For non commercial use it's free.
Well I just placed my order a week ago (est arrival on February, and yes I was hoping to get some result before purchasing but looks like it's not going to happen ) but mine is a base model, would be nice to get those with other specs to post it up as well. Since there's some discussion within the 3D CAD community on how much do the number of cores or GPUs affects the software performance.
I ordered 8 hours after launch and got the feb delivery date too. It's agonising. I guess we are all in the second batch or something.
You still have time to change your order if somebody puts up the results for you.
Wow, 8 hours after lunch and feb? Yikes! I thought only after 24 hours it became February. I hope mine arrives in Feb and not March!![]()
got my 6 core d500 today, here are the results, this is with latest fire pro drivers from amd
https://dl.dropboxusercontent.com/u/1177306/viewperf.htm
I bought Windows 8 Pro System Builder to Bootcamp Windows on my oMP 4,1, because the typical Windows DVD is only an Upgrade disc. I thought the System Builder disc is tied to, one and only, computer/logic board which meant that I would have to buy another System Builder disc for nMP, BUT searching around http://windows.microsoft.com/en-us/windows/answers?tId=23cfae6b-d992-470c-a5cb-d725013ddfc6 says I can use again the System Builder disc! See the question below:...
reset driver to default to be sure, noticed 16x AF was on, did a reboot
better now but still not super impressed https://dl.dropboxusercontent.com/u/1177306/viewperf3.htm
reset driver to default to be sure, noticed 16x AF was on, did a reboot
better now but still not super impressed https://dl.dropboxusercontent.com/u/1177306/viewperf3.htm
got my 6 core d500 today, here are the results, this is with latest fire pro drivers from amd
https://dl.dropboxusercontent.com/u/1177306/viewperf.htm
----------
About to do another test using boot camp driver
Finally! Thanks! And for comparison, here's the link for specviewperf12 offical benchmark summary
Hmm, any chance you can run a Specviewperf 11? I know it's not comparable, but TomsHardware did the test with 11, and there's more worstation GPU test made with 11 that can be found online.
Regarding the benchmark result, it looks slightly slower then AMD W series, probably cause of the clock rate but when it comes to workstation GPU, it seems Nvidia still lead the pack as damn those Kepler Quadros are fast. It could also be that they are better optimized.
The D500 compared to the W5000, seems to be on par with it being faster and slower depending on the program. It'll be interesting to see anyone with D300 as other benchmark results show there's not much performance gain between a D300 and D500. As for D700 I expect it to perform similar to a W9000 with it being slower, it's a pity there's no W9000 data in the specviewperf12 official benchmark page.
Interestingly, AMD compares W5000 to Nvidia K2000, link. On Amazon, it seems both cards are originally retailed at $599, but current price discount made the K2000 cheaper then W5000
UPDATE: I checked a few websites (link,link) and based on the spec given by Apple and AMD, they expect the cards to be
D300 - under clocked and lower memory W7000
D500 - ???
D700 - slightly under clocked and lower memory W9000
So it seems the D500 is 10fps slower then a W7000, I wonder could it because of the lower GPU memory or because of the drivers. Nonetheless, although it's slower, the specviewperf result at least proved that they are indeed FirePros and not FirePros with Radeon drivers cause if it is, the performance will be much worse
Does this test measure OpenCL or OpenGL performance?
Anyway, I updated my GPU comparison chart based on the fact we know the clock speeds now making it somewhat easier to compare cards based on specs.
The D500 is almost certainly a Tahiti Pro as in the R9 280 with some cores disabled and lower clocks.
I think the clock speed advantage of the D300 makes it an excellent choice and would certainly explain why it's on par with the D500 in most tests.
SPECviewperf 12 measures the 3D graphics performance of systems running under the OpenGL and Direct X application programming interfaces. The benchmark’s test files, called viewsets, represent graphics content and behavior from actual applications.(link)
From my understanding, people run Specviewperf to measure OpenGL performance as most 3D CAD software uses OpenGL and not DirectX except for Autodesk Inventor, but from the official Specviewperf12.
Though I'm not exactly sure how it benchmarks DirectX, but nonetheless Gaming GPU drivers does VERY poorly on a Specviewperf benchmarking test which makes it easy to know if the FirePros supplied in nMP comes with actual FirePro drivers or not.
Hmm, I thought it has been concluded that the architecture of the GPUs; D300 is Pitcairn XT, D500 is Tahiti Pro and D700 is Tahiti XT?
I'm really interested to see results from D300 and D700 as when it comes to 3D CAD (SolidWorks, Inventor, SolidEdge), from my understanding, the softwares makes little use of cores but more into processor speed, meaning a 3.7GHz Quad-Core Xeon will be faster then 3.5GHz 6-Core Xeon (I haven't seen an actual benchmark test comparing those 2 in 3D CAD environment, but I've seen a benchmarking done on a Dual-Core i7 with faster GHz versus a Quad-Core Xeon with lower GHz, the i7 actually beats it quite comfortably in many tasks).
Though this is CPU, I wonder if it's the same for GPU as well since the D300 has much faster clock speed (from the chart you've made) then D500 except the D500 has larger memory, wider bus, and higher bandwidth.
reset driver to default to be sure, noticed 16x AF was on, did a reboot
better now but still not super impressed https://dl.dropboxusercontent.com/u/1177306/viewperf3.htm
Yup, thanks a lot guys (and maybe gals?)! Now I'll be looking for anyone willing to run this test on their D300 and D700, I'll will run the benchmark test when my nMP w/ D300 arrives but it'll only be next month, hopefully.Interesting, Thanks. Anyway, glad we got this sorted out for you.
Hmm, if I'm not mistaken he did tested with standard bootcamp drivers and drivers that's downloaded from AMD, interesting result as some software sees some slight improvement while degrades in other.BTW, I don't know if you picked this up, but he says he ran the benchmark using AMDs Pro drivers (not Apple's bundled boot camp drivers). Which is interesting, because I remember suggesting early on (post #5) that if someone was successful in using the Catalyst Pro drivers with the nMP in Windows it would be as important and revealing about the underlying nature of these cards, and it seems the dude that ran the benchmark did exactly that. Had these not been workstation level cards in Windows, the Pro drivers wouldn't work at all. He would have had to use gaming drivers if they were not workstation cards.
Do we know for a fact that Apple includes the equivalent of the Catalyst Pro drivers in the Bootcamp bundle, or is that still an open question?
Nope, it didn't. I thought you're wondering about what architecture each card is based on.I agree, did my table seem to suggest otherwise? I think it supports that conclusion - in fact I'd like to think this table, which has been around for months, helped draw that conclusion![]()
Yeah, once we get specviewperf benchmark results from D300 and D700, it'll be clearer (I hope).Interesting. I'm interested too, not in these particular apps, but to try and understand Apples reasoning behind the D500... It seems to have a bit of an identity crisis as to where it makes most sense.
Hmm, I thought it has been concluded that the architecture of the GPUs; D300 is Pitcairn XT, D500 is Tahiti Pro and D700 is Tahiti XT?
D500 it's not the Pro. It's XT2 like D700.