And then there's the shit kicking the G5 gave the P4 during the Photoshop bakeoff at WWDC.
Yea, but that does not count. Bench testing is where it's at babe. Who cares about real world performance?!
I never understood this fascination with Bench testing. Theoretical scenarios testing theoretical numbers that differ from OS to OS, chip to chip, testing my chip against yours but with mine you have to do it this way if you want to see these theoretical results. G-r-e-a-t.
Amazing, so much furor over slapped together benchmarks that mean nothing. We don't even know what the testing conditions were. A couple of apps optimized (if at all) for a completly different chip running on a patched OS, and it "only" held its own.
Give it time folks, it will all come together but it might take a bit more then 1 or 2 days.
For what it is worth, this is old by a couple of weeks, but was posted on barefeats. Hopefully now that machines are actually shipping, we might get some real numbers soon once he gets his hands on a production version. Still it is probably worth more then any of the so-called benchmarks mentioned here so far.
8/9/03 -- Here's a tidbit for you weekend surfers: I tested a pre-production G5 Power Mac with dual 2GHz processors and a Radeon 9800 recently. I promised not to publish the results but I can say that it is faster than the 3GHz Pentium 4 in all but one of my famous "real world" tests (a 3D game). And that's without running Panther version of OS X or G5 optimized software.
Conclusion: If you ordered a G5, you won't be disappointed. If you are on the fence, I say jump... especially since I'm in the queue ahead of you.
As for richcigar, he is probably wisely avoiding this madness and getting to know his new G5 and what it can do in the real world now, and leaving the future to what will come.
As for me, I would be thrilled to have any of the G5's even with unoptimized apps and patched OS, cause I know it will only get better.
Far behind with poorly optimized code, yes. As one of those people on the thread mentioned, recompiling for the G4e in some cases provided 50% improvements in speed over the G4 optimized equivalent... the G5 and G4e are very different architectures so it's hard to tell what optimization will do.
Hmmm...well, either you recall incorrectly, or you saw a incorrect roadmap. Currently the G5 is running at 2 Ghz and the P4 is running at 3.2 Ghz, so there is a 60% clock speed advantage for Intel. Apple/IBM have committed to increasing clock speed to 3 Ghz within one year, which presumably means no later than Q3 2004. Intel is not committing to anything, but their official roadmap has the Pentium 4/5 at "3.4 Ghz or greater" through June 2004 (Q2 2004) (http://www.intel.com/products/roadm...rocess_roadmap&). There is some upside potential in the "or greater" part (though definitely not 1 Ghz worth of upside potential!), but the recently confirmed power dissipation problems of this processor are likely to limit the clock speed increases. Certainly 3.4 Ghz is the fastest projected speed through Q1 2004 (http://www.theinquirer.net/?article=11092), so at best we are likely to see is 3.6 Ghz in Q2 and 3.8 Ghz in Q3. The highest possible projection I have seen for Q3 2004 (from before the power dissipation problems were apparent) is 4 Ghz - that is unlikely to be attained (http://pc.watch.impress.co.jp/docs/...2/kaigai01l.gif).
So the bottom line is that in one year the G5 should be at 3 Ghz, whereas Intel should be between 3.6 to 3.8 Ghz, with 4 Ghz at the upper end of possibilities (but increasingly unlikely). So Intel's clock speed advantage will be reduced to between 20 to 33 percent (as opposed to 60 percent right now) - not bad! And of course the G5 will be getting much more done per clock cycle, so in real world performance it should have a definite advantage.
Waaah! A 1.6 is beating dual gig G4s...or have parity in unoptimized tests. WAHHHH!
Yeesh. Wonder why the 'net got a reputation for 'kneejerk' reactions. Machines just shipping. No optimised Mac benches. No optimised apps bar PS7. No Macworld reviews. No optimised OS re: Panther. And yeah, Apple just shipped a 'low end' machine that is holding its own with top end G4 machines of a little while ago.
WAAAAAAAAAH! (Throws rattle to ground...)
A low end G5 machine that boasts umpteen architectural updates that the G4s don't have. Put that in context of the G4 bumps Apple was giving last year.
And the dual 2 gigger aint shipped yet. It will give at least parity with the best PCs have. At worst, clobber the best the G4 has to offer.
And by the time we've finished complaining about the 1.6 G5's performance not pounding dual Xeons, the Rev' B will be shipping at dual 2.5 gig and soundly thrash the best Intel has to offer.
How long? Jan'. Remember that '2.5' press release. I'm sure said '2.5 gigger' will be ready relatively soon after the dual 2 gig ships. Enough to make a San Fran Jan splash 2004. I'm guessing 2.5 gig on .13 in a, well, would you believe, a over-engineered case to deal with such hot cpu bumps! Guess there was some kind of long term plan to that case...
Wah.
Lemon Bon Bon
The Future's Bright. The Future's the G5. The Future's Apple.
Did it occur to you guys that Steve might have tested the g5's at WWDC with a beta of Panther. If you look at www.hardmac.com aka macbidouille, there are benchmarks of xbench that shows the difference between 10,2,6 and 10,3. If we wait for optimized programs and we still get relativly "bad" results, we might consider the possibility of waiting for 10,3 before getting Kickass results.
Personally I love the G5 and still consider them great machines.
Is Photoshop the only application from Adobe thats ben optimised for the G5?
Actually I saw a benchmark yesterday over at macnn about a photoshop test done against an Athlon XP 2200 (1.83Ghz) where the athlon came in at 1 min for the suite of tests, and the 1.6Ghz G5 came in at 56 sec. And that was with the unoptimized Photoshop. So I'd assume with the 75-200% increase in performance that the new update brings, that there'd be some serious butt-kicking in that app. Especially if you'd extrapolate to a dual 2Ghz G5. So I guess it's in the real-world optimized apps where we're going to see our biggest increase in performance right now until these benchmarks get in line.
I thought it time for me to chime in with another perspective.
Those of you who live by benchmarks are nuts! That said, it is not entirely your fault. I said that Apple was wrong to intro these new machines with benchmark mania. Half the announcement was based on benchmarks and so was most of the press that followed. All you people who are beating up on the benchmark geeks are also caught up in a knee jerk reaction mode. Apple, in the person of SJ is the one who made benchmarks an issue and wanted them to be used as a tool for testing performance. Now that there are some real machines out in the hands of a few lucky people, it is only natural for them to do what SJ practically encouraged them to do, run benchmarks. Don't be too hard on them. This is in part how Apple sold the product. They probably should have released a suite of benchmark tools with optimizations that were similar to what SJ and Co used on stage. Apple announced the world's fastest, Pentium crushing PC based partly on benchmark scores. Stupid! Stupid! Stupid! As the Bible says, "He who lives by the benchmark, shall die by the benchmark." or something to that affect.
What a disappointing result on Cinebench. I AM MOVING TO DELL!
Serious. I use Cinema as my 3D app. Honestly when I saw the posts from the Maxon guys in Postforum I was kind of bummed to be told that the Dual G5 (even used a Cinebench that recognizes G5) still can't beat the fastest Xeon.
After seeing the demo of PS, Mathmetica, Logic, etc etc....even Renderman benchmark I really expect to see the same magnitude of speed boost in Cinema/Cinebench - much faster than G4 and even faster than PC....
Cinebench is indentical to the Cinema application. It loads all the real projects created by users and run the rendering, fly-thru etc to test the hardware performance.....pretty much that the performance on Cinebench will be the same on the Cinema application.
Looks like I better to wait for 3GHz G5 to come..... or buy couple of cheap PCs
Anyway....would be curious on how Maya and Lightwave run on G5 (of course if they are optimized)
Comments
Originally posted by Existence
Even with heavy G5 optimizations, it looks like the G5 is still far behind the x86 camp according to a Maxon employee who posted here.
The G5 is just outmatched and outgunned by current P4s. It will get worse as Athlon64 comes out next month and Intel Prescott(P5) later this year.
Compare this to a single 3.2GHz P4 that scores about 370.
Maybe not quite true:
Athlon 64 supply to be limited until 2004
"AMD Athlon 64 processors are going to be hard to come by in the months following the chips launch on 23 September."
There have been rumours about Intel delaying its Roadmap, too!
Besides: There's no Xeon 3.2GHz yet 3.06 is the max, so 500 seems fairly good!
A Dual Xeon 3GHz should be around 650!... I could live with 500, honestly! ;-)
But it's good to know Apple ist working with them!...
Originally posted by Existence
Even with heavy G5 optimizations, it looks like the G5 is still far behind the x86 camp according to a Maxon employee who posted here.
The G5 is just outmatched and outgunned by current P4s. It will get worse as Athlon64 comes out next month and Intel Prescott(P5) later this year.
Compare this to a single 3.2GHz P4 that scores about 370.
Umm... if a P4 is doing 370, and a dual G5 is doing close to 500, does the G5 not win? So how is it "far behind" if it is winning?
Originally posted by the cool gut
And then there's the shit kicking the G5 gave the P4 during the Photoshop bakeoff at WWDC.
Yea, but that does not count. Bench testing is where it's at babe.
I never understood this fascination with Bench testing. Theoretical scenarios testing theoretical numbers that differ from OS to OS, chip to chip, testing my chip against yours but with mine you have to do it this way if you want to see these theoretical results. G-r-e-a-t.
Provided richigar was on the up-and-up. It seems strange he hasn't been back, even with all the techies and their concern about benchmarks.
Originally posted by OldCodger73
<<well maybe if we hadn't scared richcigar away he would have gave us some real world performance tests.....Like browser resizing>>
Provided richigar was on the up-and-up. It seems strange he hasn't been back, even with all the techies and their concern about benchmarks.
Isn't he at Yale Divinity School? So he wouldn't be pulling our legs would he? We can individually email him from his profile. I just did.
Give it time folks, it will all come together but it might take a bit more then 1 or 2 days.
For what it is worth, this is old by a couple of weeks, but was posted on barefeats. Hopefully now that machines are actually shipping, we might get some real numbers soon once he gets his hands on a production version. Still it is probably worth more then any of the so-called benchmarks mentioned here so far.
from http://www.barefeats.com/#quick
8/9/03 -- Here's a tidbit for you weekend surfers: I tested a pre-production G5 Power Mac with dual 2GHz processors and a Radeon 9800 recently. I promised not to publish the results but I can say that it is faster than the 3GHz Pentium 4 in all but one of my famous "real world" tests (a 3D game). And that's without running Panther version of OS X or G5 optimized software.
Conclusion: If you ordered a G5, you won't be disappointed. If you are on the fence, I say jump... especially since I'm in the queue ahead of you.
As for richcigar, he is probably wisely avoiding this madness and getting to know his new G5 and what it can do in the real world now, and leaving the future to what will come.
As for me, I would be thrilled to have any of the G5's even with unoptimized apps and patched OS, cause I know it will only get better.
Originally posted by mark_wilkins
Far behind with poorly optimized code, yes. As one of those people on the thread mentioned, recompiling for the G4e in some cases provided 50% improvements in speed over the G4 optimized equivalent... the G5 and G4e are very different architectures so it's hard to tell what optimization will do.
-- Mark
EasyAnswer™ :
G5-optimization
Hmmm...well, either you recall incorrectly, or you saw a incorrect roadmap. Currently the G5 is running at 2 Ghz and the P4 is running at 3.2 Ghz, so there is a 60% clock speed advantage for Intel. Apple/IBM have committed to increasing clock speed to 3 Ghz within one year, which presumably means no later than Q3 2004. Intel is not committing to anything, but their official roadmap has the Pentium 4/5 at "3.4 Ghz or greater" through June 2004 (Q2 2004) (http://www.intel.com/products/roadm...rocess_roadmap&). There is some upside potential in the "or greater" part (though definitely not 1 Ghz worth of upside potential!), but the recently confirmed power dissipation problems of this processor are likely to limit the clock speed increases. Certainly 3.4 Ghz is the fastest projected speed through Q1 2004 (http://www.theinquirer.net/?article=11092), so at best we are likely to see is 3.6 Ghz in Q2 and 3.8 Ghz in Q3. The highest possible projection I have seen for Q3 2004 (from before the power dissipation problems were apparent) is 4 Ghz - that is unlikely to be attained (http://pc.watch.impress.co.jp/docs/...2/kaigai01l.gif).
So the bottom line is that in one year the G5 should be at 3 Ghz, whereas Intel should be between 3.6 to 3.8 Ghz, with 4 Ghz at the upper end of possibilities (but increasingly unlikely). So Intel's clock speed advantage will be reduced to between 20 to 33 percent (as opposed to 60 percent right now) - not bad! And of course the G5 will be getting much more done per clock cycle, so in real world performance it should have a definite advantage.
Waaah! A 1.6 is beating dual gig G4s...or have parity in unoptimized tests. WAHHHH!
Yeesh. Wonder why the 'net got a reputation for 'kneejerk' reactions. Machines just shipping. No optimised Mac benches. No optimised apps bar PS7. No Macworld reviews. No optimised OS re: Panther. And yeah, Apple just shipped a 'low end' machine that is holding its own with top end G4 machines of a little while ago.
WAAAAAAAAAH! (Throws rattle to ground...)
A low end G5 machine that boasts umpteen architectural updates that the G4s don't have. Put that in context of the G4 bumps Apple was giving last year.
And the dual 2 gigger aint shipped yet. It will give at least parity with the best PCs have. At worst, clobber the best the G4 has to offer.
And by the time we've finished complaining about the 1.6 G5's performance not pounding dual Xeons, the Rev' B will be shipping at dual 2.5 gig and soundly thrash the best Intel has to offer.
How long? Jan'. Remember that '2.5' press release. I'm sure said '2.5 gigger' will be ready relatively soon after the dual 2 gig ships. Enough to make a San Fran Jan splash 2004. I'm guessing 2.5 gig on .13 in a, well, would you believe, a over-engineered case to deal with such hot cpu bumps! Guess there was some kind of long term plan to that case...
Wah.
Lemon Bon Bon
The Future's Bright. The Future's the G5. The Future's Apple.
Personally I love the G5 and still consider them great machines.
So I'm with Lemon Bon Bon. G5 is the future.
Originally posted by Multimedia
Yes. Has been since Tuesday evening. Adobe Photoshop 7.0.1 G5 Processor Plug-in update for Mac OS X
Is Photoshop the only application from Adobe thats ben optimised for the G5?
(It's like we keep going around in circles)
Look kids. Big Ben. Parliment.
I loved European Vacation.
Originally posted by moox12
Is Photoshop the only application from Adobe thats ben optimised for the G5?
Actually I saw a benchmark yesterday over at macnn about a photoshop test done against an Athlon XP 2200 (1.83Ghz) where the athlon came in at 1 min for the suite of tests, and the 1.6Ghz G5 came in at 56 sec. And that was with the unoptimized Photoshop. So I'd assume with the 75-200% increase in performance that the new update brings, that there'd be some serious butt-kicking in that app. Especially if you'd extrapolate to a dual 2Ghz G5. So I guess it's in the real-world optimized apps where we're going to see our biggest increase in performance right now until these benchmarks get in line.
The G4 got only one FPU or?
The G5 got two.
Does a compiler need to know how much FPU units are out there?
Maybe we'll get twice the power
I mean the G5 scaled like a G4 at the same mhz.
Look back when there were rumors that the G5 is twice as fast as the G4. It was about 6 months ago...or?
Originally posted by moox12
Is Photoshop the only application from Adobe thats ben optimised for the G5?
No, Acrobat reader is too, smooth scrolling and all that.
Those of you who live by benchmarks are nuts! That said, it is not entirely your fault. I said that Apple was wrong to intro these new machines with benchmark mania. Half the announcement was based on benchmarks and so was most of the press that followed. All you people who are beating up on the benchmark geeks are also caught up in a knee jerk reaction mode. Apple, in the person of SJ is the one who made benchmarks an issue and wanted them to be used as a tool for testing performance. Now that there are some real machines out in the hands of a few lucky people, it is only natural for them to do what SJ practically encouraged them to do, run benchmarks. Don't be too hard on them. This is in part how Apple sold the product. They probably should have released a suite of benchmark tools with optimizations that were similar to what SJ and Co used on stage. Apple announced the world's fastest, Pentium crushing PC based partly on benchmark scores. Stupid! Stupid! Stupid! As the Bible says, "He who lives by the benchmark, shall die by the benchmark." or something to that affect.
Serious. I use Cinema as my 3D app. Honestly when I saw the posts from the Maxon guys in Postforum I was kind of bummed to be told that the Dual G5 (even used a Cinebench that recognizes G5) still can't beat the fastest Xeon.
After seeing the demo of PS, Mathmetica, Logic, etc etc....even Renderman benchmark I really expect to see the same magnitude of speed boost in Cinema/Cinebench - much faster than G4 and even faster than PC....
Cinebench is indentical to the Cinema application. It loads all the real projects created by users and run the rendering, fly-thru etc to test the hardware performance.....pretty much that the performance on Cinebench will be the same on the Cinema application.
Looks like I better to wait for 3GHz G5 to come..... or buy couple of cheap PCs
Anyway....would be curious on how Maya and Lightwave run on G5 (of course if they are optimized)