Photoshop 6.01 results at Anandtech

[AeroSnoop]

Did anyone notice the Photoshop 6.01 test at Anandtech? Photoshop with the 6.01 patch is optimized for the P4 right? It shows a dual AthlonMP 1.2ghz beating a dual Xeon 1.7ghz.

The dual AthlonMP 1.2 also beats the dual Xeon 1.7 in database transaction handling and rendering performance.

 

[Makaveli]

The 760MP setup did beat the Xeon as I was not suprised. And it has a 500mhz clock speed advantage. Can't wait to see the next clock release for the Athlon MP, also I wish they would have included a 128bit DDR Bus like the Ncrush Chipset. I believe that setup would equal or beat the Dual Rambus in the P4. To be honest I think the P4 recieves most of it's memory bandwidth edge from the 400mhz FSB.
Image a Athlon on a 200DDR bus with 128 DDR bus....oooo just the thought makes me shiver. It would destroy any P4 including The next P4 coming out

 

[Guest]

Yeah, I did.

When Anand did the test of the Dual Xeon a while ago, I was already very suprised the single Athlon 1.2 performed that well compared to a dual/single 1.7 Xeon. In spite of the SSE2 optimizations.

Also astonishing how little advantage photoshop takes from dual cpu's in general.. have a look at the table.. unless you use polar coordinates all day (who uses this anyway ?) or radial blur, a dual configuration hardly performs any better than a single cpu (goes for both xeon and athlon)

While you're looking at the table, see how well the dual P3 performs !!! Give or take a few percent, its as fast as the 2x 1.7 Xeon in most tests !!! Shame on Anand for not including it in the graph.



---- Owner of the only Dell computer with a AMD chip

 

[Kelledin]

I don't specifically remember the photoshop benchmarks--but I do remember that in everything but SysMark, the dual Athlon slapped the dual P4 around like a silly b|tch! I guess we know who's going to rule the server roost from now on...

Kelledin

"/join #hackerz. See the Web. DoS interesting people."

 

[Guest]

>While you're looking at the table, see how well the dual
>P3 performs !!! Give or take a few percent, its as fast as
>the 2x 1.7 Xeon in most tests !!! Shame on Anand for not
>including it in the graph.

They didn't include it because it failed in the polar coordinates test.

How's that for software compatibility?


<i>Cognite Tute</i>
(Think for Yourself)<P ID="edit"><FONT SIZE=-1><EM>Edited by ergeorge on 06/05/01 02:53 PM.</EM></FONT></P>

 

[killall]

interesting... a 1.2 battering a 1.7 hehe...

you do not strengthen the weak by weakening the strong

 

[rcf84]

Image a Athlon on a 200DDR bus with 128 DDR bus....oooo just the thought makes me shiver. It would destroy any P4 including The next P4 coming out

Yeah right... Northwood is going to Kick A$$. With L2 512kb cache it will crunch data like no tommorrow. Stop comparing northwood to the Current P4.

Nice Intel and AMD users get a Cookie.... Yummy

 

[AeroSnoop]

It's funny how now we have Amd beating the P4 in P4 *OPTIMIZED* apps. There goes that argument for the P4. What is your answer to this, Intel fanatics? You can't just sit quiet and hope it passes because it won't

Good to see your extra 500mhz published clock + rdram hard at work.

Btw - I don't really mean to post this directed at you bbaeyens, more for everyone in general.

 

[dhlucke]

i'm starting to think northwood is going to be very critical for intel...

if they screw up...

<font color=red>Amd or Intel? Who cares?? Not me...</font color=red>

 

[Guest]

Well good to see you give some credit to the dual P3 setup. "Hardly smp" is what you use to call it. Perhaps you could retract your statement now since it compares to the dual xeon 1.7's? You got to see a dual p3 unit in cad or max...whoooaaaa baby....One other thing ..I just tried there first test on my dual 733...I got 4.3 secs on a 53.9MB file including revert...whats happening there...freakin Anandtech. They could define there test bed a little better. Your signature...Cognite Tute (think for yourself)..have you actually tried it or was it something your boss told you? The dual P3 unit is competive in most of those tests, I'd suggest the polar co-ordinates test was mucked up by Anandtech, if not the programmers at Adobe ought to get there butts kicked. I can't see how you can discredit the p3 after watching it kick in the rest of the tests. Now that you have recognised the value of a dual P3 setup are you going to retract your statements of previous posts?

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[bhc]

This "software optimization" business is a lot of hype and spin anyway. First of all, we don't know if a lot of software can be optimized with the long pipeline or SSE2. More importantly, SSE2 can be adopted by Intel's competitors as well. The sad fact is mighty Intel could NOT get it done with raw CPU power so it has to spin this "wait till software getting optimized" line on customers. But seriously, Intel is NOT going reimburse its customers' expense to replace the old version with the new "optimized' version. So, what is the big deal?

**Spin all you want, but we the paying consumers will have the final word**

 

[Guest]

LMAO!!
tonestar, are you dyslexic, or just stupid?

I'm hardly giving credit to the dual PIII, <b>I'm pointing out its failure to complete a benchmark.</b> The only configuration that failed actually. Kind of embarrasing considering all the "AMD incompatibility" crap you hear in this forum.

And it doesn't speak to well of the dual Xeons either, particularly since they got beat by the dual TBirds for a fraction of the cost.

>The dual P3 unit is competative in most of those tests

Really? So you're math challenged also!

In the render benchmark, it got beat by 15% by a <i>single</i> 1.2 GHz TBird. Not even the Palimino or a 1.4 GHz. It got stomped 37% by the dual Pali

In the database server test, it got beat 12% by the single TBird 1.2 and 46% by the dual Pali

In the "workstation performance (Overall)" test, it beat the single TBird by a mere 1% and got smoked 27% by the dual pali.

In the Linux kernel compile, it does well vs. the single TBird, beating it by 43%. But vs. the dual Pali, it loses by 31%

Ok, I'm tired of this. I've got better things to do. You can look the rest of it up yourself. The dual PIII is barely (if at all) competative with the single TBird, and completely out of its league wrt the 760MP (which will only get faster).

Face it tonestar. Your platform is obsolete.



<i>Cognite Tute</i>
(Think for Yourself)

 

[Guest]

Take it easy on the crack ergeorge...and give your bong away.

YOU said on the last page:

>While you're looking at the table, see how well the dual
>P3 performs !!! Give or take a few percent, its as fast as
>the 2x 1.7 Xeon in most tests !!! Shame on Anand for not
>including it in the graph.

Then you go onto say:
"I'm hardly giving credit to the dual PIII"

Sounded like you were giving it lots of credit, even said what a shame it was that Anandtech didn't graph it.
So if dual 933's perform compareably to dual 1.7s xeon's you don't think they deserve some credit?

Which one is it you drug f@*ked moron?

This is the "Photoshop 6.1 Anandtech benchmark" post you donkey.

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[Guest]

You complete idiot. Is that what this is all about?

I'm quoting bbyaens you moron. Read the thread.

Look at the next line where I explain why Anandtech didn't include the PIII result in their graph. <b><i>Because the PIII failed to complete the benchmark.</i></b>

Is there anybody else that this wasn't obvious to?


<i>Cognite Tute</i>
(Think for Yourself)

 

[Guest]

They are in fact good figures. You have pointed out the polar coordinates test. WHAT I AM SAYING TO YOU SH#$THEAD is don't negate its performance over some crappy filter no one probably uses, if in fact this is the case or simply Anandtech's mistake. BByaens is quite right but being the AMD zealot you are you can't bare to give the platform any credit. If the polar coordinates test blinds anybody to the merits of this budget platform (compared to dual 1.7 Xeons) they must be a total whopper sucker. SO MY QUESTION TO YOU WHOPPER SUCKER is ARE YOU STILL SUGGESTING A DUAL P3 SETUP IS WORTHLESS?

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[Guest]

Cool down man..we're talking computers, we're not insulting your girlfriend or your mother.

As for giving the dual P3 credit.. well, depends how you look at it. I was amazed the dual P3 performed very close to the dual Xeon in PHOTOSHOP, and PHOTOSHOP only. If you want to interprete this as a thumbs up for the P3, go ahead. I'd rather see it as a blamage for the dual Xeon... or as a strange software behaviour.

For the rest of the test, Im with egeorge.. the numbers speak for themselves. Of course a dual P3 gets spanked in most of the other tests.. but if you use photoshop, you dont give a [-peep-] about linux compile times and database performance. For photoshop, you wouldnt want to ditch a dual P3 for a dual palomino or Xeon if you already have one. If you need a new setup, depending on your apps, a dual Palomino seems to make much sense.

I only wished Anand would comment on the failure of polar coordinates. Did he fail to bench it, or was the setup unable to actually finish the test ? And for what reason ? I mean.. kinda weird. The fact that "no one" uses the filter is no excuse.


---- Owner of the only Dell computer with a AMD chip

 

[Guest]

Ye true, some of those figures look a bit shonky, I tried test 1 on a file great than 50mb and on my dual 733's I got a quicker time. My annoyance is that Adobe should make that test work on a p3 (if in fact it was tested correctly) but to dismiss the rest of the figures because of that one useless plug (yes useless until someone can point out why they need to use it frequently) is .... stupid. Ergeorge went onto to say somewhere "I don't care what the figures are cos it didn't complete the benchmark" is another insight into his biased and illogical mind.

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[Guest]

Funny.. when an intel setup doesnt complete a test, intel users say "Adobe should make that test work on a p3 ".. When (if ?) an AMD setup doesnt complete it, they say "AMD is incompatible"..

Anyway. the facts remain. When looking at the Anands table I wonder what other bottleneck there is that ensures all systems perform that close in most of the tests. Disk speed ? Either way, hardcore photoshop users should think twice about spending a fortune on a dual setup..

---- Owner of the only Dell computer with a AMD chip

 

[Guest]

"Funny.. when an intel setup doesnt complete a test, intel users say "Adobe should make that test work on a p3 ".. When (if ?) an AMD setup doesnt complete it, they say "AMD is incompatible".."

The Pentium range collectively, like it or not, represents the bulk of the market, so its only sensible to target your largest demographic. Kinda like Flash and Shockwave plugs represent the majority of user side viewers in that market.


"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[Guest]

Tonedeaf, you are so full of yourself it is funny. The only thing the dual PIII kicks (as in, barely beat the competition) in was Photoshop 6.0. As you are a self- proclaimed high end 3D modeler and animator, I am sure that you spend the majority of you "artistic" talents in a 2D paint program environment for post production work instead of using After Effects, right? Oh I forgot, you are a self proclaimed mechanical engineer that just happens to be an expert in 3D modeling and animation. Yes sir, you certainly have boat loads of credibility in this forum.

By the way, how does your competitive dual PIII rig compare to either the dual AMD or PIV Xeon setup in MAX 4.0 (http://www.anandtech.com/showdoc.html?i=1483&p=14)? Cinema 4D (http://www4.tomshardware.com/mainboard/01q2/010605/760mp-08.html - sorry, no PIII comparison, but I've posted before that a dual 1G PIII gets about 20 Cinebench marks), or MAYA? So how does a dual PIII compare? It doesn't... really. Why don't you go out and buy a clue sometime...

 

[Guest]

P4Fool...how appropriate you should call yourself this. You sure a fool...no debate there.

1. I never ever said I was an animtor.
2. Yes I am a mechanical engineer, a working one too. Not some wanna be school drop out.
3. What are you trying to say about After Effects? Do you even know what after effects does? Cos if your trying to compare it to any 2D paint/imaging program you ought to change your name to P4Mule....whaaaa?
4. Before you compare the dual setups maybe you ought to get a grip on the overall cost inc. mainboard.
5. If we followed your logical (hehehe...what logic?) we'd all be comparing an old 286 to a 40 cpu render box.
6. When you get that cock out of your mouth post back.

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[Guest]

>I tried test 1 on a file great than 50mb and on my dual
>733's I got a quicker time.

But you see tonestar, you have absolutely no credibility here. I'd as soon believe the earth is flat as the statement I've quoted here.

>Ergeorge went onto to say somewhere "I don't care what the
>figures are cos it didn't complete the benchmark"

In your world, failure may be an option. That's kinda scary for someone who calls himself a mechanical engineer. Of course, maybe you just maintain the machines that make rubber dog crap.

It's not an option in my world. If it doesn't complete the benchmark, it fails. End of story. That goes for any platform.

Now frankly, I don't use photoshop. And we don't know anything about this failure. So I'm not to concerned about it.

But I won't consider the numbers on the parts of the benchmark it did complete.


<i>Cognite Tute</i>
(Think for Yourself)

 

[Guest]

Ok, ergeorge cos your such a dumb b%astard I will go through this with you step by step. I, unlike you, wasn't stupid enough to believe that Anandtech couldn't get one filter to run so rather than bend over and spread my cheeks like you have done I thought I'd test it. Anyone with a similar comparison can test this so the decision will be left to the forum:

My setupual P3 733Mhz, SCSI 10k, MSI 694D mainboard (Apollo 133a..694x)

what I did: get an image..RGB mode...IN PSP 6 go to filters/Distort/Polar Coordinates...select either....rectangular to polar..or polar to rectangular...hey presto!!! it works just fine!!!

And any dual P3 owner can test this!! go ahead and let us know!!

Cos you are sooo stupid you neglected all the figures on that test and dismissed the P3 setup, not because you are objective but because you a freakin AMD maggot. As for

"In your world, failure may be an option"...no idiot..thats why I test things....failure is an easy option for you, your life is an example of one.

You know what...even if the P3 couldn't do this test it wouldn't mater too much...apply the filter and see what you get....hmmm how often could you use that? The filter is 1/20 of the benchmark you fool. So:

"It's not an option in my world. If it doesn't complete the benchmark, it fails. End of story. That goes for any platform."

1. It didn't fail the benchmark. Anandtech have made a mistake.
2. It's too small an issue to judge the platform.
3. Anyone else with a dual setup can prove it.
4. Your one of the dumbest people in this forum.

Then you go onto to say:

"But I won't consider the numbers on the parts of the benchmark it did complete"

You freakin two bit whore.


"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!

 

[Guest]

>what I did: get an image..RGB mode...IN PSP 6 go to
>filters/Distort/Polar Coordinates...select
>either....rectangular to polar..or polar to
>rectangular...hey presto!!! it works just fine!!!

But you see tonestar, it's very simple.

<b><i>I don't believe you.</i></b>

You have a history of lying in this forum, so you saying it works means nothing to me.

And no, I didn't judge the entire platform by that benchmark. I looked at how it got buried by even the single CPU AMD solution in the other benchmarks.

How's that for price/performance ratio? Can you build a dual PIII for the price of a single CPU 1.2GHz TBird? How about for 15% less then the price of a 1.2GHz TBird? That's how much it got beat by in the render benchmark.

>2. It's too small an issue to judge the platform.

Your right, and I said as much. But I won't consider the rest of the photoshop results. I will consider the platform on the basis of the rest of the results. Anandtech seems to agree with me considering that they left the PIII of the graph.



<i>Cognite Tute</i>
(Think for Yourself)<P ID="edit"><FONT SIZE=-1><EM>Edited by ergeorge on 06/06/01 11:35 AM.</EM></FONT></P>

 

[Guest]

"I don't believe you"..ok ergeorge...lets leave it up to the forum to try it for themselves shall we? Prepare an appology for your next post.

The P3 setup is virtually on par with both the dual Athlon platforms, winning in some and falling marginally short in others. But did you see the Color halftone test? What a difference!!!

"Can you build a dual PIII for the price of a single CPU 1.2GHz TBird? " Perhaps not...but I can build one alot cheaper than a dual AthlonMP setup...and get similar performance in Photoshop. The 933 are on par with the Athlon 1.2's...strange isn't it?....hehehe...

Anyhow..whatever those numbers maybe...my point is that your a moron who automatically believes everything he reads. One of the first things you learn as an engineer is to test things...never assume. I tested. They have made a mistake. So have you.

"Cock-a-doodle-do" is what I say to my girl when I wake her UP in the morning!!