SPEC - Single-Threaded Performance

Single-thread performance of server CPUs usually isn’t the most important metric for most scale-out workloads, but there are use-cases such as EDA tools which are pretty much single-thread performance bound.

Power envelopes here usually don’t matter, and what is actually the performance factor that comes at play here is simply the boost clocks of the CPUs as well as the IPC improvement, and memory latency of the cores. 

The one hiccup for the Xeon 8380 this generation is the fact that although there’s plenty of IPC gains to be had compared to previous microarchitectures, the new SKU is only boosting up to 3.4GHz, whereas the 8280 was able to boost up to 4GHz, which is a 15% deficit.

SPECint2017 Rate-1 Estimated Scores

Even with the clock frequency disadvantage, thanks to the IPC gains, much improved memory bandwidth, as well as the much larger L3 cache, the new Ice Lake part to most of the time beat the Cascade Lake part, with only a couple of compute-bound core workloads where it falls behind.

SPECfp2017 Rate-1 Estimated Scores

The floating-point figures are more favourable to the ICX architecture due to the stronger memory performance.

SPEC2017 Rate-1 Estimated Total

Overall, the new Xeon 8380 at least manages to post slight single-threaded performance increases this generation, with larger gains in memory-bound workloads. The 8380 is essentially on par with AMD’s 7763, and loses out to the higher frequency optimised parts.

Intel has a few SKUs which offers slightly higher ST boost clocks of up to 3.7GHz – 300Mhz / 8.8% higher than the 8380, however that part is only 8-core and features only 18MB of cache. Other SKUS offer 3.5-3.6GHz boosts, but again less cache. So while the ST figures here could improve a bit on those parts, it’s unlikely to be significant.

SPEC - Multi-Threaded Performance SPEC - Per-Core Performance under Load
POST A COMMENT

170 Comments

View All Comments

  • Hifihedgehog - Tuesday, April 6, 2021 - link

    Correction: Maybe if it was released over two years ago... Reply
  • amootpoint - Wednesday, April 7, 2021 - link

    Not really. Just look at the perf numbers vs AMD. Intel is broadly winning. Reply
  • Bagheera - Thursday, April 8, 2021 - link

    which review article did you read? didn't sound like you read the same one I did. Reply
  • Unashamed_unoriginal_username_x86 - Tuesday, April 6, 2021 - link

    the 8280 core to core bounce latency image is 404ing Reply
  • Ryan Smith - Tuesday, April 6, 2021 - link

    This has been fixed. Thanks! Reply
  • JayNor - Tuesday, April 6, 2021 - link

    News release says "The platform supports up to 6 terabytes of system memory per socket".
    This story says 4TB. Is Intel wrong?
    Reply
  • ParalLOL - Tuesday, April 6, 2021 - link

    Depends on whether you agree to say that Optane is "system memory". It is mapped onto address space and is directly attached to CPU so it probably can be said to be system memory. Reply
  • DigitalFreak - Tuesday, April 6, 2021 - link

    It's like when Intel says their desktop processors have 40 PCIe lanes. It's actually 16 CPU and 32 chipset. Well 20 CPU with Rocket Lake, but still... Reply
  • brucethemoose - Tuesday, April 6, 2021 - link

    Y'all are measuring power from the socket sensors, not the wall, right?

    I think the later would be more interesting, even with the apples-to-oranges factor of different hardware surrounding the platform. After all, whole system power consumption is what one gets charged for.
    Reply
  • Jorgp2 - Tuesday, April 6, 2021 - link

    That would vary massively between systems. Reply

Log in

Don't have an account? Sign up now