14:07:48
hyc:
I wonder if we should look into making a unikernel build of this so people can boot up their own dedicated mining ASIC ... :P
14:31:30
gingeropolous:monero.social:
u think we could squeeze more h/s/w from a rig?
14:51:03
hyc:
maybe a tiny bit. mostly because there would be no other code running besides the kernel and the miner. no other processes on the box.
14:53:06
hyc:
we could make things totally turnkey. add a broadcast listener to p2pool, have the miner broadcast to find its p2pool server. just plug it in and it works.
14:53:59
DataHoarder:
like good old game servers :D
14:54:00
hyc:
distribute it on bootable USB flash or microSD cards.
14:54:46
DataHoarder:
it can even be p2pool agnostic, the broadcast listener can just be something you run on the network to send config
14:54:57
DataHoarder:
sends stratum url and args to queriers
14:55:15
DataHoarder:
target can be anything
14:55:48
hyc:
it could ... but there's no reason to support anything besides p2pool ;)
15:00:25
DataHoarder:
sweep fees might grow considerably past FCMP++ btw, so I'm trying to get some aggregation ongoing via multisig groups on the fly (with fallback txs ahead of time)
15:08:15
gingeropolous:monero.social:
id want to support local daemon mining. i don't use p2pool because of the dust. would rather just accept the variance and have larger outputs
15:09:42
hyc:
fair enough. monerod itself could answer that broadcast query
15:10:15
DataHoarder:
tbh, mining on main gets you way less "dust" as payouts unless you have high hashrate
15:10:32
DataHoarder:
compared to mini/nano etc.
15:10:35
hyc:
we should write an IETF RFC spec for this Dynamic Miner Configurartion Protocol :P
15:10:58
DataHoarder:
just do UDP broadcast on port 3333
15:11:21
hyc:
the spec part is defining the message format
15:11:30
DataHoarder:
I am suggesting the port
15:11:54
DataHoarder:
let's maybe first find if this "turnkey" solution is that much efficient :P
15:12:10
hyc:
sure. the first attempt can use a hardcoded destination.
15:12:13
gingeropolous:monero.social:
nonsense. cart -> horse
15:12:20
hyc:
LOL
15:13:25
DataHoarder:
at least we have the road defined
15:14:22
gingeropolous:monero.social:
well it could have value for ppl launching massive mining infra. i guess they could just clone images onto drives as is, or pixy boot or whatever
15:14:58
hyc:
sure. but people are doing MSR tweaks and other such stuff for tiney fraction of percent improvements
15:15:19
hyc:
maybe there's a fraction of a percent to be found here too
16:08:49
sech1:
hyc so basically ragerx for p2pool :D
16:09:02
sech1:
the one big problem with that is the lack of drivers
16:09:15
sech1:
no driver for GPU = GPU will be burning energy instead of going to the lowest energy state
16:09:45
sech1:
So need to package at least the latest AMD/NVIDIA drivers to put the GPU to idle
16:11:39
sech1:
hyc I'm still waiting for the review https://github.com/tevador/RandomX/pull/316
16:12:09
sech1:
RandomX v2 is almost done, only more documentation and PPC64 (big endian) intrinsics are left
16:55:49
plowsof:
👏
18:22:28
gingeropolous:monero.social:
the ipc from zen2 to zen4 is no joke. my monerosim run goes from 9 minutes to 3 minutes.
20:22:26
sech1:
Did a writeup on RandomX v2 changes: https://github.com/SChernykh/RandomX/blob/v2/doc/design_v2.md
20:29:52
DataHoarder:
might indeed be good to get more datapoints there to show the impact on quoted Ryzen 9 3950X
20:35:14
sech1:
Need 3950X test, but I don't have it
20:35:26
sech1:
I took the numbers from XMRig's benchmark page
20:43:34
gingeropolous:monero.social:
i got one.. where are these tests...
20:51:13
gingeropolous:monero.social:
i got em. tryna compile
20:53:50
gingeropolous:monero.social:
oh jesbut. this box is currently a space heater in a far away room thats on a bullshit wifi extender so apt upgrade is gonna be slow
21:52:48
sech1:
You will need to compile the v2 branch from my repo, then run randomx-benchmark with the same command line I used for 9950X. But first run randomx_boost.sh from XMRig scripts, and enable huge pages
21:53:26
sech1:
oh, and the command lines are here https://gist.github.com/SChernykh/6058ecf01c929883b9d19c7eeadc8809
21:54:11
sech1:
https://github.com/xmrig/xmrig/blob/master/scripts/randomx_boost.sh
23:06:38
gingeropolous:monero.social:
of course v2 builds but not v1: https://paste.centos.org/view/1ad72230
23:11:26
DataHoarder:
import via "#include <cstdint.h>" afaik
23:39:43
sech1:
what do you mean not v1? The latest code has everything, you can switch between v1 and v2 with command line
23:40:03
sech1:
old code might have some unfixed issues of course
23:50:32
eureka:
sech1: there is ppc64be JIT now?
23:51:25
sech1:
No
23:51:45
sech1:
But the interpreter mode might not work if the new intrinsics are wrong
23:55:21
eureka:
hyc back in cnv0 days I used to have a stripped down kernel with bare minimums needed for storage+ethernet, initrd just brought up network, allocated hugepages, started my custom static xmr-stak build (built w/ icc, hot funcs had runtime dispatch for various uarch from westmere - skylake)
23:55:29
eureka:
nice easy boot and mine setup
23:55:53
eureka:
would be pretty fun to do the same for randomx
23:56:09
eureka:
sech1 I have access to ppc64 host, let me see if I can build and run tests
23:59:33
eureka:
or well, my account appears to be broken on IdP, can't log in right now but I will get that fixed and then try builds