LinearTurboFold still runs strictly single-threaded, so even though the algorithm is O(n) it can’t fully use modern multi-core
machines and ends up slower than tools like mlocarna that parallelize their heavy loops. Could you add an optional multi-thread
implementation so users can specify the number of threads and get better throughput? Thanks!
LinearTurboFold still runs strictly single-threaded, so even though the algorithm is O(n) it can’t fully use modern multi-core
machines and ends up slower than tools like mlocarna that parallelize their heavy loops. Could you add an optional multi-thread
implementation so users can specify the number of threads and get better throughput? Thanks!