Ok, thank you. Do you by any chance have a link for them on amazon? I'm not from the States, so an Amazon.co.uk link would be cool
Just Google for "Mac Pro PCIe power cable" and you should find an Amazon link.
Ok, thank you. Do you by any chance have a link for them on amazon? I'm not from the States, so an Amazon.co.uk link would be cool
Just Google for "Mac Pro PCIe power cable" and you should find an Amazon link.
I have an EVGA GTX680 in my Mac Pro. I just installed the cuds driver 5.0.17.
This is what the system preferences show. CUDA-Z works fine.
What program do you guys use? Thanks!
Hi guys, what do you guys use to monitor GTX670 temp/vram usage?
Since doing the opencl hack from netkas, and installing the nvidia cuda driver, istat pro only recognizes the gpu, but no vram usage.
Atmonitor would start, and tell me the gpu processor load, but no vram usage or temp. And then it crashes 5 seconds after opening( crashing is due to the opencl hack or cuda driver, since it was working fine before I installed those).
What program do you guys use? Thanks!
I have one, maybe stupid question...
Do you have any problems with putting 670 in Mac Pro? Because Mac Pros have PCI/E 2 and GTX 670 has PCI/E 3 so (I told I am stupid in PCs ) No problem in compatibility, yea?¨
Thx
Without a new 64 bit EFI, it runs at pcie 1.1 16x (which is 97%-100%) the performance of pcie 2/3 16x
Thank for your help now I can buy gtx 670 "in peace"
But I hope last quick question... Can I put into my Mac Pro Intel Core i7 3770k e.g.? I dont think now but sometime in future... I have bought my Mac Pro last year (december) and CPU is good but it is little old :/
No. No Mac Pro supports latest generation CPU's.
The question is, is the card bottlenecked by PCIe 1.1? And the answer is, yes. Around 10-15%. You will bottleneck the card with PCIe 1.1.
http://www.hardocp.com/article/2012/07/18/pci_express_20_vs_30_gpu_gaming_performance_review/
If the workload doesn't involve a ton of bus traffic, then the impact of PCIe 2.0 or 3.0 can be quite minimal. Modern games will basically download all their resources to video memory, and once that's done, there is little traffic outside the commands to render the frame (which are tiny in comparison to the textures/model data etc).
"...most of the performance advantage had under PCIe 3.0 (as compared to PCIe 2.0) was well under 10%. This actually falls in-line with the kind of performance advantages one might expect using an Ivy Bridge CPU clock-for-clock compared to a Sandy Bridge CPU. The IPC can affect performance by as much as 4-7% in favor of Ivy Bridge easily."
It really depends on the workload, I don't think you can claim 10-15% across the board.
http://www.hardocp.com/article/2012/07/18/pci_express_20_vs_30_gpu_gaming_performance_review/
If the workload doesn't involve a ton of bus traffic, then the impact of PCIe 2.0 or 3.0 can be quite minimal. Modern games will basically download all their resources to video memory, and once that's done, there is little traffic outside the commands to render the frame (which are tiny in comparison to the textures/model data etc).
Certain GPGPU workloads are definitely dependent on a fast connection between the CPU and GPU, but as I said, it really depends on the application/workload.
(2) Sorry, this question must sound ridiculous, but how do I access windows under ML without the bootscreen? I read that bootchamp isn't working under Lion. Parallels, Virtual Machine or what?
Thank's a lot!
Sorry, this question must sound ridiculous, but how do I access windows under ML without the bootscreen? I read that bootchamp isn't working under Lion. Parallels, Virtual Machine or what?
I don't know where you picked up that little piece of information but BootChamp works wonderfully with ML. I am looking at it in the top-right of my screen now and it performs flawlessly.
The official BootChamp website is http://kainjow.com and I am sure Kevin (the author) would love a small donation.
Ok, I've to admit that you really got me hooked to buy a new card!
I read a fair amount of related threads, but some noob questions still remain, as all of this 'pro-talk' can confuse me a great deal at times.
(1) I came across this card - eVGA GeForce GTX 670 SC 4 GB RAM - and while confused because it says 500 W think it will work because of 2 6pin connectors.
I read somewhere that with >2gb you have to 'hack' around a bit, because OSX doesn't recognise it - Is that still true? I do work with Adobe CS and Maxwell Renderer and was looking at the 4GB but would be a 680 more advisable? Or even the 570 with 2.5GB as MVC wrote (I think) somewhere?
(2) Sorry, this question must sound ridiculous, but how do I access windows under ML without the bootscreen? I read that bootchamp isn't working under Lion. Parallels, Virtual Machine or what?
Thank's a lot!
I have that card. (...)
So my HD 5870 is, until better drivers are available, faster for x-plane.
For the rest Geforce GTX 680 is the champ.
I find this very strange, why isn't x-plane also faster?
One more thing:
In X-plane, turning HDR and 4XSSAA Quad Render on with the GTX 680 is not dropping the fps!
Sometimes fps are higher, this is in Win and ML. The power of the GTX 680?
The nvidia drivers are indeed beta, but I have a feeling that this is also the case for the windows drivers (?)
Thank you very much xav8tor! Very much appreciated.
Can I ask you why you chose the 670 instead of 680?
There is a new 680 UltraCharged with 4GB available (soon):
should work as well?! Sorry, if I produce spam in this thread...