AMD Helios Platform and MI400 Series Accelerators Move AI Data Centers to Double-Wide Racks

Updated

AMD Helios Platform and MI400 Series Accelerators Move AI Data Centers to Double-Wide Racks1

Advanced Micro Devices (AMD) reported blowout second-quarter 2026 results, demonstrating that its investments to challenge Nvidia's AI chip dominance are delivering accelerating returns. Driven by robust demand for EPYC server CPUs and Instinct AI accelerators, AMD’s quarterly Data Center revenue surged 107% year-over-year to a record $6.72 billion (accounting for 58% of total sales). Total revenue for the quarter jumped 50% year-over-year to $11.54 billion, beating consensus estimates of $11.28 billion, while non-GAAP gross margin expanded to 56%.

The 2027 Guidance Step-Up and Gigawatt-Scale Wins

On the earnings call, CEO Dr. Lisa Su provided a massive upward revision for AMD's long-term outlook, stating that the company expects its overall Data Center segment revenue to more than double (well over 100% growth) in 2027 off a much higher base. This acceleration is driven by the upcoming ramp of Helios, AMD's rack-scale AI platform combining EPYC Venice CPUs, MI450 series GPUs, Pensando networking, and ROCm software.

AMD announced several landmark gigawatt-scale commitments for Helios and its MI450 series:

  • Anthropic Partnership: Anthropic will deploy up to 2 gigawatts of MI450 series GPUs in Helios racks, with deployment of the first gigawatt beginning in the first half of 2027. This deal includes a multi-year engineering collaboration using Claude to optimize workloads and accelerate ROCm software development.
  • Microsoft Azure Expansion: Microsoft will deploy Helios at scale on Azure for frontier model inferencing across its internal services and Azure AI customers.
  • OpenAI and Meta: Ramping multi-generation, gigawatt-scale Instinct deployments.

Accelerating Server CPU Market Share

AMD is also capturing significant market share in server CPUs. Lisa Su noted that server CPU revenue grew over 70% year-over-year in the cloud and enterprise segments during Q2 2026. The company expects server CPU revenue to grow more than 80% year-over-year in the second half of 2026 and more than 70% for the full year 2027. This growth is backstopped by the upcoming launch of sixth-generation EPYC Venice CPUs, built on Zen 6 cores and a 2nm process, which deliver up to a 3.3x performance-per-watt advantage over leading Arm-based processors.

Long-Term Market Revisions

AMD raised its total addressable market (TAM) projections for high-performance and AI computing, now forecasting the market to approach $2 trillion by 2030 (growing ~40% annually). Within this TAM, AMD expects the data center AI accelerator market to reach $1.4 trillion (growing >45% annually) and the server CPU market to reach $220 billion. As a result, AMD expects to "significantly exceed" its long-term target of $20 annual EPS within its strategic timeframe.

Verbatim Quotes

  • Dr. Lisa Su on the 2027 Data Center ramp:

    "The data center AI TAM is certainly larger, as a result, as we ramp the large strategic customers that we have on Helios in 2027, you would expect substantial growth in that business in 2027... We believe the overall segment will grow by over 100%, more than double, due to the data center AI ramping."

  • Dr. Lisa Su on the Anthropic 2GW Helios partnership:

    "Anthropic will deploy up to 2 gigawatts of MI450 series GPUs in Helios, with deployment of the first gigawatt beginning in the first half of 2027. The partnership includes a multi-year joint engineering collaboration using Claude to optimize workloads for Instinct CPUs and accelerate ROCm software development."

  • Dr. Lisa Su on server CPU growth:

    "In server, we delivered our fifth consecutive quarter of record server CPU revenue, with cloud and enterprise sales each growing more than 70% year-over-year, exceeding the outlook we provided last quarter... We now expect server revenue to grow more than 80% year-over-year in the second half of 2026 and more than 70% for the full year 2027 off a much higher base."


  1. An instance of Scaling agentic AI clusters breaks traditional rack dimensions and networking structures. — The extreme physical demands and performance scaling of next-generation accelerators are forcing data center designs to transition to double-wide rack architectures. ↩︎

Part of

This finding is an example of a pattern recurring across your work:

Revision history

  • Update AMD's AI platform and Instinct roadmap with blowout Q2 2026 earnings, massive 2027 data center segment doubling guidance, 2GW Anthropic Helios contract, and EPYC Venice 2nm CPU details.
    · by the agent
  • Update AMD's AI platform and Instinct roadmap with blowout Q2 2026 earnings, massive 2027 data center segment doubling guidance, 2GW Anthropic Helios contract, and EPYC Venice 2nm CPU details.
    · by the agent
  • Update AMD's AI platform and Instinct roadmap with blowout Q2 2026 earnings, massive 2027 data center segment doubling guidance, 2GW Anthropic Helios contract, and EPYC Venice 2nm CPU details.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent
  • Updated without a stated reason.
    · by the agent