6 ms·
I’m more impressed by Xeon Gold vs EPYC perf. The fact that Power 9 is currently slower on workloads that heavily benefit from hand-optimized code specific to I
by throwaway84742 9y ago
I’m more impressed by Xeon Gold vs EPYC perf. The fact that Power 9 is currently slower on workloads that heavily benefit from hand-optimized code specific to Intel CPUs should not really surprise anybody. It’s likely that this code just uses a “portable”, slow implementation on Power9, just because nobody has the systems to work with yet. Additionally it’s also likely that most software in these tests can’t take advantage of such a huge number of threads.
- ianai 9y agoIs there anyway to price out a power9 itself? Just for comparison.
- benchaney 9y agoDo you mean just the processor? RaptorCS sells Processors + main board bundles.
- asfgionio 9y agoPOWER doesn't have that many more threads. The top-of-the-line is 22-core 88-thread, compared to 32-core 64-thread from AMD and 28-core 56-thread from Intel. I have no idea what having that many threads per core means for performance.
- sp332 9y agoOffhand, I would guess that it's good for I/O-bound tasks, where you can have lots of threads waiting for input that don't need CPU time. A busy database maybe.
- lultimouomo 9y agoYou don't need CPU threads for threads that don't need CPU...
- sp332 9y agoThe 8-core version only has 8 (each) ALUs, LSUs, and vector units. https://en.wikipedia.org/wiki/POWER9#Core https://en.wikipedia.org/wiki/POWER9#Core If each core has 4 threads "running" on it, some of them are not going to be executing.
- Joky 9y agoStill: the kernel won’t schedule threads that are waiting on I/O i believe.
- sp332 9y agoOk that's an interesting point. So they'd have to be waiting on a fetch instruction for the time not to be totally wasted?
- blattimwind 9y agoBlocked tasks are not scheduled until unblocked.
- ajross 9y agoExtra threads means that there's more likely to be something else available to schedule when one thread's pipelines are stalled due to a fetch from DRAM or MMIO, basically. There's very much a diminishing returns beyond two, but for some workloads (big mostly-RAM-resident data sets with poor locality of reference -- consumer databases say) it's worth it. It's unclear to me that 4-way multhreading is going to help any of the benchmarks in this test.
- Symmetry 9y agoIn the specific case of POWER 8 and 9 the cores are seriously overprevisioned with execution resources and you really need at least 2 threads running in order to make full use of them.
- ajross 9y agoCould be. But the point was more that there's a very thin regime between "waiting for DRAM latency too often" (where more threads can help) and "bound by DRAM bandwidth" (where they won't). The DRAM isn't nearly as parallel as the cores are and saturates really fast.
- gnufx 9y agoKnights Corner was another case where you normally needed multiple threads.
- rbanffy 9y agoIn that case it was even more serious. Like the PPUs of the Cell processors, each core runs one instruction every other cycle. If you have only one thread, you effectively have half the throughput.
- rbanffy 9y agoAll diminishing returns really depend on your workload. We normally would say that for caches, with no point having enormous caches on desktop and typical x86 server processors, but IBM's mainframe CPUs have tons of L3 and even more tons of L4 caches, as well as dedicated cores (in the form of secondary processors) for offloading all kinds of tasks from the main CPUs, each with its own cache architecture.
- zrm 9y ago> I’m more impressed by Xeon Gold vs EPYC perf. The hardware used in this benchmark isn't very well chosen. All the systems have a different number of cores (Power 16, AMD 32, Intel 40), and the AMD system has the additional disadvantage over the others of being 1P instead of 2P, which approximately halves the available cooling and correspondingly limits clock speeds. The results aren't very surprising given that. Making comparisons like this can make sense when one of the competitors doesn't make an equivalent product, but they all do make products with an equivalent number of sockets and cores.
- mping 9y agoI believe the fair comparison would be performance per dollar - comparing similarly priced components. Same sockets and cores can have wildly different costs.
- zrm 9y agoComparing like with like can give you a better idea of what's going on even if the prices aren't the same. It's a matter of isolating the architecture from the other factors. If a 40 core 2P Intel system is faster than a 32 core 1P AMD system, you don't know if the difference is architectural or just the extra cores and sockets, so you don't know whether to expect a 48 core 2P AMD system to be faster or slower than the Intel system. And the tested systems aren't equivalently priced here either.
- inasio 9y agoPer dollar of CPU, CPU + power, CPU + motherboard, TCO?
- rch 9y agoI still need to see raw performance: single core, multi-core single socket, and multi-socket.
- rbanffy 9y agoIt depends on what you want to compare. If it's performance at a certain price point, you select systems at that point. You can compare on performance per watt or per volume or more abstract comparisons (ideally unbound by price or size) of performance per core, socket or thread at various load factors.