I know most gamers wouldn't bother doing this, but occasionally I do CAD and 3D visualization work whose performance is greatly improved on a Quadro. Has anyone tried soft-/firm-modding a GeForce 580M with success? The current version RivaTuner 2.24c doesn't seem to know what to do with the Device IDs of the newest GPUs and how to deal with the latest drivers.
Thanks!
-
-
The 580m uses the GF114 core. The highest mobile quadro is a 5010m with almost the same specs: same shaders and memory bandwidth but lower clocks. However it uses a GF100 core.
They use different core models. If the cores were the same, usually a simple vbios flash is all it takes. However, they are different and will not work.
If you go about editing the ID strings, chances are no drivers at all will install, not even modded ones and worse case is that you brick your card. -
A 6970M/6990M can be flash to a FirePro M8900.
[Softmod] AMD 6970m -> FirePro M8900 *UPDATE* 6990m works too! -
Right, don't want to brick my new computer.
I'll take a hit on outright fps to trade for FP-crunching. Back to my little cubicle hole... Aside from fairly minor annoyances, I really do love my R3. My former Fujitsu-Siemens Amilo had serious unresolved issues (WLAN BSODs) but I kept it for 4 1/2 years!
@GeoCake: Grrr, I'm envious (sort of). AMD has no current equivalent to Optimus (which I consider indispensable) and what few engineering apps I'm using now that exploits the GPU are presently CUDA-only. Waiting for OpenCL ports and I'll likely go AMD next time around! -
Star Forge Quaggan's Creed Redux!
Which makes me wonder why you never went and got a Precision M6600 in the first place? The Quadros 3000M in the most part is way more optimized and better in simulation work than a 580M, despite its actual hardware is one-tier lower in performance. The pain of nVidia is that they want to truly keep their Gaming and their Professional lineup separate and over the last few years, they are trying really hard to prevent cross-modding the two series to work like one of the other so to speak. -
katalin_2003 NBR Spectre Super Moderator
Hello, Dave here's a good read if you ever decide to keep this system and plan on buying a Quadro:
http://forum.notebookreview.com/alienware-m18x/632619-have-m18x-2xgtx580-need-quadro-cad.html -
SlickDude80 Notebook Prophet
AMD allows you to manually switch to integrated graphics. So i decide when to do this. i can get 4-5 hrs on battery on integrated Intel graphics -
Star Forge Quaggan's Creed Redux!
-
I did have an M6600 just recently and I traded up for this. I do 3D CAD only occasionally but for what I do, the Quadro 3000M falls far short of raw GPU computing power to the GTX580M and, AFAIK as far as CUDA/OpenCL is concerned, it does not matter whether it's a GeForce or Quadro. In fact, in terms of single-precision MFLOPS, active cores, clock speeds, memory GB/s, etc., the 3000M is quite far below even a 470M.
I did look for an M6600/4000M but was way out of my budget. It didn't help that the M6600 I had only had a 1600x900 screen. -
SlickDude80 Notebook Prophet
SuperDave, the feeling in the community is that CUDA will be less and less important as we move forward. People like open standards like openCL. Its just a matter of time
i should think back to what i did and write a guide lol -
Anyone tried softmodding to a Quadro - or is there a "firm-mod" to do this?
Discussion in 'Alienware 17 and M17x' started by superdave643, Feb 14, 2012.