Well now that the conditions that gave birth to MW_GPU have reoccured I think its time to START MW_GPU, everyones time n' money is being wasted when they cant get or even report WUs! Time for action, the fix for the problem is OBVIOUS! Please do what you said you were going to do this time. We do really need the 2nd site now, I'm sure keeping things as they are is going to( and already has) cause people to go elsewhere.

---

Well now that the conditions that gave birth to MW_GPU have reoccured I think its time to START MW_GPU, everyones time n' money is being wasted when they cant get or even report WUs!

I don't think that there's a need for the MW_GPU project at all.
They just need to increase the allowed amount of WUs for the GPUs, that's all.

Time for action, the fix for the problem is OBVIOUS!

Yeah, but he won't give up control over the project, even though it's obvious that he can't handle it ...

---

Join Support science! Joinc Team BOINC United now!

Well now that the conditions that gave birth to MW_GPU have reoccured I think its time to START MW_GPU, everyones time n' money is being wasted when they cant get or even report WUs! Time for action, the fix for the problem is OBVIOUS! Please do what you said you were going to do this time. We do really need the 2nd site now, I'm sure keeping things as they are is going to( and already has) cause people to go elsewhere.

Second that. My 7 Ati 4870 happely crunhing Collatz.

---

I don't think that there's a need for the MW_GPU project at all.
They just need to increase the allowed amount of WUs for the GPUs, that's all.

I didnt know if they could set the srever up to distinguish between GPU and CPU when giving out new work, if they can then that might solve the problem, but there is a lot of traffic, maybe it wont help enough, but definately worth a try. If we can get somebody to start the server up again!!!
;-p

---

Well now that the conditions that gave birth to MW_GPU have reoccured I think its time to START MW_GPU, everyones time n' money is being wasted when they cant get or even report WUs! Time for action, the fix for the problem is OBVIOUS! Please do what you said you were going to do this time. We do really need the 2nd site now, I'm sure keeping things as they are is going to( and already has) cause people to go elsewhere.

Second that. My 7 Ati 4870 happely crunhing Collatz.

I was running Collatz fairly heavy before last weekend but since then I've been running Collatz exclusively & haven't missed MWay a bit with all it's Hiccups. Collatz has Hiccups too but at least you can have a decent Cache there & Slicker does a good job of keeping on top of things even on the Weekends.

I don't think that there's a need for the MW_GPU project at all.
They just need to increase the allowed amount of WUs for the GPUs, that's all.

Again, that will either starve out those of us who do not have GPUs, or will end up at a state where no work is available because those of us with CPUs have enough work on hand to where it causes new work creation to slow down too much.

If there was a separate project or if the tasks that GPUs processed TOOK LONGER (same credit/time ratio), that would satisfy EVERYONE, not just you folks... I can't process a GPU Grid task with my CPU. I can't process a SETI CUDA task with my CPU. I can't process a Collatz GPU task with my CPU. Why is it that we can't have you all with GPUs process a SEPARATE TYPE OF TASK from the rest of us here???????????

I don't think that there's a need for the MW_GPU project at all.
They just need to increase the allowed amount of WUs for the GPUs, that's all.

Again, that will either starve out those of us who do not have GPUs, or will end up at a state where no work is available because those of us with CPUs have enough work on hand to where it causes new work creation to slow down too much.

No it won't. Since the GPUs will return results quite a bit faster. In fact if one would get 150 tasks per gpu, the turnaround time for the 150 task would be the same as one or two Cpu task.

If there was a separate project or if the tasks that GPUs processed TOOK LONGER (same credit/time ratio), that would satisfy EVERYONE, not just you folks... I can't process a GPU Grid task with my CPU.

Well you got to blame GPUgrid for that one. It's their choice not to release a CPU app.

I can't process a SETI CUDA task with my CPU.
I can't process a Collatz GPU task with my CPU.

Yes you can since there's no difference. All WUs are the same "size", both seti and collatz.

Why is it that we can't have you all with GPUs process a SEPARATE TYPE OF TASK from the rest of us here???????????

That's basically the same as sending your CPU one task and 150 to a GPU...
The only difference is that the 150 WUs would be combined to a single big one.

So what difference would that make compared to what i suggested ?

---

Join Support science! Joinc Team BOINC United now!

Assuming the rough scaling between a MW task and the same on my GPU it is between 60-100 something faster. Assume 100 for round numbers... a typical GPU Grid task right now runs about 9 hours ... assuming that they did port to the CPU that would imply tasks that take 900 hours to run. In that GPU Grid is, like MW, a stream oriented project where the next work is based on the return of current work this would add significant delays to them reaching their goals.

In effect, GPU Grid has divided the work into smaller packages already and the actual task is multiple issues to us ... in a way that is helpful.

You probably already knew this ... but it might be of interest to others ...

Anyway, I can see I am queuing up MW tasks slowly but surely and when I have run off the Collatz and GPU Grid tasks currently running I will be back to running off MW tasks again ... in the mean time I did get 82K on Collatz and an increase in my pending list ... :)