23 Seconds
Mac Pro 3.1 (Octophallus) 2.8GHz
18GB Ram (6 Dimms)
Apple RAID5 - empty scratch partition
CS3
OSX 10.5.5
Mac Pro 3.1 (Octophallus) 2.8GHz
18GB Ram (6 Dimms)
Apple RAID5 - empty scratch partition
CS3
OSX 10.5.5
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
'Real World' Photoshop Benchmark - Speed Test
- Thread starter BakedBeans
- Start date
- Sort by reaction score
21 Seconds. Not bad for a Hack. I'll bet I could get it into the teens if I removed 4GBs of ram. My system has performance issues with 8GBs of ram running faster than 800MHz. When I only have 4GBs in there I can get the ram over 1000MHz which gives me better performance in things like Xbench and Geekbench...
With that said, I'll give it a shot...brb
With that said, I'll give it a shot...brb
241 seconds
That was on my PowerMac G4:
Dual 1.8GHz G4 processor (from giga designs)
1.5GB RAM
ATi 9800Pro
single 120GB IDE hard drive
OS X Tiger 10.4.11
CS3
Slow and old, but I don't do a lot of PS work. Much of the time I'm converting videos for DVDs. Yes that, too, is slow but I can set to work and walk away.
That was on my PowerMac G4:
Dual 1.8GHz G4 processor (from giga designs)
1.5GB RAM
ATi 9800Pro
single 120GB IDE hard drive
OS X Tiger 10.4.11
CS3
Slow and old, but I don't do a lot of PS work. Much of the time I'm converting videos for DVDs. Yes that, too, is slow but I can set to work and walk away.
Do you use Handbrake? If you do, how many FPS does your G4 gets when converting a DVD to iPod Hi-Rez?
18.5 seconds on Hackintosh Leopard 10.5.5; Photoshop CS3.
Time taken using Apimac's Timer for Mac OS X. I clicked the timer to start first, then I started the benchmark after. The real time is lower than 18.5 seconds, but not sure how low.
Benchmark done with Intel Quadcore Q9550 - 4Ghz.
There is a better Photoshop benchmark which do not rely on human reaction time, but uses the timer built inside Photoshop. Driverheaven's Photoshop benchmark is better to use.
http://www.driverheaven.net/photoshop.php
I'm going to run these benchmarks in Windows to see which is faster.
Time taken using Apimac's Timer for Mac OS X. I clicked the timer to start first, then I started the benchmark after. The real time is lower than 18.5 seconds, but not sure how low.
Benchmark done with Intel Quadcore Q9550 - 4Ghz.
There is a better Photoshop benchmark which do not rely on human reaction time, but uses the timer built inside Photoshop. Driverheaven's Photoshop benchmark is better to use.
http://www.driverheaven.net/photoshop.php
I'm going to run these benchmarks in Windows to see which is faster.
I took 4GBs out and upped the Ram speed to 1000MHz but my benchmark results didn't change at all in any of the benchmark programs I use (Xbench, Geekbench, Photoshop). In 10.5.1 I got a pretty nice 10% bump or so when upping the Ram speed. I wonder if it's because when I moved to 10.5.5 I installed the 800MHz specific BIOS and my 200MHz overclock isn't even registered in Leopard.
I actually go the other way with videos. Many times I'm using Toast to burn tv shows recorded from my TIVO. A while ago, I did try creating some videos from DVDs using handbrake and it wasn't able to encode at realtime.
OSX 10.5.5 = 18.5 seconds (Hackintosh)
Vista 64 = 16.5 seconds
Tested with Quadcore, Q9550 - 4Ghz.
Full system specs:
http://browse.geekbench.ca/geekbench2/view/79985
Vista 64 = 16.5 seconds
Tested with Quadcore, Q9550 - 4Ghz.
Full system specs:
http://browse.geekbench.ca/geekbench2/view/79985
It's painfully obvious that CS3 isn't all that impressed by the number of cores a person has in their machine. If CS3 can take advantage of more than 4 cores this benchmark isn't designed to show that. Higher clocks are winning out every time.
...Mac Pro is good for multi threaded applications since it has 8 cores and basically, it is the cheapest workstation in the market, even against you building your own workstation!
I don't think that's true. Mac Pros may be as cheap or cheaper than comparable workstations from manufacturers like Dell or HP but I've been pricing out a personally built 8 core workstation comparable to a Mac Pro offering and in every case I come out significantly cheaper even factoring in the cost of a retail copy of Leopard (for a Hack pro) and a retail copy of Vista. Building certainly isn't for everyone since not everyone has the know-how to support their own machines or just don't want to spend the time learning or doing it. However, for those who have the know-how and are willing to spend the time it's quite a substantial savings.
Any CS3 vs. CS4 tests?
I'm anxiously waiting to see if there's any speed up on the mac pro. I'd also like to see how fast CS4 Leopard compares with Vista64. Somebody please do all the tests, especially the one on driverheaven:
http://www.driverheaven.net/photoshop.php
I'm anxiously waiting to see if there's any speed up on the mac pro. I'd also like to see how fast CS4 Leopard compares with Vista64. Somebody please do all the tests, especially the one on driverheaven:
http://www.driverheaven.net/photoshop.php
I did some OS X vs Vista 64 Photoshop tests here. In short, CS3 in OS X is 32% slower than Vista 64, using the driverheaven test.
https://forums.macrumors.com/posts/6458916/
I expect CS4 on Vista to be even faster than OS X, possibly even greater than 32% in the Driverheaven test. I'll find out when I can get the trials for them.
Exman,
I did notice your revealing tests. I assume the CS3 tests were done as fairly as possible--the same number of history states, ram allocation, scratch disks, etc.?
I think this all boils down to Adobe spending more time and resources on optimizing PS for Windows. Also all video drivers are better optimized on Windows. I don't think we will see the Mac getting even or surpassing until Snow Leopard, Nehelem processors and CS5 which will be in Cocoa and 64-bit. Still, it must say something about the speed of the OS if handbrake is slower on OSX as well. What do you think accounts for the speed difference? I suspect that intel chips are always optimized for windows code or intel shares a lot more proprietary info with Microsoft.
I am interested to see if any of the Open GL stuff can improve performance in PSCS4 on the mac, though I suspect it mostly improves display zooming. But it might also benefit certain photoshop operations.
I still await CS4 speed tests. My guess is that CS4 is at best 10% faster than CS3 on the mac with the same hardware. Adobe has not advertised any kind of improved performance, so I'm guessing there ain't much or it's a wash, the new features aside.
Paul
I did notice your revealing tests. I assume the CS3 tests were done as fairly as possible--the same number of history states, ram allocation, scratch disks, etc.?
I think this all boils down to Adobe spending more time and resources on optimizing PS for Windows. Also all video drivers are better optimized on Windows. I don't think we will see the Mac getting even or surpassing until Snow Leopard, Nehelem processors and CS5 which will be in Cocoa and 64-bit. Still, it must say something about the speed of the OS if handbrake is slower on OSX as well. What do you think accounts for the speed difference? I suspect that intel chips are always optimized for windows code or intel shares a lot more proprietary info with Microsoft.
I am interested to see if any of the Open GL stuff can improve performance in PSCS4 on the mac, though I suspect it mostly improves display zooming. But it might also benefit certain photoshop operations.
I still await CS4 speed tests. My guess is that CS4 is at best 10% faster than CS3 on the mac with the same hardware. Adobe has not advertised any kind of improved performance, so I'm guessing there ain't much or it's a wash, the new features aside.
Paul
Hi Paul,
I used the same Photoshop settings in both tests: History state:1, Cache Levels:4, RAM Usage: 2.8GB. I've got my 640GB SATA HDD divided into two partitions, one running Vista 64 and the other OS X, they are off the same drive.
There is nothing which leads me to believe that Intel is sharing something with Microsoft about their CPU's that Apple don't know about; Prior to the mid 2000's, Intel may have been in bed with MS (hence, the nickname Wintel), but an Intel CPU engineer have recently said everyone is now competing against one another for a piece of the pie, so those days of Wintel are well and truly gone. What they also don't want to do is upset Apple (who probably have a few secret AMD Macs in their R & D labs).
I am not a veteran of the Mac platform, so I don't have a good understanding of it's technology as other people around here, but I suspect OS X isn't all that well optimized for the Intel platform (yet), afterall, the current version was built to support two architectures (PPC & x86Intel).
This maybe the reason why OS X lags behind Windows (both XP and Vista 32bit/64bit) in a few of these benchmarks.
It isn't just in Photoshop or Handbrake, my other tests show it is slower than Vista (and XP even faster than Vista) in Video rendering too (Adobe Premier and Avidemux). A recent test on arstechnica also shows Flash 9 and 10 is many times faster (or more efficient) in Windows here. Unlike my tests which was run on a Hackintosh, this review was based on a real Mac Pro with Windows, using Boot Camp. Compare the CPU utilisation (lower = better) and FPS (higher = better).
http://arstechnica.com/news.ars/post/20081017-benchmarking-flash-player-10.html
Despite Apple's marketing poking fun at PC's, Windows is a very fast platform and personally, I've rarely encountered errors or crashes in all my years as a user (though, my love and passion is really in Linux, it is a pity I cannot run Photoshop on Linux).
I think the next version of Leopard will open up more of OS X's underlying power, and it is really exciting stuff.
Good reply, Exman
Exman,
Thanks for your lengthy reply. I'm sure it might stimulate some discussion. I don't really understand the ins and outs of OS design, so I can't comment on that stuff. But like you said, it may be that since apple only recently is using Intel, that they are not optimized for it. Maybe with an all cocoa 64-bit Snow Leopard, they will. But my cynical hunch is that the Intel-MS relationship is more important to Intel because of volume and that MS gets some extra bonuses, ditto for Adobe. There was a time when PPC (Motorola and IBM) was a great platform, that the Mac OS actually ran some things faster than Windows, but those days are over.
Anyway, I don't mind paying a premium for apple products because I prefer the aesthetics of the hardware and OS. I just don't want to lose out that much on performance as well.
Paul
PS--Somebody out there put PS CS4 through the tests already, please!
Exman,
Thanks for your lengthy reply. I'm sure it might stimulate some discussion. I don't really understand the ins and outs of OS design, so I can't comment on that stuff. But like you said, it may be that since apple only recently is using Intel, that they are not optimized for it. Maybe with an all cocoa 64-bit Snow Leopard, they will. But my cynical hunch is that the Intel-MS relationship is more important to Intel because of volume and that MS gets some extra bonuses, ditto for Adobe. There was a time when PPC (Motorola and IBM) was a great platform, that the Mac OS actually ran some things faster than Windows, but those days are over.
Anyway, I don't mind paying a premium for apple products because I prefer the aesthetics of the hardware and OS. I just don't want to lose out that much on performance as well.
Paul
PS--Somebody out there put PS CS4 through the tests already, please!
I guess my post got deleted. Fine. Here is my "guess" about CS4 performance in OSX on my machine. My CS3 performance is provided for comparison...to my CS4 guess.
CS3 - 21s
CS4 - 24s
I can't wait till I get a CS4 trial so I can actually confirm my CS4 "guess"...
P.S. I just like italicizing and quoting things completely at random. My italicization and quotations don't entail anything beyond my love for randomly italicizing and quoting things...Oh and my eye rolling emoticon was selected completely at random.
CS3 - 21s
CS4 - 24s
I can't wait till I get a CS4 trial so I can actually confirm my CS4 "guess"...
P.S. I just like italicizing and quoting things completely at random. My italicization and quotations don't entail anything beyond my love for randomly italicizing and quoting things...Oh and my eye rolling emoticon was selected completely at random.
Paul,
While these benchmarks are interesting in an educational kind of way, many of us base our decisions on more factors than just raw speed from A to B. If I drove a BMW 7 series (I wish) and some Japanese-hot-up-boy-racer pulls up to me and say "I can smoke you" I would reply "Good for you, but it ain't no Bimmer..." .
sneezymarble,
.S. I just like italicizing and quoting things completely at random. My italicization and quotations don't entail anything beyond my love for randomly italicizing and quoting things...Oh and my eye rolling emoticon was selected completely at random.
Run that by me again
While these benchmarks are interesting in an educational kind of way, many of us base our decisions on more factors than just raw speed from A to B. If I drove a BMW 7 series (I wish
) and some Japanese-hot-up-boy-racer pulls up to me and say "I can smoke you" I would reply "Good for you, but it ain't no Bimmer..." .sneezymarble,
.S. I just like italicizing and quoting things completely at random. My italicization and quotations don't entail anything beyond my love for randomly italicizing and quoting things...Oh and my eye rolling emoticon was selected completely at random.
Run that by me again
sneezymarble,
.S. I just like italicizing and quoting things completely at random. My italicization and quotations don't entail anything beyond my love for randomly italicizing and quoting things...Oh and my eye rolling emoticon was selected completely at random.
Run that by me again
Yesterday I posted performance numbers on an unreleased photo editing software trial and my post was deleted. So I made another posted about what I thought my performance would be like.
some results on the adobe forums
I posted the same question on the Adobe PS forums and got a few answers. Looks like the speed up on MacPros is around 10-15% in CS4 vs. CS3. Not fantastic, but admirable.
http://www.adobeforums.com/webx?128@@.59b6d299
I posted the same question on the Adobe PS forums and got a few answers. Looks like the speed up on MacPros is around 10-15% in CS4 vs. CS3. Not fantastic, but admirable.
http://www.adobeforums.com/webx?128@@.59b6d299
Hmm. I wonder I'm not getting the same performance increase. Maybe it has to do with the fact that I'm using a modified kernel...
EDIT: I went back and ran some more tests. The problem was that I had lowered my system voltage a few days ago while running some unrelated tests. The Photoshop test was putting my cores under a near full load and I think the system was choking a bit because the voltage was too low. I upped the voltage to where I usually keep it and now I'm getting 19s in CS4 as opposed to 21s in CS3. That's about a 10% improvement.
I think the performance gains look better on the Apple forum because most of the benchmarks are on slower systems and so the spread is higher.