Jukan
Semiconductor & AI Infrastructure Research at Citrini
Broadcom's roughly 60% margin looks unsustainable because its custom-chip design role is not irreplaceable. Google's TPU v8i has already succeeded, and it is being handled by MediaTek, whose margin is at best in the 30% range. That undercuts the claim that customers need Broadcom's design capability to compete with Nvidia, especially when Broadcom's margin is not far from Nvidia's own. x.com x.com x.com x.com x.com x.com
Broadcom and MediaTek risk losing chip ownership and margin as hyperscalers pull custom silicon in-house for platform control and margin. @jukan05 relays a rumor that Google will move to customer-owned tooling from TPU v9, ordering wafers directly from TSMC like AWS's Annapurna and reducing Broadcom and MediaTek to IP and design-service roles at lower per-chip margins. Even if not now, it will happen eventually, because the end state is all about optimization. x.com
Broadcom's top growth driver, Google TPUs, can be captured elsewhere, such as through Vicor, while its OpenAI exposure through the Jalapeño chip is unwanted in light of today's news, even if that proves a nothingburger. Broadcom should still do great and may be worth re-entering later, but the capital can find better ways to outperform it. x.com x.com
Broadcom's integrated optics, electronics and packaging could lower custom-chip system costs because separate suppliers defend their margins. At similar valuations, it is preferred to Nvidia and has more room to beat expectations. Google TPU v8 shipments are estimated at 4.0M in 2026 and 9.1M in 2027, assuming Broadcom supplies the remainder after MediaTek's allocation. Anthropic financing adds value only if chip profit exceeds likely support payments and capacity costs. x.com x.com
Broadcom is depending on a massive Anthropic and OpenAI ramp over the next two years as Google becomes less and less important to it. CEO Hock Tan gave exact gigawatt figures for every customer except Google, which got only a vague tens of billions, a bad sign that MediaTek has taken share. x.com
Broadcom's valuation is likely to price in margin compression even if its actual margins hold, because squeezing leading suppliers' margins is the core aim of its four big customers. Chips are now production tools for cloud providers, so saving money is making money, and they shift to lower-margin second suppliers wherever they can. Real shortages make that hard near term, but the market prices it in early, as with Nvidia's 12x P/E. Trading can diverge from fundamentals. x.com x.com x.com x.com
Broadcom's custom AI chips for major customers position it as a clear winner in the continuing AI buildout. Kris Patel's case that efficiency and total cost of ownership leave Broadcom room to gain share from incumbent Nvidia is persuasive. Broadcom has not seen peers' upside and is 30% below its highs, making the stock an opportunity. x.com x.com
Broadcom's Jalapeño chip deal with OpenAI is already priced in and margin dilutive. OpenAI guides to about 30 GW by 2030; at 30-40% Jalapeño, that is 9-12 GW, matching the 10 GW deal signed in October 2025. Selling finished chips, Broadcom books full ASPs but fronts wafers, HBM and other costs, about $15 billion per GW or $25-45 billion a year, at 50-60% margins at best. FY27 guidance already tops $100 billion; the bull case is maybe $120 billion. Broadcom even lends OpenAI the build money. … Show more
MediaTek has ample ASIC opportunity as Google's TPU ramp accelerates and Broadcom's pricing becomes a problem for Google. Per Mizuho, TPU shipments could top 35M in 2028, about 8x 2026, while Broadcom's CoWoS plus N3 package has become expensive. MediaTek's potential challenges are 224G SerDes IP integration and EMIB-T packaging yields, though checks by @sssjeffpu put EMIB-T above 95%, so yield should not be a major problem now. Intel and Broadcom also have ASIC opportunities. x.com
Broadcom's custom AI chips and networking could drive earnings beyond what its valuation implies over 2-3 years. It is the favorite semiconductor opportunity now, preferred over NVIDIA without replacing it. AI semiconductor revenue is projected at about $58B in FY2026, $115B in FY2027 and potentially $230B in FY2028. At about 17x FY2027 consensus adjusted earnings, risk/reward is compelling if growth materializes; customer commitments add visibility, but delays and financing remain risks. x.com … Show more
NVIDIA, AMD, Broadcom, Marvell, MediaTek and Cadence could not look more bearish, because hyperscalers don't want to work with them and are incentivized to cut them out. NVIDIA in particular is being disintermediated by Google's TPU program while priced like a monopoly; a post the author replies to notes it is now worth more than the GDP of every country except the US and China. x.com
