AMD X2 Photoshop Test

[Griffinhart]

I saw this thread at |H|ard OCP. Apparently it started out on an Apple forum.

http://www.hardforum.com/showthread.php?t=926660&page=1&pp=20

http://forums.macrumors.com/showthread.php?t=136593

Using Photoshop CS or CS2 (or PS 7)
==================================================

1.) Download the test image from http://www.quicklance.com/test.jpg

2.) Save it to the computer and then open it up in Photoshop

3.) From there please apply a 'radial blur' with the settings at:

Amount = 100
Blur Method = Spin
Quality = Best

Using a stop watch / ps timer see how long it takes to apply this filter

I just want to see what these new cpu's can really do.

Please list your cpu details, amount of ram and the time it took for the test, just so we can get an idea of how the different cpu's handle this task.

================================================

After seeing the results on both forums, I think it goes a long way to dispelling the idea that PC's aren't as good at this sort of job. Plus I found it cool. Especially since it's a test that can be fairly run on both platforms.

My X2 4400+ (stock Speed, 2GB of cheap crucial RAM (512MBx4) 7200RPM WD HD's with 8MB cache (Raid 1)) completed the job in 36 seconds. From clicking OK to finish.

edit: You can download a trial version of Photoshop CS2 from Adobe.

 

[Refractoryman]

My system:
ASUS A7V266-E
Athlon XP2400+
512mb ram

I ran this test in 2 minutes 40 seconds / (280 seconds). Is this good for my configuration. Thinking about up-grading to the 3800 x2. Comments?

 

[Griffinhart]

From what I have seen, that's pretty good for your system. Especially being a non AMD64 and single core.

As for the 3800 X2, I've been hearing nothing but good things about it.

 

[Markfw]

Is it IO dependent ? I will try it on my raid array if so.....

 

[Griffinhart]

with 2GB of RAM, it may not be. But, it would be interesting to see your dual opteron vs your X2 score.

 

[jpeyton]

I got 28 seconds, 4400+ @ 2.7GHz in CS2.

Make sure to use CS2. CS is a little slower, and PS7 is much slower.

 

[asicman]

You have to perform this test with CS2 for best performance. My P4 3.4 machine with 2 GB and RAID 0 finishes in 28 seconds with CS2. Your X2 should beat my P4. Try CS2.

 

[jpeyton]

I did use CS2. I'll try again. But keep in mind I had 1GB RAM and a 40GB Seagate PATA.

 

[Griffinhart]

less than 30 seconds on a single core p4? Are you sure you put all the right settings? Single cores should be getting near the 60 second mark or longer.

 

[asicman]

28 seconds clocked at 3.9 GHz

 

[jpeyton]

Originally posted by: asicman
28 seconds clocked at 3.9 GHz

Interesting. I should try it on my 660 @ 4.0 and my 3.46EE @ 3.7. See what the speed and the L3 cache does for this test.

P4s have always been stronger in PS, but I have a hard time believing that HT would speed them up so much.

 

[asicman]

Originally posted by: jpeyton

Originally posted by: asicman
28 seconds clocked at 3.9 GHz

Interesting. I should try it on my 660 @ 4.0 and my 3.46EE @ 3.7. See what the speed and the L3 cache does for this test.

P4s have always been stronger in PS, but I have a hard time believing that HT would speed them up so much.

I'm putting together an Athlon 64 3700+ San Diego machine right now. I've recently read benchmarks that indicate the A64's are besting the P4s in CS2. PSU and heatsinks due to arrive today, and then the build will begin.

 

[Bona Fide]

Hah...that test would take minutes on my 1.6 Williamette...I'll wait till my new setup ships.

 

[Markfw]

Can somebody get me a link where to download the CS2 ? I want to make sure I am using the EXACT program we are all talking about, as I am a NOOB in PS.

 

[Duvie]

It took me 31-32sec at my current 2.563mhz X2...the best I could do with my watch....no stop watch here...that was with 2 instances of FH running....

Also asic you are doing something wrong....Turned off FH to see task manager and this takes 100% cpu usage of 2 REAL cores versus your HT core which would be 20% of a real core.....I ran a test like this back with my A64 at 2.664ghz and I was beating P4's in the 3.8 to 3.9ghz range...It would have not beaten this X2 so I think your time is bogus...not calling you a liar, just wrong...

 

[toyota]

Originally posted by: Refractoryman
My system:
ASUS A7V266-E
Athlon XP2400+
512mb ram

I ran this test in 2 minutes 40 seconds / (280 seconds). Is this good for my configuration. Thinking about up-grading to the 3800 x2. Comments?

isnt that 160 seconds not 280 or am i missing something??

 

[Duvie]

Originally posted by: toyota

Originally posted by: Refractoryman
My system:
ASUS A7V266-E
Athlon XP2400+
512mb ram

I ran this test in 2 minutes 40 seconds / (280 seconds). Is this good for my configuration. Thinking about up-grading to the 3800 x2. Comments?

isnt that 160 seconds not 280 or am i missing something??

you are right....4min 40sec would be 280sec...so i wonder if he made an error at the min and secs or the overall seconds...

 

[asicman]

Originally posted by: Duvie
It took me 31-32sec at my current 2.563mhz X2...the best I could do with my watch....no stop watch here...that was with 2 instances of FH running....

Also asic you are doing something wrong....Turned off FH to see task manager and this takes 100% cpu usage of 2 REAL cores versus your HT core which would be 20% of a real core.....I ran a test like this back with my A64 at 2.664ghz and I was beating P4's in the 3.8 to 3.9ghz range...It would have not beaten this X2 so I think your time is bogus...not calling you a liar, just wrong...

Not sure what to tell you. I use photoshop all the time and know what I'm doing. It's running in 28 seconds with my P4 at 3.9 with dual DiamondMax SATA 16mb drives in RAID0 and 2GB DDR667 at 2-2-2-4.

I plan to run my A64 3700+ at 2.8 GHz and see how it compares. I expect it to do even better.

 

[Duvie]

Originally posted by: asicman

Originally posted by: Duvie
It took me 31-32sec at my current 2.563mhz X2...the best I could do with my watch....no stop watch here...that was with 2 instances of FH running....

Also asic you are doing something wrong....Turned off FH to see task manager and this takes 100% cpu usage of 2 REAL cores versus your HT core which would be 20% of a real core.....I ran a test like this back with my A64 at 2.664ghz and I was beating P4's in the 3.8 to 3.9ghz range...It would have not beaten this X2 so I think your time is bogus...not calling you a liar, just wrong...

Not sure what to tell you. I use photoshop all the time and know what I'm doing. It's running in 28 seconds with my P4 at 3.9 with dual DiamondMax SATA 16mb drives in RAID0 and 2GB DDR667 at 2-2-2-4.

I plan to run my A64 3700+ at 2.8 GHz and see how it compares. I expect it to do even better.

Unless it is IO limited which scaling of scores above dont suggest your score is illogical....Dont know what to tell you there....I dont think ram has much of anything to do with it either, as PS has always liked cpu raw power first and foremost..

I also dont see you beating a dual 2.5ghz mac by 21 seconds or nearly 50%....

The fact the program used 100% or both cores tells me the single core A64 should not come close to beating it....

 

[Duvie]

Is there a way you are caching this thing and the reaso it runs faster??? I get same time running it over and over again....

Also do you have a full blown copy and/or not the same version we are running???

Logic would tell me if this app ran say 50-60% cpu usage i could buy a 3.9ghz P4 could hange with a 2.7ghz X2, but not when it uses 100% for effective 2 cores....

 

[Griffinhart]

I saw one post that a p4 clocked at 3.3 got 65.2 seconds. I just can't see how one clocked at 3.9 is twice as fast. The only thing I can think of is the test wasn't run at 100 bluring and best quality for the radial blur test. a better than 100% increase in performance with an 18% clock speed increase just doesn't make any sense at all for a single core processor.

and for those that asked:

http://www.adobe.com/products/tryadobe/main.jsp#product=39

That's the link where you can get the trial version of CS2

 

[asicman]

OK my bad. I was using the wrong image. There is a different site with a different image doing the same radial blur test, and I didn't notice the images were different.

Sorry 'bout that.

 

[ProviaFan]

I had a very strange anomality. The first time I tried with the test image, the filter ran very slowly, almost as if PS was only using one core. I started Photoshop again, the second time loading the image through the File/Open command instead of launching PS by the Open With entry in the context menu on the image file itself, and the filter completed in 37 seconds (the second time, I watched CPU utilization in the Task Manager, and PS pegged at 99% the whole time).

My relevant specs are as follows:

Adobe Photoshop CS2 on Windows XP x64 edition
AMD Athlon 64 X2 4400+ at stock speed
2GB PC3200 DDR (Transcend) at normal timings (not sure what they are specifically)
ASUS A8N-E
WD Raptor 36GB OS and Programs
Seagate 300GB 7200RPM disk (holds scratch file)

Also note that not once did I see the hard disk indicator light flash during this process; with 2GB of RAM, assuming you don't have loads of crap in the background and your PS memory settings are good, the hard disk should not be a bottleneck in this test.

 

[Scitex]

Just for fun I did this with my AXP 1700+, 1Gb and got 271.2 sec. Also, for the ones using a stopwatch....click on the little triangle at the bottom of the picture in PS and select "Timing" and it will time it for you

 

[Brunnis]

2m 3s

AthlonXP 2.2GHz (11*200)
1GB @ 200MHz 11-3-3-2

2m 18s

Pentium-M 1.7GHz (17*100)
512MB @ 166MHz 2-2-2-5