Apple M6 chip architecture reveals 2nm process and AV1 encoding omission

The transition to TSMC’s 2nm process via the M6 establishes the performance baseline for the iPhone 18, though thermal constraints may force a single neural engine for the A20 Pro. For the streaming ecosystem, the continued omission of AV1 hardware encoding in Apple’s latest silicon suggests the company is not yet prioritizing this royalty-free codec for its mobile roadmap. This technical choice forces developers to continue relying on older standards or software-based decoding, potentially impacting battery efficiency for high-resolution playback. Watch for whether Apple maintains this encoding gap when the A20 Pro specifications are finalized next year.

TSMC's 2nm process, the foundation of Apple's M6 chip, is drawing significant industry attention as the foundry races to scale production for multiple clients. TSMC began risk production of its N2 node in late 2025 and is targeting volume manufacturing in the second half of 2026, with Apple historically serving as the lead customer for each new process generation. The N2 node introduces gate-all-around (GAA) nanosheet transistors, replacing the FinFET architecture that has dominated advanced logic since 2011. This transistor-level shift promises roughly 10-15% performance gains or 25-30% power reduction at equivalent performance compared to N3E, figures that directly affect thermal budgets for sustained video encode and decode workloads on mobile devices.

Apple's decision to omit AV1 hardware encoding from the M6 has implications for the broader codec licensing landscape. The Alliance for Open Media, which governs AV1 specifications, has seen adoption accelerate among streaming platforms including YouTube, Netflix, and Amazon Prime Video since 2023, yet Apple's continued reliance on HEVC and H.264 for encode-side hardware means iOS content creation workflows will not natively produce AV1 streams. This creates a split ecosystem: Android flagships from Samsung and Google already ship with AV1 encode support via dedicated silicon, while Apple's installed base of over 1.2 billion active iPhones remains encode-limited to older codecs. For streaming services targeting Apple devices, the decode path (which Apple does support in hardware for AV1 since the A17 Pro) is less of a concern than the absence of encode, which affects user-generated content pipelines and live streaming from mobile.

The competitive silicon landscape underscores how unusual Apple's codec choices remain. Qualcomm's Snapdragon 8 Elite, announced in late 2024, includes dedicated AV1 hardware encoding alongside its Adreno GPU, and Google's Tensor G4 in the Pixel 9 series similarly supports AV1 encode at up to 4K60. MediaTek's Dimensity 9400 also added AV1 encode capability. Apple's M6 and the anticipated A20 Pro therefore represent the only major mobile silicon families still without hardware AV1 encode, a gap that may persist until Apple's next architectural refresh. For streaming engineers, this means bandwidth optimization strategies targeting Apple hardware must continue to account for HEVC as the primary encode target, even as AV2 codec ecosystem gains traction across Android and web-based playback environments.