LIVE
CHALLENGING NVIDIA

AMD Instinct accelerators and Broadcom custom AI ASICs capture fresh hyperscaler orders as clouds diversify beyond Nvidia

Capitalizing on Nvidia’s leadership transition and strict allocation freezes, AMD and Broadcom lock in multi-billion-dollar cluster expansions with top cloud providers.

UpdateNext Staff
Editorial team
5 min read
1 view
AMD CEO Lisa Su holding Instinct AI accelerator chip and Broadcom networking silicon
AMD CEO Lisa Su holding Instinct AI accelerator chip and Broadcom networking silicon. Photo via DigiTimes & Wall Street Semiconductor Desk / UpdateNext

For three years, trying to pry data-center market share away from Nvidia was the hardest job in technology. In mid-November 2026, the combination of exploding enterprise inference demand and governance turbulence at Nvidia has opened the widest window yet for AMD and Broadcom.

According to supply-chain checks by DigiTimes and institutional hardware analysts, cloud procurement teams at Microsoft, Meta, Google, and Oracle have accelerated second-source purchase orders for AMD’s Instinct GPU lineup and Broadcom’s custom XPU silicon.

ADVERTISEMENT
Get the daily AI briefingSubscribe

How AMD and Broadcom Are Splitting the Opening

✓ WE KNOW
  • AMD and Broadcom are emerging as the primary beneficiaries of hyperscaler supply diversification following Nvidia’s governance shakeup and allocation review.
  • AMD’s Instinct MI350 and upcoming MI400 series have secured expanded inference and training commitments from Microsoft, Meta, and Oracle.
  • Broadcom’s custom AI ASIC division — which co-designs TPUs for Google and custom accelerators for Meta, ByteDance, and OpenAI — projects record 2027 networking and XPU revenue.
? STILL UNCLEAR
    “Every cloud CFO now mandates a dual-source silicon strategy: merchant GPUs from Nvidia and AMD paired with custom Broadcom-built ASICs.”
    —
    SOURCES:DigiTimes & Wall Street Semiconductor Desk: AMD & Broadcom AI Share Gains ↗
    > how_we_reported_this
    • ✓Sources: written from primary sources and official statements, linked above.
    • ✓Fact-check: names, dates and claims checked against the original sources.
    • ✓AI use: researched with AI tools; written and reviewed by our editors. Our AI policy
    • ✓Independence: UpdateNext has no financial relationship with the companies covered.
    1 view
    > react:
    UpdateNext Staff
    The UpdateNext newsroom covers AI and tech from primary sources, every day. Every story is checked against official statements.
    Write for us →
    Comments
    Comments are moderated · Sign in to comment
    > read_next
    MEMORY WALL BREAKTHROUGH
    TSMC draws up plans to enter 3D DRAM stacked memory manufacturing at Nvidia’s request to break the data-center memory wall
    HBM4 SUPERCYCLE
    SK Hynix, Samsung, and Micron report HBM4 memory sold out into 2027 as custom 3nm logic base dies reshape GPU packaging
    FOUNDRY PRICE HIKES
    TSMC and OSAT partners hike 3nm, 2nm, and CoWoS advanced packaging quotes by 10% to 20% as AI accelerator demand triples
    What's next in AI.

    The daily briefing on AI news, tools and breakthroughs.