William Blair upgraded AMD to Outperform from Market Perform on Thursday, Investing.com reported at 5:26 a.m. New York time on 12 November. The firm projects AMD is positioned for more than 80% revenuegrowth next year, more than double its long-term model, and said AMD’s data center AI revenue is set to surpass its server CPU franchise in the fourth quarter, according to the report.
AMD traded at $576.45 before the opening bell at 10:43 UTC, 4.3% above its $552.90 close, on Yahoo Finance’s pre-market quotes. On the 1,640,087,251 shares on its latest 10-Q cover, that is about $38.6 billion of market value, by our arithmetic. Nvidia traded 5.9% lower at $163.50.
Business Pill · GPUs Explained
An evergreen Business Engineer explainer on GPUs and AI hardware. It teaches the general idea only and does not discuss AMD, Nvidia or William Blair.
The key insight: William Blair moved AMD to Outperform and projects more than 80% revenuegrowth next year, Investing.com reported. AMD’s own 10-Q shows Data Center at 63.9% of revenue, up from 47.0%, and before the open AMD traded at 35.9 times next-year consensus against Nvidia’s 10.4 times, by our arithmetic.
More on this: the customer warrants. At AMD’s $552.90 close on 11 November, the 320 million AMD shares that OpenAI and Meta can earn under their warrants come to about $176.9 billion if fully vested, about $14.7 billion per gigawatt on average, by our arithmetic: OpenAI, Meta’s AMD Warrants Are Worth $177B if Fully Vested.
The firm said all four factors that led to its previous Market Perform rating have moved in AMD’s favor since its July initiation, Investing.com reported. According to the report, Intel’s Xeon roadmap “continues to face delays and supply shortages”, while the shift to Arm-based hosts “has proven less decisive than anticipated”.
On valuation, the report says AMD’s multiple “has declined significantly as estimates have moved higher, now roughly in line with its 3-year forward median”. The firm noted that Nvidia “faces increased scrutiny and a potential leadership vacuum”, Investing.com reported. Helios, AMD’s rack-scale design, is ramping up through 2027, according to the report.
William Blair sees competition from custom chips, or ASICs, as a risk to AMD’s Instinct GPUs, particularly at Meta and OpenAI, but views custom programs as an opportunity for CPU attachment, the report says. AMD’s 10-Q lists warrants issued to OpenAI and Meta that are exercisable for up to a combined 320 million shares of its common stock, subject to vesting and exercise conditions.
Pre-market prices at 10:43 UTC on 12 November (Yahoo Finance) divided by next-year consensus earnings per share as used in our 11 November valuation piece (Nasdaq). The multiples are our arithmetic.
What AMD’s 10-Q Shows
AMD’s Data Center segment, which the filing says primarily includes AI accelerators and server CPUs, had net revenue of $8,546 million in the quarter ended 26 September, up 97% from $4,341 million a year earlier, according to the 10-Q filed on 4 November. Total net revenue was $13,377 million, up 45%.
That puts Data Center at 63.9% of AMD’s revenue, up from 47.0% a year earlier, by our arithmetic. The filing credits strong demand for EPYC processors and the launch of its Instinct MI450 Series GPUs. It also says AMD supplies certain products used in Helios rack-scale solutions but does not manufacture or sell the completed racks. William Blair’s AI-over-CPU call for the fourth quarter is the firm’s estimate.
AMD at 35.9 Times, Nvidia at 10.4 Times
On the consensus earnings figures in our valuation piece, AMD’s $576.45 pre-market price is 35.9 times its $16.04 next-year consensus, Broadcom’s $354.50 is 20.3 times its $17.46, and Nvidia’s $163.50 is 10.4 times its $15.75 for the fiscal year ending January 2028, by our arithmetic. AMD’s multiple is about 3.5 times Nvidia’s, up from about 3.1 times at the 11 November closes.
Nvidia’s 5.9% pre-market fall is about $246.5 billion of market value on 24.1 billion shares, by our arithmetic. Five brokers have cut their Nvidia targets or stepped back from the stock since 10 November, most recently Stifel and Bernstein. AMD’s 11 November close was already 36.7% above its 5 November close, as we reported.
Investing.com’s report also lists earlier AMD targets from Oppenheimer ($600), HSBC ($540), StoneX ($685), DA Davidson ($600) and TD Cowen ($675). Their average is $620, and AMD’s pre-market price sits 7.0% below it, by our arithmetic. On software, a slide in Anthropic’s IPO roadshow video put the bring-up of AMD GPUs at about three days with Mythos against about six to 12 months with the traditional approach, for prompt processing on one model, as we reported.
The Structural Read
As we read the note as reported, William Blair is paying for growth rather than a lower multiple: it says AMD’s multiple has come down to about its 3-year forward median because estimates moved higher, while AMD still trades at about 3.5 times Nvidia’s next-year multiple, by our arithmetic.
As we read the 10-Q, the upgrade rests on the segment that already carries the company: Data Center revenue rose by $4,205 million in a year, by our arithmetic, and the filing ties the growth to EPYC demand and the MI450 launch.
William Blair, as reported by Investing.com, 12 Nov 2026
“has declined significantly as estimates have moved higher, now roughly in line with its 3-year forward median”
Three Implications
FOR AMD HOLDERS AMD’s $576.45 pre-market price is 7.0% below the $620 average of the five earlier targets Investing.com lists, by our arithmetic.
FOR THE GPU MARKET William Blair sees competition from custom chips, particularly at Meta and OpenAI, as a risk to AMD’s Instinct GPUs, according to Investing.com.
FOR NVIDIA Nvidia traded 5.9% lower before the open, about $246.5 billion of market value, by our arithmetic, as five brokers have stepped back since 10 November.
The framework reads: “Inference is a revenue engine: continuous, distributed across millions of applications and agents, incurring marginal cost with every query.”
As we read it, William Blair’s case is a bet on that engine running on more than one supplier’s chips: GPUs ramping through Helios and CPUs attached to custom programs. The framework predates the note and does not discuss it.
AMD net revenue by segment, three months ended 27 September 2025 and 26 September 2026, from AMD’s Form 10-Q. Shares of revenue are our arithmetic.
What Is Not Established
We have not read William Blair’s note. The rating change, the growth projection, the fourth-quarter crossover and the risks come from Investing.com’s report, which says it was generated with the support of AI and reviewed by an editor. The 80%-plus growth and the AI-over-CPU call are the firm’s projections, not AMD guidance. Pre-market prices come from thin trading and can change before the 9:30 a.m. New York open; they are not closes.
This Week In The AI Supercycle
The Great Convergence of Capital, Power, and Geopolitics
This week’s issue: for two years the AI build-out was priced as a supply story, and last week as a credit story. This week it was priced as a permission story, and the answer arrived as a cost, not a ban.
William Blair moved AMD to Outperform, saying all four reasons for its Market Perform rating have turned in AMD’s favor and projecting more than 80% revenuegrowth next year. AMD’s Data Center segment was 63.9% of its September-quarter revenue, and its shares traded at 35.9 times next-year consensus before the open, against 10.4 times for Nvidia, by our arithmetic.
A note on sourcing. William Blair’s rating change, projections and risk view come from Investing.com’s report published at 5:26 a.m. New York time on 12 November 2026, which we read in full; MarketScreener carried an MT Newswires headline on the upgrade at 5:24 a.m. We have not read William Blair’s note. AMD’s segment revenue, share count, Helios and warrant disclosures are from its Form 10-Q for the quarter ended 26 September 2026, which we read on SEC EDGAR.
Pre-market prices are Yahoo Finance quotes read at 10:43 UTC; consensus earnings figures are those used in our 11 November valuation piece, from Nasdaq’s forecast pages; Nvidia’s share count is from its latest 10-Q cover. Percentages, multiples, averages and market values are our arithmetic.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur (for example, How to Digitize a Family Business in Europe) and HackerNoon. Find him on LinkedIn and Substack. See how we source.
