Silicom's existing designs are still ramping, so FY27 should be strong even from today's wins, though more design wins are needed. A roughly flat sequential Q4 guide does not make the quarter poor: Silicom posted a double beat, raised guidance and is close to profitability. It has tailwinds beyond its partial AI exposure at about 3x sales, but the market remains unmoved. x.com  x.com  x.com 

Silicom is a recovering custom-networking supplier with real but early option value in AI inference networking, post-quantum cryptography and white-label switching. 2025 revenue rose to $61.9m, Q4 grew 17% and Q1 2026 guidance implies about 18% growth; it won 8 design wins in 2025 and targets 7-9 in 2026, several worth $1m-$10m a year. It runs at ~41% of 2022 revenue and loses money, and inventory at 84% of revenue risks write-downs. With ~$74m of cash and securities and no debt, EV is ~$69m. … Show more

Silicom's 58% year-over-year revenue growth is what drove the stock, and the next leg depends on that revenue turning into earnings. It still loses money, with a -20% operating margin last fiscal year, so the margin line now decides the case. The earlier view rested on 3.4 times sales on 33% growth, its first AI inference production order and a blue-chip design win. x.com 

Silicom adds quantum-security exposure, and the positive news looked like a good buying opportunity. Per the FT, a customer selected a Silicom FPGA SmartNIC after evaluating an off-the-shelf product that showed the performance and reliability needed for its advanced encryption, including post-quantum cryptography. x.com 

Semiconductors could capture scarce spot compute's value if its forward curve holds. Compute trades in deep contango, while chip stocks price in little of that value; near-term moves are unpredictable. That makes the sector very bullish. The earlier lab-funding risk looks less acute: Anthropic should hold nearly $200B cash after its IPO and is operating-cash-flow profitable. Even with 2029 capex cuts, chips look cheap if future buyers do not default. x.com  x.com  x.com  x.com  x.com 

Financing announcements point to more accelerator demand, supporting semiconductors. About $200bn in announced Meta, Anthropic/Broadcom and SpaceX/Nvidia financing likely drives $400bn in XPU purchases. Data centers burn cash until lit, then cash flow 20-40% annually; AI spend rising from 0.1% to 0.5-1% of current payroll could balance the buildout if agents replace roles. Sustainable token demand is the bet, but smaller local models are the bigger risk, though not an immediate concern. x.com  … Show more

High open-source inference margins imply a need for much more compute supply, supporting chipmakers and at least 2027 consensus estimates. DeepSeek is earning roughly 90% gross margins with ten-month GPU paybacks. Those returns look too high for a competitive, low-switching-cost product; more supply should increase competition and move pricing toward returns near the cost of capital, though margins need not fall. Labs face the demand question today; the supply chain probably faces it from 2028. … Show more

VeriSilicon is China's Broadcom and should take off as custom inference ASIC orders concentrate at mature ASIC design service providers. Jalapeño's 9-month tape-out shows inference ASIC design barriers are lower than assumed, because the slow work of place-and-route, timing signoff and foundry adaptation sits with turnkey platforms. Every hyperscaler wants custom inference chips but won't build back-end teams, so pipelines at Broadcom, Marvell, MediaTek and VeriSilicon should thicken. x.com  … Show more

Broadcom could roughly triple to $900+ per share if 30x is applied to Hock Tan's expectation of over $30 of EPS in FY2028, versus about $350 today. ASICs are expected to take about 25% of the AI accelerator market by 2030 as inference explodes, and management has consistently delivered. There is no guidance beyond 2028, and near-triple-digit AI chip growth is hard to sustain, but growth should remain strong. x.com  x.com 

The market still underestimates Broadcom because its AI ramp is limited by how fast data centers come online, not by demand. Broadcom says it has a "high degree of confidence" in shipping $350B over two years even if the full 30 GW pipeline is never built; supply is secured through 2028 and demand exceeds the FY27 outlook. AI semis grew 221% with 68% operating margins, helped by networking sales. Customer concentration and financing guarantees are real risks. x.com