BearHand

BearHand

Antnode offloader reduced cpu load by up to 80% especially on non-ghash accelerated cpu or simply turn it off. Using matrix instead of just scalar with GCC15.1(bleed), ~20x performance increased. Developing new CE for offload similar antnode tasks under pcie2.0 protocol, under $10/board 2242 - abandon
Updated: no longer works after 0.4.1, abandoned project & account
Unwilling to fix already fixed bugs and issue, asking for binary 0.4.0 for testing welcome