3 ms·
I wonder if there were NUMA effects at play since the Intel server was dual-socket and the EPYC was not. It's not entirely surprising that a CPU-heavy workload
by davidmr 8y ago
I wonder if there were NUMA effects at play since the Intel server was dual-socket and the EPYC was not. It's not entirely surprising that a CPU-heavy workload reading from filesystem cache (i.e. memory) could cause a substantial amount of nonlocal reads that the single-socket server wouldn't have to deal with.
That said, that's just the first thing that popped into my head to explain the result. Either way, the results here are extremely encouraging!
- wmf 8y agoA "single" EPYC is eight 4-core processors so there are even more NUMA effects there.
- h4b4n3r0 8y agoI think it’s the other way around: four 8-core processors
- blattimwind 8y agoIt's four dice with two groups of cores of four each connected by a callosum. In NUMA-speak, each group is (probably) a node with neighbouring nodes having a lower distance.
- davidmr 8y agoWe're now well out of my knowledge base, but your comment certainly makes things interesting. If they have similar NUMA latencies (big if!), I guess maybe the edge would go to whichever chip had higher NUMA bandwidth? Just sort of thinking out loud... Or perhaps it has absolutely nothing to do with memory speeds, but I can't think of what else would have such a dramatic effect. I'd love to see the same test with a single Xeon.