This month I want to highlight the recent launch of Apple’s M1 Ultra chip via their press release, Apple unveils M1 Ultra, the world’s most powerful chip for a personal computer. M1 Ultra is the fusion of two M1 Max chips, which includes an innovative way to connect the two chips such that each chip can access the resources of the other chip uniformly, and without problematic latency. The M1 Ultra demonstrates true creative thinking; instead of inventing a faster chip, Apple innovated faster communication pathways between chips, which let them realize unparalleled performance via connected chips. While solving Non-Uniform Memory Architecture (NUMA) is not a new problem in Computer Science, Apple created a new solution that simplifies the space. I love breakthroughs of this nature, as they unlock new paths for technology to progress.

The Computerphile YouTube channel wonderfully breaks down the M1 Ultra in a Computer Science-y ELI5 manner, which I find helpful to wrap my head around the topic. I recommend subscribing to Computerphile if you have any interest in Computer Science, as they frequently describe current events and deconstruct how those events occur (such as the Facebook outage on 10/04/2021).

“The most common way to scale performance is to connect two chips through a motherboard, which typically brings significant trade-offs, including increased latency, reduced bandwidth, and increased power consumption. However, Apple’s innovative UltraFusion uses a silicon interposer that connects the chips across more than 10,000 signals, providing a massive 2.5TB/s of low latency, inter-processor bandwidth — more than 4x the bandwidth of the leading multi-chip interconnect technology.”