CNv2 and General: SQRL Miner loads bitstreams onto the Acorn and will get updated to handle various things. The Acorn in particular works on a part of CN that isn’t much affected by the fork.
Nevertheless they are claiming all FPGAs will be 4x slower on CNv2. Is this true for the Acorns?
That 4x slower statement is only true when an FPGA is running the whole algo.
Acorns assist GPU mining by taking over the parts that GPUs are less capable at.
So its not definite one way or the other in the case of Acorn acceleration,
IF there is any part of CNv2 that can be done faster on an FPGA, then an Acorn can improve the GPU miner.
So far from what I've read about CNv2 it has parts in the algo that ASIC or FPGA can not process as well as a CPU or GPU.
That will make implementing CNv2 on ASIC or FPGA not viable, economically.
Or at the very least no massive performance gain over a GPU, levelling the field for GPU miners.
That still leaves the possibility that GPU/FPGA combo might have an advantage over any single technology implementation.
How far it is possible to exclude FPGA from mining is unknown at the moment, I think the competition has only just begun.
Baz
