AVX-512 Makes Ryzen 9 7950X Geekbench 5 Results Look Good — Too Good
Ryzen 7000 is launching later this month and we are getting to see official Ryzen 7000 chip benchmarks in action. But with the introduction of AVX512 in Zen 4, these new benchmarks have inflated his average Ryzen 7000 performance estimates against Zen 3 chips to ridiculous levels.
@benchleaks I shared something new on Twitter Ryzen 9 7950X Geekbench 5 Score This shows a very impressive advantage of the Zen 4 architecture over Zen 3.
The test bench for this new Geekbench 5 result features a Ryzen 9 7950X 16-core Zen 4 chip with a reported maximum frequency of 57.38 GHz, 32 GB of 6000MT/s DDR5 RAM, and an Asus ROG Crosshair X670E Extreme motherboard. I was.
The test results show a staggering 2,217 points for the single-core score and 24,396 points for the multi-core score.
For comparison, Recent Geekbench 5 Benchmarks Ryzen 9 5950X 16-core Zen 3 processor (running on a B550 Aorus Elite motherboard with 32GB of 3600MT/s DDR4 memory). The 5950X’s test results show a low single-core score of 1,725 points, dramatically 17,069 point multicore score drop. Based on these results, the 7950X beats his 5950X by 29% in single-core tests and 43% in multi-core tests.
These scores look great for the 7950X, but they don’t tell the whole story. To do that, we need to look at the different workloads that Geekbench 5 uses in its benchmarks. Geekbench 5 splits three types of workloads across 21 benchmarks, including integer, floating point, and AVX-heavy crypto workloads.
The story becomes clearer when we see how both chips perform in each of these different workload categories.
Workload – single thread | Risen 9 5950X | Ryzen 9 7950X | |
integer | 1467 | 1806 | 23.1% |
floating point | 1874 | 2288 | twenty two% |
Crypto | 4188 | 7140 | 70.5% |
Workload – Multithreaded | |||
integer | 16735 | 24273 | 45% |
floating point | 19304 | 26652 | 38% |
Crypto | 8006 | 12465 | 55.7% |
While the floating point and integer results are impressive on their own, the Ryzen 7000’s biggest performance boost comes from the chip’s significantly higher crypto benchmark results. That’s almost 71% higher than his 5950X result in single-core performance.
This unexpected performance leap can be attributed to Ryzen 7000 adding the AVX-512 instruction set that Ryzen 5000 lacks. AVX-512 is one of the newer instruction sets found in modern processors and can significantly improve performance for apps that support it.
The problem with AVX-512 is its very low adoption rate by current standards. Despite being available for over 5 years, few apps currently support it. As a result, only a few power users and content creators can use the AVX-512’s features.
In contrast, integer and floating point workloads are the most common workloads on today’s processors. Gaming, multitasking, production, and just about anything else uses some form of integer instructions or floating point calculations.
In short, the Ryzen 7000’s Geekbench 5 performance results are misleading for the majority of consumers. The crypto score is incredible to look at but realistically he only affects a small percentage of Ryzen 7000 users plus this plus Geekbench 5 puts that average score into his AVX specific benchmarks are known to place more emphasis on
Unfortunately, this is an ongoing problem with synthetic benchmarks. But with the introduction of his AVX-512 in newer processors, including Intel’s 11th Gen Rocket Lake platform, these wide-ranging performance deviations between workloads have become much more pronounced, making overall benchmarks generally making accurate performance estimates nearly impossible.
That said, the Ryzen 7000’s real-world performance is very good in Geekbench 5, with integer and floating point scores far superior to the Ryzen 5000. But when you add AVX-512 and run AVX-512 intensive benchmarks, the Ryzen 7000’s performance looks ridiculous. This is good, but it’s not. that good.