There's a note about a server shutdown on the main page, and in that note it is written that single precision GPU users are not wanted here. I've recently picked up with MW and started receiving GPU wu's and have processed an enormous amount of them. The trouble is, I have an Nvidia card, and I was led to believe that it was only single precision. Imagine my surprise when I was able to d/l wu's for the Nvidia gpu's and then imagine more surprise when I found out that this work may be interfering with the normal operation of this project. I'm not trying to scam anybody. In fact, I've done nothing out of malice, so I need a determination. Can I use my Nvidia gpu's for this project or not? I've been with SETI for years, as well as Einstein and I was looking for a change. I can always take my compute cycles somewhere else if that is what is called for. Someone please let me know.

On a side note: I had a CPU wu that took just over 21 hrs. I haven't had a wu from anybody in nearly a decade that took that long. My 486 used to take that long, but I'm running a quad core now with loads of ram in 64 bit mode. What happened?

Nope, your 260 is capable of double precision.

It is the lowest 200 version that does.

---

Don't worry - unless you actively tried to scam the system, this won't affect you :)

On a side note: I had a CPU wu that took just over 21 hrs. I haven't had a wu from anybody in nearly a decade that took that long. My 486 used to take that long, but I'm running a quad core now with loads of ram in 64 bit mode. What happened?

You must not be familiar with Climate Prediction. Shortest are about 2 weeks on my quads.

-Dave

---

That kinda' takes the fun out of it, doesn't it? I was just curious. I seem to be blowing through most MW cpu wu's in less than 8 hrs, and the CUDA wu's in about 30 mins or less. That was the only one I'm aware of that took that long. Thanks for all the responses.

By the way, the CUDA WUs are just as valuable to the project as the CPU WUs - so you might want to consider using your CPU for other projects and your GPU for Milkyway, as your CPU isn't contributing that much compared to your GPU.

I do run SETI as well, although I haven't found a reliable way of saying MW use gpu only and SETI use cpu only. I'm not sure how I did it, but before I got back in here, I was crunching Einstein on the cpu's, and SETI on the gpu's. Now, I've dropped Einstein and it seems there's a fair amount of MW cpu wu's running with the occasional SETI wu sneaking into the cpu's and gpu's. I wish there was a better way of managing this.

I figured out how to regulate who gets what. I'm now set up to run Einstein on the cpu's, and MW on the gpu's. Both apps seem to favor the cruncher they're destined for.

I figured out how to regulate who gets what. I'm now set up to run Einstein on the cpu's, and MW on the gpu's. Both apps seem to favor the cruncher they're destined for.

I've got Seti and Climate Prediction sharing the CPU cores and MW and Seti sharing the GPU cores reliably (using Boinc 6.10.32).

F.

---