I must not be trying hard enough but I haven't found a failure case for mfakto-0.14-MGW on the RX480 yet. What are some quick tests that have a higher probability of failing?

Just to make sure we are talking about the same failures, I've only seen the failures of the selftests with 0.15pre6 (as noted by James) but not the hang or BSOD noted by ixfd64.

Also, the R7 M460 isn't a "new" 14nm GPU is it? So cards other than the RX4xx are also affected? (I think that was noted in the other [URL="http://www.mersenneforum.org/showpost.php?p=439092&postcount=53"]thread[/URL])

Just a FYI. You should probably check the failed results just to make sure if this is the case or if it's something else.
[QUOTE=Bdot;392302]mfakto version 0.14 did not have these errors because the failing tests were part of another 2300 tests that I only added for 0.15. I built a version of 0.14 with just these tests added, and they are failing too.

This means that these tests were unreliable:
[LIST][*]run 81-82, 82-83 or 87-88 bits on any GCN GPU (other GPUs would select a different kernel)[*]out of 262 test cases in these 3 ranges, 9 are failing (~3.5%). Due to the small number of tests, the error percentage is rather inexact.[/LIST]Please do not use mfakto 0.14 for tests beyond 80 bits on GCN. I will reduce the bitlevels of these kernels to a safe level for 0.15 and add another kernel that has the additional code to safely run to 88 bits.

I'm sorry.[/QUOTE]

[QUOTE=James Heinrich;447628]So I ran the more extensive -st and got:[/QUOTE]Probably doesn't add much info, but just for fun I ran the even-longer -st2 and got:[code]Selftest statistics
number of tests 335250
successful tests 334633
no factor found 617
selftest FAILED!
ERROR: selftest failed, exiting.[/code]Which is roughly the same ~0.19% missed factors as -st, which isn't huge but I still don't think I should be saying that exponents certainly don't have factors when I'm not 100% certain they don't :unsure:

[QUOTE=James Heinrich;447628]So I ran the more extensive -st and got:[code] successful tests 33961
no factor found 65
selftest FAILED![/code][/QUOTE]For what it's worth, using the latest drivers (17.1.2) I now get:[code]Selftest statistics
number of tests 34026
successful tests 34026
selftest PASSED![/code]

I updated the driver to version 17.3.3 and am no longer getting self-test failures. However, the computer occasionally crashes with the following error:

[CODE]The computer has rebooted from a bugcheck. The bugcheck was: 0x000000ea (0xffffda0f1ca5c7c0, 0x0000000000000000, 0x0000000000000000, 0x0000000000000000). A dump was saved in: C:\WINDOWS\MEMORY.DMP. Report Id: e97b2994-dcbe-4a1e-8723-7545811caec6.[/CODE]

I had a crash with the same codes (0x...000ea, a random one, and 3 0x...000) on 17.3.2. I wasn't running mfakto.
I only saved the event log, and named it threaddevicestuck... probably because on the bsod screen there was something similar to threaddevicestuck.
[I]Have you tried to analyze MEMORY.dmp with e.g. Debugdiag ? <- i realized it's a poor program for diagnostic afterwards[/I]

Hey all, I have a question. I know I posted a thread about it. I may need it deleted, I don't know yet.

I have an Intel Graphics Card (Yes, I know), and I found the download page for OpenCL for Intel Cores. I downloaded the SDK and the Runtime, restarted my computer, tried to run mfakto, and ended up with the following error:

"ERROR: init_CL(3, 0) failed"

Do I always have to have AMD/ATI to run this? I want a way to be able to run it on Intel. Thanks!

[QUOTE=Yoshi24517;455884]Hey all, I have a question. I know I posted a thread about it. I may need it deleted, I don't know yet.

I have an Intel Graphics Card (Yes, I know), and I found the download page for OpenCL for Intel Cores. I downloaded the SDK and the Runtime, restarted my computer, tried to run mfakto, and ended up with the following error:

"ERROR: init_CL(3, 0) failed"

Do I always have to have AMD/ATI to run this? I want a way to be able to run it on Intel. Thanks![/QUOTE]

We did have something working for intel iGPU's a while back.. but it really wasn't worth it. I think we got less than 20 GHz/days if i recall correctly and not much more was done on that.

Hi kracker!

[QUOTE=kladner;457160]Hi kracker![/QUOTE]

Hi kladner!

[SIZE="1"][COLOR="Silver"](been busy)[/COLOR][/SIZE]

[QUOTE=kracker;457153]We did have something working for intel iGPU's a while back.. but it really wasn't worth it. I think we got less than 20 GHz/days if i recall correctly and not much more was done on that.[/QUOTE]

Ah. I did try changing GPUType to CPU and SieveOnGPU to 0, but now I'm getting an out of resources error. :down: