Apple quietly added FlexCache to M6, echoing Qualcomm's same architectural bet
clattner_llvm · x · 2026-10-01
Per SemiAnalysis, Qualcomm announced FlexCache in its Snapdragon 8 Elite Gen 6 a month ago, but Apple already implemented it in the M6 without fanfare. The M6 is Apple's first chip to combine heterogeneous cores in a shared L2 cache domain, enabling a third core type (branded P, identified as M) without a separate M cluster. Chris Lattner commented that two CPU leaders making the same bet is a great sign for an open ecosystem, despite SemiAnalysis framing it as controversy.
More from Infra
- Free course built from Cornell's GPU architecture workshop now shared publicly — idanbeck · 2026-10-01
- Qwen Flash Next MTP work resumes with official GGUF quants and llama.cpp PR — jacek2023 · 2026-10-01
- Undocumented Strata tip: set default sampling params via a sampling block in run config — KissMyShinyArse · 2026-10-01
- Meta's Loop Scaling Laws: Sparsity Gives ~3x Active-Param Efficiency, Recurrence ~2x on Reasoning — facebook · 2026-10-01
- 87GB Qwen3.8 Flash Next runs at 120-150 t/s on a single RTX 5090 — CurieuxExplorer · 2026-10-01
- When evaluating GPU infra, "it passed testing" says little until you know what was tested — AccBalanced · 2026-10-01