Saved
Semiconductors · Hardware

Read-only memory

Read-only memory isn’t about limitation — it’s about guaranteeing what cannot change.

Worth your time?4.5/5Yes.Study the whole thing.What works, what does not, who it is for ↓
9:15in productionCh. 1 · Immutable by physics/ 9:15 · ceiling 15 min
4 chapters · 9:15Ceiling 15:00

The brief is written and checked. The video joins this page when it is cut.4 chapters · 1 min read · 9:15 planned

Read the brief

ROM fixes data in hardware. It trades updateability for reliability. Its evolution — from mask to PROM to EPROM — reflects successive attempts to widen the window for user control without sacrificing non-volatility. The core constraint remains: no electrons may alter what physics has set.

Chapters & takeaways4
  1. 0:59
    Immutable by physics

    ROM is non-volatile and unchangeable by design — not a flaw, but its defining feature.

  2. 2:36
    Wired in silicon

    Mask ROM encodes data in transistor wiring defined by photomasks — no electrons can rewrite it.

  3. 4:19
    One-time programmability

    PROM lets users burn in data once — physically, not electronically — using high-voltage pulses.

  4. 5:33
    From mask to erase

    The 1967 floating gate idea made erasable ROM possible — and turned ROM from static to reusable.

1 more chapterFewer chapters
Worth your time?

Yes. Study the whole thing.

4.5/ 5
What works
  • Guarantees boot code integrity
  • Eliminates power-loss corruption risk
  • Enables predictable timing in real-time systems
What does not
  • Support runtime writes
  • Enable field updates without physical intervention
  • Replace RAM for general-purpose data storage
Study it if
  • Firmware engineers
  • Bootloader developers
  • Semiconductor designers
Skip it if
  • Application developers needing mutable persistent storage
  • Cloud infrastructure teams managing dynamic configs
Who should study it, and who should skipHide who it is for
The written brief1 min read

What it is and the problem it solves

ROM is non-volatile memory whose contents are fixed at manufacture. It solves the problem of storing essential instructions — like CPU boot code — that must survive power loss and resist accidental overwriting.

How it works

ROM stores data in physical structures — diode matrices or transistor switches — wired during semiconductor manufacturing. Mask ROM uses photomasks to define connections between word lines and bit lines. PROM is programmed once by blowing fuses with high-voltage pulses. EPROM builds on the 1967 floating gate proposal to enable erasure via UV light.

What works

Mask ROM delivers deterministic timing and zero runtime failure risk. PROM gives users one-time field programming. EPROM enables reuse via UV erasure — a direct consequence of the 1967 floating gate proposal.

What does not

ROM does not support electronic modification after manufacture. Mask ROM cannot be altered at all. PROM cannot be reprogrammed. None of these store volatile state or respond to software write commands.

What it changes

ROM establishes a hardware-enforced boundary between code and data that must persist across power cycles. It shifts bootstrapping from external media to integrated silicon. It makes firmware distribution a manufacturing step, not a runtime operation.

Is it worth your time

Yes — if you need immutable, power-independent storage for boot code or firmware. No — if you require frequent updates or runtime writes. Its value lies in predictability, not flexibility.

Same field · Semiconductors4 of 53
9:47
Static random-access memorySRAM is a volatile memory technology built from six-transistor latches. It delivers speed and stability without refresh — but at steep cost and density penalties. It enabled CPU caches and displaced magnetic-core memory, yet never challenged DRAM’s dominance in volume storage. Its mechanism is elegant, its trade-offs absolute, and its role fixed: fast memory for the few bits that matter most — right now.
10:09
X861978x86 began as a 1978 workaround — memory segmentation to stretch 16-bit addressing. It endured not by shedding complexity, but by layering it: 32-bit with the 80386, 64-bit via AMD’s x86-64, and finally failing to simplify with X86S in 2024. Its mechanism is dependency, not design.
9:18
Microprocessor1971The microprocessor is not a new kind of computer. It is the CPU—arithmetic, logic, and control—reduced to one or a few integrated circuits using VLSI. It runs machine code on binary data via clocked registers and digital logic. It works reliably because it removes dozens of soldered joints from earlier TTL designs. It fails as a standalone device: no memory, no I/O, no power regulation. It changed computing by making CPUs cheaper, smaller, and more dependable—not smarter or faster. Its value lies in its mechanism, not its promise.
9:45
Multigate deviceThe multigate device is a MOSFET with more than one gate on a single transistor. It solves short-channel effects that arise when planar transistors shrink below ~32 nm. More gates wrap the channel — improving electrostatic control. This suppresses off-state leakage and boosts on-state drive current. N-channel FinFETs reached 17 nm in 1998. A 25 nm FinFET operating at 0.7 V was demonstrated in December 2002. Bulk FinFETs enabled mass production in 2004. It does not eliminate short-channel effects — only mitigates them at the cost of fabrication complexity.
Up next in Technology

Satellite Internet access

· 9:51

Satellite Internet doesn’t replace broadband—it patches its geography with physics, not policy.

9:51