You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
The logo of SoftBank Group Corp is displayed at SoftBank World 2017 conference in Tokyo, Japan, July 20, 2017. REUTERS/Issei Kato
11 Nov 2025 10:28PM(Updated: 11 Nov 2025 10:55PM)
SoftBank’s decision to sell a $5.8 billion stake in Nvidia has rattled investors and reignited fears of an AI-driven market bubble. The Japanese conglomerate, led by Masayoshi Son, reportedly trimmed its holdings as Nvidia’s valuation soared past $3 trillion, making it the world’s most valuable chipmaker. The sale comes amid concerns that AI-related stocks have surged too far, too fast — echoing the exuberance of the dot-com era.
SoftBank, which once suffered massive losses on speculative tech bets through its Vision Fund, appears to be taking profits while enthusiasm around artificial intelligence remains red-hot. Nvidia’s dominance in AI computing has made it the face of the boom, but its meteoric rise — more than 200% in a year — has drawn warnings from analysts who see parallels with previous market manias.
The move signals a more cautious stance from SoftBank, which has been rebuilding its balance sheet and exploring a new AI-focused investment push. However, investors interpret the timing as a potential red flag that even early believers in AI’s promise are hedging their exposure. As markets digest SoftBank’s sale, the question remains whether this marks the start of a broader cooling in the AI rally or just prudent profit-taking.
SoftBank Group's $5.8 billion sale of its Nvidia stake jolted stock markets on Tuesday, stoking fears that the frenzy around artificial intelligence may have peaked, especially after recent warnings from Wall Street bank chiefs and a famed short seller.In its quarterly results, the Japanese tech...
Agreed. Sofbank is taking profits as Softbank normally does. The title is pure clickbait but since there is no punishment for clickbait we will continue to be a clickbait society. Non semiconductor people will get dumber which will make us experts more popular. I am definitely an upside guy. It is in my DNA.
My concern is the Intel stake. Hopefully Softbank holds on to that one for a bit longer. I am hearing more chatter about Intel and Tesla doing a deal. I am also hearing more positive notes on 18A. 2026 should be a good year for Intel, absolutely. Hopefully Reuters will stop calling Intel "beleaguered", "struggling", "faltering" or "embattled".
but this is a bit more than just taking profits - they sold 100% of their stake. They might be looking at this as "Nvidia has more competition coming on hardware, so the (stock price) gains are more likely to be on the software side going forward".
My concern is the Intel stake. Hopefully Softbank holds on to that one for a bit longer. I am hearing more chatter about Intel and Tesla doing a deal. I am also hearing more positive notes on 18A. 2026 should be a good year for Intel, absolutely. Hopefully Reuters will stop calling Intel "beleaguered", "struggling", "faltering" or "embattled".
Intel news and PR might be positive, but the indicators are not good. External Foundry revenue dropping from trivial to non-existant. Little to no growth planned in the greatest boom of 20 years. 18A loadings and Fab 52 loadings are low until 2027 at least (peaks in 2030). Stock has skyrocketed. some people might say selling is a good idea.
The improvements in margin are listed as related to headcount cuts in the 10Q (Those were good decisions to cut headcount by both CEOs)
Side note: when the smoke clears, I expect Tesla to have a "deal" with multiple companies and then decide who really gets money. If they buy a factory off of Intel, that would be a huge Intel win.
Intel news and PR might be positive, but the indicators are not good. External Foundry revenue dropping from trivial to non-existant. Little to no growth planned in the greatest boom of 20 years. 18A loadings and Fab 52 loadings are low until 2027 at least (peaks in 2030). Stock has skyrocketed. some people might say selling is a good idea.
The improvements in margin are listed as related to headcount cuts in the 10Q (Those were good decisions to cut headcount by both CEOs)
Side note: when the smoke clears, I expect Tesla to have a "deal" with multiple companies and then decide who really gets money. If they buy a factory off of Intel, that would be a huge Intel win.
It will be another year or two before the Lip-BU Tan effect kicks in. The wafer agreements he negotiates this year will not see meaningful revenue for 2-3 years. Lip-Bu has said the agreements need to be signed, sealed, and delivered, before he goes public. Set expectations then beat them, right? Quite the difference from what Pat Gelsinger was doing.
Elon has a much better chance dealing with Lip-Bu than Samsung. I have experienced both sides of this and can tell you Samsung is a nightmare to work with from an American business standpoint.
The only angle I am not seeing yet which I would like to see is JD Vance promoting the Intel fabs in Ohio. JD is from Middletown, Ohio, a small industrial city located between Cincinnati and Dayton. The Ohio One Intel campus is Licking County, Ohio which is not far from Middleton. Politics is all about influence peddling and Vance only has 3 more years to do get Ohio One built and scaling. He really needs to get started!
