the problem is that pools won't implement it - they simply don't care about anything but their own profits

it was the same with self-select

this is madness: sinovate.io/x25x.pdf

let's call RandomX "X10^512R"

no, 2^512, whatever

X13407807929942597099574024998205846127479365820592393377723561443721764030073546976801874298166903427690031858186486050853753882811946569946433649006084096R

what a bunch of morons

chip densities continue to double every 18 months

which means that if you want to keep adding N algorithms to make it too expensive to fabricate, you have to keep doubling the number of algorithms

considering our price has basically doubled in the past few months, surprised hashrate hasn't increased further

profitability should be way up right now

hyc bitinfocharts.com/comparison/monero-mining_profitability.html#3m

it's at the same level as it was on December 10th

and network was at 875 MH/s on December 10th

its gonna get interesting when moon happens

maybe in my free time i'll make a thing to launch a pool from scratch

i've always gotten like halfway with Snipa's scripts before something distracts me

but i imagine when moon happens they'll be a lot of AWS / any cloud stuff

be best if that kind of hashrate gets on its own pool

you used to be able to share AWS images

gingeropolous: it's probably best to use something like ansible to provision and update everything

yeah see thats where im like damnit i don't wanna figure out another thing with a l33t name

but if u know how......

jtgrassie: are you mining alt chain on monerop.com?

mainnet alt chain*

I've finished JIT implementation for Navi and it's very memory bottlenecked

I could reduce core clock from 1750 to 900 MHz and hashrate only dropped from 549 to 522 h/s

btw it's 522 h/s @ 55 W

5500 XT running at 900/1860 MHz (core/mem)

I can't change voltage, I think it can be tuned more to reduce power

you need to transplant the HBM from your vega :P

the performance is similar to what RX550 used to do with CN

5500 XT is still 25-30% better than RX 570 8GB at the same bandwidth/tflops

so Navi is a better architecture for RandomX

I think much better because GPU core can handle this hashrate at only 900 MHz

but memory is a bottleneck

if only Navi had big instruction/data caches...

Big Navi :P

cohcho: yeah looks like something up with monero daemon. it can't find any peers

jtgrassie: monerop.com is your pool?

yes

but the daemon it's using isn't syncing right now

height looks OK to me

yeah it does now, i unwedged the daemon

10H/s/W

not earth shattering