Hi Andreas, Thanks for the quick response! This is on a Linux machine, and apparently one of the tests the user did was to take a machine where this error was happening on both Titan XP cards that were installed, and swap one of them for a 1060. After this, the error did not occur on the 1060, but did continue to occur on the Titan XP.. I'm not quite sure what to conclude from that actually. I will ask the user to try updating/downgrading the driver version and see what happens.
Thanks! Ali On Wed, Aug 14, 2019 at 12:13 AM Andreas Kloeckner <li...@informa.tiker.net> wrote: > Dear Ali, > > Ali Punjani <apunj...@structura.bio> writes: > > We develop scientific software for molecular biology applications using > > pyCUDA. A user is having a very strange issue where the same code works > > perfectly fine on the same machine (with CUDA 10.1) with GTX 1060 cards, > > but not with Titan XP cards. The error message is > > > > *LogicError: cuMemHostAlloc failed: OS call failed or operation not > > supported on this OS* > > > > Is there any explanation of what this could mean? Initial googling does > not > > return many reports of anything similar. > > > > A more detailed report of the issue is here: > > > https://discuss.cryosparc.com/t/logicerror-cumemhostalloc-failed-os-call-failed-or-operation-not-supported-on-this-os/3278 > > I'm sorry to say that I've never seen or heard of this error > message. One thing that comes to mind is that this might be an issue of > PCIe versioning. The 1060 might be PCIe3, while the XP might be PCIe2 > (guessing, might be better to check), and driver support might > differ. Is this on a Windows or Linux machine? > > Andreas >
_______________________________________________ PyCUDA mailing list -- pycuda@tiker.net To unsubscribe send an email to pycuda-le...@tiker.net