advanced packaging
In plain wordsWays of joining several chips very closely inside one package — side by side on a silicon base or stacked on top of each other — so they act almost like one bigger chip.
Techniques that integrate multiple dies — logic chipletschiplet, n.: One of several smaller chips that are made separately and then joined inside a single package to work as one processor, which is cheaper than making one giant chip and lets makers reuse parts., HBMhigh-bandwidth memory, n.: Very fast memory made by stacking memory chips on top of each other and placing them right next to the processor, so AI chips can be fed data quickly enough. stacks, input/output dies — within one package using wafer-fab-like processes, with far denser and shorter connections than a circuit board allows. The main forms are 2.5D integration, in which dies sit side by side on a silicon interposer or embedded bridge (TSMC's CoWoS, Intel's EMIB); 3D stacking, in which dies are bonded face to face or connected through through-silicon vias (TSVs), as in HBM, AMD's 3D V-Cache and TSMC's SoIC; and fan-out wafer-level packaging. Every leading AI accelerator depends on 2.5D packaging to place HBM next to compute, and TSMC's CoWoS capacity was a principal bottleneck on AI chip supply in 2023–2025. Co-packaged optics, which places optical transceivers inside switch or processor packages, extends the approach to interconnectsinterconnect, n.: The wiring and networking that connect chips to each other — inside a server, across a data center, or within a single chip — so they can share data and work as one..
Origin
Multi-chip modules date to mainframes of the 1980s; silicon-interposer integration reached volume with TSMC's CoWoS, first used for Xilinx's Virtex-7 FPGAs (c. 2011–2012), and with AMD's HBM-equipped Fiji GPU (2015). 'Advanced packaging' gained policy prominence in the 2020s, including a dedicated program under the CHIPS and Science Act.
Related
- chipletchiplet, n.: One of several smaller chips that are made separately and then joined inside a single package to work as one processor, which is cheaper than making one giant chip and lets makers reuse parts.
- high-bandwidth memoryhigh-bandwidth memory, n.: Very fast memory made by stacking memory chips on top of each other and placing them right next to the processor, so AI chips can be fed data quickly enough.
- semiconductor foundrysemiconductor foundry, n.: A factory company that manufactures chips designed by other companies. Taiwan's TSMC, the largest, makes nearly all of the world's most advanced AI chips.
- photonic computingphotonic computing, n.: Using light instead of electricity to compute or move data. Light can carry huge amounts of information with little energy, but building full computers from it is still very hard.
- interconnectinterconnect, n.: The wiring and networking that connect chips to each other — inside a server, across a data center, or within a single chip — so they can share data and work as one.
- Moore's lawMoore's law, n.: Gordon Moore's prediction that chipmakers could pack twice as many transistors onto a chip roughly every two years — a pace that for decades made computers steadily faster and cheaper, and has slowed but not stopped.
Mentioned in
Moore's lawMoore's law, n.: Gordon Moore's prediction that chipmakers could pack twice as many transistors onto a chip roughly every two years — a pace that for decades made computers steadily faster and cheaper, and has slowed but not stopped., high-bandwidth memoryhigh-bandwidth memory, n.: Very fast memory made by stacking memory chips on top of each other and placing them right next to the processor, so AI chips can be fed data quickly enough., interconnectinterconnect, n.: The wiring and networking that connect chips to each other — inside a server, across a data center, or within a single chip — so they can share data and work as one., semiconductor foundrysemiconductor foundry, n.: A factory company that manufactures chips designed by other companies. Taiwan's TSMC, the largest, makes nearly all of the world's most advanced AI chips., TSMCTSMC, prop. n.: The Taiwanese company that manufactures most of the world's advanced computer chips, including nearly all of the specialized chips used to train and run AI., chipletchiplet, n.: One of several smaller chips that are made separately and then joined inside a single package to work as one processor, which is cheaper than making one giant chip and lets makers reuse parts., extreme ultraviolet lithographyextreme ultraviolet lithography, n.: The chip-printing technique used for the most advanced chips, which draws circuit patterns with ultra-short-wavelength light. A single Dutch company, ASML, makes all the machines, each costing well over $100 million., memory wallmemory wall, n.: The growing gap between how fast chips can calculate and how fast they can fetch data from memory. Processors increasingly sit idle, waiting for numbers to arrive., photonic computingphotonic computing, n.: Using light instead of electricity to compute or move data. Light can carry huge amounts of information with little energy, but building full computers from it is still very hard., wafer-scale enginewafer-scale engine, n.: A single giant processor made from nearly a whole dinner-plate-sized silicon wafer instead of the small chips normally cut from it, so data rarely has to leave the chip.