AMD Ryzen 9 7950X with AVX-512 shows pretty good performance on Geekbench 5

Later this month, Ryzen 7000 will be made available, and benchmarks of these official Ryzen 7000 chips are already appearing online. However, the new Geekbench scores exaggerate Ryzen 7000’s average performance estimates over Zen 3 chips to levels that seem ludicrous in light of the advent of AVX512 on Zen 4.

A new Ryzen 9 7950X Geekbench 5 score posted by @Benchleaks on Twitter demonstrates very substantial improvements in the Zen 4 architecture over Zen 3. A 16-core Ryzen 9 7950X Zen 4 processor with a maximum reported clock of 5.738GHz, 32GB of 6000MT/s DDR5 RAM, and an Asus ROG Crosshair X670E Extreme motherboard made up the test bench for this new Geekbench 5 score.

The test results for the 5950X reveal a significantly lower multi-core score of 17,069 points and a significantly lower single-core score of 1,725. According to these outcomes, the 7950X performs 29% better in the single-core test than the 5950X and 43% better in the multi-core test.

These results appear to be excellent for the 7950X, however, they do not fully convey the situation. To do that, we must examine the various workloads that Geekbench 5 employs as a benchmark. Integer, floating-point, and AVX-intensive cryptographic operations are among the three types of workloads divided among 21 benchmarks by Geekbench 5.

While the floating point and integer results are excellent on their own, the Cryptograph benchmark results, which are about 71% better than the 5950X results in single-core performance, are where Ryzen 7000’s biggest performance benefits are found.

The AVX-512 instruction set, which the Ryzen 5000 lacks, is added in the Ryzen 7000, which is the cause of this surprise speed increase.

One of the more recent instruction sets found on contemporary CPUs is AVX-512, which can significantly improve speed in apps that use it. The issue with AVX-512 is that it has a very low adoption rate by today’s standards. Even though AVX-512 has been around for more than five years, only a small percentage of power users and content creators are able to utilise its features.

In contrast, the most prevalent workloads on processors today are integer and floating-point workloads. Integer instructions or floating-point computations are used in almost everything, including gaming, multitasking, production, and other activities.

In other words, it’s possible that the Geekbench 5 performance results for the Ryzen 7000 are slightly exaggerated. Even though they are amazing to behold, only a small percentage of Ryzen 7000 users will actually be impacted by the crypto scores. Additionally, Geekbench 5 has a history of weighting average scores to prioritise particular subtests over others.

Unfortunately, this has been a problem with various subtests for a while; the Core i9-Geekbench 11900’s results specifically showed this problem. However, these significant performance differences between workloads have grown much more pronounced with the advent of AVX-512 on later processors, particularly Intel’s 11th Gen Rocket Lake platform, making broad performance assessments with some synthetic benchmarks challenging.

Also Read:

• NVIDIA GeForce RTX 4070 GPU is rumoured to come with 7680 Cores and 12GB of GDDR6X Memory

Source