Cisco has begun a new run after a massive technical breakout from more than 25 years below its previous all-time high, though it likely won't move as fast as AI infrastructure, chip and memory names. Its active venture arm is expanding aggressively into AI and security, including the $28 billion Splunk acquisition, the $400 million purchase of AI security startup Astrix and $1 billion invested in startups such as Cohere, Mistral AI and Scale AI. x.com 

Cisco looks attractive as a winner in the AI revolution outside the hyperscalers. x.com 

Cisco's networking supercycle is no longer a pitch but booked business. Cisco guided fiscal 2027 AI infrastructure revenue to $7.5bn, almost double fiscal 2026, and that revenue is already booked. x.com 

Cisco is drawing aggressive bullish options flow and institutional accumulation ahead of earnings. Someone sold $15 million of January 2028 puts that are $28 in the money, and institutions have accumulated around $117 since the last earnings report broke the stock out of a nearly three-decade base. x.com 

Cisco's business is on fire, with soaring purchase commitments pointing to strong demand ahead. Per Wells Fargo, fiscal Q4 purchase commitments reached $16.033B, up 59% quarter over quarter and 153% year over year, accelerating from 73%, 56% and 47% in prior quarters. Management said silicon supply is secured through 2026 with 2027 negotiations under way, and its direct management of silicon design gives it end-to-end control. x.com  x.com 

Cisco's all-time-high stock price overstates how expensive it is, because it ignores more than 20 years of massive buybacks. Even after today's surge, Cisco's market cap is still below its all-time high. x.com 

Nvidia in 2026 is not Cisco in 2000: it grows over 100% a year at far lower multiples, making it still an amazing buy and likely the first $10T company, possibly by 2028. Cisco peaked near 200x earnings, 30x sales and 176x FCF on 55% growth; Nvidia trades at 30x, 18x and 45x with 75% gross margins. Forward, it is about 20x earnings and 22x FCF, near all-time lows, and supply constraints and contracts let it guide a full year ahead for the first time. A $6T cap would not mean overvalued. x.com  … Show more

Credo Technology sits right in the middle of the AI networking, connectivity and photonics theme, and Wall Street has just handed over an easy opportunity in it. The AI connectivity thesis keeps playing out across photonics, with FY26 revenue up 206% at 68% non-GAAP gross margins. x.com 

Nvidia is undervalued, not the poster child of an AI bubble. Cisco traded near 200x earnings at the dot-com peak, while Nvidia trades around 20x, about a tenth of that; even after Cisco's roughly 90% crash, its P/E was about where Nvidia's is today. Nvidia just delivered 106% year-over-year revenue growth and guided next year's growth above 70%, so the AI boom is not over. Unprofitable $100 billion-plus companies are better bubble candidates. x.com  x.com  x.com 

Nvidia looks more like Apple than Cisco in 2000, because its demand is more durable and its lock-in stronger. Cisco had a few months of cancellable backlog and over-levered telco customers such as WorldCom and Global Crossing that went bankrupt, and its boxes commoditized. Hyperscalers now hold over $2.3T of cloud backlog stretching years, funded by very strong balance sheets. CUDA, accelerators, networking and systems integration create switching costs behind margins nearly twice Apple's. x.com  … Show more