Saved
Semiconductors

Application-specific integrated circuit

ASICs lock logic into silicon—brilliant for scale, brutal for change.

Worth your time?4.5/5Yes.Study the whole thing.What works, what does not, who it is for ↓
9:46in productionCh. 1 · One chip, one purpose/ 9:46 · ceiling 15 min
6 chapters · 9:46Ceiling 15:00

The brief is written and checked. The video joins this page when it is cut.6 chapters · 1 min read · 9:46 planned

Read the brief

ASICs are custom silicon built for fixed tasks. They deliver unmatched efficiency but zero runtime adaptability. Their evolution—from 1974 CMOS gate arrays to modern SoCs—is defined by scale, not flexibility.

Chapters & takeaways6
  1. 0:54
    One chip, one purpose

    ASICs are not general-purpose chips—they are built for one job and one job only.

  2. 2:06
    From lab to living room

    Gate arrays launched in 1974 and powered mass-market computers by 1981–82.

  3. 3:11
    More than logic: full systems on chip

    ASICs now hold over 100 million gates and embed full processors—as SoCs.

  4. 4:08
    How you build what cannot be changed

    They are made with MOS technology and described in Verilog or VHDL—not soldered or scripted.

  5. 5:24
    The cost of certainty

    They trade flexibility for efficiency: no reprogramming, no compromise on performance.

  6. 6:26
    Built by companies, not conferences

    Their history is commercial—not theoretical: ZX81, ZX Spectrum, IMI, Fairchild, Motorola.

3 more chaptersFewer chapters
Worth your time?

Yes. Study the whole thing.

4.5/ 5
What works
  • delivers superior power efficiency for fixed workloads
  • enables integration of processors and memory as SoCs
  • achieved commercial success in 1981–82 consumer computers
What does not
  • support field reconfiguration
  • allow repurposing after fabrication
  • scale economically below high volume
Study it if
  • chip architects
  • high-volume device manufacturers
  • power-constrained embedded systems designers
Skip it if
  • prototypers
  • low-volume OEMs
  • AI researchers requiring iterative hardware changes
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

An ASIC is a custom integrated circuit built for one use—not general-purpose. It solves the problem of inefficiency in generic hardware by hardwiring logic for a specific task.

How it works

ASICs are fabricated using MOS technology. Designers describe functionality in HDLs like Verilog or VHDL. The design flows through RTL, synthesis, placement, routing, and sign-off. Gate arrays emerged in 1974; standard-cell MOS technology arrived in the 1970s.

What works

ASICs deliver superior power efficiency, speed, and density for fixed workloads. Complexity grew from 5,000 to over 100 million gates. Modern ASICs integrate full microprocessors and memory blocks as SoCs. Gate arrays powered the ZX81 (1981) and ZX Spectrum (1982).

What does not

ASICs do not support field reconfiguration. They cannot be repurposed after fabrication. Their high non-recurring engineering (NRE) cost makes them uneconomical below scale.

What it changes

ASICs shift computation from programmable general-purpose chips to immutable, application-locked silicon. They enable SoCs—but at the cost of design rigidity and long lead times.

Is it worth your time

Yes—if you need high-volume, low-power, fixed-function silicon and can absorb high NRE costs. No—if you require flexibility, rapid iteration, or low upfront investment.

Same field · Semiconductors4 of 53
9:26
DDR4 SDRAM2012DDR4 SDRAM is a JEDEC-standardized memory technology finalised in September 2012 and released in 2014. It increases per-DIMM capacity to 64 GB, reduces voltage to 1.2 V, and introduces bank grouping to improve bandwidth concurrency—all while breaking full compatibility with DDR2 and DDR3. Its launch was phased: ECC-only in Q2 2014, non-ECC in Q3 2014. It does not use deeper prefetch; it relies on higher command rates and bank group isolation instead. Its value lies in capacity and efficiency scaling—not raw speed leaps or plug-and-play upgrade paths.
9:30
Mellanox Technologies1999Mellanox Technologies was an Israeli-American fabless semiconductor company founded in May 1999 in Yokneam Illit, with financial offices in Santa Clara. It specialised in high-speed, low-latency interconnects—primarily InfiniBand and Ethernet—delivering adapters, switches, software, cables and silicon for HPC, data centres, cloud, storage and financial services. Its products enabled RDMA and RDMA over Converged Ethernet. By 2011, its InfiniBand solutions were embedded in many TOP500 supercomputers. It expanded into Ethernet-based storage area networks from 2011. A $64 million venture round in 2002—backed by Intel, IBM, Sequoia Capital and U.S. Venture Partners—confirmed early market confidence. Nvidia acquired Mellanox for $6.9 billion in March 2019; the deal closed in April 2020 after antitrust approvals. Mellanox was integrated into Nvidia’s networking division, and the brand was discontinued for new products.
11:19
DDR5 SDRAMDDR5 delivers higher capacity, lower voltage, and scalable bandwidth—but not lower latency. Its gains depend on supporting infrastructure, not just the standard itself.
9:41
Graphics processing unitThe GPU is a specialised electronic circuit for digital image processing and computer graphics acceleration. It emerged in the 1990s with internal calculation capabilities—unlike earlier video display controllers and blitters—and evolved to support shaders, 3D functionality, and massively parallel computation. Its parallel architecture enabled adoption beyond graphics, notably in AI and scientific computing. The term 'GPU' was coined by Sony in 1994 for the PlayStation's Toshiba-designed chip.
Up next in Technology

Blu-ray

Blu-ray Disc Association · 2006 · 10:08

Blu-ray won the format war—but only because it shipped first, not because it was inevitable.

10:08