SoftBank builds Nvidia, TSMC stakes under Masayoshi Son's focus on AI gear
The Japanese technology investor raised its stake in Nvidia to about $3 billion by the end of March, up from $1 billion in the prior quarter, according to regulatory filings. It bought around $330 million worth of TSMC shares and $170 million in Oracle Corp., they show.
That's while SoftBank's signature Vision Fund has monetized almost $2 billion of public and private assets in the first half of 2025, according to a person familiar with the fund's activities. The Vision Fund prioritizes its returns on investment and there is no particular pressure from SoftBank to monetize its assets, said the person, who asked not to be named discussing private information. A representative of SoftBank declined to comment.
At the heart of SoftBank's AI ambitions is chip designer Arm Holdings Plc. Son is gradually building a portfolio around the Cambridge, UK-based company with key industry players, seeking to catch up after largely missing a historic rally that's made Nvidia into a $4 trillion behemoth and boosted its contract chipmaker TSMC near a $1 trillion value. 'Nvidia is the picks and shovels for the gold rush of AI,' said Ben Narasin, founder and general partner of Tenacity Venture Capital, referring to a concerted effort by the world's largest technology companies to spend hundreds of billions of dollars to get ahead. SoftBank's purchase of the US company's stock may buy more influence and access to Nvidia's most sought-after chips, he said. 'Maybe he gets to skip the line.'SoftBank, which reports quarterly earnings Thursday, should've benefited from that bet on Nvidia — at least on paper. Nvidia has gained around 90% in market value since hitting a year's low around early April, while TSMC has climbed over 40%.That's helping to make up for missing out on much of Nvidia's post-ChatGPT rally — one of the biggest of all time. SoftBank, which was early to start betting on betting on AI long before OpenAI's seminal chatbot, parted with a 4.9% stake in Nvidia in early 2019 that would be worth more than $200 billion today.
Crippling losses at the Vision Fund also hampered SoftBank's ability to be an early investor in generative AI. The company's attempts to buy back some Nvidia shares, alongside those of proxy TSMC, would help Son regain access to some of the most lucrative parts of the semiconductor supply chain. The 67-year-old SoftBank founder now seeks to play a more central role in the spread of AI through sweeping partnerships. These include SoftBank's $500 billion Stargate data center foray with OpenAI, Oracle and Abu Dhabi-backed investment fund MGX. Son is also courting TSMC and others about taking part in a $1 trillion AI manufacturing hub in Arizona.As Arm's intellectual property is used to power the majority of mobile chips and is increasingly used in server chips, SoftBank could carve out a unique position without being a manufacturer itself, according to Richard Kaye, co-head of Japan equity strategy at Comgest Asset Management and a long-time SoftBank investor.'I think he sees himself as the natural provider of AI semiconductor technology,' he said. 'What Son really wants to do is capture the upstream and the downstream of everything.'Investors have cheered Son's audacious plans, while analysts say they expect SoftBank to report a swing back to a net income in the June quarter. SoftBank shares marked a record high last month. SoftBank's planned $6.5 billion deal to acquire US chip firm Ampere Computing LLC and another $30 billion investment in OpenAI are further encouraging investors who see the stock as a way to ride the US startup's momentum.Son, however, remains dissatisfied, according to people close to the billionaire. Son sees the big projects in the US as having the potential to help SoftBank leapfrog the current leaders in AI to become a trillion-dollar or bigger company, they said.The stock continues to trade at a roughly estimated 40% discount to SoftBank's total assets — which includes a roughly 90% stake in the $148 billion-valued Arm. SoftBank's market capitalization stands at around $119 billion, a fraction of Nvidia's $4.4 trillion valuation and that of other tech companies most closely associated with AI progress. Son, who in the past has seen Washington hamper or derail merger plans like the union of Arm and Nvidia, seeks to leverage his relationship with Donald Trump and is arranging frequent meetings with White House officials. Those efforts are now critical as AI and semiconductors become geopolitical flash points. SoftBank's plan to buy Ampere is facing a probe by the Federal Trade Commission. Attention at its June quarter earnings will be on what other assets SoftBank might sell down to help it secure the liquidity it needs to double down on hardware investments. The Japanese company has so far raised around $4.8 billion through a sale of some of its T-Mobile share holding in June. Its Chief Financial Officer Yoshimitsu Goto has cited the company's end-March net asset value of ¥25.7 trillion ($175 billion), saying the company has ample capital to cover its funding needs.In the business year ended March, the Vision Fund's exits included DoorDash Inc. and View Inc., as well as cloud security company Wiz Inc. and enterprise software startup Peak, even as SoftBank bought up the stakes in Nvidia, TSMC and Oracle. 'We're after AI using an array of startups and group companies,' Son told shareholders in June. 'We want to become the organizer of the No. 1 platformer in the artificial super intelligence era.' Elevate your knowledge and leadership skills at a cost cheaper than your daily tea. Can Coforge's ambition to lead the IT Industry become a reality?
BlackRock returns, this time with Ambani. Will it be lucky second time?
Amazon is making stealthy moves in healthcare, here's why!
The trader who blew the whistle on Jane Street
Stock Radar: Globus Spirits breaks out from 9-month consolidation; check target & stop loss for long positions
Weekly Top Picks: These stocks scored 10 on 10 on Stock Reports Plus
These large-caps have 'strong buy' & 'buy' recos and an upside potential of more than 25%
Stock picks of the week: 5 stocks with consistent score improvement and upside potential of up to 36% in 1 year
Hashtags
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Hindustan Times
10 minutes ago
- Hindustan Times
OpenAI introduces free customizable AI models: Price, features and more
OpenAI announced two new open-weight AI language models, gpt-oss-120b and gpt-oss-20b, on Tuesday, August 5. These models offer similar capabilities to the company's O-series, as reported by TechCrunch. This marks the first time since OpenAI rolled out GPT-2 in 2019. OpenAI releases two new AI models.(Reuters) The company stated that the text-only models have been specially designed to be used as lower-cost options, which can be easily run and customized by developers and researchers. OpenAI releases lower-cost models OpenAI's latest release comes after tech giants like Meta and Mistral AI, backed by Microsoft, as well as Chinese startup DeepSeek, unveiled their open-weight models in the past few years. According to CNBC, the AI model is considered open weight when all its parameters or elements that have been utilized to improve the output and prediction at the time of training are publicly made available. Such models not only offer control and transparency in the respective field, but also remain quite different from the open-source models, where the full source code is made available to customers to modify and use. OpenAI President Greg Brockman said it is "exciting" to see the developing ecosystem and that the company is looking forward to contributing to it. "...we are excited to contribute to that and really push the frontier and then see what happens from there,' the outlet quoted Brockman as saying. Also read: Google's AI tool 'Big Sleep' flags 20 security flaws in open-source software without human input All about OpenAI's new models Among the two models, gpt-oss-120b is large and holds the capability to run on a single Nvidia GPU, while the gpt-oss-20b model is the lighter version to run on a consumer laptop having 16GB of memory. The two models, which have been dubbed "state-of-the-art," remain freely available to download from the Hugging Face platform. For this project, OpenAI partnered with Nvidia, Advanced Micro Devices, Cerebras, and Groq to see how the models work on various chips. In a statement, Nvidia CEO Jensen Huang said that OpenAI has shown "what could be built on Nvidia AI — and now they're advancing innovation in open-source software.' During a briefing, the company highlighted that its new open models are capable of sending complex queries to AI models in the cloud. This means whenever it is unable to perform a specific task, then the developers will be able to connect it to the more capable closed models of the company, per TechCrunch. The two new AI models, gpt-oss-120b and gpt-oss-20b, underwent a coding test on Codeforces and scored 2622 and 2516. They have outperformed DeepSeek's R1 model. The release of OpenAI's open weight models comes after the company delayed the launch on a repeated basis. Last month, OpenAI CEO Sam Altman informed in an X post that it required extra time to 'run additional safety tests and review high-risk areas.' FAQs: 1. How to try new OpenAI models? As per the company, users can download them from Hugging Face and GitHub under the Apache 2.0 license. These are being made available to run on PCs via LM Studio and Ollama. 2. What are the new AI models from OpenAI? These include open-weight AI language models, gpt-oss-120b and gpt-oss-20b. 3. What's the price for new OpenAI models? They are free to download and users can customize them accordingly.
Economic Times
38 minutes ago
- Economic Times
BDO appointed receiver for Evergreen Power's Mauritius arm
Synopsis In Mauritius, receivers are typically appointed when creditors make claims on a company. "As a policy, Evergreen Power does not comment on market speculation or matters involving internal corporate structure or investor arrangements," the company said in a statement in response to ET's queries on the matter. ET Bureau NEW DELHI: Private equity (PE)-backed Evergreen Power has appointed BDO as a receiver, said people in the Power's Indian operations have been recipients of capital through a Mauritius-based entity where the receiver has been appointed, they Mauritius, receivers are typically appointed when creditors make claims on a company. "As a policy, Evergreen Power does not comment on market speculation or matters involving internal corporate structure or investor arrangements," the company said in a statement in response to ET's queries on the matter. Evergreen Power is executing tenders worth 1 gigawatt for state-backed entities such as Solar Energy Corporation of India (SECI), NTPC and SJVN. Indian banks have provided bank guarantees for projects that Evergreen Power has committed to execute, as per those in the know. "We remain committed to the development and execution of clean energy projects in India and continue to work closely with our partners and stakeholders," the company said. Evergreen Power is backed by UK-based Devonport Capital. The company has won tenders to set up electricity generation through solar, wind and hybrid methods. It has competed with well-known names like Adani Green, Sembcorp, Jindal Renewables and ReNew Power, among others, for these tenders. Evergreen Power was set up in 2012 as per its website, which states it has headquarters in Chicago and Mumbai. It developed its first renewable energy project, a 23 MW solar photovoltaic plant located in Telangana, in a joint venture with Hilliard Energy (US) in April 2016.
Time of India
40 minutes ago
- Time of India
ISMA launches AI–ML based national program to boost sugarcane yield and farmer income
Pilibhit: In a significant move to transform sugarcane farming and uplift rural livelihood in the country especially in three top sugar producing states – including Uttar Pradesh, Maharashtra and Karnataka – the Indian Sugar & Bio-energy Manufacturers Association (ISMA) has launched an artificial intelligence (AI) and machine learning (ML) technology based National AI-ML Network Program in collaboration with Agriculture Development Trust (ADT). A memorandum of understanding (MoU) in thisb regard was signed between ISMA, ADT and Map My Crop (MMC) – an agri-tech platform specialising in AI-driven farm monitoring. The scientific program aimed at enhancing sugarcane productivity, crop quality, sustainability and farmers' incomes, involves satellite monitoring through an IoT (internet of things) device, and is expected to benefit over 50 lakh sugarcane growers' families and around 30 lakh hectares of sugarcane cultivation area in the state. ISMA has so far coordinated with over 30 sugar mills in UP, 20 in Maharashtra and 10 in Karnataka to disseminate the advanced AI-ML technology among the farmers. Officials said in the initial phase, the program is to be launched as a pilot project in an area of 100 acres per sugar mill. "We have set a target to enhance sugarcane productivity by at least 100 quintals per hectare and sugar recovery rate of 11% or more through scientific advancement", said ISMA's managing director Deepak Ballani while speaking to TOI. by Taboola by Taboola Sponsored Links Sponsored Links Promoted Links Promoted Links You May Like No annual fees for life UnionBank Credit Card Apply Now Through ADT's collaboration with Oxford University and Microsoft, a groundbreaking project titled "application of AI, computer vision and machine learning for game-changing improvements in sugarcane crop farming practices" was started last year under which, over 1,000 farmers across the three major sugar producing states were empowered with AI-driven solutions, including machine learning and computer vision tools, to enhance their sugarcane farming practices. "The outcomes reflected a 40% increase in crop yield, 35% reduction in labour costs, 30% savings in water usage, considerable reduction in input costs and overall expenditures, 35% improvement in harvesting efficiency and significant reduction in greenhouse gas (GHG) emissions through precision farming", said Ballani, adding that ISMA already partnered with the Indian council of agriculture research's (ICAR) sugarcane research institutes to promote high-yielding varieties and efficient farm practices. "The program will deliver practical farmer-centric benefits – including location-specific precision farming modules, satellite-based soil health and fertility mapping, real-time pest and disease alerts through remote sensing, image-based weed detection and control, custom weather advisories and water stress monitoring, IoT-enabled precision irrigation for water efficiency, variable rate application (VRA) for optimal fertilizer use and reduction in both input costs and carbon footprint," Ballani added. "The field demonstrations and capacity-building sessions are also planned to promote rapid adoption of the program by farmers in broader aspects", he said.
