Bernstein Server OEMs: Agentic AI Driving Next Leg of AI Infrastructure Build
Summaries are public. PDF access requires an active membership; all members have the same access. Sign in
Bernstein asserts that agentic AI will significantly increase CPU-intensive workloads, driving a massive expansion in the traditional server total addressable market (TAM) and benefiting server OEMs. The report estimates the traditional server TAM could grow 2.5 to 4.5 times by 2030 from approximately $213 billion in 2025. Furthermore, Super Micro Computer's (SMCI) recent compliance issues may trigger market share losses, potentially benefiting Dell and HPE.
Agentic AI inference relies more heavily on CPUs for multi-step task execution, retrieval, and verification loops. AMD projects server CPU TAM growth exceeding 35% annually, potentially surpassing $120 billion by 2030. Intel notes the inference CPU-to-GPU ratio has compressed from 1:8 to 1:4.
SMCI's FQ3'26 non-GAAP EPS of $0.84 beat estimates by 34%, but this was driven by lower-margin enterprise clients rather than AI servers. Dell is identified as the optimal beneficiary due to its structural advantages in scale and balance sheet financing capabilities, positioning it to absorb SMCI's displaced cloud and sovereign clients.
Consequently, Bernstein raises Dell's target price to $280 with an Outperform rating and HPE's target to $35 with a Market-Perform rating. SMCI maintains a Market-Perform rating with a $37 target, reflecting a downside risk-reward profile amid ongoing compliance challenges.
Key exhibits
Exhibit 1
Exhibit 2
Exhibit 3
Exhibit 4
Exhibit 5
Exhibit 6
Exhibit 7
Exhibit 8
