Posts by Stef

1) Message boards : Cruncher's Corner : AMD Fury GPU (Message 142215)
Posted 7 days ago by Profile Stef
I think those high-end GPUs need a lot of units in parallel to benefit from their massive amount of ALUs, regardless what the GPU usage says.

Would be nice if you could compare them with 4x to 8x.
2) Message boards : Cruncher's Corner : GPU computing benchmark site tip? (Message 141664)
Posted 35 days ago by Profile Stef
Computerbase (german) did some computing Benchmarks; Musemage, CompubenchCL, Computemark, Folding@home, GPUPI and luxmark:

http://www.computerbase.de/2015-06/amd-radeon-r9-fury-x-test/8/
3) Message boards : Cruncher's Corner : GPU computing benchmark site tip? (Message 141648)
Posted 36 days ago by Profile Stef
It depends very much on the application. Application A can run faster on one card and application B on the other.
The new R9 Fury X should be very fast with double precision calculations, at least compared to the NVidia.

I'm very curious how it performs on einstein (it doesn't use double precision though).
4) Message boards : Problems and Bug Reports : BRP4 Intel GPU app feedback thread (Message 141443)
Posted 51 days ago by Profile Stef
I did a "echo 0 > /sys/module/i915/parameters/enable_cmd_parser" and patched the Kernel (4.0.4) with https://01.org/zh/beignet/downloads/linux-kernel-patch-hsw-support
But I only ran one WU, which turned out invalid.
5) Message boards : Problems and Bug Reports : BRP4 Intel GPU app feedback thread (Message 141434)
Posted 51 days ago by Profile Stef
I compiled beignet from the latest git. It compiled and worked, but it did not verify against BRP6-cuda32-nv301.
So the latest version of beignet known to work with einstein is 0.9.2.
6) Message boards : Cruncher's Corner : Times (Elapsed / CPU) for BRP5/6/6-Beta on various CPU/GPU combos - DISCUSSION Thread (Message 140865)
Posted 86 days ago by Profile Stef
Well, luckily I just sold the 750 Ti for exactly the same price I bought it 10 months ago - because of currency loss. So all I paid was the energy, which I would have to pay anyway to heat the house.
But I just found out I'm in the wrong thread. So I'll start a new one in september/october :)
7) Message boards : Cruncher's Corner : Times (Elapsed / CPU) for BRP5/6/6-Beta on various CPU/GPU combos - DISCUSSION Thread (Message 140841)
Posted 87 days ago by Profile Stef
Heat isn't necessarily a problem. I'm planning to heat the living room with e@h next winter. 300W would be decent, I hope the R9 390/X will deliver a good performance. But I will have to see and compare what's on the market in October. And then sell it once it gets warmer outside.
That's the plan :)
The biggest problem is noise. I like it quiet. And dissipating 200-300W quietly isn't that easy without water cooling, which is too expensive.

The 750Ti provided a little too less heat past winter. The whole rig drew only about 110W including CPU tasks.
8) Message boards : Technical News : Fermi LAT Gamma-ray pulsar search #4 "FGRP4" (Message 140766)
Posted 89 days ago by Profile Stef
On a second thought I guess this won't go the other way round, once the fft size is defined.
9) Message boards : Technical News : Fermi LAT Gamma-ray pulsar search #4 "FGRP4" (Message 140760)
Posted 89 days ago by Profile Stef
The memory requirement is basically due to the FFT size. We could reduce this, but then would either loose resolution (sensitivity) or largely extend the computing time by ^2 (half size = quad time).


Any chance to make this user-configurable?
For me a thread could take 2-4GB, so it could run a lot faster at the same resolution, if I understood you right.
10) Message boards : Problems and Bug Reports : BRP4 Intel GPU app feedback thread (Message 140030)
Posted 115 days ago by Profile Stef
So they're working on a patch, should be finished by next week. I'm curious if it will work with e@h.

Could a forum moderator move beignet related postings to a separate thread?
Thank you.


Next 10

Home · Your account · Message boards

This material is based upon work supported by the National Science Foundation (NSF) under Grants PHY-1104902, PHY-1104617 and PHY-1105572 and by the Max Planck Gesellschaft (MPG). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the investigators and do not necessarily reflect the views of the NSF or the MPG.

Copyright © 2015 Bruce Allen