AMD and Nvidia consume the entirety of TSMC’s CoWoS packaging processes until 2025

KazeoHin said:
They're entire Highly-Specialised, Niche, Whole Wafer System that very few others would have the need or the engineering expertise to utilise....

Not "All of TSMC's packaging capacity"

Fix your sensationalised title.
Click to expand...
The CoWoS process is specialized but it was formerly used by a lot of companies… Apple, Tesla, Microsoft, Intel, Amazon, Meta, etc… If this is true it completely screws their datacenter AI plans, as well as a lot of other things.
 

TSMC struggles to meet demand for CoWoS packaging, holding back AI and HPC chip production: report​


TSMC, is struggling to keep up with the demand for its chip-on-wafer-on-substrate (CoWoS) technology, TrendForce reports, citing Commercial Times.

The larger size of interposers required for the latest AI and HPC processors from Nvidia and AMD means fewer interposers can be obtained from each 300-mm wafer, straining CoWoS production capacity. Additionally, the number of HBM stacks integrated around GPUs is increasing too, adding to the production challenges. As the interposer area grows, the capacity to meet GPU demand diminishes. This has led to a persistent shortage in TSMC's CoWoS production capacity.

TSMC aims to increase its monthly production to 40,000 units by the end of 2024

https://www.tomshardware.com/tech-i...ing-back-ai-and-hpc-silicon-production-report
 
Is this news? There's a 6 month lead time on capacity that is typically purchased 18-36mos in advance? Is water still wet?
 
You must log in or register to reply here.
Back
Top