[firefly]

Considering the power of the faster Athlons, I just wanted to ask whether a dual 800 G4 is faster than an AthlonXP 1800+ in Photoshop? Does Apple still have the lead? I sure hope so...

[Scotttheking]

off to art and graphic design forum.

[godzookie2k]

uhmmmm sure. Then again we haven't seen a pshop bake off in a while.


Nick

[r-0X#Zapchud]

Heh, I doubt that the Atlon can compare to the 2xG4's.

This weekend, I was at a LAN at one of my friends, who just got a new machine, 512 MB DDR RAM, Athlon 1700+, else it was pretty much the same as my G4 466 w/ 1GB RAM (350 megs allocated to PS). Ok, we ran only 1 filter, but I was very amazed by the results. The Athlon machine's picture was in 1024x768, and my pic (G4) were in 1280x1024 (60% difference). We took the Radial Blur filter, set to best quality and set the slider to 100.

We finished in about the same time

Ok, this filter has to be very well Altivec-optimized, and other filters are not. But it was damn amazing, since the Athlon has 1000 more MHz, and twice as fast RAM (though only half the amount).

[r-0X#Zapchud]

Mffhh...just have to add that we were both playing MP3's in the background(WinAMP & iTunes). This should give the Athlon even greater advantage, since it then has even more processor-cycles left. Operating systems: Windows 2000 and Mac OS 9.2.1!

This is true!

[KellyHogan]

Originally posted by firefly:
<STRONG>Considering the power of the faster Athlons, I just wanted to ask whether a dual 800 G4 is faster than an AthlonXP 1800+ in Photoshop? Does Apple still have the lead? I sure hope so...</STRONG>

NO it isn't. The Dual 800 is about the speed of a 1300Mhz Athlon XP (1600+). But the price is more than double.

[KellyHogan]

Originally posted by r-0X#Zapchud:
<STRONG>Mffhh...just have to add that we were both playing MP3's in the background(WinAMP & iTunes). This should give the Athlon even greater advantage, since it then has even more processor-cycles left. Operating systems: Windows 2000 and Mac OS 9.2.1!

This is true!</STRONG>

Sure. But my Athlon 1400 toasts all three G4's I have had. My Dual G4-450 was HALF the speed of the Athlon. Now I have published benchmarks with video several times. I wish more people would do that before making such claims.

[Anand]

and if you don't agree with KellyHogan than get the heck off these boards!

(:

[r-0X#Zapchud]

Ok, I agree, but I dont believe it before I see it!
But my G4 WAS faster @ Radial Blur

[Fyre4ce]

Duh, one machine is going to be better at some things and the other at others. IT DEPENDS ON WHAT YOU ARE DOING! Neither machine has an across-the-board lead. Admittedly, the Athlon is cheaper. Here's a way to think about it: Athlon = riced out honda civic, G4 = BMW. They both have comparable 1/4 mile times and skidpad performance, but the BMW looks better, is more elegant, is more fun to drive, and is more comfortable. The Athlon is just a box. The G4 is a pleasure to use. Just because it's cheaper to buy a Civic and rice it out than it is to buy a BMW, doesn't mean that someone who can afford both is stupid for buying the BMW.

Originally posted by Fyre4ce:
<STRONG>Duh, one machine is going to be better at some things and the other at others. IT DEPENDS ON WHAT YOU ARE DOING! Neither machine has an across-the-board lead. Admittedly, the Athlon is cheaper. Here's a way to think about it: Athlon = riced out honda civic, G4 = BMW. They both have comparable 1/4 mile times and skidpad performance, but the BMW looks better, is more elegant, is more fun to drive, and is more comfortable. The Athlon is just a box. The G4 is a pleasure to use. Just because it's cheaper to buy a Civic and rice it out than it is to buy a BMW, doesn't mean that someone who can afford both is stupid for buying the BMW.</STRONG>

That is crap excuse. If you are using Photoshop, it doesnt matter what the box looks like or the OS. The apps features and the speed is all that matters along with stability. On all counts a fast, cheap Athlon running Windows 2000 will beat a G4 running OS 9.X

I've been testing Wintel and Apple machines for a number of years now. a good rule of them to predict the speed of the G3 and G4 is as follows:

-A G3 clocked the same as a Pentium III is 10-15% faster at the same clock speed. Against an Athlon the G3 is about 10% or a bit less faster.

-A G4 clocked the same as a Pentium III is 20-25% faster at the same clock speed. Against an Athlon the G4 is about 20% or a bit less faster.


To be 15-25% faster than a competing chip is a very good achievement. The problem lies in the price. But that is only to be expected if only one manufacturer supplies G-series machines.

When it comes to SMP, I'm not very impressed with Adobe's programming. A Dual CPU box doesn't seem to offer an impressive improvement. Not like a program like Lightwave where the improvement can be 50% or more.

[slipjack]

My Dual G4/450 smears the PIII 800Mhz boxes at work all over the place in performance, especially when working on a Network. (yet another often untested area.) I'm sure the 1.7 Athon will be pretty snappy compared to the fastest Macs, but just wait till next month!!!

Originally posted by slipjack:
<STRONG>My Dual G4/450 smears the PIII 800Mhz boxes at work all over the place in performance, especially when working on a Network. (yet another often untested area.) I'm sure the 1.7 Athon will be pretty snappy compared to the fastest Macs, but just wait till next month!!! </STRONG>

This is pretty remarkable. My Dual G4-450 is slower than my PIII-800 that I upgraded to the Athlon. In a cross examination in Photoshop is was about equivalent to a PIII-700-750. In fact, my Ti-Book 550 is faster than the G4-DP!! This may be due to the revised G4 having more than one AltiVec unit.

[r-0X#Zapchud]

What KellyHogan?
Are you sure multiprocessing is actually working?
I mean, a DP 450 should beat the Ti550 easely!

Originally posted by r-0X#Zapchud:
<STRONG>What KellyHogan?
Are you sure multiprocessing is actually working?
I mean, a DP 450 should beat the Ti550 easely!</STRONG>

Apparently not. Adobe's SMP is crap and that is how it has been for years. the Ti-550 is only better than the 450DP by such a small fraction I might as well say it is the same. So that means Photoshop is getting an extra 100Mhz out of the second CPU in the Dual configuration. Not to bad by Adobe standards, certainly better than what they can squeeze out of a second x86 chip.

[r-0X#Zapchud]

I visited barefeats.com today, and found some interesting tests & results. The old 500 MHz G4 reduced it's testing time from 51 -&gt; 31 when being DP. That is about 60% increasement, which is waaaaay more than you are getting, actually the 450 should play as well as a SP 720 MHz, under the exact same circumstances.

I still doubt the Athlon 1,3 GHz is as good as a DP 800, I'd rather believe it is more the equalient of the SP 867....but of course, it depends alot of which filters and functions you test...
...since my 466 actually beat the 1,46 GHz athlon...

[KellyHogan]

Originally posted by r-0X#Zapchud:
<STRONG>
I still doubt the Athlon 1,3 GHz is as good as a DP 800, I'd rather believe it is more the equalient of the SP 867....but of course, it depends alot of which filters and functions you test...
...since my 466 actually beat the 1,46 GHz athlon...</STRONG>

I have benched something like 50 machines in the last 4 years and this statement about a 466 beating an Athlon of a 1Ghz higher spec is one of the most diatripe remarks I have ever read. Please, come back when you reach 18 and stop worshipping computer brands like a religion.

[r-0X#Zapchud]

WTF is this?

Are you a troll that simply can't accept the truth?
My G4 beat the Athlon 1,46 GHz in radial blur by a healty margin, and it is not at all anything I dreamt, or something I make up just to annoy you "all-knowing" folks. If you are trying to stop me from claiming the truth, you will fail every time.

Ok, the Athlon will probably beat my G4 in all the other benchmarks (exept RC5) and tests, but I simply couldn't care less!

[Apple Pro Underwear]

Kelly Hogan,

i think some of your claims are questionable. I have seen your benchmarks before and i'm not impressed.

PC mag who is non-biased one way or the other, benchmarks macs along with wintels machines all the time. They are consistantly surprised by mac performance despite low mhz ratings. I'm not going to elaborate on this article because i don't have it and can't post it. They are out there though.

Your claims are ridiculous.

MPs are faster than single processor machine.

Top of the line current G4 macs do not get out performed by bargain basement wintel counterparts. If they did, why does the entire graphics industry and almost all major deign houses and studios use macs? Plus, why would they pay more for a slower machine?

KH, if what you say is true, Apple would have went out of business long ago.

[Apple Pro Underwear]

Originally posted by KellyHogan:
<STRONG>
Please, come back when you reach 18 and stop worshipping computer brands like a religion.</STRONG>

Also, i appreciate a good burn when i see one, but i won't condone genuine insults meant to hurt other people. I've noticed you do this in a lot of threads.

So as a senior member of Macnn, i request you stop it. you're ruining the overall enjoyment of this forum.

[Fyre4ce]

Geez Kelly, who lit the fuse on your tampon?

[r-0X#Zapchud]

Originally posted by KellyHogan:
<STRONG>
stop worshipping computer brands like a religion.</STRONG>

btw. what is the difference of worshipping computer brands like a religion and stating remarkable results?
everything

I've only touched a Mac twice in my life. But I have done some research on them because I'm considering buying one. Here's what I can offer on this comparison:

r-0X#Zapchud, what chipset is in your friend's Athlon box? If it's a VIA chipset, that could affect the performance you're seeing. You mentioned that your friend's PC is running Windows 2000. It's a known fact that Win2K prior to SP2 did not like VIA chipsets.

Another area where the Mac can be faster than PCs is the fact that some Macs have L3 cache, whereas PC's do not. Since we're on the topic of memory, r-0X#Zapchud: what bus speed is your friend's Athlon CPU? He may be using DDR RAM, but if the front side bus is just 100 MHz, that's about the same as a Mac.

Photoshop is a very poor benchmark to use in comparing Macs and PCs. I've applied a filter onto an image, undo, reapply it, and get totally different results each time.

As far as DP performance goes, Adobe is only partially to blame for lousy DP performance. Photoshop has yet to be released for Mac OSX, and I'm guessing multi-threading has been improved in OSX, allowing the programmers at Adobe to do a better job at writing multi-threaded code.

[r-0X#Zapchud]

Yes, he were using a VIA chipset, but he had SP2 installed, so the conflict should have been gone. The processor were a 266 MHz FPB one.

The thing that surprises me here, is that in the past I have seen some (but not many) reviews/tests of PMG4's versus athlons in PS. In one (I think it was by SFGate), the SP 450 were ~60% faster than an Athlon 1000. Of course that may have been an Athlon on 100 MHz SDRAM, but so was the SP450. That it becomes impossible (for KellyHogan & some more) that a G4 466 beats an Athlon 1,46GHz in one filter seems kind of weird to me...

(this paragraph did not make that much sense, I know, but at least try to understand).

[Cipher13]

OT: Does PhotoShop support network rendering?
If so get an Athlon renderslave... that's what I plan on doing for my Bryce/video work...

This topic has been done to death... I'll refer you to this thread from five months ago (there was a similar thread here as well).

The gist being SP athlon-C's were walking all over SP G4-867's in PSBench and PS6.01 (with the recent x86 plugins updater). Now they'd have to contend with dual athlon XP 1900+'s and the just released northwood based P4 2.2GHz...

Here's some examples of the final results in seconds from that thread -- lower obviously being better:
867MHz G4 (OSX under Classic) &#0124;&#0124; 418.2
1 GHz Athlon (space monkey) &#0124;&#0124; 412.1
2x 533 MHz G4 (Ashby) &#0124;&#0124; 382.3
2x 700 MHz PIII (bash666) &#0124;&#0124; 376.7
2x 500 MHz G4 (BEIGE) &#0124;&#0124; 362.3
1.4 GHz TBird (c4zp3rgh0zt) &#0124;&#0124; 295.5
1.5 GHz TBird (c4zp3rgh0zt) &#0124;&#0124; 278.0
1.55 GHz TBird (TTSam) &#0124;&#0124; 264.4
2x 1.2 GHz TBird (anthrax) &#0124;&#0124; 245.6

If you look through the thread you can see which filters each platform is better at based on the itemized results.

I'd imagine 2x Athlon XP 1900+'s would score in the low 100 seconds as will dual northwood based systems when they become available -- especially after the transition to the 533MHz FSB in Q2.

[smacintush]

The Athlon is just a box.

Amen brtoher!

It's amazing the SOOOO many people define their computer by which processor they have.

[tallscot]

What's all this talk about single versus dual? Why not get a dual Athlon XP from Polywell?

Go to www.dv.com (DV Magazine) and read their "Duel of the Duals". The Mac is the slowest dual computer among many dual Xeons and Athlons.

Go to Chris' Lightwave Benchmarks and you can see a dual Athlon beats a dual 800 G4 by a large margin, and this is with MP and AltiVec optimization.

I hope Apple has some seriously jacked up Macs coming next week...

[godzookie2k]

Yeah as much as I hate to say it, at this point the mhz gap between ppc and x86 chips is so wide that the megaherz myth *does* apply. The mhz myth *doesn't* apply only when clock speeds are remotely close to one another. i.e. 1.2 ghz vs 867 mhz. or 450 mhz vs 600 mhz but in a case of dual 1.8's vs a dual 800, well there is a whopper of a gap there, much like comparing a 200 mhz 604e vs a G4 867.

IMHO.
Nick

[KellyHogan]

Originally posted by godzookie2k:
<STRONG>Yeah as much as I hate to say it, at this point the mhz gap between ppc and x86 chips is so wide that the megaherz myth *does* apply. The mhz myth *doesn't* apply only when clock speeds are remotely close to one another. i.e. 1.2 ghz vs 867 mhz. or 450 mhz vs 600 mhz but in a case of dual 1.8's vs a dual 800, well there is a whopper of a gap there, much like comparing a 200 mhz 604e vs a G4 867.

IMHO.
Nick</STRONG>

Exactly, man. The G4 is pretty good when compared to an x86 chip clocked 20% higher but with the gap increasing so much there simply is no megaherz myth. I think the G5 will do very well though but will cost too much.

[godzookie2k]

&lt;/hopepray&gt;