01:55:39 <hyc> zen3 has slower L3 cache
01:56:56 <hyc> zen2 avg L3 latency 39 cycles, zen3 avg is 46 cycles
01:57:59 <hyc> which I suppose is unavoidable since they unified it across the 8 cores and doubled the size
06:01:35 <hv-bridge> <xx> how do we beat intel hyc bro
06:02:10 <hv-bridge> <xx> https://xmrig.com/benchmark/1
06:05:24 <hyc> that's per-core? not really a useful number
06:06:01 <hyc> since AMD has far more cores now than Intel
06:06:21 <hyc> the multi-thread result is the only thing that matters. https://xmrig.com/benchmark
06:06:35 <hv-bridge> <xx> yes but its actually the most fun to bench
06:06:51 <hv-bridge> <xx> we do know that hyc
06:07:12 <hv-bridge> <xx> that 10900kf btw is 5.6GHz intel with 5ghz ring/cache
06:07:21 <hv-bridge> <xx> ddr4 4200 16-16-16-34
06:08:01 <hv-bridge> <xx> hyc: single thread wise, zen2 is still faster than zen3
06:08:06 <hyc> the ryzen 5000 numbers aren't looking too impressive yet
06:08:16 <hyc> but I'd guess an immature BIOS is part of that
06:08:17 <hv-bridge> <xx> that 3600 and 5600X of sech1 is both on 4.7GHz
06:08:54 <hv-bridge> <xx> theres this weird thing on zen3's bios
06:09:01 <hyc> but yeah, zen3 L3 cache is slower than zen2, so this may be as good as it gets
06:09:01 <hv-bridge> <xx> i dont know if itll matter
06:09:02 <hv-bridge> <xx> 0xC86E             One Of: Enhanced REP MOVSB/STOSB, VarStoreInfo (VarOffset/VarName): 0x1C3, VarStore: 0x5000, QuestionId: 0x1C, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 41 00 42 00 1C 00 00 50 C3 01 10 10 00 01 00}
06:09:35 <hyc> that's not an instruction that randomX uses
06:09:56 <hv-bridge> <xx> useless then
06:10:14 <hv-bridge> <xx> https://cdn.discordapp.com/attachments/725316395428085831/774516042683777024/8E38A88A-C267-4131-A8CD-C0BC80A24CB5.txt
06:10:24 <hv-bridge> <xx> common bios options for zen3
06:10:50 <hv-bridge> <xx> how about this
06:10:51 <hv-bridge> <xx> One Of: RdRand Speedup Disable, VarStoreInfo (VarOffset/VarName): 0x294, VarStore: 0x5000, QuestionId: 0x22, Size: 1, Min: 0x0, Max 0x1, Step: 0x0 {05 91 4E 00 4F 00 22 00 00 50 94 02 10 10 00 01 00}
06:10:59 <hv-bridge> <xx> i guess zen3 will have this working
06:11:10 <hv-bridge> <xx> they disabled this on zen2 shit was buggy
06:24:52 <hyc> we don't use rdrand either, the mining nonce is our RNG seed
07:50:27 <sech1> https://www.reddit.com/r/MoneroMining/comments/jpmw5l/zen3_vs_zen2_clock_for_clock_comparison/
08:18:41 <Inge-> sech1: the 3700x - you can't get the RAM there to run well > 3200MT ?
08:18:51 <Inge-> I had that issue on asus strix B450-i
08:18:58 <sech1> No, it's not stable
08:19:07 <sech1> Well, my newer kit ran at 3600 MHz
08:19:25 <sech1> and even my old kit could run (but unstable) with the same timings/voltage. Both are b-die
08:19:28 <Inge-> what board chipset?
08:19:40 <sech1> But my old kit doesn't like high voltage, so I have to run it at 1.4V and lower speed
08:19:48 <sech1> B450M-K
08:20:23 <Inge-> right. B-die should be happy with a good deal of voltage. But haven't tried my Falre-X over 1.4 either
08:20:28 <Inge-> Flare-x
08:21:25 <sech1> Yes, but my old kit seems to be highly temperature sensitive. It errors out at 1.5V even if I run at 3200 MHz with stable timings that worked at 1.4V
08:23:48 <Inge-> hm maybe I should try pushing my 3600MHz bdie more
08:24:16 <Inge-> Not sure why the per-core h/s falls so much when I'm using more cores - temps are under control, clock seems stable
18:49:15 <gingeropolous> total miners: 44591