log in |
1)
Message boards :
News :
A bad day happened
(Message 9170)
Posted 17 Oct 2023 by k3ack3r Wish you the best for a fast recovery! |
2)
Message boards :
Number crunching :
Looking to optimize kernel load orders with RX 6xxx
(Message 7878)
Posted 30 Aug 2021 by k3ack3r For the most part RDNA and RDNA2 kernel orders seem similar with the results I got. Thanks everyone for the benchmarks. |
3)
Message boards :
Number crunching :
Looking to optimize kernel load orders with RX 6xxx
(Message 7872)
Posted 28 Aug 2021 by k3ack3r Could I also have the output of 2.log? You can also PM me the output. Thank you for the test! |
4)
Message boards :
Number crunching :
Looking to optimize kernel load orders with RX 6xxx
(Message 7864)
Posted 21 Aug 2021 by k3ack3r Looking for anyone willing to run some benchmarks on windows for the RX 6xxx cards. I know this is late and I perhaps should be posting this in the TF thread. https://ufile.io/o4g8qeae Just run the bat file and post/PM/whatever the output of the *.log files. Thanks |
5)
Message boards :
Number crunching :
Trial Factoring
(Message 6251)
Posted 17 Apr 2020 by k3ack3r mfakto new version, maybe people can sanity check it. http://www.filedropper.com/mfakto015pre7win64 |
6)
Message boards :
Number crunching :
mfakto --perftest
(Message 6250)
Posted 17 Apr 2020 by k3ack3r Thanks! Decent speedups from using the MUL32 kernels. 69-70 15% 71 20% 72-73 28% 74 31% 75 36% 76 36% 77 31% 78 27% |
7)
Message boards :
Number crunching :
mfakto --perftest
(Message 6244)
Posted 16 Apr 2020 by k3ack3r Did you try the second one I posted later? |
8)
Message boards :
Number crunching :
Trial Factoring tests
(Message 6148)
Posted 13 Apr 2020 by k3ack3r Even though your APU is not recognized, I think selecting APU will be slower than "AUTO" - kernels used on the APU settings were tested on much older, different APU's (if i recall correctly, that was tuned for A8-3850... released 2011.) I'm hoping I can test the IGP of the Ryzen 2400g that I get occasional access to... |
9)
Message boards :
Number crunching :
mfakto --perftest
(Message 6147)
Posted 13 Apr 2020 by k3ack3r All I need is data from R VII, once I get that most GPU's should be running on "good" kernels with no unrecognized warning. Also... thank you to everyone who has posted their results! :) |
10)
Message boards :
Number crunching :
mfakto --perftest
(Message 6106)
Posted 11 Apr 2020 by k3ack3r Well, appearently "skipped if empty" is skipped only when the whole entry/line is gone... really hoping this one works/doesn't crash the GPU. http://www.filedropper.com/mfaktokernelbenchmark |
11)
Message boards :
Number crunching :
mfakto --perftest
(Message 6102)
Posted 10 Apr 2020 by k3ack3r Also, ignore the errors on console and post perftest*.txt Thanks :) |
12)
Message boards :
Number crunching :
mfakto --perftest
(Message 6101)
Posted 10 Apr 2020 by k3ack3r http://www.filedropper.com/mfaktowinbenchmark Just run the benchmark script - edit it and add -d xx if you need... |
13)
Message boards :
Number crunching :
Trial Factoring
(Message 6100)
Posted 10 Apr 2020 by k3ack3r Here is a list of how the GPU's perform: Notice: it's not 100% accurate, and I've found AMD gpu's usually perform better than listed here. https://www.mersenne.ca/mfaktc.php |
14)
Message boards :
Number crunching :
Trial Factoring tests
(Message 6078)
Posted 10 Apr 2020 by k3ack3r Also- if anyone is trying to debug mfakto errors make sure to set Verbosity to 3 in the ini to show everything. |
15)
Message boards :
Number crunching :
Trial Factoring
(Message 6077)
Posted 10 Apr 2020 by k3ack3r Thanks. Still could use results from any recent GPU(Vega 56/64, Radeon VII, etc) Think there's a pretty big speedup possible for the VII - maybe. Radeon_5500XT 75bits 9% speedup 76bits 8% speedup 77bits 7% speedup |
16)
Message boards :
Number crunching :
Trial Factoring
(Message 6075)
Posted 9 Apr 2020 by k3ack3r That's what I needed. Thank you! |
17)
Message boards :
Number crunching :
Trial Factoring
(Message 6072)
Posted 9 Apr 2020 by k3ack3r That's weird - are you using an edited mfakto.ini or removed TestExponents from it? I have TestExponents=2000093,39000037,66362159,74000077,78000071,332900047,999900079,2001862367,4201971233 in mine as default.. need the TF benchmarks. |
18)
Message boards :
Number crunching :
Trial Factoring
(Message 6070)
Posted 9 Apr 2020 by k3ack3r No output file, just the text output of the program itself. |
19)
Message boards :
Number crunching :
Trial Factoring
(Message 6067)
Posted 9 Apr 2020 by k3ack3r Thought I'd post here with likely more success - I'm trying to add support for relatively newer GPU's on mfakto(AMD GPUs) - basically a correct kernel ordering... could be decent speedups. Could I have the output of --perftest and GPU name for any GPU later than RX 5xx? |