The first CPU and GPU benchmarks of Apple’s M2 are now out, and while there is a slight performance improvement compared to the M1 in both single-core and multi-core applications, there is a substantial leap in graphics performance. Let’s take a closer look at these figures.
Apple’s M2 Delivers CPU Gains Over M1 – But GPU is 50% Slower!
Apple’s M2 received a single-core score of 1919 and a multi-core score of 8928 on Geekbench. The device that was tested has 16GB of unified LPDDR5 RAM and a CPU that supposedly ran at a base frequency of 3.49GHz. The M2 performs 12 percent faster than the M1 in the single-core test, with the multi-core results indicating a 19 percent improvement. We have included the scores received by both chipsets below to provide a better comparison.
- M2 – single-core 1919, multi-core 8929
- M1 – single-core 1720, multi-core 7474
Remember that the CPU configuration of both the M1 and M2 is the same, with four cores dedicated to performance and the other four dedicated to power savings. Regardless of the number and type of cores, Apple stated during its WWDC 2022 keynote that the improved design of the M2 would result in higher performance for users, which is exactly what we see in the images below.
The M2 may be configured with up to 10 GPU cores, making Apple’s latest proprietary silicon 50 percent quicker than the M1 (20440). Apple claims the memory controller can work at 100GB/s, making it 50% faster than the M1.
The current SoC is mass-produced on TSMC’s second-generation 5nm technology, whereas the M1 was fabricated on the first-generation 5nm node. The M2’s CPU performance rose 18% compared to the M1, although the multi-core saw a 19% rise.
Synthetic workloads are used by Geekbench, which may not be entirely accurate. If the unique silicon in the M2 is any indication, it could outperform the M1 or let down new MacBook series users. Keep an eye out for new benchmark results.