PC World: Apple G5 Gets Trounced By Athlon 64 1063
StewedSquirrel writes "PC World magazine has published an article comparing the AMD Athlon 64 and Opteron versus Apple's G5 processor, both 64-bit contenders for the title of 'fastest desktop processor.' Apple has made many claims to be the first, fastest and only 64-bit processor for the desktop and workstation market, but (not mentioning the fact that Opteron beat the G5 to market by over 4 months) the benchmarks should speak for themselves. Of note is the 3.2GHz Pentium 4, coming in competitive with the G5, but significantly behind the Opteron and Athlon 64 systems."
Compatibility Issues? (Score:5, Interesting)
"But upgrading to XP 64 could mean giving up functionality without getting much in return. In fact, XP 64 looks like a throwback to Windows past: Its interface mirrors that of Windows 2000 or even Win 98. Microsoft has not disclosed what else will be in the OS, so it is possible that you'll still get most of XP's other features.
XP 64 won't have the 32-bit XP's support for DOS apps at all, nor will it run 16-bit apps (but it should have no trouble with 32-bit software). More important, 64-bit drivers for common hardware, such as printers, will be scarce when the OS debuts."
In moving from a Dual 1GHz G4 (Quicksilver 2002) to a Dual 2GHz G5, I have yet to find any software incompatibilities - everything works just fine.
This may change once my copy of Panther shows up, but my printer and other hardware continue to work for now.
Re:uhm... (Score:1, Interesting)
Eh, I'm still buying a G5 (Score:5, Interesting)
Re:uhm... (Score:3, Interesting)
The Opteron 140 and 240 series are workstation-class chips. Put an Opteron or two in a box with a bunch of hard drives--it's a server; put it in a box with a $1,000 graphics card--it's a workstation.
Premiere is a worthless benchmark. (Score:5, Interesting)
If you want to do a proper test, you'd use a crossplatform product that runs equally well on both platforms and is highly optimized for dual processors, like Discreet Cleaner or Combustion.
There's only one benchmark I can think of that is more worthless than Premiere, the "MSWord scroll test." For some stupid reason, some benchmarkers think it's a useful test to see how fast the can scroll to the end of a long Word document with the arrow key. Unfortunately, Word has a delay loop built into the scroll function, it even changes the delay loop depending on the speed of the CPU. The results are totally useless.
Re:How Tested (Score:2, Interesting)
out side Photoshop (which the Mac won hands down) the rest of the apps are plain stupid choices.
Another Worthless Benchmark (AWB) (Score:3, Interesting)
It's too bad that no rigor is applied to 99% of the benchmarks that are applied.
Raw CPU benchmarks like SPEC end up being compiler tests rather than processor tests.
"Real-world" app tests like this one are better, but only if the apps used are representative of apps used by the person reading the benchmark. They are not a realistic measure of holistic system performance.
Adobe Premiere? Come on! Does anyone on the Mac use that at all anymore? Is it even OSX native? Since Final Cut 1.0 came out 2+ years ago, anyone who considered that dog Premiere deserves what they get. Isn't there a better editing package on the PC, or is the Mac just that much better for video editing?
Microsoft products should be excluded from benchmarks on Microsoft's OS. Of COURSE Microsoft optimizes performance of their apps on their OS more than their apps on other OSs. That test is pointless.
The Quake test would be valid, except as many people here have pointed out, it's a 32-bit app, so it's not using any of the 64-bit capabilities of these boxes. When we get a native, 64-bit version and can compare it to two boxes with the same ATI video card, then it will be a valid test.
Re:Trounced? With this kind of comment? joke... (Score:4, Interesting)
I tried to be as fair as possible. However, it was usually difficult or impossible to get matching optical drives (as in SuperDrive), and many models not only didn't offer Gigabit ethernet, they didn't offer Ethernet *at all.* They had modems, though. Optional (I didn't add them, so they'd cost you extra.) Good grief.
So lacking some of the features that the G5 comes with standard, the Athlon-based PCs came in more expensive. Same old game on the "PCs are cheaper" front. The Athlons may be faster, but they'll cost you.
Now how is that for fair? "Faster costs more money." That sounds like reason to me. You can have a really fast processor, but at the expense of giving up a few things you might want in order to be productive. Like the internet.
Re:idiot (Score:3, Interesting)
Also, there are proven benchmarks that show that the opteron indeed runs faster, though due to the many differences between the opteron and the K7/P4, it's hard to narrow down exactly where the performance increase derives. The key feature, as stated is the doubling of the number of registers... The x86 has NEVER been able to perform register optimizations (like loop unrolling, etc) because there just arn't enough registers for even trivial book-keeping operations. The only way of really speeding up such apps is to have smart CPU's which rename explicit registers into remappable internal registers while instructions are running in parallel.. It's very hard however (and EXTREMELY cpu-specific) to have the compiler write code that can take advantage of this.
So having more regs means at least all the book-keeping can be explicitly stated by the compiler, and MANY explicit memory load/stores can be avoided (including CISC: add [mem], [mem], [mem] operations).
BUT, here's the misleading part.. The G5 has been RISC for a LONG time.. And one of the key tenents is 32+ registers. So while the x86-64 nicely ups the anti to a whopping 16 registers, don't get too excited, we're still living in the dark ages.
The main reason is that we're still doing register renaming, so to have 32 explicitly named regs you'd have to have nearly a thousand renamable registers.. This means slower addressing time (more propagation levels for a greater power-of two number of addressable registers). Is this additional cost warrented? The alternative is removing the power of the existing renaming facility, which may hurt more than help. So, given that CISC operations can still efficiently use memory addresses as virtual registers (e.g. the CPU keeps them in high-speed caches), it's a cheaper / faster over-all solution to just stick with 16 registers.
It still doesn't solve the problem of loop-unrolling however, so I'm not particularly happy with the decision (Being a compiler whore myself). AMD's not going to get another opportunity to radically alter the x86 line.. Intel is going to do this next and AMD will have to play catch up again.. And it's possible that the x86 won't last another major iteration (e.g. x86-64-AI?)
Re:Compatibility Issues? (Score:3, Interesting)
One of the limitations AMD put into the AMD64 series is that when you put the processor into long mode (64 bit mode) it no longer supports virtual real mode. The virtual real mode is how windows supports DOS apps. So in this way, it really is the chip, not windows, that is preventing DOS support.
That said, I don't see why they can't support 16 bit windows apps. That support is still there in the chip. I suppose you can always dual boot into 32 bit windows, and then you do get DOS support as well as 16 bit wondows support (such as it is).
Apple's G5 is a 32-bit platform (Score:2, Interesting)
Re:Compatibility Issues? (Score:4, Interesting)
Having said that few people need a full 64-bit OS and Panther is aimed squarely at consumers. I expect 64 bit pointers to come eventually, but it is probably better to stick with 32 bit pointers and keep better compatibility.
And, let's be honest. Most people running the Athalon-64 will be using it just as a fast Athalon running WindowsXP.
Re:idiot (Score:2, Interesting)
Also, the modern x86 processors have logic to help ease the problem of a limited register set. It's called register renaming. I am not sure the extra registers will have a profound effect on the performance.
Fairness (Score:3, Interesting)
If you want to compare the performance of CPUs with different ISAs, then you need to have the benchmark source code to compile it to the target ISA. This also brings the performance of the compiler into question but there's just no other way to do it if you want to compare CPUs with different ISAs.
One benchmark that people in the industry use to measure relative CPU performance across different ISAs is the SPEC benchmark. Just about every single computer maker from Dell to HP to Sun have submitted scores. Apple has not. This is in an of itself very telling. What is Apple afraid of? I'm hoping IBM releases a computer based on the G5 so that we can get some idea of the SPEC performance. Apple seems to believe it has something to lose by submitting a SPEC score.
Don't like SPEC? Please suggest some other CPU intensive applications to benchmark that have source code and publish your results.
Re:You're MISSING a point (Score:3, Interesting)
I.e., the BOXX based opteron workstation that shipped in June of this year [com.com] beats the pants off of Apple G5's shipping today, and is the first (serious) 64bit desktop to ship (workstations are desktops, BTW). So even AMD's own statements to the contrary (that Opterons were meant for servers) is irrelevant to the issue.
I.e., there has never ever been a point in time when the Apple G5 was shipping and ever has been the fastest desktop. The Opteron has been eclipsing it for its entire lifetime.
Furthermore, the Apple G5 was paper launched . It took them months to ship, unlike the Athlon 64s which shipped immediately upon launching (AMD's track record for doing this is remarkable -- they do this in order to underscore the fallacy of Intel's paper launches.) There may have been at most a two week window where G5's were shipping while AMD was not yet shipping Athlon 64s (but were shipping Opterons.)
If you read the article you would see that they did it in cooperation with their sister publication MacWorld. There is also nothing in their disclosures that raises any eyebrows (unlike the ridiculous Veritest/Apple Spec CPU 2000 disclosures.)
Re:This is a joke of a benchmark (Score:3, Interesting)
Ars Discussion of Athalon-64 vs. G5 [infopop.net]