Apple introduces M4 chip

Original Article ↗ Hacker News Discussion ↗

Apple’s new M4 chip debuts in the latest iPad Pro, promising roughly 1.5× the CPU performance of M2 and double the performance‑per‑watt, a much faster Neural Engine for on‑device AI, and support for OLED “tandem” displays. Commenters are impressed by the silicon and power efficiency, but question Apple’s marketing claims, the usefulness of so much compute on iPadOS, and the continued pairing of high prices with 8 GB RAM and tightly coupled storage tiers. Many see the launch as a preview of upcoming M4‑based Macs and a signal that Apple’s AI strategy leans heavily on running models locally on consumer hardware.

Where M4 Ships and Roadmap Speculation

  • M4 debuts only in the new iPad Pro; Macs remain on M2/M3 for now.
  • Many expect M4-based MacBook Air/Pro, Mac mini, Studio/Pro later this year, possibly around WWDC.
  • Several comments tie the fast move from M3→M4 to TSMC’s shift from problematic N3 to cheaper, higher‑yield N3E.

Performance, Efficiency, and Marketing Claims

  • Apple claims: ~1.5× CPU vs iPad M2, 2× performance per watt vs M2, and big Neural Engine gains.
  • Some see this as in line with industry efficiency trends; others view Apple’s charts and selective M2 comparisons (not M3) as classic marketing spin.
  • Many note unchanged “up to 10 hours” battery spec, inferring CPU savings are spent on the new OLED display and thinness, not runtime.

iPadOS Limitations and macOS Wishes

  • Large contingent calls iPadOS “gimped”/“toy-like” relative to the hardware; main asks: macOS dual‑boot, real dev tools (Xcode, terminal, virtualization), or at least better multitasking.
  • Others argue the iPad excels as an “appliance” for drawing, music, media, and focused work, and that desktop UIs on touch are inherently poor.
  • Skepticism that Apple will ever allow macOS or full Unix-style access, for both UX and business/segmentation reasons.

Memory, Storage, and Product Segmentation

  • iPad Pro M4 ships with 8 GB RAM on 256/512 GB models, 16 GB on 1/2 TB, plus a core disabled on lower tiers.
  • Many criticize tying RAM to storage and 8 GB on a “Pro” device; others note most users won’t hit that limit.
  • Explanations offered: logistics/SKU simplification, yield salvage, and deliberate price discrimination/“ladder” to push buyers up.

AI, Neural Engine, and “AI PC” Comparisons

  • Apple touts 38 TOPS and “most powerful Neural Engine” vs current “AI PCs”.
  • Some clarify this targets integrated NPUs in Intel/AMD laptop SoCs, not discrete GPUs like RTX 4090, which still dominate raw AI throughput.
  • Broad agreement that Apple is leaning into on‑device inference as part of a privacy/latency story; others see “privacy‑first” as partly marketing and note lack of transparency.

Use Cases and Value of iPad Pro vs Macs

  • Many commenters can’t justify upgrading from older iPads (even pre‑M1) because their usage is mostly video, reading, light note‑taking.
  • Others report iPad Pros are excellent for digital art, music production, and some video workflows, where extra GPU/CPU and RAM matter.
  • A recurring tension: for people who need Xcode, terminals, VMs, or serious dev tools, a MacBook is still mandatory, making the iPad an expensive secondary device.

Battery, Displays, Environment, and Repair

  • Thread notes display and radios, not CPU, dominate light‑use power; OLED can save power on dark content but may draw more at high brightness.
  • Efficiency gains are praised for aggregate energy savings, though some expect them to be “spent” on brighter screens and thinness rather than dramatically longer battery life.
  • Concerns raised about soldered components, limited repair paths, and devices becoming e‑waste when logic boards or single USB‑C ports fail.