The custom AI ASIC state of play (May 2026) — Broadcom deals, Google TPUs, Meta MTIA & beyond
ASIC-based AI server shipments are projected to reach 27.8% of the market in 2026, the highest share since 2023, and custom ASIC shipments will grow 44.6% year-over-year.
The rapid growth in AI compute demand is driving increased investment and development in custom silicon solutions, as specialized architectures offer significant performance and efficiency gains over general-purpose hardware.
The surge in custom AI ASICs reflects a significant architectural shift in AI infrastructure, impacting competition, supply chains, and the strategic capabilities of major tech players in the AI race.
The market share of custom and ASIC-based AI servers is rapidly increasing, indicating a move away from purely off-the-shelf GPU solutions towards more specialized, vertically integrated AI hardware strategies.
- · Broadcom
- · Meta
- · Custom ASIC designers
- · General-purpose GPU manufacturers (without ASIC offerings)
- · Cloud providers reliant solely on commodity hardware
Demand for specialized ASIC design and manufacturing capabilities will intensify, leading to increased R&D and investment in these areas.
Companies with strong internal ASIC development will gain a significant competitive advantage in AI performance and cost efficiency, potentially consolidating market power.
The proliferation of custom ASICs could fragment the AI hardware ecosystem, making standardization more challenging but also fostering greater innovation on the silicon level.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at Tom's Hardware