← The AI Systems Map
L2 · ChipsAcceleratorUnited StatesMRVL

Marvell Technology

Second major custom-AI-silicon (ASIC) and interconnect design partner

Marvell is the other big merchant custom-silicon partner (alongside Broadcom), co-designing AI accelerators for hyperscalers — notably AWS Trainium/Inferentia and Microsoft — and supplying optical DSPs, interconnect, and networking IP for AI data centers. Its custom-compute and electro-optics businesses grew sharply through 2025 on AI demand. Fabless on TSMC, Marvell is a core enabler of hyperscalers' in-house accelerator strategies and of the optical interconnect that stitches AI clusters together.

Custom ASIC

AWS Trainium/Inferentia, Microsoft

Also supplies

Optical DSPs, interconnect IP

Fab

TSMC (fabless)

How it fits the stack

Marvell Technology with what it depends on (above) and what it feeds (below). The figure renders as a crawlable diagram and upgrades to an interactive 3D graph as it scrolls into view.

depends onpartners withcompetes withpartners withdesignsMarvell TechnologyChipsTSMC (TaiwanSemiconductorManufacturing Company)chokepointAWS Trainium /InferentiaBroadcomMicrosoft Maia
Marvell TechnologyDepends on ↑Related

Marvell Technology in the AI stack. Marvell Technology with its immediate upstream dependencies (top) and downstream dependents (bottom) in the AI value chain. Hover a node in 3D, or read the full relationships below.

Graph data (text) — 5 entities, 5 relationships
  • Marvell Technologydepends onTSMC (Taiwan Semiconductor Manufacturing Company)
  • Marvell Technologypartners withAWS Trainium / Inferentia
  • Marvell Technologycompetes withBroadcom
  • Marvell Technologypartners withMicrosoft Maia
  • Marvell TechnologydesignsAWS Trainium / Inferentia