Apple's M6 MacBook Pro claims 24 hours of battery and a memory bandwidth of 170 GB/s
The 14-inch model has a 12-core CPU, a 12-core GPU with a Neural Accelerator in each core, and ships with macOS 27 from November 6.
Against M5, Apple claims graphics up to 30 percent faster, AI workloads up to 25 percent faster and a Neural Engine up to 1.9x faster. Memory is 16 GB standard, 32 GB maximum. The quoted bandwidth became the thread's sticking point.
Owners of older machines used it to say they are not upgrading.
The announcement in numbers
Apple's newsroom post says the 14-inch MacBook Pro with M6 pairs a 12-core CPU with a 12-core GPU, plus a dual 16-core Neural Engine. Versus M5 it claims up to 30 percent faster graphics, up to 25 percent faster AI workloads and up to 1.9x on the Neural Engine; versus M1, up to 11x graphics and 13x AI. Battery life is up to 24 hours. Memory is 16 GB, configurable to 32 GB, at up to 170 GB/s. Pre-orders open October 28 and machines arrive November 6, in space black and silver.
The thread's objection: 170 GB/s
A commenter noted the M5 Max has bandwidth up to 600 GB/s and asked whether 170 GB/s is the real limit for M6, and whether that would slow LLM processing.
A former big-tech engineer disliked the naming:
"I have to say, I really dislike the thing where the base M(n+1) is always in the market alongside M(n) Pro/Max."
He allowed that the chips are probably quite distinct, but called the naming confusing.
Waiting out the cycle
A founder keeping an M1 Pro alive said he is not eager to pay $6000+ for a decent-spec dev computer until he has to. Another said all he needs is a laptop with a nice enough screen to ssh into a beefy Linux box, and his 48 GB 16-inch will last 15+ years.
For LLM work, the answer to the bandwidth question is on the spec sheet: the base chip is not the one to buy for it, and the Pro and Max variants are not announced yet.
Get breakdowns like this every day
The top Hacker News stories, unpacked — by email, Telegram, or RSS.
Lean has checked all 43.1 million lines of the finite simple groups classification, and found one gap
The proof spans thousands of journal pages across 715 papers and books, and its formalization found a flaw in a published argument for the Unbalanced Group Theorem.