"Huge Pages, also known as Large Pages (on Windows) and Super Pages (on BSD or macOS) is very important thing for almost all supported CPU mineable algorithms, typical hashrate increase is 20-30% when huge pages used, for RandomX it can increase performance up to 50%" from https://xmrig.com/docs/miner/hugepages
On Linux, it is controlled by the kernel setting vm.nr_hugepages .
We should attempt to use huge pages, if we can't, then fallback to not using huge pages. Currently huge page is disabled. We need to research what an optimal value for vm.nr_hugepages is for our flavor of RandomX. For XMR, vm.nr_hugepages=1300 or vm.nr_hugepages=1280 is recommended, but that seems to be too small for us and doesn't work.
"Huge Pages, also known as Large Pages (on Windows) and Super Pages (on BSD or macOS) is very important thing for almost all supported CPU mineable algorithms, typical hashrate increase is 20-30% when huge pages used, for RandomX it can increase performance up to 50%" from https://xmrig.com/docs/miner/hugepages
On Linux, it is controlled by the kernel setting `vm.nr_hugepages` .
We should attempt to use huge pages, if we can't, then fallback to not using huge pages. Currently huge page is disabled. We need to research what an optimal value for `vm.nr_hugepages` is for our flavor of RandomX. For XMR, `vm.nr_hugepages=1300` or `vm.nr_hugepages=1280` is recommended, but that seems to be too small for us and doesn't work.
"Huge Pages, also known as Large Pages (on Windows) and Super Pages (on BSD or macOS) is very important thing for almost all supported CPU mineable algorithms, typical hashrate increase is 20-30% when huge pages used, for RandomX it can increase performance up to 50%" from https://xmrig.com/docs/miner/hugepages
On Linux, it is controlled by the kernel setting
vm.nr_hugepages.We should attempt to use huge pages, if we can't, then fallback to not using huge pages. Currently huge page is disabled. We need to research what an optimal value for
vm.nr_hugepagesis for our flavor of RandomX. For XMR,vm.nr_hugepages=1300orvm.nr_hugepages=1280is recommended, but that seems to be too small for us and doesn't work.Latest code attempts to use large/huge pages and falls back to not using it if it fails. Closing