Skip to main content

Leaderboards / L2 Silicon & Memory / #5

AMD

Active product Chiplet die-to-die interconnect (UCIe / 2.5D-3D chiplet partitioning)·MI300 / MI325 / MI350 Instinct accelerator family — the dominant merchant alternative to NVIDIA in AI training silicon. ROCm software stack maturity is the primary watch-point.

Rank in layer

#5

Confirmed customers

–

Recent engine activity · 90 day window

Customer wins

0

Partnerships

0

Capital events

2

Hiring (60d)

2

Capital events

Funding, earnings filings, and IPO-related signals

  • funding

    2026-05-15

    AMD entered into a new five-year, $5.0 billion unsecured revolving credit facility with JPMorgan Chase Bank, N.A. as administrative agent, replacing an existing agreement.

    ↗ sec.gov

  • funding

    2026-05-15

    AMD increased the maximum aggregate amount of its unsecured commercial paper notes program from $3.0 billion to $5.5 billion.

    ↗ sec.gov

Hiring focus

2 roles posted in last 60 days

engineering

2 active postings

Active in

Products this company is committing to per engine analysis

  • Chiplet die-to-die interconnect (UCIe / 2.5D-3D chiplet partitioning)L2
  • AMD Instinct datacenter GPU / CDNA acceleratorL2
  • Chiplet AI Accelerator with 2.5D/3D Advanced Packaging/ai-asics
  • Non-NVIDIA AI accelerator ASIC (LPU / RDU / wafer-scale / Instinct class)L2
  • Chiplet die-to-die interconnect (UCIe-based)L2
  • AMD Instinct Datacenter Accelerator (CDNA chiplet)/ai-asics
  • 2.5D/3D Chiplet Advanced Packaging with UCIe + HBM/ai-asics
  • Chiplet / UCIe Die-to-Die InterconnectL2
  • AMD Instinct Datacenter Accelerator (CDNA, chiplet, UCIe/HBM)/ai-asics
  • Chiplet / Advanced Packaging stack for AI accelerators (2.5D/3D + UCIe + HBM)/ai-asics

Share provenance

Notes

Layer L2 product mentions: Chiplet die-to-die interconnect (UCIe / 2.5D-3D chiplet partitioning), AMD Instinct datacenter GPU / CDNA accelerator.

Leaderboard updated 2026-06-08 · Methodology