SPEC CPU - Multi-Threaded Performance

Moving onto multi-threaded SPEC CPU 2017 results, these are the same workloads as on the single-threaded test (we purposefully avoid Speed variants of the workloads in ST tests). The key to performance here is not only microarchitecture or core count, but the overall power efficiency of the system and the levels of performance we can fit into the thermal envelope of the device we’re testing.

It’s to be noted that among the four chips I put into the graph, the i9-11980HK is the only one at a 45W TDP, while the AMD competition lands in at 35W, and the i7-1185G7 comes at a lower 28W. The test takes several hours of runtime (6 hours for this TGL-H SKU) and is under constant full load, so lower duration boost mechanisms don’t come into play here.

SPECint2017 Rate-N Estimated Scores

Generally as expected, the 8-core TGL-H chip leaves the 4-core TGL-U sibling in the dust, in many cases showcasing well over double the performance. The i9-11980HK also fares extremely well against the AMD competition in workloads which are more DRAM or cache heavy, however falls behind in other workloads which are more core-local and execution throughput bound. Generally that’d be a fair even battle argument between the designs, if it weren’t for the fact that the AMD systems are running at 23% lower TDPs.

SPECfp2017 Rate-N Estimated Scores

In the floating-point multi-threaded suite, we again see a similar competitive scenario where the TGL-H system battles against the best Cezanne and Renoir chips.

What’s rather odd here in the results is 503.bwaves_r and 549.fotonik_r which perform far below the numbers which we were able to measure on the TGL-U system. I think what’s happening here is that we’re hitting DRAM memory-level parallelism limits, with the smaller TGL-U system and its 8x16b LPDDR4 channel memory configuration allowing for more parallel transactions as the 2x64b DDR4 channels on the TGL-H system.

SPEC2017 Rate-N Estimated Total

In terms of the overall performance, the 45W 11980HK actually ends up losing to AMD’s Ryzen 5980HS even with 10W more TDP headroom, at least in the integer suite.

We also had initially run the suite in 65W mode, the results here aren’t very good at all, especially when comparing it to the 45W results. For +40-44% TDP, the i9-11980HK in Intel’s reference laptop only performs +9.4% better. It’s likely here that this is due to the aforementioned heavy thermal throttling the system has to fall to, with long periods of time at 35W state, which pulls down the performance well below the expected figures. I have to be explicit here that these 65W results are not representative of the full real 65W performance capabilities of the 11980HK – just that of this particular thermal solution within this Intel reference design.

SPEC CPU - Single-Threaded Performance CPU Tests: Office and Science
POST A COMMENT

229 Comments

View All Comments

  • Alistair - Monday, May 17, 2021 - link

    no, that's not right, Tiger Lake is mainly clock speed improvements, not IPC, running Tiger Lake or Comet Lake in a desktop at the same 4.8GHZ all core would get you almost exactly the same performance Reply
  • heickelrrx - Monday, May 17, 2021 - link

    are u drunk? Reply
  • laduran - Tuesday, May 18, 2021 - link

    No Reply
  • mode_13h - Tuesday, May 18, 2021 - link

    > Tiger Lake is mainly clock speed improvements, not IPC, running Tiger Lake or Comet Lake
    > in a desktop at the same 4.8GHZ all core would get you almost exactly the same performance

    You're confusing Comet Lake with Ice Lake. IPC of Tiger Lake isn't much improved above that of Ice Lake.
    Reply
  • Qasar - Wednesday, May 19, 2021 - link

    can you blame him ? its time intel started using different names for their cpus. Reply
  • mode_13h - Thursday, May 20, 2021 - link

    I didn't mean that in a pointed way. If anything, I thought it would partially validate the statement.

    And yes, I wish Intel would've left the Lakes behind with 14 nm, but I guess there are just too many "lake" names for them to part with it for the mere sake of naming consistency. More annoyingly, they're even using "lake" names for some things that aren't CPUs.
    Reply
  • Qasar - Thursday, May 20, 2021 - link

    it seems they are staying with lake and cove names so unless you are using a slide ruler and a decoder ring, you have no idea which cpu is what :-) i gave up trying to keep track of their names are which cpu line a while ago. Reply
  • mode_13h - Friday, May 21, 2021 - link

    > it seems they are staying with lake and cove names

    Sapphire Rapids is the next server CPU. Also, the little cores seem to be Monts, while the big cores are Coves.

    I think the biggest groaner is "Lakefield", which is that mobile CPU with one big core, 4 little cores, and a chunk of eDRAM. MS used it in a Surface model.
    Reply
  • Qasar - Friday, May 21, 2021 - link

    mode_13h, still doesnt change the fact that the names intel gives its cpus, are confusing has he double hockey sticks :-) Reply
  • mode_13h - Sunday, May 23, 2021 - link

    Oh, for sure. It was so nice when *Bridge = LGA 1155, *Well = LGA 1150, and *Lake = LGA 1151.

    ...then, it seemed like nearly everything became a Lake!
    Reply

Log in

Don't have an account? Sign up now