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.
2009 Mac Pro in hand - unboxing etc.
- Thread starter WonderSausage
- Start date
- Sort by reaction score
Have you seen where the Photoshop benchmarks came from ?
Lol, I guess he finally received his machine.
Updated Nehalem/Penryn Benchmarks
The 2.26 was unfairly underrated early on, but it still won't outperform a 3.2 Penryn. It's more or less on par with the 2.8 Penryn.
But, like countless others have mentioned, if your tasks are more toward single-threaded operations, it might be worth it to see about a 2.8 or 3.2 over the new 2.26.
The 2.26 was unfairly underrated early on, but it still won't outperform a 3.2 Penryn. It's more or less on par with the 2.8 Penryn.
But, like countless others have mentioned, if your tasks are more toward single-threaded operations, it might be worth it to see about a 2.8 or 3.2 over the new 2.26.
Have you seen where the Photoshop benchmarks came from ?
??? Me? No. I dunno anything about the PS benchmarks.
To me the two PS benchmark downloads are probably the worst way to BM any machine. For one PS on mac is in a bit of a development crisis. Adobe coded the thing for 64-bit Carbon and just before their release date Apple pulled the plug on 64-bit Carbon and went with 64-Bit Cocoa. Instead of spending the one or 2 years it would take to recode everything for 64-Cocoa Adobe just released a 32-bit app.
On top of all that Photoshop scales probably the worst of any app of it's kind across multiple cores. There's like maybe 5 filters and one or two "tools" in photoshop that are coded to take advantage of multiple cores. Welp, that sucks! It's certainly not going to be a very good indication of a multi-core machine's abilities!Anyway, all this to say that I haven't been looking at any of the PS benchmarking posts.
Geekbench is another one I do not trust AT ALL. People are using it so, whatever. On my machine I get WILDLY different results from week to week on Geekbench. My scores range from 8000 to 8600 without any changes to the hardware or running environment. Just different temperatures and different times. <shrug> So I don't put much faith in Geekbench either.
On Mac right now all we have that I understand well and can trust to any reasonable degree is Cinebench10 so I've spent all my time on that.
I'm very surprised that Mac doesn't have more good benchmarking utilities! Specint or something like it for mac would be wonderful!
Oh-well, wish in one hand...
Grimace,
Something is wrong with your graph.
It says 25644 but the bar extends to the 29500 mark,
It says 23945 but the bar extends past the 24000 mark and almost touches the 27000 mark.
Etc.
What's up with that?
EDIT: Ah.. it looks like it's adding the two scores together instead of overlaying them.
Grimace,
Something is wrong with your graph.
It says 25644 but the bar extends to the 29500 mark,
It says 23945 but the bar extends past the 24000 mark and almost touches the 27000 mark.
Etc.
What's up with that?
EDIT: Ah.. it looks like it's adding the two scores together instead of overlaying them.
In fact all the lines are exagerated... it appears the single and multi thread scores are being added together to determine the length of the bar.
Yup. The numbers on the bars are correct but the graph's purpose is a little lost.
or just get rid of/ disregard the x-axis scale....all the bars are correct in proportion to one another.
Maybe they should combine this grimace's graph with the other guys yellow graph together??
Sorry!
Epic X-Axis fail on my part! The numbers are fine, the Axis was just showing the sum of the two scores -- that axis is removed now if you refresh!!
Epic X-Axis fail on my part! The numbers are fine, the Axis was just showing the sum of the two scores -- that axis is removed now if you refresh!!
Epic X-Axis fail on my part! The numbers are fine, the Axis was just showing the sum of the two scores -- that axis is removed now if you refresh!!
Nice Grimace.. now where's the hamburglar!!
Everytime I see your avatar, I want some southern style chicken sandwich!
They are all correct in proportion to like-colors; the only technical flaw would come from the proportion of the total length, which is (under the hood) based on the sum of both. (Ex. if the multi-thread for two cpus were both 15,000 but the single thread values were vastly different, the overall bars would be different lengths, even though the multi-thread values were the same. The standard deviation was consistent enough that it didn't make any noticeable difference.)
I had them side by side and then one bar below the other, and the total proportional difference was really negligible so I decided to keep one bar per computer, just for clarity sake. The numbers are the focus, and whether the blue bar extends a millimeter further than it should doesn't overshadow the actual numbers. We show different things in our graphs, and I like both for different reasons.
I'm too old for games (maybe XPlane?) but I was around our two college web interns this afternoon who were looking at a LolzCats site with Epic this and that. I smiled at their screens and told them to get back to work.
Received a 2.26GHz 8-core Mac Pro loaner for testing. It comes with the stock 640G drive which turns out to be a WD6400AAKS Caviar Blue.
TBH, I can't wait to see the real-life bench mark tests of the '2.26GHz 8-core 2009' VS. the '2.8GHz and the 3.2GHz 8-core 2008' once 'Snow Leopard' is installed.
Should be interesting...
Should be interesting...
I wouldn't expect anything particularly ground breaking. I think people seem to think that somehow Snow Leopard will make single threaded applications magically run multithreaded - that's not going to happen. The big changes as far as I can tell in regards to Snow Leopard are that it should be much easier to program code to use more cores. Yes Snow Leopard has some things that will make a difference to the user in regards to multithreading, i.e. multithreaded Finder (at last) but its not going to magically make Photoshop, Safari or StuffIt expander run faster.
Oh dear...
I really hate to post this in public, but I have to report the first dead MP!
Remember my power supply, the one that smelled so awful? Well, it finally gave in today. The smoke came out of the back when it blew itself up. It started with weird noises like the power connector wasn't properly fitted (yeah, those good old PC days) and then BOOM and all light went out.
What a sad moment.
I really hate to post this in public, but I have to report the first dead MP!
Remember my power supply, the one that smelled so awful? Well, it finally gave in today. The smoke came out of the back when it blew itself up. It started with weird noises like the power connector wasn't properly fitted (yeah, those good old PC days) and then BOOM and all light went out.
What a sad moment.