We are now past the reported GeneferCUDA b limit of 995k for GFN262144 work on PRPnet (we are over 996k now). However, I am still completing CUDA work with no errors.

Looks like we have a little "b" wiggle room, better get them while they last if you want to do some GPU work on this sub-project!

We are now past the reported GeneferCUDA b limit of 995k for GFN262144 work on PRPnet (we are over 996k now). However, I am still completing CUDA work with no errors.

Looks like we have a little "b" wiggle room, better get them while they last if you want to do some GPU work on this sub-project!

We are now past the reported GeneferCUDA b limit of 995k for GFN262144 work on PRPnet (we are over 996k now). However, I am still completing CUDA work with no errors.

Looks like we have a little "b" wiggle room, better get them while they last if you want to do some GPU work on this sub-project!

We are now past the reported GeneferCUDA b limit of 995k for GFN262144 work on PRPnet (we are over 996k now). However, I am still completing CUDA work with no errors.

Looks like we have a little "b" wiggle room, better get them while they last if you want to do some GPU work on this sub-project!

Strange. I'm not sure why the server was sending out some numbers that high. It looks like, for the most part, it's really only up to about 867K.

We are now past the reported GeneferCUDA b limit of 995k for GFN262144 work on PRPnet (we are over 996k now). However, I am still completing CUDA work with no errors.

Looks like we have a little "b" wiggle room, better get them while they last if you want to do some GPU work on this sub-project!

Strange. I'm not sure why the server was sending out some numbers that high. It looks like, for the most part, it's really only up to about 867K.

Well, we have reached the highest b value in the current work (999998), and now I get a no new work available on the server message on the GFN262144 port despite there being apparently more than 6,000 units left on the port. Looks like something is broken on this one.

Do you have the other genefer apps unchecked on prpclient.ini file?

Do you have the other genefer apps unchecked on prpclient.ini file?

I only have GeneferCUDA checked in the prpclient.ini file:

geneferexe=genefercuda.exe //geneferexe=genefx64.exe //geneferexe=genefer.exe //geneferexe=genefer80.exe

I think that if the candidates on the server are above genefercuda limits, you will not get any work with those settings.

Do you have the other genefer apps unchecked on prpclient.ini file?

I only have GeneferCUDA checked in the prpclient.ini file:

geneferexe=genefercuda.exe //geneferexe=genefx64.exe //geneferexe=genefer.exe //geneferexe=genefer80.exe

I think that if the candidates on the server are above genefercuda limits, you will not get any work with those settings.

No.

The candidates on the server are about 100K or more below the limit of GeneferCUDA.

Do you have the other genefer apps unchecked on prpclient.ini file?

I only have GeneferCUDA checked in the prpclient.ini file:

geneferexe=genefercuda.exe //geneferexe=genefx64.exe //geneferexe=genefer.exe //geneferexe=genefer80.exe

I think that if the candidates on the server are above genefercuda limits, you will not get any work with those settings.

No.

The candidates on the server are about 100K or more below the limit of GeneferCUDA.

I've been in PM contact with Lennart. There appear to be some server problems in sending work that he is looking into.

And just to verify that it is not the CUDA only selection as an issue, I also unchecked genefx64 and still could not get any 262144 work.

Any updates? I still can't get any 262144 work for my gpu's.

I moved this discussion over from Genefer B Limits because this has more to do with a problem on the server than with limits.
____________
My lucky number is 75898⁵²⁴²⁸⁸+1

I have found a possible solution to getting work. I have been able to get work by making active all of the genefer applications. In other words, your code should look like:

geneferexe=genefercuda.exe geneferexe=genefx64.exe geneferexe=genefer.exe geneferexe=genefer80.exe

I am verifying this on one more machine with a different client version. Obviously, this is not ideal for those of us who wish to run only the geneferCUDA application (due to other things working on the CPU), but it will work as needed only on the GPU unless a PRP is actually found on your machine (which would be good news anyway).

I have e-mailed Lennart about this fix so that he can continue investigating what is going on with this on the server side of things.

GFN262144 will soon stop sending work again, because b<1M is almost finished. :)
____________

Thanks for loading more work. Unfortunately, it looks like we're finally passing GeneferCUDA's b limit, so most of the new work has to be done on CPU.

My GTX 460 crashed twice trying to test 994560^262144+1, which is close to the b=995k limit reported by genefercuda -l.

On GTX 660Ti, genefercuda -l reports b=985k as limit. It was able to test 1000176^262144+1 though, but failed on anything beyond that.
____________