Solar Designer

Right. So should have few _sequential_ MULs (a code generator constraint) and/or clamp to short range of initial exponent, which I think isn't that bad if our alternative is building...

So I experimented with FP in a ProgPoW hack a little bit. I am currently leaning towards the following design aspects _if_ we actually switch to FP: - Instead of...

JFYI: Staying with integers but maximizing the ratio of arbitrary 32-bit MULs, I am able to get up to ~630G MUL/s on the Vega 64 while also doing the cache...

@olalawal It is unhealthy to the project if/when we have to self-censor our sharing of potential issues out of concern that these would be looked at from a political ([inadvertently]...

One way to look at the attack I described is that it doesn't let attackers eliminate memory-hard computation (unlike #51), but it gives miners the (unintended) flexibility to partially replace...

Here's an extension of the attack: During initial cache filling, also use the Keccak ASICs to search nonces such that e.g. upper 20 bits of seed would be all 0's....

[This comment previously included corrections to the initial revision of my comment above, but I've since edited the comment above to reflect my latest understanding. I am not deleting this...

Thank you for reporting this @gnd. My main two guesses as to what could be causing this are: 1. Too frequent kernel integrity verification, which LKRG by default does not...

I think Adam meant not the number of those processes, but whether they're the ones actively running on CPU (e.g. per `top`) during the load spikes. Anyway, you show that...

Thanks @gnd. This is puzzling. We really need the `WCHAN` field to hopefully figure it out. I don't know why exactly it is empty for you, but perhaps you need...