guess. I do. I find benchmarks that don’t test features using the same codebase silly. Who knows what hidden optimizations are going on behind the scenes? For either platform.
I *know* the Apple systems are going to perform the worst in the lot. I’m just genuinely curious beyond the fuzziness of
of saying they’re slower. As an ex-Apple employee I *know* we’ll never get true comparison numbers from pretty much any company. They have to play the C.Y.A. game, which is acceptable. However, I want to know how the different systems can be rated. Where does your dualie celeron box go
in processor and cache performance versus your cube?
And I never said you said anything! I’m sorry, I must have been unclear (as I often am when posting
– I wasn’t attempting to indict you personally. I think OS X is slow compared to Win2K running on my gigger thunderbird. But that has much less to do with the hardware than the software. I thought we were talking about hardware – I was posting regarding the article, not your comments 
OS X will get better. Already has – Jaguar is worth the money. It’ll go
on my TiBook 667 when it’s released and I have to pay for it.
Just like Win2K professional will go on my P4 2 GHz Netlux laptop – the best OS for the hardware I’ll be running.
‘Rithm
PS1 – C’mon, no takers? I’ll post results if it’ll get the ball rolling…
PS2 – Eugenia, you like XP? Is it better than Win2K pro?
I’ve used it for a few days here, and had to blow it away – it requires too much customization to be useful to me. And it felt slower on my hardware – but is it worth enduring for
any other reasons? Honest questions, not troll-bait.
I wouldn’t complain about the speed difference if Apple would come clean with their customers about it. I mean, I just surf the web and watch DVD’s, I don’t need blazing speed to get where I want to go. HOWEVER, Apple pisses me off by LYING about their fabulous performance which isn’t even close to reality. If Steve said, “Look, I know we’re a bit slower, but we’re working on it, and we think our cool OS is enough to compensate by giving a better user experience” I’d be fine. But NO. He always lies like a politician. And we always believe him. What’s really funny is that the same way Windows people didn’t know that Macs have GUI’s, the Mac people don’t know that x86 PC’s are double the speed! They believe whatever Steve tells them. After seeing Enron, Worldcom, Arthur Anderson, etc, I think IT’S TIME FOR HONESTY IN BUSINESS AGAIN.
And I’m STILL pissed off about BeOS being gone. Microsoft burned down my home and left me wandering.
Mattman
> PS2 – Eugenia, you like XP? Is it better than Win2K pro?
Yes, I like Windows XP PRO, it is my main operating system these days. It is fast and *very stable* on my Celerons (note: I turn off my machine every night. I only had a *single* crash many months ago, when my CD-R died, it was a hardware fault that caused XP to crash), I think that with the additional patches Microsoft made available last year, its speed is absolutely comparable, if not better, than Windows 2000 on the same machine here. Loading the OS for example, is much-much faster than Win2k.
Keep in mind, “Megahertz Don’t Matter”.
No, but raw speed matters though. And x86 is clearly faster. And that’s a shame both for Apple and their marketing campaign…
I like XP and seem to run fine on my BP6 2 X 400Mz Celerons now. However, when I first installed it, on a new IBM hard drive, using the HPT366 controller it trashed the hard drive.
Now this could have been coincidence ; but, since it happened I’ve heard of 3 other Hard drives that have been trashed by ‘XP’, from online buddies. I’m just wondering if this a common occurrence with the BP6 and XP and if peeps should be advised to install it on any controller other then the HPT366?
Sorry, if this seems off topic, its just it’s been a real pain in the butt
Never, EVER use the HPT366 controller on any motherboard that carries it!! I have stated this on a previous discussion too on osnews, I really hope most people read it!! ALWAYS use the standard IDE controlers on these mobos, not the HPT366 ones. These controllers are buggy beyond belief!
Even the company that makes them DO NOT support them anymore and they had a beta driver for WindowsXP which later they retracted. **There is no native support for WinXP for these controllers (it does not even list in XP’s hardware compatibility matrix).** Even Linux has major problems with this controller, and BeOS too. A bit confidentially speaking ;-), a Be engineer spent months trying to go around the bugs before the release of R5, but the hardware was so buggy, especially if you added a second hard drive to the controller, that it would trash the hard disks. Be decided to never ship that driver (a third party driver attempt is available at BeBits, but still not to be trusted with important data).
So, the problem is not just wiht XP, it is with any OS that uses these unbelievably buggy controllers.
I use the standard controllers on my BP6 and they work perfectly.
I really agree with the part about Apple not telling the truth.
We actually briefly considered using OS X for our computational research, since Apple and Genentech have been boasting about their altivec-accelerated version of the BLAST bioinformatics package ‘running up to 5 times faster on a 1GHz G4 than a 2 GHz P4’. That would offset the more expensive hardware, and we would get a nice GUI to go with it.
Then we visited Apple on a briefing session and saw the detailed benchmark results. The truth is this:
At the recommended (best for normal searching) BLAST wordsize
of 11, the altivec version is slightly *slower* than the normal blast running on a PC. If you increase the wordsize to insane values like 40 (useless for most people) it can be up to two times faster than the x86 version. They didn’t show a single result justifying the ‘up to 5 times faster’.
Frankly, the results for non-default wordsizes is nonsense; they claim to speed up our program by changing the algorithm to one that destroys our result.
So, in practice the machines are 2–3 times more expensive,
and with handcoded altivec code then can _almost_ match a non-optimized x86 version.
This is also a question of principles: No way are we going to spend $1M with a company that is lying to us to make the sale.
Ars Technica Benchmark [Ver: 2.0]
Model: PowerMac3,4
RAM (MB): 768
CPU Type: PowerPC 7400
CPU Count: 2
CPU (MHz): 533
FSB (MHz): 133
L1 iCache (KB): 32
L1 dCache (KB): 32
L2 Cache (KB): 1024
L3 Cache (KB): 0
Individual Results:
Mean Peak Type Time Test
99.7% 239.4% INT 2.40s Complex Bitfield Operations
123.9% 156.1% INT 2.52s Blowfish Block Cipher Encryption (448 bit key)
67.0% 74.3% INT 3.99s File Allocation Table Manipulation
59.0% 64.1% INT 10.05s Dhrystone 2.1 (500k Iterations)
128.3% 140.3% FP 5.74s FP Divide
105.2% 113.0% INT 2.90s Encode G.723-40 Audio 100k
130.8% 161.5% MEM 2.16s LinPack (500×500)
144.4% 194.5% INT 4.04s LZ77 (Lempel-Ziv) Data Compression
328.0% 177.2% INT 0.54s LZ77 (Lempel-Ziv) Data Decompression
55.2% 57.7% INT 10.03s Towers of Hanoi
141.6% 153.1% INT 2.27s Queens Puzzle Solver
171.5% 344.8% INT 2.01s RGB to CMYK with Color Correction
129.2% 149.9% FP 1.16s Fast Fourier Transform 16k
67.6% 79.0% FP 4.09s Spherical Harmonics Legendre Polynomial
165.4% 174.4% MEM 6.22s Sieve of Eratosthenes (Prime Generator)
123.3% 54.3% FP 0.44s Solve Vandermonde Equation
44.3% 48.8% FP 6.62s Whetstone 1.2 (1000k Operations)
138.5% 175.9% FP 3.81s NTSC YIQ -> PAL RGB
109.5% 112.5% INT 13.77s TSCP 1.71 (Chess Benchmark)
103.5% 263.8% SIMD 5.10s NTSC YIQ -> PAL RGB (Altivec)
58.1% 119.4% FP 4.11s NTSC YIQ -> PAL RGB (Double)
217.6% 239.4% SIMD 1.10s Complex Bitfield Operations (Altivec)
130.7% 160.0% MEM 1.53s Fast Fourier Transform 64k
121.0% 138.8% INT 2.58s Blowfish Block Cipher Decryption (448 bit key)
Aggregate Results:
Aggregate Integer Floating Point Cache SIMD
Peak 152.1% 109.7% 165.3% 251.6%
Final 127.2% 98.5% 142.3% 160.6%
SMP Results:
# Threads Time Charged Time Score
2 56.1s 107.4s 179.7%
BTW, My Mac seems faster then my AMD 2100+ in everyday use. Most of my windows friends think so also after using it.
I have a mac..i love OS X,,,,,well i have windows also on my pc. OS X is my favorite os to this date. But the fastest mac isn’t as fast as the fastest PC. it just isn’t….ppc’s are lagging behind. I think Apple is doing the right thing by considering going to intel or amd.
I MEan IBM can do *any* order at *any* quantity. I’m sure that they would love to be the official (read: exclusive) supplier of PowerPC chips for Apple. IBM was *very* happy to supply Nintendo with the Gekko processors.
Screw Motorola. Get IBM to make those 2GHz chips so the PCs user would shut up already.
But then again, I’m still waiting for companies to put out low-cost dual P4 mobo.
Resizing that page took about 1 second on my 50 MHz Amiga. How long
does it take on the Mac?
A lot. It does not really resizes smoothly at all. The other non-smooth resizes on the Mac that everyone is unhappy about, should be considered fast, compared to this one.
I don’t think this article told us anything we didn’t already know. There are some still under the spell of Apple’s marketing that believe the outrageous performance claims, but I think those of us in the real world have pretty much known for quite some time that Macs are generally quite far behind PCs performance-wise, and the situation is getting worse.
I think what this performance stagnation will do for the Mac is weed out most of those who need a fast machine from the Mac user base, leaving those who don’t jump ship to enjoy OS X, the iApps, etc. What I’m seeing, though, is something I’ve never seen before – a lot of people seem to be content paying more money for a slower machine, and would actually rather use that slower machine than a faster one for (insert task here). (Disclaimer: I’m one of these people
Now that pretty much any computer one buys – Mac or PC – is easily capable of fulfilling everyday casual desktop computing demands, I think there are quite a few people who don’t strictly care about MHz or performance anymore because they know whatever they buy will suit them just fine and last for a long time.
The performance gap will close a little bit once the new Power Macs are released, but Apple zealots should realize that even dual 1.4GHz G4 Macs with DDR memory will not be enough to catapult ahead of PCs again. I’d like to see quad processors in the next top-end Power Mac, and the G5 as soon as is realistically possible. I do not believe it would be wise for Apple to change processors at this point; it would be too disruptive. Way behind x86 or not, they’re still profitable, and their user base (in the US at least) is still growing, albeit very slowly…
Alex
About BLAST. I take it that the apple version was optimized heavily for altivec. Was the P4 version optimized at all? For these kinds of applications SSE2 can make a really huge difference. I wouldn’t have been surprised to see the P4 run 4x faster with optimizations. If you seriously want to use BLAST and put a lot of money into running a large system, I hope you guys demand a proper compile for you hardware.
Since PPC is obviously slower than PCs today (they should have tried a dual P4 Xeon as well, that should have painted an interesting picture), and since a shift in hardware won’t take place for some time for apple, what they need is to get honest with the public about their hardware, lower the price, and conduct a massive effort to optimize MacOS X as much as they ever could.
They will not be able to run computationally heavy software faster than the x86 machines, which means that they will have to find other ways to be competitive. Unfortunatly they are doing the opposite, trying to compete where they will fail. Altivec won’t save them from SEE2 after all.
But most importanlty, Apple has to give people a reason to buy a new computer and make it one of theirs. I wouldn’t trade my faster WinXP machine for a very expensive, much slower, Mac with less software to choose from.
> I wouldn’t have been surprised to see the P4 run 4x faster with optimizations.
Exactly. And you will need the latest versions of Visual C++ or GCC 3.x for that. Did the ArsTechnica guy compiled it with these compilers that “understand” SSE-2? Because he did so for the G4’s Altivec you see…
Not even the software tested in the article’s benchmark have code for Pentium4 in it possibly while (especially Photoshop) their Mac counterparts do have Altivec optimizations mostly because Altivec is older than SSE-2. Today, all high end software for example, do have optimizations for MMX and maybe SSE-1, but not for the latest stuff, like SSE-2.
Here is some articles from recent benchmark tests…
Apple smoking Dell on the Server spectrum;
http://www.xinet.com/benchmarks/benchmarks.2002/index.html
Apple Power Mac beating a Sony Vaio in numerous Photoshop tests;
http://www.usatoday.com/life/cyber/ccarch/2002/07/12/steinberg.htm
more comparisons;
http://www.usatoday.com/tech/columnist/2002-07-19-steinberg_x.htm
Sun Microsystems verify the MHz Myth reality spectrum… a myth in which only the PCdom believe in like a religion!
http://www.sun.com/executives/realitycheck/headsup020314.html
But at the end of the day, it’s not going to matter… Eugenia says that she is not impressed with her G4 Cube, I of course work around the same PCs that you guys describe in your posts on here everyday, or play with my friends (also my co-workers) machines… I just had the luxury to play with a homebuilt AMD Athlon 2100+ running Red Hat Linux 7.3 with KDE 3, though I was impressed with the new version of RH Linux & KDE, the machine (though not slow) wasn’t impressively fast. I also work/program on a Dual AMD Athlon 1.6 GHz machine (that I love) running Red Hat Linux 7.2 and I like the performance, but that is 1.6 x2 equalling 3.2 GHz, which should be impressive. On the other hand, we have these useless Compaq DeskPro PCs with Pentium IIs/IIIs at 400 MHz running either Windows NT/2000, and they are dog slow, especially compared to my iMac G3 at 400 MHz running Mac OS 9 and/or Mac OS X (10.1.5)! In my mind it isn’t what the fastest Mac can do compared to that fastest PC of today, but when running 2 machines with identical configurations, the Mac (pretty much) always wins, this is where the ‘clock for clock’ techno comes in! Of course if you decide to throw in another slant to the already wild and complicated situation, throw BeOS, Mac OS, Windows and Linux into the same scenario (from a software perspective) and BeOS wins everytime from a performance point of view, but guess what, that is (and was) not enough to get users to switch OSes or platforms for that matter!
Don’t you people get it?! I have been on both sides of the fence, and I like this side very well. Apple will deliver what the Mac user community expects sooner or later (though not everybody will be satisfied). I am about to purchase a new Mac as well soon and will be waiting for DDR and a faster System Bus because I know that is what is holding the G4 behind (other than the turtle of Motorola)!
In response to the whole multiprocessing issue. OS X is built on Mach, and Mach was designed from the beginning with fine-grained locking for proper SMP support. Thus, if dual G4’s aren’t performing, its the CPUs, not the OS.
Not flamebait, legitimate question. What is it about OS X and Apple machines that makes Mac users so loyal to Apple? Hard data, not some stuff about Steve’s aura. As someone who’s only experience with Apple machines is with crappy PowerMacs at school, I’d like to know what’s so nice about “the other side of the fence.”
Apple Power Mac beating a Sony Vaio in numerous Photoshop tests;
I read this article a while ago and researched it further. It’s actually another lame Photoshop filter benchmark. It put a top-of-the-line (dual GHz) Power Mac up against a 2.2GHz P4. There is no other app tweaked for the G4 and AltiVec as much as Photoshop, and one thing the article neglected to mention was that the Vaio costs $800 less than the Mac. (Well, the faster Vaio that has since replaced the model mentioned in the article does, anyway.) So, yes, on a benchmark that puts a top-of-the-line Mac up against a lower-middle-of-the-line PC that’s $800 cheaper, I sure HOPE the Mac comes out ahead!
more comparisons;
This is actually just the same author defending his above PC-Mac duel. He fails to address why he put a more expensive Mac up against a slow-and-expensive-relative-to-the-rest-of-the-PC-world PC.
Sun Microsystems verify the MHz Myth reality spectrum… a myth in which only the PCdom believe in like a religion!
Sun is right, but that article talks about SPARC machines, not PPC machines, and although the MHz Myth can be true to some extent, it can’t be applied to the large MHz gap between Macs and PCs.
You seem to think that the “real” myth is that of the slow Macintosh, perpetuated by “PC drones.” I’m a Mac fan too, but I freely admit that my computer is substantially slower than a comparably priced PC at just about everything except RC5 and a few Photoshop filters. This is a commonly held and, I believe, mostly accurate view held by many people who have not only read articles such as the one in the title of this thread, but have seen the quite vast performance differences first hand. Sure there are exceptions, but in most cases, the PC really is substantially faster than the Mac. I think it would do more good for Mac fans to openly accept this rather than making laughable denials of poor Mac performance a la Steve Jobs…
In my mind it isn’t what the fastest Mac can do compared to that fastest PC of today, but when running 2 machines with identical configurations, the Mac (pretty much) always wins, this is where the ‘clock for clock’ techno comes in!
Of course, such a comparison is useless, since with the numbers equalized, a Mac will be up to twice as expensive. It’s only relevant to compare similarly priced machines, so that you’re comparing value. Who cares if a G4 does a bit more per cycle if you have to pay for more than you get?
It used to be really cool to dispel the “MHz myth.” It was accurate too. Nowadays it has ceased to be accurate or even pertinent–it takes an awful lot of gumption to still be spouting that nonsense. Of those disagreeing benchmarks you provided, one was so far from being relevant to desktop use that it was laughable, 2 were from USA Today, which is not a publication that I would trust for computer news, and one had literally nothing to do with PowerPC. Do you think we’re stupid? All of the relevant and trustworthy benchmarks indicate the truth: AMD/Intel processors are faster for your money.
That doesn’t mean you’re not allowed to like your Mac better if want to. There are lots of good reasons to prefer Apple. It’s just that speed is no longer one of them.
Replace your PC(s) with a mac for a year. Then switch back to the PC and you’ll be amazed how crappy the PC interfaces are. Nuff’ said.
You can find a PR rating chart for AMD’s barton here
http://www.xbitlabs.com/news/story.html?id=1025349410
With the hammer optimizations that AMD is adding the Barton should gain another 600+ PR rating for chips clocked at the same speed.
If you just look at the pictures you might be led to beleive that there is a huge difference between the mac and intel/amd boxen. When looking at the numbers most of the time the intel/amd box is twice as fast but who cares when you are talking about something that takes 2 seconds on the intel/amd boxen takes 4 seconds on the mac. If you are really worried about 2 seconds that much you are far too busy and should probably take a vacation. When things are going to take a few minutes like some of the tests I would usually go get a cup of coffee or step out for a smoke anyway, so once again the time isn’t that big of an issue, at least for me.
Any computer that you can go out and buy right now probably has enough power to do what 95% of the people will use it for. They will all function fairly well, they will all have bugs, they will all do things that will drive you completely batty.
I think that we have nearly reached a point where unless you are a gamer chances are you won’t notice much difference between a low end system and the fastest thing out there. Marketing has done a great job of making us think we need the fastest thing out there but people are starting to realize the truth. Why do you think the PC market is in the tank and has been for a couple years? The 500MHz machine that people bought 2 or 3 years ago is still working fine for them, most of them don’t need anything faster and won’t buy anything until the old one breaks.
So I ask again does it really matter?
PS I use/own both macs and pcs so I am not really trying to stick up for Apple here just asking a question. Albeit a very long winded question.
>So I ask again does it really matter?
well, does 15fps and 30fps matter?
“>well, does 15fps and 30fps matter?”
If you are a gamer, but then I said that didn’t I?
to my mind, it just points to the poor SMP implementation underlying macosx and its Mach implementation.
(and more curious as to why as Jordan said, they have little interest in the SMPng and KSE in freebsd 5.0)…
>So I ask again does it really matter?
Might not matter in seconds, but some things take hours to render.
I’m basically a Mac person. I used to be a zealot until I found out how much fun it is to use other OS’s. I still love Apple and am very glad Jobs saved it.
Having said that, my experiences have been pretty much the same as Eugenia’s and some others. You don’t need Ars Technia benchmarks or any benchmarks other than your own eyes. All you have to do is surf the web and launch programs. Apple has got to do something. Yes, improved bus speeds and RAM is part of the answer. But ultimately, the processor is the problem. If the PPC can make some big strides in the near future, fine, But, if it cannot…Jobs is looking at options. He has to, that is his job and also the future of the Mac and Apple. He is the hype-master of all time to the public, but he is a hard nosed realist to himself. He would be crazy if he wasn’t looking for options at this point in time.
the other thing people need to understand in these types of bench marking is, everyone seems to equip the same amount of memory, which is dodgy, since x86 is a densly packed architecture, 512mb is a lot, on a PPC which is sparsly packed, 512mb is less code space than 512mb on an x86.
the same goes for cpus like UltraSparc’s + MIPS
Yes, apple is right, mhz’s don’t matter, benchmarks do,
which is why you should go with AMD, not Apple. HAHAHAHAHAHA
Don’t you have a cheaper option? Like I said, I was looking for something concrete, not the “aura” stuff I always get.
>the other thing people need to understand in these types of >bench marking is, everyone seems to equip the same amount of >memory, which is dodgy, since x86 is a densly packed >architecture, 512mb is a lot, on a PPC which is sparsly >packed, 512mb is less code space than 512mb on an x86.
>the same goes for cpus like UltraSparc’s + MIPS
But the code size difference is less then 10MB (Likely also less then 2MB). Remember that all that memery that photoshop use is used for data structures not code. Just try to take a look at the size of the photoshop executeable. It is not THAT big.
Martin Tilsted
>”Might not matter in seconds, but some things take hours to >render.”
When things take hours to render do you sit and wait for it or go work on something else? Most people I know when they know something will take hours they save it until the end of the day, start it running, and then go home.
it appears that outdated OS architecture was a major factor in Mac OS X losing.
what if adobe ported a ground up, rewritten, and highly optimzed version of photoshop for beos, which then ran on the exact same hardware that was stated in the review.
would we exepct be to win? if not, then it’s no wonder be inc. died.
Goober, please. This is laughable. When you got design houses that they need the job done “NOW” by a customer, every minute counts. Especially counts if the job won’t take whole hours, but minutes.
If anybody out there thinks Apple is doing something ‘wrong’, whether ethically or product-wise, please write them a letter through their website. I have, I don’t know whether it’s going to help or not, but really, these Apple guys seem to have their heads in the clouds. The only time they change something is when there’s a large public outcry (spring-loaded folders, etc.). At 2.4% market share, they HAVE to listen to customer requests. XP is solid (though I still hate it), and Gnu.Linux is getting better daily. If Apple sucks, it’s our duty to take our money and walk. That’s the Capitalist system.
One personal beef I have is the continued absence of the G5’s. I get the feeling that Apple will continue selling G4’s as long as people are willing to buy them. That’s common sense. They have *no* incentive to release a faster processor unless we lean on them. What do you think their cost is, selling 3 year old RAM? They complain RAM prices are going up, but they must be buying their parts dirt cheap. They’re throwaways to the PC world. And they still charge a premium, for one reason – BECAUSE THEY CAN. It’s time for Mac users to stand up for themselves. Write a letter, if nothing else. Don’t just go “Yay! They bumped up the iPod’s HD to what the Rio’s had since 1998! Yay! Apple is great, they really CARE about me!!!”
Wake up and stop kissing Steve’s lily-white butt. Make a difference. Complain about what you don’t like, don’t thank them for their flaws. “You do not have, because you do not ask.”
Mattman
>what if adobe ported a ground up, rewritten, and highly optimzed version of photoshop for beos
That would never happen for obvious reasons. But even if they did that, you forget that BeOS has the same crappy compiler than OSX currently have. An old GCC (although the new GCC 3.1 has better support for OSX). *Especially* for BeOS, the gcc version used is very old and it is known to create slow code.
On OSX Photoshop is saved not because of the compiler, but because a lot of the altivec code has been written by hand. On x86, on a Windows machine, the compiler would do the optimizations for you, but on BeOS, to use good SSE-1/2 code, you would need to write assembly by hand, which is something that Adobe would never do. So no, I do not believe that Photoshop would be extremely faster on BeOS than on OSX or Windows. It might be from comparable, to much slower. But not faster.
> Goober, please. This is laughable. When you got design >houses that they need the job done “NOW” by a customer, every >minute counts. Especially counts if the job won’t take whole >hours, but minutes.
“That job is gonna take 8 minutes instead of four? To hell with you I’ll take my business eslewhere.”
I hear stuff like that all the time.
>>I read this article a while ago and researched it further. It’s actually another lame Photoshop filter benchmark. It put a top-of-the-line (dual GHz) Power Mac up against a 2.2GHz P4. There is no other app tweaked for the G4 and AltiVec as much as Photoshop, and one thing the article neglected to mention was that the Vaio costs $800 less than the Mac.<<
The usual conspiracy theory, and as the guy said before… the PC was the top of the line not so long ago, and the Dual G4 has been out for months as well as you should know that Macs don’t get updated as often as the average PC!
Here was his jargon on the subject:
“The Power Mac G4 I used for these benchmarks retails for $2,999, although a rebate program is in effect if you buy one with an Apple monitor. The Sony VAIO, now discontinued, listed for $2,499, but other high-powered PCs are much cheaper.”
Where does the $800 difference come in?
>>Of those disagreeing benchmarks you provided, one was so far from being relevant to desktop use that it was laughable, 2 were from USA Today, which is not a publication that I would trust for computer news, and one had literally nothing to do with PowerPC. Do you think we’re stupid? All of the relevant and trustworthy benchmarks indicate the truth: AMD/Intel processors are faster for your money.<<
Okay, so you’re not impressed with USA Today’s piece on the issue, but you have not commented on the Xinet’s benchmark tests! These we’re released a less than a month ago, but I guess these are invalid too?
The conspiracy theory on Adobe tweaking Photoshop for Macs over PCs is really lame, if you haven’t noticed, these 2 are not playing as friendly as you might think, especially with Apple stepping in Adobe’s turf these days!
You people seem to think that the only valid benchmarks are the ones where the PC is the winner! Seems pretty bias to me! I mean you go as low to say Ars Technica fudged a benchmark in favor of the Mac, and that site happens to be well respected, well until the PC was shown not to win at everything, or reality was more vivid than fiction!
> I mean you go as low to say Ars Technica fudged a benchmark in favor of the Mac
I do not know if someone else said that, but I certainly did not say that, and please do not put words in my mouth for things I did not say. GCC 3.1 and VC++.NET are new. It is very possible that the ArsT developer did not have access to either a Pentium4 neither its compilers.
Therefore, he did not optimize his benchmarks for the new line of CPUs for x86. Nobody said that ArsTechnica fudged anything. They just created an application with the tools they had available. No more, no less. Even if that might hurt new x86 cpus making the benchmark not “current” anymore. But we never said that ArsT fudged anything.
I didn’t put words in your mouth, I was just going by what you said.
Anyways, I have to go to bed… I am heading to the Eastern portion of Europe in the morning for a week holiday (or vacation as they say in the US) and will let you folks keep the PC vs. Mac bashing on course!
Later…
🙂
MattMan, I agree with you about Mac zealotry. My God, when the OS X public beta was out, I wrote Apple so many times…and have continued, even a flurry of them while on vacation at the beach this past week (because of the MacWorld announcements)! I also agree that, usually, Apple will do nothing if people keep buying. The only thing that makes me suspect differently is that Apple is buying up all these digital companies, etc. If for no other reason than his own ego, Jobs has to see that the Mac needs more power or it will become a laughingstock. That ego of his won’t allow that <g>. If, next month, they announce new Power Macs with 1.2 GHz, they’re going to be in trouble.
Quality of this benchmark is very low.
Any pro quality benchmarks put into account I/O bandwidth.
Thus for benchmarks to be able to compare CPUs all other components such as number and type HDs NICs etc.
Plus different OSes make it even harder. (Not even talking about different hardware architectures and different compilers used to produce underlying software)
Ralf, email me or IM me please (you can find more IM contact IDs on my web site).
the main thing learned from the ars technica test bench was performance of a platform is HEAVILY modified by the quality of the compiler used
ive been involved with the ars bench since it was started, and thats really the main conclusion one can reach from the data. yes you can get a general performance trend for a system using a certain compiler by comparing to other systems you have avail in which you know what compiler was used… but its not the end all and be all of cpu benchmarks nor was it designed to be.
Eugenia can take care of herself, but there is no call for personalizing things – speculating on her motives, etc. I see her trying to be as objective as possible and provide some talking points to boot. I am a Mac person and even I see that there are areas that Apple *must* attend to. That’s all it comes down to. I love OS X, but it’s too slow, it needs speed, applications need speed. It is the most obvious thing in the world.
Jay, thank you.
hmm
the only good benchmark is getting GNU on ppc and on x86, same distro, same stuff, same benchmarks
and see if is really faster the x86.
To be sincere I’m ok with my athlon 850 and I just apply one rule:
cpu speed matter less compared to ram.
I bet that a system with 1.5GB ram will burn away a system with the current usual ram (256M? 512M?)
I didn’t say anything about Ars Technica, so that must have been somebody else.
I did address the Xinet benchmark. I said it was laughably irrelevant to desktop performance, so far as I could tell. Anyway, assuming that it’s at least somewhat valid, it’s still pretty pathetic that you only supplied one out of four that was even worth the electrons transmitted.
You harp on the disparity between the Sony and the Mac as being $500 instead of $800. $500 is still a lot of money. If Macs are not updated so frequently as PCs that is yet another of their many flaws, but prices may be updated to match the market whenever Apple likes. Sorry, but your explanation was thin even by your standards.
The reason why I think that the only valid benchmarks show x86 winning is that all of the benchmarks I’ve seen where Apple wins have been transparently, demonstrably unreasonable. However, all of the sound looking benchmarks indicate the same result. You imply false causality–you immediately assume that I only believe the ones I want to believe. This is typically irrational, but even you must see that it is easy to determine whether a benchmark is worth its salt, and the ones Apple wins simply have not been.
And, like I said before, this is not necessarily a reason to dislike your Mac. I have a Mac, and I like it a lot. I don’t pretend that it’s speedy when it’s not, though.
Every article I’ve seen that uses After Effects as a benchmarking app has shown that the Windows version is faster than the Mac version. This does not necessarily translate into P4s are inherently great and G4s inherently crap. What it does mean is that the Windows version is better optimized for the Intel architecture. Unless the algorithms used in AE have been written to take advantage of vector math (which I suspect most of them haven’t) then the G4 will always come out behind since the most powerful part of the G4 CPU (the 128bit DSP part) isn’t being used. Given that Apple will soon be moving into the 2D compositing space (I’m speculating given Apple’s acquisition of Shake and Chalice) I doubt Adobe will remedy this shortcoming any time soon. If you make your living working in AE and have no interest in what Apple will do with its most recent acquisitions then your choice is pretty clear.
While I agree faster is always better in a CPU and wish Motorola would produce a faster version on a more regular basis, I am saddened by the overwhelming “group think” that believes that scalar math is the only valid approach to programming computers.
Mac is slow, because both the hardware and the software are slow. Yet Apple continues to lie to potential customers, to current customers, to the media, etc.
“The Dual 1Ghz Power Mac is a twin-engined supercomputer with a phenomenal peak performance of 15 gigaflops.”
“The dual 1GHz PowerPC G4 processors — with a combined performance of 15 billion floating point operations per second, or 15 gigaflops — put this fearsomely fast Power Mac G4 squarely in the lead as the ultimate high-end graphics workstation. The first Power Mac to blast through the 1GHz barrier, the new twin-engined G4 runs professional applications like Adobe Photoshop up to 68 percent faster — and crunches digital video over 300 percent faster — than a 2GHz Pentium 4-based PC.”
In the 933MHz and dual 1GHz Power Mac G4 models, faster-than-light processor speed gets an additional boost with an advanced cache memory architecture that provides ultrafast, dedicated memory with massively enhanced throughput. Accessing data from main memory is significantly faster than accessing data from the hard drive, and in these two models the system architecture takes this concept one step further with an even faster level of memory called L3 cache.
“Professional applications like Adobe Photoshop”
What professional applications OTHER THAN Photoshop?
Apple hasn’t done a performance demo without Photoshop for the past three years.
They continually hype their machines as a “supercomputer”, but we know this is a complete lie as there is nothing about any Apple machine that makes it a supercomputer.
Oh yes, give me “faster than light processor speed” Steve. Wait. My web page still doesn’t scroll fast. I suppose the Mac OS is a sort of bose-einstein condensate slowing down the speed of light, then, eh?
It seems like Apple is digging their own grave by delivering a continual stream of bullshit.
If Apple were serious about making the best personal computer, they’d also make their machines fast.
When you spend your top dollar on a Porsche, you get performance. If a Porsche were slow, you’d say “what the fuck?!?”.
Apple tells you their machine is a Porsche. When in reality, it is a high-priced slow luxury car, something like a Cadillac built with old car parts like three year old memory.
So the bottom line:
If Apple wants to stop the bleeding and increase their market share, they need to:
Stop the lies.
Deliver on their promises of performance.
Switch to a non dead-end CPU line.
Reduce their greed and lower prices.
Create a culture based on reality, not on hype.
#m
> What professional applications OTHER THAN Photoshop?
Well, there is always Maya, Peak, Nuedo, LightWave, more Adobe apps and a few other imaging, 3D, sound and video apps available for OSX and more are coming. They do have some important applications right now and on pipeline for the future.
I agree to most of your points, but not on the above one. MacOSX does have some good software ports going on there…
What applications does Apple use for showing the speeed of their platform vs. other platforms?
Maya is cross-platform
Nuendo is cross-platform
Lightwave is cross-platform
Adobe apps are cross-platform
Logic Audio is cross-platform (well until Sept 30)
Out of all the cross-platform apps that have been available for Mac, I believe only Photoshop has been shown for benchmarking and platform comparison purposes.
#m
But I think that kind of marketing hype is pretty common. I know Intel does it (the P4 “accelerating the Internet” or whatever), as just one of many potential examples. Apple’s marketing jingo is right up there with the worst, and from browsing various Internet discussion boards, apparently some people actually fall for it. It’s up to the consumer to educate him/herself before making a purchasing decision. I can’t imagine what they could do differently…
“The Power Macintosh G4… It’s not all that much slower than Pentium 4-based systems.”
“The PowerBook: It might be slower than a Toshiba, but at least it’s all cool and silvery-looking, look, see?!?”
“The iMac: Yes, it uses 100MHz memory, but just think about that for a second! 100MHz… that’s 100 MILLION HERTZ! Who says that’s slow?!? 100 million, that’s a LOT!”
This is just a theory, but I think Apple’s marketing department could be concocting an elaborate comedy routine designed to increase sales… I bought my Mac from the Apple store, and I was smiling all the way through reading that Powerbook marketing spooge. It made me smile, it put me in a good mood, it made me actually feel good to click the “buy now” button.
“Well, Steve, you didn’t convince me, but you DID make me collapse in hysterics, and because you tried so hard, I’ll give ya my money anyway.”
There’s been a lot of “brand fanaticism” messing around with comments on benchmark results lately – AMD vs. Intel, nVidia vs. ATI, Mac vs. PC, etc. I’m tired of reading the same idiotic blanket statements over and over again no matter which tech web site I visit. The most frustrating blanket statement is this one:
Benchmarks don’t matter.
Argh! This one really grates. People take the time to install a clean system, defragment, run the same benchmarks over and over again, write articles, make theories, research the topic involved, and formulate conclusions. Benchmarks DO have use and meaning in the real world. It’s just LIMITED, that’s all.
The benchmarking of PCs vs. Macs in commonly used digital video editing applications produces results that indicate which brand is likely to be better for digital video editing with the tools that are currently available. D’oh! You cannot just blindly dismiss the article because it has *gasp* benchmarks. They are perfectly valid results when considered in context.
Sometimes benchmarks on specific applications -may- highlight a more general problem, like insufficient bandwidth, that affects many other applications. People who do not agree with the general conclusions other posters draw from the specific benchmarks make the mistake of disregarding benchmarks on the basis of…well, absolutely nothing. They merely declare the benchmarks invalid. What they should be doing is arguing about whether or not one can properly draw such general conclusions from the specific benchmark results.
Another problematic reaction is to take one critical factor and declare it completely unimportant, thus rendering any objective performance measurement impossible. In this case, the “critical factor” is the CPU speed, aka the MHz myth. But the real world problem is that CPU speed can never be arbitrarily thrown out of the picture, because at some point it really does matter. The question people should be asking is, “When? When does it matter?”
At some point, the GIGAhertz myth – the one propagated by APPLE – will shatter. Hypothetically, if we were to benchmark a dual 3.5 GHz P4 PC against the same old dual 1 GHz G4 Mac that is currently available, the Intel box would mop the floor with the Mac. It is simply not possible for Apple to scale IPC as fast as Intel can scale clock speed. At some point, Apple MUST start bumping clock speed STEADILY. The P4 architecture will scale to 10 GHz through various manufacturing process changes in the next few years.
What will Jobs say to that when it happens? “Um, look I know Intel’s processors run 4x as fast as ours, but…erh…we have this new OmniVec thingy that will harness processing power from some alternate dimension where clock speed is irrelevant.”
Even when we understand that the top-of-the-line P4/Xeon’s trounce the top-of-the-line G4’s in most applications except those few, isolated applications specifically optimized to take advantage of the G4, there is another problem which has always tied Macs down: price. When a system performs worse AND costs more, it’s very hard on the Mac fanatics. They can easily set up benchmarks that have the Mac winning price OR performance, but only in a few isolated cases will it ever win the price/performance ratio.
When Mac-heads finally admit that performance isn’t the Mac’s strong suit, they declare performance to be irrelevant. (Getting tired of these blanket “X is irrelevant” statements yet?) They hold that there is always something special about owning a Mac, whether it be some kind of “coolness” factor, a prettier desktop, or whatever. I can’t argue with that – lots of people are into SFF PCs and other things that aren’t terribly focused on importance. But the problem is that at some point, price/performance DOES matter.
Most people use their computer and don’t wait for it to finish calculating some horrendously huge program (rendering, scientific). What would speed any computer up is an improvment to the User Interface.
Isn’t this the real speed advantage Mac has over x86? The interface requires less time to do tasks?
“Um, look I know Intel’s processors run 4x as fast as ours, but…erh…we have this new OmniVec thingy that will harness processing power from some alternate dimension where clock speed is irrelevant.”
Please delete null_pointer_us’s post, as it contains proprietary information pertinent to the future of our products. Our interdimensional OmniVec technology has not yet been declassified and is not yet available to the public. Our legal department will be contacting you.
🙂
I agree, proper benchmarks certanly do have a place. What I was saying up there is that the (subjective) benchmark of my own eyes is telling me that Apple has a problem <g>. And here I am, a long time Apple booster and Mac user. This is why I think Jobs really is looking at options, at least for the high end. He knows his hype and RDF are running out of gas as far as the PPC (and high end) is concerned. I can’t believe that he’s buying up all these digital companies and, at the same time, thinks that the current state of the PPC is going to complement the fruits that come out of those purchases. I can’t wait until next month – I want to see how much faster 10.2 is and, assuming new Power Macs come out, what the specs are. It sounds as if 10.2 will be significantly faster, which is good. But, if they come out with a tiny incremental speed boost in the processor, that will be a pitiful sight. I don’t know, maybe there will also be faster RAM and a faster bus, but if they don’t really deliver and Steve starts blowing hot air, I don’t think many will fall for it.
The odd thing is that OS X seems perfect for general businesses. While I’d never really want one as my sole home machine, it’s great for users who don’t need the mass of Windows apps, and for whom usability counts more than computation speed.
This is why these benchmarks seem good. They’re measuring how much time their work takes. What’s the bottleneck? Pure computation. If the bottleneck instead were stability, maintainability, or usability, the Mac would be a strong contender.
well eugenia, that’s interesting but you forget to state that beos is pervasively multithreaded, and so it can more efficiently multiprocess than either os x or win xp.
it is often said that beos can harness 95% of the potential of 2 processors, whereas windows xp and os x does only 60%
>what if adobe ported a ground up, rewritten, and highly optimzed version of photoshop for beos
That would never happen for obvious reasons. But even if they did that, you forget that BeOS has the same crappy compiler than OSX currently have. An old GCC (although the new GCC 3.1 has better support for OSX). *Especially* for BeOS, the gcc version used is very old and it is known to create slow code.
On OSX Photoshop is saved not because of the compiler, but because a lot of the altivec code has been written by hand. On x86, on a Windows machine, the compiler would do the optimizations for you, but on BeOS, to use good SSE-1/2 code, you would need to write assembly by hand, which is something that Adobe would never do. So no, I do not believe that Photoshop would be extremely faster on BeOS than on OSX or Windows. It might be from comparable, to much slower. But not faster.
as i’ve said before, this is really funny. you folks can fight forever whether or not athlon/intel is faster than mac (which it is). but what it all comes down to is what the majority of people out there need. most people want something that is easy to use. they want something where to install software means simply dragging a directory or file to the Applications directory, and run it, not having to deal with DLL-Hell or dependencies-hell. it’s like saying, “hey my fridge is colder than your fridge” or “hey my tricycle has three wheels and your bicycle has two wheels” or “hey my car has 500 bhp and yours has 160 bhp.”
speed matters to those whose life is all about computers and tweaking and gaming. framerate matters to the hardcore gamer. for those gamers, don’t even bother looking at any other platform than the pc. but in reality its the graphics card that counts. try running warcraft 3 or grand theft auto 3 on a 2 ghz pc with a savage4 32mb and you will understand.
speed matters to those who do video rendering. somewhat. just because it takes hours on a mac doesn’t mean it will simply take seconds on a pc. that would be absurd. it would still take hours, probably a little less. but then again, i know people who do rendering work with photoshop, bryce, etc, and for really huge work that they already know takes hours, they’ll start the rendering process when they are about to go home. they don’t do it early in the morning when they come in. they’d be fired for it. even people who do this stuff at home do it before sleeping, not when they wake up knowing they have other things to do.
so does it really matter? to those that need each and every little speed tweak it does. to those who play games for a living, to those who work with 2d and 3d graphics it does. to those who compile applications for a specific processor and platform it does. but does it matter to those will be using it to surf the net, do a little typing, play music, burn cds, watch videos, download porn, chat on IMs, do their financing? i highly doubt it. i’m looking at the computer these days, and it seems it has reached its peak. we already have stable OS’s (WinXP, Linux, MacOS X), we have systems faster than most people need. we have wonderful graphics cards. we have all the applications we need. so what’s next? that’s what i’m waiting for. i don’t want a 3ghz pc because that would waste too much power. what i need is something that i can turn on, and use.
i’ll get a new mac ibook, keep my 500mhz p3 system winxp system, keep my 166mhz system for linux stuff, and wait for the moment that foldable electronic paper becomes a reality.
The benchmarks of the last month are important for two reasons. The first reason is that is shows that the current “workstation” line is not as fast as the workstation line offered by the PC crowd. Most Photoshop benchmarks, all After Effects benchmarks, and pretty much all Lightwave benchmarks show that the G4, even with AltiVec, are now completely behind the times. I have seen benchmarks for hybrid uses of the vector units on G4’s, P4’s and Athlons (to do memory fetches and actual double precision math) and the G4 doesn’t do well here either. On top of this, the OS on Mac’s requires more horsepower. If it didn’t, Apple wouldn’t have invested all that time and money into Quartz Extreme.
That being said, the second factor to remember is the Xserve benchmark that CattBeMac mentioned. It showed that using DDR and L3 caches blew the old G4 performances out of the water. This means that *if* the G4’s get faster, and updated hopefully, and they can use the 166MHz DDR Ram instead of 133MHz DDR Ram, then things can be golden again at Apple.
These benchmarks had better be a wake up call for Apple and those trying to justify buying the overly priced towers that Apple sells. Considering the recent RDF field intensity coming out of Apple, I wouldn’t be surprised if this just causes them to spew out more propaganda instead of actual hardware innovation.
As a Mac user, I’ve been frustrated by the onslaught of the Apple apologists. I want to see Apple with a 30% market share. I want to see Microsoft with a 40% market share, and the rest filled with Linux and other OS’s. This would be OS nirvana for me. However if the Mac users of the world, me include, don’t critically analyze our own platform, then we are doomed to congratulate ourselves into our own demise.
Don’t think that anyone is convinced by continuing to tow the party line recited by Apple. This mentality is what broke communism. While it is true that if you repeat something enough times it appears true, that apparition is temporary. Eventually truth shines through, and the apologists are left holding their ideology in their hands and nothing else.
Please, stop saying things like “speed doesn’t matter, it’s user experience”. When someone is sitting in the store, and wants a reason to buy a computer that 90% of the world doesn’t use, their experience with it helps, but if it doesn’t blow their doors off in all other areas, it becomes a hard sell. In the case of Apple it is worse than that because the systems actually appear to work slower, as well as having slower MHz speeds. Apple can not rely on pretty pictures and ease of use issues. They need to blow the doors off of Intel performance, and also offer features that aren’t offered by the Windows platform. If they can’t, they are doomed to their 4% market share forever. In their current state, Apple execs are buying their PR faster than the consumers are. This doesn’t bode well for the long term increase in the number of Macintosh users.
Here we go again with this insane argument.
… goes to the entire computing experience. really now, raw speed is only one part of the equation.
when we want to get from point A to point B in our computing tasks, we have other very important concerns akin to autos.
is it easy to drive?
how does it handle?
how smooth is the ride?
how reliable is it?
how often do i have to troubleshoot?
how often do i have to overhaul it?
does it get me from point A to point B?
do i enjoy using it?
my “fastest” computer is my dual athlon with win2k pro, but the “car” i prefer to drive is my extremely reliable mac with osx.
I don’t place much weight on benchmarks. I like Linux regardless of how it performs against OSX or Windows. I like AMD processors because they are cheaper than Intel; that they are faster is just frosting on the cake. I like the look and feel of OSX much better than Microsoft’s The Sound of Music XP OS, but Mac hardware is too expensive so I don’t use it.
I buy and use what I like. I don’t care if it’s the best or the worst. As long as it fits my tastes and needs, I’ll use it. I think everyone would be happier if they ignored benchmarks and bought what they like.
Anyone had trouble with repartitioning under XP on a VIA KT133 based mobo? My one hard drive that I partitioned wioth partitionmagic 7 is fine, but the 3 I have done with XP’s disk management tools all died within 2 weeks… I could understand ONE duff HDD but not 3..
All had identical probs.. complete lockup of OS after ~50 minutes running, and increased booting time.. running Chkdsk /f would lock up the machine on boot.
Old, PM7 partitioned drive boots perfectly and runs for weeks on end without a reboot.
Any similar experience?
I just saw this video on the web.
Windows movie file format (6 MB) here:
http://l0g1c.iamgod.org/media/apple_gamer.wmv
Quicktime format (14 MB) here:
http://www.drunkgamers.com/switch0001.shtml
Pretty funny…
That is the best post I have ever read on this board yet. I couldn’t agree with you more
“Zork… Breakout… Super-Breakout… *PhotoShop*”
LOL
Thanks Eugenia for pointing us this, it make my day!
🙂
this one’s from Epinions, ;>D
http://www.ihateapple.com/av/Epinions256.asx
Hehehe, bashing time.
I declared finalized this power comparison with a mdronline.com report like four years ago, PPC is slower than Intel’s and AMD’s CPUs. Apple OS interfaces are nicer. Apple is easier. Apple empties your wallet twice a year. And… Apple OS interfaces are nicer. Shot it.
http://news.com.com/2100-1001-945310.html?tag=fd_top
Industrial Light and Magic has joined the empire, hardware wise.
The technical effects studio has switched from using RISC-Unix workstations from SGI to using Intel-based Dell systems running Linux for the bulk of its animation and special effects work, said Cliff Plumer, ILM’s chief technology officer. As part of the conversion, ILM recently deployed 600 Pentium 4 workstations.
“The Intel workstations that were deployed were probably 20 percent of the price of SGI workstations we bought a few years ago,” Plumer said. “Performance-wise, they are about three times as fast.”
~
I hope Apple makes a move soon. They need to switch to a high-performance mainstream processor.
#m
Screw Motorola. Get IBM to make those 2GHz chips so the PCs user would shut up already.
I didn’t know IBM has 2GHz processors. And if they do have them in the first place, I doubt Apple would void all of the “G4-is-better-than-G3” marketing :-).
> I wouldn’t have been surprised to see the P4 run 4x faster with optimizations.
Exactly. And you will need the latest versions of Visual C++ or GCC 3.x for that.
ICC, from Intel, accroading to benchmarks I have seen, does much better in P4-optimized code.
Apple smoking Dell on the Server spectrum;
http://www.xinet.com/benchmarks/benchmarks.2002/index.html
Compare the prices of all the machines there. If you actually take machines that are priced the same as the Dell, or a Dell that is priced the same as the rest of the machines, the results would differ greatly.
Apple Power Mac beating a Sony Vaio in numerous Photoshop tests;