Posts by Beyond
|
1)
Message boards :
Number crunching :
Amicable is this weekends FB Sprint
(Message 1159)
Posted 30 May 2019 by  Beyond
Post: Once we reach the end of this stage and begin the search up to 10^21 the GPU applications should once again be very efficient (I was at approximately 4,000,000 credit per day with an RTX 2070 before it became CPU-bound), and a single task should fully utilize any GPU with only a fraction of a CPU core to support it. About when is that due to happen? |
|
2)
Message boards :
Number crunching :
GPU Utilization
(Message 1093)
Posted 20 Feb 2019 by Beyond
Post: What should we reduce it to?You can reduce kernel size in your project preferences: https://sech.me/boinc/Amicable/prefs.php?subset=project That wasn't the question. Please read the question. |
|
3)
Message boards :
Number crunching :
GPU Utilization
(Message 1090)
Posted 19 Feb 2019 by Beyond
Post: Reduce kernel size, run as many WUs per GPU as possible. 8 WUs per GPU should be good enough to load it. What should we reduce it to? |
|
4)
Message boards :
Number crunching :
Amicable Numbers has slowed down
(Message 1077)
Posted 3 Feb 2019 by Beyond
Post: WUs seem to slowing down more and more quickly in the last week or so. |
|
5)
Message boards :
Number crunching :
Auto-ban faulty hosts
(Message 1076)
Posted 29 Jan 2019 by Beyond
Post: It's not harsh, because extremely few computers actually get banned here. And that's a problem for the project and the rest of us that are still processing WUs. There's a mechanism in BOINC that will lower the number of WUs a bad host gets per day until it only receives 1/day. Please institute this mechanism. If the host starts producing valid WUs their allocation will increase. Right now the situation is ridiculous, with some hosts spewing out hundreds of errors. This causes valid users to receive erroneous "Completed, can't validate" messages. All because of good WUs that get flagged as "Too many errors (may have bug)" when they are fine. Example 1: https://sech.me/boinc/Amicable/workunit.php?wuid=9905159 My machine is listed as "Completed, can't validate" because of "Too many errors (may have bug)". In reality EVERY other machine is one that produces almost all errors. Example 2: https://sech.me/boinc/Amicable/workunit.php?wuid=9880120 Once again, my machine is listed as "Completed, can't validate" because of "Too many errors (may have bug)". In reality the 6 other machines are ones that produce almost all errors. This is just crazy. Please get these bad machines under control. |
|
6)
Message boards :
Number crunching :
Massive Improvement to Runtimes With New Nvidia Driver
(Message 856)
Posted 2 Jun 2018 by Beyond
Post: A little over 10% improvement in both Win 7 and 10 compared to various earlier driver versions. I'm pleasantly surprised. |
|
7)
Message boards :
Number crunching :
Massive Improvement to Runtimes With New Nvidia Driver
(Message 855)
Posted 1 Jun 2018 by Beyond
Post: Thanks to both of you for the heads up. |
|
8)
Message boards :
Number crunching :
Massive Improvement to Runtimes With New Nvidia Driver
(Message 853)
Posted 1 Jun 2018 by Beyond
Post: What version driver? What OS? |
|
9)
Message boards :
Number crunching :
GPU version requirements for the search up to 10^20
(Message 550)
Posted 2 Jul 2017 by Beyond
Post: Fixed it, version 2.02 should work fine. It's fixed for me. Thanks Sergei! |
|
10)
Message boards :
Number crunching :
GPU version requirements for the search up to 10^20
(Message 540)
Posted 2 Jul 2017 by Beyond
Post: Looking through the list of top hosts, many (but not all) AMD GPUs are having the same issue. They ran v1.21 fine but error with v2.01. |
|
11)
Message boards :
Number crunching :
GPU version requirements for the search up to 10^20
(Message 539)
Posted 2 Jul 2017 by Beyond
Post: Looks like an OpenCL problem. SDERR output: <core_client_version>7.6.33</core_client_version> <
©2024 Sergei Chernykh