technologybriefs
10:09in productionCh. 1 · The 1978 Workaround/ 10:09 · ceiling 15 min
Semiconductors

X86

1978

x86 isn’t enduring because it’s good — it’s enduring because every fix creates new dependencies no one dares break.

x86 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.

SemiconductorsComputingHardwareInventionComputing across the network →
Chapters & takeaways4
  1. 1:00
    The 1978 Workaround

    Segmentation wasn’t a design choice — it was a workaround for 16-bit address limits.

  2. 2:38
    The 32-Bit Shift

    The 80386 didn’t just add bits — it displaced an entire generation of 16-bit chips in general-purpose computers.

  3. 4:27
    The AMD Takeover

    AMD’s x86-64 was the first major x86 extension not started by Intel — and the only one Intel adopted.

  4. 6:15
    The 2023 Abandonment

    X86S failed not because it was wrong, but because the cost of removing legacy modes exceeded the value of simplification.

Worth your time?

Yes. Study the whole thing.

4.5/ 5
What works
  • enabling backward-compatible evolution across four decades
  • providing a stable target for operating system development
  • allowing AMD to challenge Intel’s architectural authority
What does not
  • X86S was implemented
  • Intel led the 64-bit extension
  • x86 eliminated legacy modes
Study it if
  • systems programmers
  • firmware engineers
  • security researchers working on CPU-level exploits
Skip it if
  • application developers targeting high-level frameworks
  • AI model trainers focused on GPU acceleration
  • cloud platform users abstracted from instruction sets
The written brief1 min read

What it is and the problem it solves

x86 is a CISC instruction set family originating with Intel’s 1978 8086 microprocessor. It solved the problem of addressing more than 64 KiB of memory on a 16-bit processor.

How it works

The x86 architecture uses memory segmentation to extend addressing beyond 16 bits. It evolved from 16-bit to 32-bit with the 80386. AMD extended the 32-bit architecture to 64 bits between 1999 and 2003. Intel proposed X86S in 2023 to remove legacy execution modes and instructions.

What works

Memory segmentation worked as a stopgap for 16-bit addressing limits. The 80386’s 32-bit transition gradually replaced earlier 16-bit chips in computers. AMD’s x86-64 extension succeeded where Intel’s own 64-bit Itanium failed — it ran existing 32-bit code natively.

What does not

X86S did not survive. The 2023 simplification proposal was abandoned in December 2024 after Intel and AMD formed the x86 Ecosystem Advisory Group. Legacy execution modes and instructions remain embedded in every x86 chip shipped today.

What it changes

It established a de facto hardware standard that forced software to adapt across decades — not through elegance, but through layered compatibility. It shifted architectural leadership: AMD’s x86-64 extension was the first major x86 advancement initiated by a non-Intel vendor and accepted by Intel.

Is it worth your time

Yes — if you work on systems software, firmware, or compatibility layers. The persistence of segmentation, mode switching, and backward support defines real-world constraints on performance, security, and portability. Its inertia is operational infrastructure, not legacy baggage.

Same field · Semiconductors4 of 51
9:51
Hardware for artificial intelligenceAI hardware is a set of purpose-built chips—Lisp machines, GPUs, TPUs, NPUs—that accelerate AI workloads by optimising for parallelism, memory bandwidth, and low-precision arithmetic. It delivers real speed and efficiency gains, but at the cost of flexibility, portability, and architectural transparency. Its dominance reflects engineering necessity—not inevitability.
9:20
ChipletChiplets are modular ICs—functional fragments designed for assembly on an interposer. They enable heterogeneous integration, pre-test validation, and IP reuse. But they shift complexity from fabrication to integration, and require standards compliance to deliver on their promise.
9:27
ECC memoryECC memory adds redundancy and decoding logic to catch and fix single-bit memory errors before they propagate. It works via SECDED Hamming or Hsiao codes, using 8 extra bits per word—64+8 for DDR–DDR4, 32+8 per subchannel for DDR5. Mandatory on-die ECC in DDR5 and LPDDR6 addresses rising soft-error rates from shrinking process nodes. It delivers deterministic single-bit correction in servers and critical infrastructure—but offers no protection beyond that, and no benefit where silent corruption is tolerable.
9:33
EUV lithographyEUV lithography replaces transmissive deep-UV optics with reflective Mo/Si mirrors, tin-plasma light sources, and vacuum operation to pattern sub-7 nm semiconductor features. It works — but only after decades of co-development across continents, and only where photon budget, thermal drift and stochastic noise can be managed. It is necessary, not optional, for leading-edge logic — but it does not generalise, simplify or cheapen.
Up next in Technology

Go (programming language)

Robert Griesemer, Rob Pike, Ken Thompson · 2009 · 10:03

Go doesn’t make programming easier—it makes large-scale, concurrent systems less likely to collapse under their own complexity.

Play nextBack to Semiconductors
10:03