PC to Powerbook: processor speed

[matticus008]

That's funny, I can move the mouse across not one, but two spanned displays without picking up the mouse.

Not at a normal speed on a normal-size mousepad (~8-9" wide). Using OS X's acceleration, you can force the mouse to cover more distance by moving it faster, and adding more displays scales the range, also thanks to acceleration. If spanned displays didn't scale, it would be damn near impossible to use a 30" cinema display or 2 20" or whatever other configuration you might have. NO Mac (and most especially those with touchpads) has adequately fast mouse movement out of the box, even on the fastest setting. It's a known limitation, and the reason why at least 3 good products exist to override OS X's ridiculous defaults (except with regard to acceleration).

Works just fine for me on my G5 and my PowerBook 1.67.

Yes, it does work fine, but it is slower than on Windows. The scroll matches pace and location on other platforms and there is a delay on Macs. It's simply not true that scrolling is as fast. Try them out side by side, you WILL see the difference, OS X simply is not as responsive. If you can't see the difference, then obviously it won't matter to you.

This doesn't happen to me at all on any Mac.

My Power Mac G5 has 2GB RAM, my PowerBook G4 has 1.5 GB, and the iMac G5 has 1 GB RAM. Maybe you need more RAM?

I shouldn't need more RAM. That's the whole point--OS X is more resource intensive. That visual quality comes at a price. You shouldn't have to have poorly-timed system notifications; this hasn't been a problem for any platform since the days of 16MB to 32MB of system RAM. OS X uses all the available RAM, which is great, but it needs to have a better memory manager so that there are enough resources available to do basic things. I'm not playing a 5GB movie or doing renderings all the time. When I'm just browsing the internet, I fully expect something as minimal as an alert beep to play at the correct time, not half a second or a second late. 512MB should be more than enough to handle OS X, Mail, and Firefox.

I'm not going to address the rest of your comments, because they are all "works fine for me." I never said that they didn't work acceptably. I said they are slower. Just slower. Not necessarily pronounced slowness, just noticeable. I use Linux, Windows, and OS X every day. And I notice most of these things every day. Do you use other platforms regularly, constantly switching and being frustrated by minor annoyances on an otherwise great system? Scrolling, Mouse issues, and redraw/windowing system issues (because of a much more resource-intensive GUI) are all KNOWN limitations with OS X. Sometimes they have even been conscious choices. I'm sorry if some of you find this to be a list of "fluff;" I didn't mean to offend the shrine of the holy Mac. The fact remains that there are many things that OS X does more slowly or more resource-intensively than other platforms. This translates into slower behavior on systems. You can throw an extra gig of RAM at it if you want, and that alleviates many of the problems, but if you go back to other platforms with a matching amount of RAM, you'll still find behavioral differences. Much of it you may not notice until your system is under load. You might think your Mac is a fast as can be; that's great, but it is not perfect.

Even a PowerMac with 2GB of RAM has a poor scrolling implementation, and mouse behavior is badly designed and not user-adjustable enough, regardless of hardware. Its GUI is much more resource intensive and Apple hardware often handles OpenGL with less robustness than PC-compatible hardware. Apple is working hard to solve these issues, but you can bet that they are there. I use multiple platforms and multiple UIs and I've tried multiple systems on each. I've spoken with developers (and even am one in my spare time), and I've given an objective and honest review of what OS X does not do as well as the competition. If you don't like that it's not perfect, get over it. It's not, and some of it never will be because of intentional design choices (mainly the GUI).

 

[hazooka]

P4 vs G5 benchmark

Someone of you guys asked for a benchmark, so here it is, fresh cinebench 2003 results :

P4 2,4Gh , 1Gb RAM
1processor render
2min 5 sec

dual G5 2Gh , 1 Gb RAM
1processor render
92 sec

2processor render
53 sec

I can see that you all argue about performance. Well, no comments
But one thing I must say, everybody gets what hi wants, if somebody loves benchmarking, running faster and faster, it' ok. I prefere to do my job once with Mac, not twice, even on the fastest PC in the world. Coz no matter how fast Pentium/Amd is, the problem will still be in the system itselve -> Windows. First Gates have to rewrite his system from scratch and let go all of it's **** inside, and get rid of thousands of component manufacturers before you can compare it to Mac. And we all know he can't afford it
End of story.
Thank you.

 

[ssrbrad]

Someone of you guys asked for a benchmark, so here it is, fresh cinebench 2003 results :

P4 2,4Gh , 1Gb RAM
1processor render
2min 5 sec

dual G5 2Gh , 1 Gb RAM
1processor render
92 sec

2processor render
53 sec

I can see that you all argue about performance. Well, no comments
But one thing I must say, everybody gets what hi wants, if somebody loves benchmarking, running faster and faster, it' ok. I prefere to do my job once with Mac, not twice, even on the fastest PC in the world. Coz no matter how fast Pentium/Amd is, the problem will still be in the system itselve -> Windows. First Gates have to rewrite his system from scratch and let go all of it's **** inside, and get rid of thousands of component manufacturers before you can compare it to Mac. And we all know he can't afford it
End of story.
Thank you.

Hazooka—

I have come to question your eminence’s greatness as the world’s sole arbiter of solid computing standards. Your authoritative argument is based on bad statistics and poor logic.

The 2.4-GHz Pentium 4 was originally released in March 2004. Since then, significant performance advancements have been made—specifically regarding its multithreading capabilities and Front Side Bus speed. While the latest P4s are unquestionably power-hungry and loud, they are also powerful processors.

The dual 2-GHz PowerMac G5 was released in November 2003, 8 months later. Today, an only-slightly improved version is available (the dual 2.5-GHz system). While Intel and AMD have been adding features over the course of the past year and a half (i.e. Hyperthreading, 64-bit architecture, FSB speed), IBM’s work on the G5 desktop processor has been relatively stagnant.

Alas, there are even more errors in your methods. The last time I checked, it took system resources to run the Operating System. While the 1-processor render only uses one processor in the rendering process, the other processor may be running the operating system and background processes. We’ll never know whether or not this is actually the case, since you provided incomplete and useless statistics.

If you wanted to provide useful data, you would have used a system that’s more comparable to the Mac system. For instance, you could compare a dual 2.2-Ghz Athlon 64-based system to the dual 2-Ghz G5. Those processors were released in the same time period, and both have a 64-bit architecture. While I’m not sure what those ratings would reveal, I don’t profess to know the results.

So what are the problems with your benchmarks?

1) You compared two technologies that have an age difference of 8 months.

2) Both of those technologies are already more than a year old.

3) You compared a dual-processor system to a single-processor system.

4) You compared 64-bit processors to a 32-bit processor.

5) You used your useless benchmark to launch into an unsupported self-ingratiating rant. You’re obviously superior to anyone who’s ever had the misfortune of using Windows.

I don’t even like PCs. I like them more than I like idiots, though. If you want to defend Macs, do it properly. They are intrinsically better; you don’t need to fudge statistics and practice deceit to prove it. In addition, your lack of acumen (as evidenced by your inability construct an intelligible post) does a disservice to people who support Macs and actually know what they’re talking about. In the future, I’d advise against publicizing your rants before running your writing by a teacher or two. You have time for that in Junior High, right?

And for the record, if anyone notices that my introductory sentence bears resemblance to a recent decision written by that jackass Scalia, know that I don’t like him either.

 

[arribadia]

Well, a fairer comparison would be using the latest AVAILABLE hardware on each platform and making a comparison.

How does a dual 2.5Ghz fair vs a dual Opteron 252 system?

Nevermind the fact that there may be a 3Ghz G5 in the works. I AM NOT INTERESTING WHAT STEVE IS DOING WITH HIS MARKETING DROIDS AND WITHHOLDING FROM US TO SUCKER AS MUCH OF OUR CASH AS POSSIBLE. The key is what is available NOW.

Most assuredly, the Opteron will utterly wreck the G5.

 

[Platform]

Well, a fairer comparison would be using the latest AVAILABLE hardware on each platform and making a comparison.

How does a dual 2.5Ghz fair vs a dual Opteron 252 system?

Nevermind the fact that there may be a 3Ghz G5 in the works. I AM NOT INTERESTING WHAT STEVE IS DOING WITH HIS MARKETING DROIDS AND WITHHOLDING FROM US TO SUCKER AS MUCH OF OUR CASH AS POSSIBLE. The key is what is available NOW.

Most assuredly, the Opteron will utterly wreck the G5.

Yes the opteron is amazing......they are even in oct options

 

[arribadia]

Yes the opteron is amazing......they are even in oct options

Don't get me wrong, I do like Apple a lot, but I certainly see AMD as a no nonsense company that does what they are supposed to do well.

Apple on the other hand... let's just say they make really beautiful hardware married with fine software... but they always give me the feeling that a lot about Apple revolves around their marketing too.

When was the last time AMD hyped something? When was the last time Apple hyped something?

Case closed.

 

[Mord]

http://www.barefeats.com/pentium4.html

tada

the dual 2Ghz G5 owns the dual 2GHz opteron in two tests and is narrowly beaten in three

the k8 and the g5 are roughly the same clock for clock and AMD only has a 100MHz advantage at this time,

also the G5 has scaled better than the athlon and the p4

p4: 3.2GHz -> 3.8GHz 18.8%
athlon 2.2GHz -> 2.6GHz 18.8%
g5 2GHz -> 2.5GHz 25%

the g5's bus scales with clock speed so has seen the same 25% increase

the athlons HTT bus has gone from 800MHz-1000MHz a 25% gain.

the p4's bus has gone from 800MHz - 1066MHz + 33% but they are only useing it on a few p4's and not others .

the only thing apple falls down on is memory it's stayed the same btu thats because apple cant afford to change motherboard that often and will only do so when there is a major update which is comming pretty soon

the g5 has not lagged if you look how slowly the x86 world is going

 

[CorvusCamenarum]

sheesh, i didnt have an argument like this since elementary.

problem solved.

Oh yeah...well...well...more angels can dance on the head of my pin than yours. Oh yeah and my oscar mayer weiner is bigger too cuz i have a MAC!!!!111!!!!!11!!!! and we all know if you overclock your just overcompensating...and my Steve can beat up your Steve. *End satire and poor spelling/grammar*

OK seriously, did I miss the memo that said that the Mac/Windows debate had to become the worst petty fight since the Crusades? It's worse than listening to people debate religion, and I live in the South....

 

[Abstract]

Apple sells computers. Intel and AMD make the cpus.

Take it for what you will.

 

[Mav451]

http://www.barefeats.com/pentium4.html
tada

the dual 2Ghz G5 owns the dual 2GHz opteron in two tests and is narrowly beaten in three

the k8 and the g5 are roughly the same clock for clock and AMD only has a 100MHz advantage at this time,

also the G5 has scaled better than the athlon and the p4

p4: 3.2GHz -> 3.8GHz 18.8%
athlon 2.2GHz -> 2.6GHz 18.8%
g5 2GHz -> 2.5GHz 25%

the g5's bus scales with clock speed so has seen the same 25% increase

but the athlons HTT bus has stayed the same at 800MHz bidirectional

the p4's bus has gone from 800MHz - 1066MHz + 33% but they are only useing it on a few p4's and not others .

the only thing apple falls down on is memory it's stayed the same btu thats because apple cant afford to change motherboard that often and will only do so when there is a major update which is comming pretty soon

the g5 has not lagged if you look how slowly the x86 world is going

AMD hasn't been stuck at 800Mhz HTT. They moved to the much faster 1000Mhz HTT link around June 1st 2003 when the nForce3 Ultra MCP chipset was released. MSI Neo2 Platinum anyone?

Additionally, while you believe that AMD has been, "stagnant", they have constantly been tuning the core technology on their A64 line. Heat output? (vs. Intel it's not even close. Running at load < ANY recent P4 at idle) They have seamlessly transitioned from 130nm to 90nm over the past 10 months.

Memory support? OCing headroom? That's all been improved.

• Costs? (one of the biggest, most important factors in the PC world)
  Its gone down. The 3400+ New Castle (2.4Ghz A64 core), is now less than $200. Yes, coming from over $400 we're talking about things getting ALOT more affordable.
  
• And affordability is a VERY important issue. Tell me how many people have a Dual G5 system vs. an A64 processor, at say 2.2Ghz. That includes the 754 3200+ NewCastle cores or the 939 3500+ cores (and that's either 130nm or 90nm). I personally don't know a single Apple user that has a G5 system (i.e. my age, within 20's). I know a ton more people with A64's.

Additionally, they also put that 90nm technology to use on the Sempr0n line. You should take a look at how well the $75 and $110 Semprons 2600+ and 3100+ OC with that 90nm technology. (2.5Ghz @ 303HTT for the 2600+, 2.68Ghz @ 298 HTT for the 3100+)
http://www.xbitlabs.com/articles/cpu/display/sempron-3100e_5.html

Stagnant? Sorry, I don't think so.

*edit*
And here's that wonderful link showing the heat output of A64 vs. Prescott:
http://anandtech.com/cpuchipsets/showdoc.aspx?i=2389
(even at LOAD, A64 takes less).

 

[Mord]

sorry i have not kept up to date on the amd scene recently

my last AMD was in slot A

 

[hazooka]

I like them more than I like idiots, though. If you want to defend Macs, do it properly. They are intrinsically better; you don’t need to fudge statistics and practice deceit to prove it. In addition, your lack of acumen (as evidenced by your inability construct an intelligible post) does a disservice to people who support Macs and actually know what they’re talking about. In the future, I’d advise against publicizing your rants before running your writing by a teacher or two. You have time for that in Junior High, right?

Well, well. You call me an idiot, but it is YOU who don't know of fact that it is possible to turn off one processor on dual mashines Very nice.....

The 2.4-GHz Pentium 4 was originally released in March 2004. Since then, significant performance advancements have been made—specifically regarding its multithreading capabilities and Front Side Bus speed. While the latest P4s are unquestionably power-hungry and loud, they are also powerful processors.

The dual 2-GHz PowerMac G5 was released in November 2003, 8 months later. Today, an only-slightly improved version is available (the dual 2.5-GHz system). While Intel and AMD have been adding features over the course of the past year and a half (i.e. Hyperthreading, 64-bit architecture, FSB speed), IBM’s work on the G5 desktop processor has been relatively stagnant.

Boy, you must have been very angry writing this, because year 2003 is before 2004, not after, right ?
Never mind 32/64 bit architecture, as long as the system is 32-bit the computer runs 32-bit, and both windows and mac os x are 32-bit systems. Oh, but you know it of course, right ?
Hypertheading is a benefit of P4. And even with this it is slower.

If you want to discuss instead of being rude dude, there is still a question of stability. And in this subject the main problems of PC's are both Windows and incompatibility of components, no doubt. And, as you are so smart, you should not even try to convince people othervise.
Oh, sure you can say that HP, IBM and other brand titles are very good, fast, stable ( still windows inside ) but unfortunetely more expensive than Mac.
And one last thing, do you really need so super fast computer ? what do you need this speed for, image editing, doom3 ? or maby you are a president of a big rendering/animation studio ? in that case I agree. Or you need to prove something ?
Be good or mom won't buy you a pie
Best wishes.

 

[arribadia]

As crazy as it sounds... I do hope AMD makes the Power/GX chips for Apple instead of Freescale...

They are really quite good at what they do.

AMD hasn't been stuck at 800Mhz HTT. They moved to the much faster 1000Mhz HTT link around June 1st 2003 when the nForce3 Ultra MCP chipset was released. MSI Neo2 Platinum anyone?

Additionally, while you believe that AMD has been, "stagnant", they have constantly been tuning the core technology on their A64 line. Heat output? (vs. Intel it's not even close. Running at load < ANY recent P4 at idle) They have seamlessly transitioned from 130nm to 90nm over the past 10 months.

Memory support? OCing headroom? That's all been improved.

• Costs? (one of the biggest, most important factors in the PC world)
  Its gone down. The 3400+ New Castle (2.4Ghz A64 core), is now less than $200. Yes, coming from over $400 we're talking about things getting ALOT more affordable.
  
• And affordability is a VERY important issue. Tell me how many people have a Dual G5 system vs. an A64 processor, at say 2.2Ghz. That includes the 754 3200+ NewCastle cores or the 939 3500+ cores (and that's either 130nm or 90nm). I personally don't know a single Apple user that has a G5 system (i.e. my age, within 20's). I know a ton more people with A64's.

Additionally, they also put that 90nm technology to use on the Sempr0n line. You should take a look at how well the $75 and $110 Semprons 2600+ and 3100+ OC with that 90nm technology. (2.5Ghz @ 303HTT for the 2600+, 2.68Ghz @ 298 HTT for the 3100+)
http://www.xbitlabs.com/articles/cpu/display/sempron-3100e_5.html

Stagnant? Sorry, I don't think so.

*edit*
And here's that wonderful link showing the heat output of A64 vs. Prescott:
http://anandtech.com/cpuchipsets/showdoc.aspx?i=2389
(even at LOAD, A64 takes less).

 

[hazooka]

Yes Arribadia, I definitely agree with you. Having the fastest computer in the world is their mission I suppose - madness !

And, Skatesc go for it , buy powerbook, don't let yourselve get caught by this insanity !!!

 

[ssrbrad]

Well, well. You call me an idiot, but it is YOU who don't know of fact that it is possible to turn off one processor on dual mashines Very nice.....

Of course I knew that. I wasn't sure that you did. I mean, your post was such a pack of **** that it seemed unlikely that you knew what the hell you were talking about. Now I see that you were less a dumbass then I thought, and you were instead being intentionally misleading. In any event, comparing a dual-processor system with one processor turned off to a single-processor system is still utterly useless. What were you trying to prove--that a Mac can beat a handicapped PC? Remind me again why you didn't use a dual-processor PC?

Boy, you must have been very angry writing this, because year 2003 is before 2004, not after, right? Never mind 32/64 bit architecture, as long as the system is 32-bit the computer runs 32-bit, and both windows and mac os x are 32-bit systems. Oh, but you know it of course, right ?
Hypertheading is a benefit of P4. And even with this it is slower.

Wow, you caught a typo. Good work. For the record, I'll list a few spelling errors from your reply post:

mashines
Hypertheading
othervise
unfortunetely

Also, nice job ignoring reality and pretending that when you're running a 64-bit system that the 32-bit Software functions identically to the way it would on a 32-bit system.

Oh, sure you can say that HP, IBM and other brand titles are very good, fast, stable ( still windows inside ) but unfortunetely more expensive than Mac.

One of the best things about PCs is that you can put exactly what you want in them. To be quite honest with you, I haven't bought a Manufactured PC in years (and I use both PCs and Macs every day). If you know what you're doing, instead of buying the filth available on the consumer market, you can build yourself a stable system. Cheaper, too.

And one last thing, do you really need so super fast computer ? what do you need this speed for, image editing, doom3 ? or maby you are a president of a big rendering/animation studio ? in that case I agree.

What a dolt. You're the one who posted the results of his rendering test. Hmm, what were you trying to prove about Macs?

Once again, I'd like to remind the world that I hate defending PCs.

 

[stevey500]

Which steve are we talking about here? Me.. or jobs?? jk!

 

[Mac-Xpert]

Yes, it does work fine, but it is slower than on Windows. The scroll matches pace and location on other platforms and there is a delay on Macs. It's simply not true that scrolling is as fast. Try them out side by side, you WILL see the difference, OS X simply is not as responsive.

Well, I have a PC (2.4 Ghz P4 -750 MB-Windows XP) right next to my 2.5 Ghz G5 here. Scrolling on the PC in fast when using the scroll bar but lags when using the scroll wheel on the mouse. On the G5 I can basically scroll through this, or any other page for that matter, instantly. There is no noticeable delay whatsoever. Its not in any way slower than the PC.

About OS-X not being as responsive, try this:
Delete 1000+ small files (about 15 Mb in total) on your PC from a windows server network share. It takes several minutes to complete, and the PC can get very sluggish when you try to do anything else at the same time. Deleting the same files in OS-X, from a network share on a Mac-OS-X server takes less than 3 seconds.

Copying the files on the PC from the windows-server to the desktop takes 25+ seconds

Copying the same files from a OS-X server to my G5 takes less than 5 seconds.

So much for PC's being faster on everything...

 

[darkwing]

Which steve are we talking about here? Me.. or jobs?? jk!

Hey Steve, congratulations on your switch from a fellow Steve.

Steve

 

[JCheng]

http://www.barefeats.com/pentium4.html

tada

the dual 2Ghz G5 owns the dual 2GHz opteron in two tests and is narrowly beaten in three

the k8 and the g5 are roughly the same clock for clock and AMD only has a 100MHz advantage at this time,

Okay, whats your point? The Opteron still beats the G5 in the majority of the benchmarks and absolutely slaughters it in games. In other words, sure we can find SOME programs where the G5 "owns" the Opteron (just like we can find some benchmark where the Opteron "owns" the G5). That doesn't change the fact that the Opteron beats the G5 in most programs as well as the most popular cross-platform programs (and this IS important because we know for a fact that they are well-optimized for all platforms).

And how about linking to the most recent comparison. No it doesn't really support your argument, but hey, at least we're showing a much more accurate depiction right?

also the G5 has scaled better than the athlon and the p4

p4: 3.2GHz -> 3.8GHz 18.8%
athlon 2.2GHz -> 2.6GHz 18.8%
g5 2GHz -> 2.5GHz 25%

the g5's bus scales with clock speed so has seen the same 25% increase

the athlons HTT bus has gone from 800MHz-1000MHz a 25% gain.

the p4's bus has gone from 800MHz - 1066MHz + 33% but they are only useing it on a few p4's and not others .

the only thing apple falls down on is memory it's stayed the same btu thats because apple cant afford to change motherboard that often and will only do so when there is a major update which is comming pretty soon

the g5 has not lagged if you look how slowly the x86 world is going

Do you really have any idea of what your talking about? The G5 gets a 25% increase in clockspeed from going to a 90nm manufacturing process and using liquid cooling. The fastest Athlon 64s are built on a 130nm manufacturing process and air-cooled but still clocks higher (and I do believe a 2.8 GHz version is coming out soon). In terms of design, AMD definitely takes the cake, and obviously, it does better in the real-world as well (or at least, that is the general case).

 

[Mord]

Okay, whats your point? The Opteron still beats the G5 in the majority of the benchmarks and absolutely slaughters it in games. In other words, sure we can find SOME programs where the G5 "owns" the Opteron (just like we can find some benchmark where the Opteron "owns" the G5). That doesn't change the fact that the Opteron beats the G5 in most programs as well as the most popular cross-platform programs (and this IS important because we know for a fact that they are well-optimized for all platforms).

And how about linking to the most recent comparison. No it doesn't really support your argument, but hey, at least we're showing a much more accurate depiction right?



Do you really have any idea of what your talking about? The G5 gets a 25% increase in clockspeed from going to a 90nm manufacturing process and using liquid cooling. The fastest Athlon 64s are built on a 130nm manufacturing process and air-cooled but still clocks higher (and I do believe a 2.8 GHz version is coming out soon). In terms of design, AMD definitely takes the cake, and obviously, it does better in the real-world as well (or at least, that is the general case).

did you even bother reading my post?, the dual 2.5GHz G5 dose not require watercooling, it's just to keep the thing quiet, they still consume less power than most x86 chips, you point out that the fastest AMD's are still 130nm, so your saying when they went to the 90nm prossess they could not even get a boost in clock speed they had to stick to 130nm until they get there San Diego chips out. http://www.barefeats.com/g5op.html thats a better benchmark page as it focuses on the G5 and opteron, with 3d games it had a 9800XT Vs the G5's underclocked 9800 pro which gave it a disadvantage, the G5 would be allot closer if the opteron used a 9800 pro instead.

these are the only real world benchmarks out there, i know there are many many dual 2.5GHz G5 owners on these forums but finding someone with a dual opteron is a tad harder.

i'd like to see full comparisons of equal clock for clock G5 Vs opteron

photoshop CS2
after effects
vue 5
bryce
shake
cubase
also a range of open GL games.

as most people who own opterons and G5's are not gamers as gamers do not buy expensive dualie systems and open gl performance is impotent for pro aps that are not cross platform like motion and 3d studio max.
you want to find some better benchmarks for the opteron go ahead, everything i ever found just linked back to barefeats.

stop talking crap, AMD make some good x86 chips but they are not the end all of cpu's.

that page dose not suprise me, the opteron won two benchmarks the G5 won two and the xeon won two (even though it was the only one compared to the G5 in maya) thats what i'm trying to say they are roughly the same clock for clock i'm not saying the G5 is the fastest in everything i'm just saying that it's still pretty competitive compared to the opteron and by no means is apple falling far far behind in the benchmark wars.

games really do not matter for this class of rig, most of those games were direct X and on quake 3 the g5 did pretty well as there was a hell of allot of effort optimising it equally for the mac and pc back when it came out.

 

[JCheng]

did you even bother reading my post?, the dual 2.5GHz G5 dose not require watercooling, it's just to keep the thing quiet,[ they still consume less power than most x86 chips,

You clearly have no idea what your talking about. The entire purpose of liquid cooling in ANY design is to allow for far more efficient cooling than traditional heatsinks. Thats even stated on the apple website and as would be obvious to everyone to but yourself, a liquid-cooled system could easily be clocked faster (and with better stability) than a similar air-cooled one (just look at the liquid cooled 4+ GHz P4s some specialized computer manufacturers are making). The fact that its quieter in many cases is merely a secondary benefit that has absolutely nothing to do with the fact that it provides better cooling.

you point out that the fastest AMD's are still 130nm, so your saying when they went to the 90nm prossess they could not even get a boost in clock speed they had to stick to 130nm until they get there San Diego chips out.

Once again sheer ignorance on your part. ANY decrease in manufacturing process (whether it be due to higher yields, smaller die size, more advanced manufacturing techniques, etc) is going to lead to better scaling, thats exactly why companies like IBM and Intel switched to 90nm despite higher heat dissipation. The very fact that the 90nm San Diego cores will boast higher clock speeds only proves my point. Seriously, you really don't know anything about computers if you don't think a liquid-cooled 90nm Athlon64 can be clocked faster than a 130nm aircooled model.

these are the only real world benchmarks out there, i know there are many many dual 2.5GHz G5 owners on these forums but finding someone with a dual opteron is a tad harder.

http://www.barefeats.com/g5op.html thats a better benchmark page as it focuses on the G5 and opteron, with 3d games it had a 9800XT Vs the G5's underclocked 9800 pro which gave it a disadvantage, the G5 would be allot closer if the opteron used a 9800 pro instead.

i'd like to see full comparisons of equal clock for clock G5 Vs opteron

photoshop CS2
after effects
vue 5
bryce
shake
cubase
also a range of open GL games.

as most people who own opterons and G5's are not gamers as gamers do not buy expensive dualie systems and open gl performance is impotent for pro aps that are not cross platform like motion and 3d studio max.
you want to find some better benchmarks for the opteron go ahead, everything i ever found just linked back to barefeats.

And once again, an hopelessly biased distortion of reality on your part. Hector, I'll say this only once; you can bring up all the outdated and obsolete comparisons you want but the fact of the matter remains that as of currently the fastest Opteron beats the fastest G5 in almost every current up-to-date benchmark/real-world program. The only cross-platform benchmarks the G5 takes are the ones that are never used in the PC world and by all indications not all that well optimized for the pc platform.

stop talking crap, AMD make some good x86 chips but they are not the end all of cpu's.

that page dose not suprise me, the opteron won two benchmarks the G5 won two and the xeon won two (even though it was the only one compared to the G5 in maya).

Actually (if you bothered to count) the Opteron won three benchmarks and it would probably win Maya as well if Barefeats had bothered to include it in the comparison. A Dual 3.6 GHz Xeon system would also beat the G5 in Cinebench. Games of course, are a clear sweep for Opteron. Of the two benchmarks the G5 won, Bryce is all but unknown in the pc community and is never used by PC benchmark sites.

And when did I say that Opterons are end all cpu's?

thats what i'm trying to say they are roughly the same clock for clock i'm not saying the G5 is the fastest in everything i'm just saying that it's still pretty competitive compared to the opteron and by no means is apple falling far far behind in the benchmark wars.

games really do not matter for this class of rig, most of those games were direct X and on quake 3 the g5 did pretty well as there was a hell of allot of effort optimising it equally for the mac and pc back when it came out.

Actually, I agree with you that the G5 is competitive with the current x86 offerings. The fact of the matter is however that your using false statements, useless comparisons and outdated benchmark to make your point (and to this effect, you distort the truth in ways that I have already pointed out). Ignorance begets ignorance, why not tell things as they are?

 

[Mord]

You clearly have no idea what your talking about. The entire purpose of liquid cooling in ANY design is to allow for far more efficient cooling than traditional heatsinks. Thats even stated on the apple website and as would be obvious to everyone to but yourself, a liquid-cooled system could easily be clocked faster (and with better stability) than a similar air-cooled one (just look at the liquid cooled 4+ GHz P4s some specialized computer manufacturers are making). The fact that its quieter in many cases is merely a secondary benefit that has absolutely nothing to do with the fact that it provides better cooling.



Once again sheer ignorance on your part. ANY decrease in manufacturing process (whether be due to higher yields, smaller die size, more advanced manufacturing techniques, etc) is going to lead to better scaling, thats exactly why companies like IBM and Intel switched to 90nm despite higher heat dissipation. The very fact that the 90nm San Diego cores will boast higher clock speeds only proves my point. Seriously, you really don't know anything about computers if you don't think a liquid-cooled 90nm Athlon64 can be clocked faster than a 130nm aircooled model.


these are the only real world benchmarks out there, i know there are many many dual 2.5GHz G5 owners on these forums but finding someone with a dual opteron is a tad harder.



And once again, an hopelessly biased distortion of reality on your part. Hector, I'll say this only once; you can bring up all the outdated and obsolete comparisons you want but the fact of the matter remains that as of currently the fastest Opteron beats the fastest G5 in almost every current up-to-date benchmark/real-world program. The only cross-platform benchmarks the G5 takes are the ones that are never used in the PC world and by all indications not all that well optimized for the pc platform.



Actually (if you bothered to count) the Opteron won three benchmarks and it would win Maya as well if Barefeats had bothered to include it in the comparison. A Dual 3.6 GHz Xeon system would also beat the G5 in Cinebench. Games of course, are a clear sweep for Opteron. Of the two benchmarks the G5 won, Bryce is all but unknown in the pc community and is never used by PC benchmark sites.

And when did I say that Opterons are end all cpu's?



Actually, I agree with you that the G5 is competitive with the current x86 offerings. The fact of the matter is however that your using false statements, useless comparisons and outdated benchmark to make your point (and to this effect, distort the truth in ways that I have already pointed out). Ignorance begets ignorance, why not tell things as they are?

What you don't realise is that watercooling is used because it makes less noise per watt of heat dissipated the 2.5GHz 970FX dissipates only 50w of heat compared to the 105w of heat a 3.8GHz P4 dissipates and the P4 has a roughly twice as big die size so thats about the same amount of heat per sq mm it's still allot, apple prides itself on quiet machines and a 3.8GHz p4 cooler is anything but quiet, they would have another MDD G4 situation on there hands so they went with water cooling not because they had to but because it was the better option.

The 2.5GHz G5 is not overclocked, it's run at manufactuers spec with a silent cooling solution, just like this solution is for silenceing pc's http://www.zalman.co.kr/eng/product/view.asp?idx=63&code=021

You Ignore the point that i made that AMD went 90nm when intel and IBM did but did not get a clock speed advantage, all they got was lower power on a smaller die.

Cooling is not usually the bottleneck when overclocking the core voltage is and when you bump the core voltage you need better cooling, sure if it runs too hot it crashes and that can limit the overclocking but unless it's running above about 60 degrees it's ok. the core volatge on the dual 2.5GHz G5 is at what it is sold to run at 1.4V, as i stated the watercooling is not what got that 25% speed increase it's the work IBM did.

Hahahahah, you say the tests are not optimised for the pc, it's the oposite, if more of your users are useing pc's who are you going to optimise more for the pc version or the mac version, you tryed to make out that AMD own in the 3d tests where the tables favor the pc's even more, the benchmarks done were done on commonly used cross platform software, the benchmarks used at toms hardware or any other pc site are not available for the mac.

I tell things exaclt how they are, as someone thats built pc's and maintained and upgraded macs for the last 5 years of my life and as an apple certified technician you are plain wrong.

Proper benchamarks are not out there and the ones we have say the opteron and g5 are roughly neck and neck i know the opteron has a 100MHz clock speed advantage but thats quite small in the scheme of things.

(sorry about bad spelling i'll edit is latter i'm in a rush)

 

[JCheng]

what you don't realise is that watercooling is used because it makes less noise per watt of heat dissipated the 2.5GHz 970FX dissipates only 50w of heat compared to the 105w of heat a 3.8GHz P4 dissipates and the P4 has a roughly twice as big die size so thats about the same amount of heat per sq mm it's still allot, apple prides itself on quiet machines and a 3.8GHz p4 cooler is anything but quiet, they would have another MDD G4 situation on there hands so they went with water cooling not because they had to but because it was the better option.

The 2.5GHz G5 is not overclocked, it's run at manufactuers spec with a silent cooling solution, just like this solution is for silenceing pc's http://www.zalman.co.kr/eng/product/view.asp?idx=63&code=021

Another blatant distortion of the truth, you took (what I suppose) is the typical dissipation of the G5 and went on to compare it with the maximum dissipation of the P4. If you want to compare maximum dissipations, the 2.5 GHz G5 dissipates about 80W (this has been proven to be consistent with the rate that dissipation has increased relative to G5 clockspeed scaling) while the P4 dissipates about 118W and the Opteron dissipates about 105W. The G5 thus has a w/mm^2 of 1.3W/mm^2, which is about 24% higher than the P4's W/mm^2. You do know why Intel hasn't released an air-cooled 4 GHz P4 thusfar right? (-even though its been proven that Prescott could easily scale to that speed and yes, before you ask, it does have something to do with heat)

you ignore the point that i made that AMD went 90nm when intel and IBM did but did not get a clock speed advantage, all they got was lower power on a smaller die.

Are you just trying to be ignorant? I've already shown that 90nm 2.8 GHz A64s will come out soon (and the only reason AMD hasn't released them already was because of lack of competition). You have no argument here, clearly, the 90nm A64s will scale higher than the 130nm version, thats just not debatable.

cooling is not usually the bottleneck when overclocking the core voltage is and when you bump the core voltage you need better cooling, sure if it runs too hot it crashes and that can limit the overclocking but unless it's running above about 60 degrees it's ok. the core volatge on the dual 2.5GHz G5 is at what it is sold to run at 1.4V, as i stated the watercooling is not what got that 25% speed increase it's the work IBM did.

You clearly have no idea of what your talking about, might I refer you to this article? Gee, and I wonder why everyone for the past year or so have been talking about single-core cpus hitting a "thermal wall"? Could it be because their getting too hot for conventional cooling solutions?

hahahahah, you say the tests are not optimised for the pc, it's the oposite, if more of your clientel are useing pc's who are you going to optimise more for the pc version or the mac version, you tryed to make out that AMD own in the 3d tests where the tables favor the pc's even more, the benchmakrs done were done on comenly used cross platform bits of software, the benchmarks used at toms hardware or any other pc site are not available for the mac.

Wow, are you trying to troll or something? Doesn't it strike you as funny that the Opteron wins all the cross-platform benchmarks that are popular on both platforms while the G5 wins the benchmarks that are all but unheard of on the pc side? Seriously, how many times do I need to repeat myself?

i tell things exaclt how they are, as someone thats built pc's and maintained and upgraded macs for the last 5 years of my life and as an apple certifyed technician you are plain wrong.

Hmm, thats weird from what I can get from your profile, your a 16 year old teenager whose probably in some high school in london. Your seemingly lack of knowledge for computer hardware and poor writing technique seems to reinforce this notion.

And if you want to "tell things exactly as they are", than why are you using outdated comparisons to make your point when you could just bring up the most current comparison?

proper benchamarks are not out there and the ones we have say the opteron and g5 are roughly neck and neck i know the opteron has a 100MHz clock speed advantage but thats quite small in the scheme of things

(sorry about bad spelling i'll edit is latter i'm in a rush)

The ones we have say that the Opteron beats the G5 in every popular cross-platform benchmark, 3/5 of the CPU benchmarks (4/6 if you include maya) and every game. Nice job distorting the truth again.

 

[840quadra]

Can we please keep this Technical and not personal ?

 

[Mav451]

Now play nice fellas. I think Hector's just a lil fed up with the onslaught of AMD-heads on this board. I think the #'s speak for themselves, enough that we don't HAVE to talk about it or push it in their face. If you see my comments in the other thread, I try to calm jiggie down from going "omg omg omg" ala Homer and jelly-creme donuts



Yeah dual core is nice, but let's be realistic here. Neither the X2/x75 or the new revision of the G5 are in anyone's hands rite now. When everything is TANGIBLE to the normal man here, then let's talk about it.