Jeff Pu
Technology Equity Analyst at GF Securities
Agentic AI demand should drive AMD's 2027 data-center growth well over 100% year on year. Server CPU revenue is forecast to double, with over 5 million Venice units shipping in 2027; Helios racks should ramp mostly in Q1 2027, with over 20,000 expected that year, while MI450/455X volume remains limited by end-2026. Projected 2027 sales are $103bn, with 2027 EPS revised up 15% and a new $720 target at 40x 2027 earnings. x.com
AMD's business is tracking well ahead of its long-term model, with management expecting data-center revenue to more than double in 2027 across CPUs and GPUs. That lifts 2026/27 EPS estimates 8%/23% and the price target from $640 to $700, on 45x 2027 earnings. Q2 revenue rose 50% to $11.5bn and Q3 guidance of $13bn beat consensus, though 56% gross margin was soft. Helios racks ramp in Q3-Q4 with over 20k racks planned, and MI500 should beat Nvidia's Rubin Ultra; AI GPU margins stay below average. x.com
AI data-center spending should reach chips first, then racks, power and cooling, benefiting Nvidia, AMD, Vertiv, Bloom Energy, Credo, Marvell, Lumentum and Micron. BofA forecasts the global data-center market at about $1T in 2026, $2.1T in 2028 and nearly $3T by 2030, a 37% annual rise; roughly 80% of a new AI site's $57M per MW cost goes to servers, networking and storage. x.com
AMD's AI thesis is intact, and the current weakness is temporary macro noise. Raymond James upgraded AMD to Strong Buy with a $641 target, expecting the server CPU market to reach $201 billion by 2030 on a 44% compound annual growth rate as hyperscalers scale for agentic AI. Per Raymond James, AMD offers the cleanest mix of earnings leverage, pure-play data-center exposure and share gains against legacy compute. Hardware demand is accelerating. x.com
AMD risks missing large advance rack orders because Helios is still unproven at volume manufacturing. It looked credible on paper at July's Advancing AI, but this is AMD's first rack-scale system; meaningful volume is expected only in 2H27, two quarters past AMD's window. Hyperscalers order racks a year or two before seeing them, and custom inference chips are squeezing AMD from the other side. The photographed Helios rack's cabling underscores how hard it is to build. x.com x.com
AMD is no longer just NVIDIA's backup, because every AI agent needs its own sandbox, which drains CPU and memory capacity. At Citi's TMT conference, AMD pointed to server CPU growth of 80% year on year in the second half and 70% next year, raised its 2030 CPU market estimate from $60 billion to over $220 billion, and targets half of it, a $100 billion business. MI450 GPU revenue starts in Q3, ramps in Q4 and scales in 2027; the constraints are racks, liquid cooling, HBM and packaging. x.com
On-the-ground data, including OCP rack vendors' reports of what they hear from customers, gives a clearer picture of AMD's CPU outlook. AMD's CPUs and GPUs are both in a growth cycle, leaving it on track to outgrow the market. The 14 million GPU figure through 2030 was always above the team's models, as it was for many others; the 2027 GPU sales estimate raised in the discussion was Morgan Stanley's, not theirs. x.com x.com x.com x.com x.com
AMD's strong Q2 set up a classic sell-the-news reaction against a stock that had priced in perfection. Capex hit $808M, nearly 3x what many modeled, compressing free cash flow. The roughly $13B Q3 guide beat the $12.5B consensus but not whispers near $14B, with growth slowing to about 41%. Management guided no near-term margin expansion, server CPU supply is constrained, and Helios and the multi-gigawatt Anthropic, OpenAI and Meta deals are mainly a 2027 story. x.com
AMD's Helion rack design could benefit from agentic-AI demand beyond GPUs as more advanced agents require more CPU compute. Built for agentic AI, it can support up to 32,768 EPYC CPU cores in a rack, with configurations that also integrate Instinct GPUs. Helion configurations are planned for November, making this a very bullish development for AMD's long-term data center opportunity. x.com
AMD's server CPUs have become a general compute supercycle in their own right, and that is under-appreciated. Management's 2030 server CPU market estimate rose from about $60 billion at Investor Day to over $120 billion and now about $220 billion; such revisions deserve some skepticism, but the business supports them. Server CPU revenue hit a fifth straight record, with cloud and enterprise both up over 70% on higher units and prices, and data-center growth is producing operating leverage. x.com … Show more
AI suppliers' demand may not justify today's investment. Roughly $3 trillion in revenue absent last year is needed to support the spending, yet GDP growth is weak and many companies may be seeing only incremental gains. If the estimate that Anthropic uses two-thirds of its compute for training is right, it could triple inference supply. The most likely outcome for many companies, roughly half the S&P 500, is slowing growth, compressed margins and multiples; when the bubble ends is uncertain. … Show more
AMD has been repriced as the second AI infrastructure incumbent. Meta and OpenAI at 6GW each give it a roughly 12GW MI450 pipeline, both are launch customers for its Helios rack-scale system, Samsung HBM4 supply is secured, and its server CPU TAM estimate doubled to $120B by 2030. Morgan Stanley raised its target from $255 to $360 the morning of earnings. The thesis is intact. x.com
