arm "neon"

Torbjorn Granlund tg at
Sat Feb 23 18:26:55 CET 2013

Richard Henderson <rth at> writes:

  On 2013-02-23 06:06, Niels Möller wrote:
  > Not sure what the bottlenecks of your loop are though; instruction
  > decoding, load/store, or the recurrency chain (but at least it shouldn't
  > be multiplier throughput, right?).
  Yeah, neither am I.  I can't find any info on what latency of neon
  insns should be.
I always found experientation to work best for that, using a tiny loop.


More information about the gmp-devel mailing list