Apple’s M2 Ultra is a powerful processor with 24 general purpose cores and up to 76 integrated GPU cores. It’s also probably faster than Intel’s 28-core Xeon W found in the 2019 Mac Pro, so owners of new Mac Pro workstations will see a significant performance boost. However, according to purported benchmarks, this CPU can’t beat AMD and Intel’s competitors. geek bench 5 (via @ Vadim Yuriev) because they either boast very high clocks or have a staggering number of cores.
in that an endurance workstation-grade processor should (like all client processors) provide both extremely responsive performance and consistently high performance even under heavy workloads. , unlike desktop and server CPUs. This means it must have high instruction performance per clock, high clocks, high core count, memory load support, and PCIe lane functional load. AMD’s Ryzen Threadripper Pro W5995X and Intel’s Xeon W9-3495X comply with these requirements with 64 and 56 cores respectively. In a nutshell, use server silicon configured to deliver very high clocks when needed.
Apple’s M2 Ultra consists of two M2 Max system-on-chips designed primarily for MacBook Pro and Mac Studio machines. These workstations have moderate power consumption and do not support scalability. The M2 Max wasn’t exactly designed with high clocks and expandability in mind, as compact PCs can only fit a lot of memory and storage. The M2 uses built-in dedicated accelerators rather than overclocking when compute-intensive workloads occur. It was also not designed for extreme core counts due to power and cooling limitations. The two M2 Max’s look mighty on paper, but they can’t boast the core counts of Intel’s Core i9-13900K or AMD’s Ryzen Threadripper Pro W5995WX.
| Header Cell – Column 0 | Header Cell – Column 1 | Zeon W9-3495X | Ryzen Threadripper Pro 5995WX | Core i9-13900K |
|---|---|---|---|---|
| General specifications | 24C/24T, up to 3.68GHz | 56C/112T, 1.90GHz to 4.60GHz, 105MB L3 | 64C/128T, 2.70GHz to 4.50GHz, 256MB L3 | 8P+16E/32T, 3.0 GHz to 5.80 GHz, 68MB L2+L3 cache |
| Single Core | Integer | 1793 | 1522 | 1316 | 2016 |
| Single Core | Floating | 2149 | 1815 | 1719 | 2464 |
| Single Core | Crypto | 2912 | 3926 | 3832 | 5860 |
| Single Core | Score | 1956 | 1730 | 1563 | 2343 |
| row 5 – cell 0 | row 5 – cell 1 | Row 5 – Cell 2 | Row 5 – Cell 3 | row 5 – cell 4 |
| multicore | integer | 24532 | 59183 | 46049 | 28379 |
| Multicore | Floating | 32195 | 55393 | 49414 | 31320 |
| Multicore | Cryptography | 46817 | 36466 | 44987 | 22280 |
| Multicore | Score | 27945 | 56910 | 47005 | 28956 |
| Link | https://browser.geekbench.com/v5/cpu/21305974 | https://browser.geekbench.com/v5/cpu/211458582 | https://browser.geekbench.com/v5/cpu/19923348 | https://browser.geekbench.com/v5/cpu/20655426 |
That’s why benchmark results show Apple’s M2 Ultra can’t beat Intel’s Core i9-13900K in single-threaded workloads, and even Geekbench 5’s multi-core workloads. Intel’s desktop products support simultaneous multithreading and can handle up to 32 threads. Immediately. Compared to real-life rivals AMD’s Ryzen Threadripper Pro W5995X and Intel’s Xeon W9-3495X, it easily outperforms in single-threaded workloads, but slows dramatically when more cores are needed.
Some might argue that Geekbench 5 is a synthetic benchmark that doesn’t reflect real-world application performance, and that’s a valid argument. But this gives you an idea of what to expect in terms of CPU computing power without a dedicated accelerator. And this brings us to the fact that Apple’s M2 SoC has a lot of accelerators inside. Therefore, you don’t need high clocks or extreme core counts to deliver good performance for many workstation-grade workloads.
And with so much computing power inside, AMD and Intel workstation processors are designed to handle even the most demanding workloads. That said, it remains to be seen whether Apple’s Mac Pro can actually beat workstations with AMD or Intel CPUs in workstation applications.
