Meet Intel(R) Xeon(R) Gold


Limp Gawd
Jul 2, 2016
Meet Skylake-EP - Intel(R) Xeon(R) Gold

2x Intel(R) Xeon(R) Gold 6150 CPU @ 2.70GHz (18C 36T 3.7GHz, 2.4GHz IMC, 18x 1MB L2, 24.75MB L3)

- Processor Multi-Media: 4040.67Mpix/s
- Processor Multi-Core Efficiency: 168.96GB/s

New cache structure in place. This is a 18C/36T 2.7 GHz (3.7 GHz Turbo) model in 2P configuration.

Twice Broadwell-EP's score (AVX-512 being used):

2x Intel(R) Xeon(R) CPU E5-2697 v4 @ 2.30GHz (18C 36T 3.59GHz, 3.6GHz IMC, 18x 256kB L2, 45MB L3)
- Processor Multi- Media: 2099.19Mpix/s 72T
Holy crap its fast! Really interesting to see what the 1MB L2 cache does when it gets more isolated in tests so we can compare with the 256KB consumer.
I'll reserve judgement until I see TDPs, feature set (TBM Max 3.0?), and all core turbo speeds. I'll bet that the benchmark would have been a fair bit lower had it been run repeatedly, as the short term power limit of the chip would have kicked in and doubtlessly would have throttled it's clock speed considerably. The new nameplate of Xeon "Gold" is quite intriguing and offers some hope that Intel has improved their product lines for Skylake-EP/W in terms of features, raw performance and higher TDPs.
Geekbench 4 numbers for Skylake-EP.

Skylake-EP 58 cores, 112 threads 1.8Ghz.


Compared to Naples 64 cores, 128 threads 1.45Ghz.

64 cores, 128 threads gave AMD 15620 MT (1141 ST)
56 cores, 112 threads gives Intel 37556-38983 MT (3323-3678 ST)
Compared to Naples 64 cores, 128 threads 1.45Ghz.
There is every bit in the book to suspect the Naples result was at 1Ghz or whatever other mean can reduce performance by about 40%, man. Including, for example, the last geekbench 3 against the first geekbench 3, that did leak.

I have to notice that single core turbo on that 28C Skylake-EP is about 3.5Ghz, by the way. Oh, and that cache change does not do any noticeable work.