Having downloaded and crunched a large number of VLAR tasks with CUDA, I think it's safe to say that it's no longer needed to worry about having a VLAR crash your video driver. However I have noticed that VLAR tasks, while stable, are slow as hell. Performance of my computer is reduced to the same level as when I run GPUGrid...even the mouse pointer has a framerate. :)

Non-VLAR Seti tasks cause a minor performance hiccup but nothing nearly as bad as the VLAR ones. Until this is fixed I will simply suspend VLAR tasks until I know I'll be away from my computer for awhile.

Here are two examples of VLAR tasks that I currently have suspended until later:

http://setiathome.berkeley.edu/result.php?resultid=1149404128

http://setiathome.berkeley.edu/result.php?resultid=1149404132

---

You will be assimilated...bunghole!

The slowdown is known to the developers. They're trying to fix that now, but the initial fix they wanted to give was overall stability, less hangs and crashes especially in the VLAR tasks. So they do run slow, but at least they do not crash.

---

Good to know. I would much rather prefer a slow task than one that crashes. :)

---

You will be assimilated...bunghole!