13:30:24 <m-relay> <a​ck-j:matrix.org> https://www.thucloud.com/zhenhua/papers/[ASE'25]%20Democratizing%20Cryptocurrency%20Ecosystem.pdf
13:30:34 <m-relay> <a​ck-j:matrix.org> Democratizing the Cryptocurrency Ecosystem by
13:30:35 <m-relay> <a​ck-j:matrix.org> Just-In-Time Transformation of Mining Programs
14:06:28 <DataHoarder> "Aik to revive web crypto mining to democratize it"
15:42:33 <sech1> Interesting. Merging similar RandomX instructions into a single SIMD instruction is similar to what I did in RandomX CUDA code to utilize more parallelism of the GPU.
15:44:13 <sech1> https://github.com/SChernykh/RandomX_CUDA/blob/master/RandomX_CUDA/randomx_cuda.hpp#L424
15:44:43 <sech1> init_vm function, and the "execution_plan" table in it.
16:26:08 <sech1> Wow, 1395 h/s on pixel 10 pro xl https://p2pool.io/u/1b3b868bb14e9052/Screenshot_20251127-172057.png