Back to News
Market Impact: 0.28

These 2 stocks are set to benefit as Agentic AI lifts CPU intensity

NVDABACDELLHPESMCIAPP
Artificial IntelligenceTechnology & InnovationAnalyst InsightsCompany FundamentalsAnalyst Estimates
These 2 stocks are set to benefit as Agentic AI lifts CPU intensity

Bank of America raised Dell's price target to $246 from $205 and HPE's to $38 from $32, reiterating Buy ratings on both names. The firm argues agentic AI will drive more CPU-intensive, sequenced inferencing workflows, boosting demand for AI servers, infrastructure solution stacks, and storage. BofA estimates Dell has about 12% share of total AI server revenues, while HPE is projected to generate $6.5 billion in AI server revenues this year.

Analysis

The market is likely underappreciating that agentic AI is not just a demand story for accelerators, but a mix-shift toward system-level bottlenecks where CPU, memory, networking, and storage capture more of the wallet. That matters because the incremental dollar in these builds is typically less winner-take-all than GPUs: more vendors can participate, gross margins can be defended through integration, and the revenue pool broadens beyond the chip layer. In that setup, Dell and HPE are better positioned than pure hardware peers because they monetize the architecture transition through both server refreshes and higher-value solution stacks. The second-order effect is on the competitive set. If agentic workloads force more orchestration and state management per task, then storage and networking suppliers should see a lagged but persistent tailwind, while lower-end commodity server OEMs risk margin pressure as buyers spec up configurations. The real medium-term beneficiary may be the ecosystem around enterprise deployment rather than the AI model layer itself, because enterprise buyers care more about throughput per workflow than raw model benchmark wins. The main risk is timing: the thesis is structurally positive over 12-24 months, but near-term revenue inflection could lag the narrative by several quarters if agentic adoption remains experimental. There is also a valuation risk if the market already starts discounting the next upgrade cycle before bookings convert to shipments. Conversely, if AI spend broadens from training/first-pass inference into production agents, the upside could persist longer than consensus expects, especially in the Neo Cloud and enterprise refresh channels.