Intel 5th Gen Xeon "Emerald Rapids" AVX-512 Performance

Written by Michael Larabel in Processors on 5 January 2024 at 10:35 AM EST. Page 2 of 5. 18 Comments.
miniBUDE benchmark with settings of Implementation: OpenMP, Input Deck: BM2. Emerald Rapids: AVX-512 On was the fastest.
miniBUDE benchmark with settings of Implementation: OpenMP, Input Deck: BM2. Emerald Rapids: AVX-512 On was the fastest.

While AVX-512 started out years ago with much criticism over the power and thermal impact, the performance difference of leveraging AVX-512 can be very worthwhile especially on newer generation processors. AMD Zen 4 processors too have also proven to be worthwhile for AVX-512 with their approach.

miniBUDE benchmark with settings of Implementation: OpenMP, Input Deck: BM2. Emerald Rapids: AVX-512 On was the fastest.
miniBUDE benchmark with settings of Implementation: OpenMP, Input Deck: BM2. Emerald Rapids: AVX-512 On was the fastest.

When AVX-512 was being used for the miniBUDE HPC benchmark, the Xeon Platinum 8592+ didn't show much of a difference in the peak CPU frequency and CPU temperature... Much better than the Skylake days with very stark differences.

miniBUDE benchmark with settings of Implementation: OpenMP, Input Deck: BM2. Emerald Rapids: AVX-512 On was the fastest.
miniBUDE benchmark with settings of Implementation: OpenMP, Input Deck: BM2. Emerald Rapids: AVX-512 On was the fastest.

The CPU power consumption of the 2P setup wasn't also that different when engaging AVX512, which in turn meant a nice boost to overall power efficiency.

Embree benchmark with settings of Binary: Pathtracer ISPC, Model: Crown. Emerald Rapids: AVX-512 On was the fastest.
Embree benchmark with settings of Binary: Pathtracer ISPC, Model: Asian Dragon. Emerald Rapids: AVX-512 On was the fastest.
Embree benchmark with settings of Binary: Pathtracer ISPC, Model: Asian Dragon Obj. Emerald Rapids: AVX-512 On was the fastest.
Embree benchmark with settings of Binary: Pathtracer ISPC, Model: Asian Dragon Obj. Emerald Rapids: AVX-512 On was the fastest.
Embree benchmark with settings of Binary: Pathtracer ISPC, Model: Asian Dragon Obj. Emerald Rapids: AVX-512 On was the fastest.
OpenVKL benchmark with settings of Benchmark: vklBenchmarkCPU ISPC. Emerald Rapids: AVX-512 On was the fastest.

Intel's own creator software packages with their oneAPI software suite continue to be notable users of AVX-512 and performing very well when engaging that code path.

OpenVKL benchmark with settings of Benchmark: vklBenchmarkCPU ISPC. Emerald Rapids: AVX-512 On was the fastest.
OpenVKL benchmark with settings of Benchmark: vklBenchmarkCPU ISPC. Emerald Rapids: AVX-512 On was the fastest.
OpenVKL benchmark with settings of Benchmark: vklBenchmarkCPU ISPC. Emerald Rapids: AVX-512 On was the fastest.
OpenVKL benchmark with settings of Benchmark: vklBenchmarkCPU ISPC. Emerald Rapids: AVX-512 On was the fastest.

Running with AVX-512 utilized continued to not have the bad side effects as Intel Xeon generations from years ago.


Related Articles
Ampere Computing 2024 Roadmap Update: 256 Core 3nm CPU In 2025
Benchmarking The First RISC-V Cloud Server: Scaleway EM-RV1 Performance
AMD Announces Ryzen PRO 8840 & PRO 8000G Series CPUs
Intel Announces Gaudi 3 AI Accelerator, Intel Xeon 6 Brand
AMD Zen 4 vs. Zen 4C Performance, Zen 4C Core Scaling With Ryzen 5 8500G
Intel Core i3 14100 / i5 14500 vs. AMD Ryzen 5 8500G / 8600G In 500+ Benchmarks