I've been digging for quite some time and always hit a brick wall, when I try to estimate the FLOPs of a MIPS64 CPU series, that I'm evaluating for an embedded design. Moreover I can't seem to find how many floating point operations this CPU can do per clock, per core, which is really frustrating, because otherwise I would've been able to calculate it myself. All I can find for any MIPS32/64 CPU is the DMIPS, which doesn't help me much, because I want to compare it towards other processors, most of which haven't run dhrystone and even if they did, I'm not entirely sure it'd make any sense to scale the one result towards the other.
Can anyone shed some light over how MIPS cpu cores handle float operations and in what amount of cycles? The MIPS64 CPU I'm looking at is quad-issue quad-threaded one, should the instruction issue tips the scale either way.
Thanks in advance!
Cheers,
vlex