Author

Topic: Can I mine anything with a NVIDIA GeForce FX5200 ? (Read 4145 times)

hero member
Activity: 504
Merit: 500
Minecraft... 30FPS... Mine all the Coal, Iron, Redstone, Gold, Diamond, and Obsidian you want... Tongue

You can mine with an android phone... so I know it is possible. Just not sane!

I am sure someone out there actually knows how to program nVidia cards to do what needs to be done. But most just copy-paste "known code", and don't try anything beyond that.

I was one of the first to propose the use of the GPU's super-fast memory mods, to do complex math with one instant POP, by simply using a graphical overlay that was actually a math formula, as a layered set of images. While others were still trying to push individual "commands", without success. Funny, you can ADD, MULTIPLY, SUBTRACT, (limited divide), and SHIFT, all with a single image-buffer and an overlay, onto another image, which is actually a giant array of a bunch of numbers. Lightning fast, and great for using numbers that are longer than a 64-bit, or even 1024-bit, or 32,000-bit variable/array.

We are at the mercy of the coders... The coders are at the mercy of what they are familiar with... What they are not, they initially copy-paste (or recycle), so they do not keep reinventing the wheel. However, the wheel sometimes need to be reinvented for racing on an oval track, in the desert, on dust, by crappy drivers in sloppy cars.
full member
Activity: 194
Merit: 100
Thanks! I am going to stop spending time trying to get GPU mining to work on this computer, since based on runlinux's comment, apparently it will be impossible. (BTW, I realize this is an old computer, so I wasn't expecting to mine profitably--the computer takes 5 minutes to boot up and load Windows, so I usually leave it running 24-7 anyway. Since I'm already paying for the electricity, I figured if it even mines .0001 BTC now and then, it could help offset the electricity costs, or at least give me a few fractions of BTC to play with without the inconvenience of buying them.)

sr. member
Activity: 378
Merit: 250
You might be able to mine a POOPcoin or two.
full member
Activity: 174
Merit: 100
No, I suspect this is a troll.
hero member
Activity: 566
Merit: 500
Nope. Only starting with 8 series did nvidia support mining bitcoins (OpenCL / CUDA). That FX5200 is 3 generations too old.
newbie
Activity: 30
Merit: 0
A FX5200 has 4 dedicated pixel shaders and 2 vertex shaders. I doubt fixed function pipelines like these could handle shader computation programs. You won't even get 1 Mhash/s if anything at all making it useless.
hero member
Activity: 840
Merit: 1000
Does FX5200 even support CUDA? I don't think it does. Even if you could mine on that it would cost you way more in electricity.
full member
Activity: 194
Merit: 100
I have an old Sony VGC RA710G desktop computer I bought back in 2004. I am able to use minerd to mine litecoins on it with the CPU, but I am having problems mining anything with the GPU.

The "video & graphics" part of the specifications sheet says it has the following:
NVIDIA GeForce FX5200 w/TV-Out
128MB Video Memory (128-bit DDR)
TV-Out / VGA-Out / DVI-Out (DVI output: 1280x1024x60Hz max.)
8X AGP 3D Graphics Hardware Acceleration
Giga Pocket™ MPEG2 Realtime Encoder/Decoder board with TV Tuner

The latest thing I tried was downloading "guiminer-scrypt_win32_binaries_v0.04.zip" to try mining litecoins (and other scrypt-based altcoins) with the GPU. This worked without a hitch on my other (newer) computer, but when I try to run it on this Sony, guiminer.exe doesn't load, and it generates a guimer.exe.log file with the following contents:

Traceback (most recent call last):
  File "guiminer.py", line 20, in
  File "pyopencl\__init__.pyo", line 4, in
  File "pyopencl\_cl.pyo", line 12, in
  File "pyopencl\_cl.pyo", line 10, in __load
ImportError: DLL load failed: The specified module could not be found.

Is there a way to get mining (either SHA-256 or, preferably, scrypt) to work with the GPU in this computer, or is it just completely incompatible?

Thanks!


Jump to: