Recent tests of Apple's MLX machine Xi framework have shown that the new Apple M-series chips surpass NVIDIA's RTX 4090.
On December 6, Apple announced the release of MLX, an open-source framework designed specifically for Apple silicon to be built, tested, used, and enhanced by AI developers.
Developer Oliver Wehrens recently shared some benchmark results on the MLX framework on Apple's M1 Pro, M2, and M3 chips versus NVIDIA's RTX 4090 graphics card. It uses Whisper, OpenAI's speech recognition model.
Wehrens uses the Whisper model for speech transcription and measures the time it takes to process a 10-minute audio file. The results show that the M1 Pro chip performs slightly lower than the NVIDIA GPU, taking 216 seconds to process audio, compared to 186 seconds for the 4090 chip.
However, newer Apple chips have better performance. For example, another person runs the same audio file on an M2 Ultra with 76 GPUs and an M3 Max with 40 GPUs and finds that these chips take less time to transcribe the audio than NVIDIA GPUs.
There are also significant differences between Apple silicon and Nvidia products in terms of power consumption. Specifically, when comparing a running NVIDIA 4090 PC to its idle state, the power consumption increased by 242 watts.
In comparison, a MacBook with 16 M1 GPU cores has a smaller increase in power consumption compared to the idle state in the active state, with a difference of only 38 watts.
These results highlight Apple's progress in AI and machine Xi capabilities, and may be the beginning of more powerful capabilities for Apple products. Now, the MLX framework is open source, paving the way for widespread adoption and innovation by developers.
NVIDIA's 4090 GPU starts at $1599 and includes only the graphics card, not the PC. That's the same as the 2022 M3 MacBook Pro, but the M3 Pro and M3 Max have skyrocketed.