Apple's 18-core M5 Max trounces 96-core Ryzen Threadripper in benchmarks
Apple's new M5 Max processor has achieved record-breaking results in Geekbench 6, outperforming the 96-core AMD Ryzen Threadripper Pro 9995WX in multi-threaded workloads. While Apple has long led in single-thread performance, this new milestone demonstrates that its silicon can now dominate high-core-count desktop rivals in both categories under specific benchmark conditions. In the Geekbench 6 CPU tests, the M5 Max recorded a single-core score of 4,353 and a multi-core score of 29,644. Although the AMD Ryzen Threadripper Pro 9995WX typically scores around 26,000 in multi-threading, one specific system achieved a high of 30,170 points, slightly edging out the Apple chip. However, for the vast majority of configurations, the M5 Max secures the victory. The M5 Max utilizes a hybrid architecture with six super-performance cores and twelve performance cores, supported by an advanced cache hierarchy that delivers a 12% single-thread improvement over the previous M4 Max. Its memory subsystem also features a significant leap, offering 128 GB of LPDDR5X-9600 memory with a 512-bit interface, providing 614 GB/s of bandwidth, which surpasses the capabilities of current workstation processors. It is important to note that Geekbench 6 is a synthetic benchmark designed to mimic brief, bursty consumer tasks like image editing and PDF processing rather than sustained professional workloads. Consequently, the test may not fully utilize the extensive parallel capacity of the 96-core AMD processor, as the benchmark subtests generally scale efficiently to only 8 to 32 threads. This environment favors Apple's architecture, which relies on high per-core performance rather than sheer core count. Beyond the CPU, the M5 Max features a new GPU based on a PowerVR-derived microarchitecture developed by Apple. In Geekbench 6 GPU compute tests using the Metal API, the M5 Max scored 232,718 points, up from 204,453 for the M4 Max. When compared to discrete graphics cards, the M5 Max's integrated GPU easily defeats the Ryzen AI Max+ 395 iGPU. It also surpasses the Nvidia GeForce RTX 5070, which scored 207,061 points, but falls behind the GeForce RTX 5070 Ti and the high-end GeForce RTX 5090. While the M5 Max does not yet compete with top-tier discrete GPUs in raw compute power, achieving performance comparable to mainstream cards within an integrated chip marks a significant technological breakthrough. The results suggest that Apple's focus on memory bandwidth and efficient architecture continues to yield substantial advantages in professional workflows, even as it approaches the limits of its single-die design when facing massive core counts in extended stress tests.
