🤖 SoftBank Plans to Offer AI Compute in US at 10-Gigawatt Scale

SoftBank Group Corp. and its telecom unit will start renting AI computing resources to US companies next fiscal year, using a growing pipeline of data centers to compete with the likes of CoreWeave Inc. and Nebius Group NV.

The two SoftBank companies will set up SB Neo Inc. to offer AI chips and cloud services to big companies including hyperscalers. The so-called neocloud venture plans to ramp up its computing resources to eventually supply data center capacity at a scale of 10 gigawatts by around 2030 for large-scale AI model training and inference, according to Junichi Miyakawa, head of the telecom unit.

@QSIMedia

🇨🇦 Canada’s Carney Taps State-Owned Firm to Build Oil Pipeline to Serve Asia

Trans Mountain Corp., a pipeline operator owned by the Canadian government, has been tapped to build a new export conduit connecting Alberta’s oil sands to a Vancouver-area port, Prime Minister Mark Carney said.

The proposed 1 million-barrel-a-day pipeline will link to a deepwater port capable of receiving Very Large Crude Carriers, or VLCCs, with the goal of “meeting significant Asian demand from countries like Japan, Korea, China and India,” according to the government of Alberta.

@QSIMedia

🔥 Tokenization Can Change the World's Financial Architecture

Policy decisions on money, market infrastructure, and legal frameworks will determine whether tokenization strengthens or fragments the financial system as assets migrate onto shared digital ledgers, IMF Monetary and Capital Markets Director Tobias Adrian said.

⬇️ What really changes?

🟠Tokenization goes a step beyond simple digitization by embedding ownership and transfer directly within the asset itself.

🟠Liquidity demands materialize in real time, collateral calls can be automated, and failures can propagate faster than institutions or supervisors can respond. Risk that once were borne by the balance sheet of individual institutions behind a transaction become increasingly concentrated in the platforms and code that govern these transactions.

🧱 Tokenization enables multiple forms of digital money to circulate on shared ledgers. Three forms are emerging.

→ Tokenized bank deposits
→ Stablecoins
→ Tokenized central bank reserves

🏦 Banks will change, not disappear. Tokenization does not eliminate banks. It changes how they fund themselves, manage liquidity, and bear risk.

⚖️ Legal foundations matter just as much. Market participants must know whether tokenized records constitute definitive ownership, whether settlement finality is legally recognized, and which jurisdiction’s law applies. Without clarity, tokenization will remain fragmented and peripheral.

❗️The future of tokenized finance will be determined by a complex set of decisions that policymakers will have to make about issues such as the role of public and private money; the degree of interoperability; legal frameworks; code governance; liquidity backstops, and others.

The best outcome would be of a system that provides elements of the required public goods such as risk-free settlement assets and internationally aligned oversight, while encouraging and enabling desirable features such as interoperability. 🚀

@QSIMedia

💻 Anthropic moves to close loopholes that allow Chinese access to Claude

Anthropic is tightening its detection systems to block firms like Ant Financial and ByteDance, which have used VPNs and overseas subsidiaries to access its advanced AI models.

@QSIMedia

🤑 PwC Highlights Stellar as Enterprise-Grade Infrastructure!

➡️PwC (PricewaterhouseCoopers) is a global network of professional services firms that provides audit and assurance, tax, legal, consulting, and advisory services to businesses, governments, and nonprofit organizations

PwC works with 80%+ of the Fortune 500 and in their June tokenisation report they highlight Stellar as enterprise-grade infrastructure for institutional finance. 🚀

❗️PwC has already worked with Stellar before through U.S. Bank’s stablecoin testing.

#Stellar

@QSIMedia

🤖 SK hynix, Samsung, Micron among semiconductor industry group lobbying against government intervention on domestic memory chip supply

Government attempts to address the global memory chip shortage by influencing prices or production capacity would worsen a historic squeeze on supply driven by the AI boom, a semiconductor industry group warned the Trump administration.

In a letter to senior administration officials, the SEMI industry association urged the US instead to allow companies to continue striking long-term agreements with customers and extend tax breaks aimed at increasing output in the US.

Apple is attempting to address this shortage by lobbying the government to let it purchase memory chips from blacklisted Chinese firm CXMT. SEMI takes a different approach for a temporary solution to the crisis. Instead of allowing a new player to enter the U.S. market, it suggested that the administration should work with Congress to pass tax deductions or credits on consumer electronics, thus offsetting the rising prices caused by the chip shortage.

@QSIMedia

⚡️🇺🇸Donald Trump: "My Administration is delivering another fantastic America First Victory."

The Department of Treasury will now accept philanthropic contributions of readily tradable Publicly Traded Stock to help fund TRUMP ACCOUNTS for eligible American children. This opens the door for GREAT American Businesses, Philanthropists, and Job Creators to invest directly in the future of our children, and help build a Generation of Savers, Investors, and Owners.

@QSIMedia

🤖 New Multi-Site Computer Austria (or MUSICA) supercomputer performs 45,110 trillion calculations/sec with 1,088 Nvidia chips

Austria has just launched one of the world’s 100 fastest supercomputers, which is equipped with 1,088 Nvidia chips and can carry out 45,110 trillion calculations per second. It is reportedly among the 100 fastest supercomputers in the global TOP500 ranking. More than 400 of them run on Nvidia technology (81%).

MUSICA is equipped with 1,088 high-end Nvidia H100 GPUs and cloud servers. These are highly advanced, enterprise-grade processors specifically designed for data centers and massive AI workloads. The new system delivers 45.11 petaflops of performance. This means it can carry out more than 45,110 trillion calculations per second. It is more than eight times the computing power of the VSC-4 and VSC-5 supercomputers.

@QSIMedia

🇩🇪 German Local Banks Expand Crypto Trading To Millions Of Retail Customers

Cryptocurrency trading is set to become more accessible in Germany as the country’s cooperative and savings banks expand digital asset services, allowing millions of retail customers to buy and sell cryptocurrencies directly through their local lenders.

⬇️ The move marks a significant shift for Germany’s traditionally conservative banking sector, which had previously resisted offering crypto services to retail investors because of concerns over volatility and investor protection.

🔵Cooperative banks have already begun rolling out crypto trading through a platform developed by DZ Bank, enabling customers to trade cryptocurrencies including Bitcoin, Ethereum, Litecoin, and Cardano.

🔵Meanwhile, DekaBank is preparing to launch a similar platform for Germany’s savings banks later this year, with the rollout expected to take place in phases.

➡️Each local institution will decide whether to offer the service, though industry participants expect widespread adoption. Representatives from DZ Bank said interest among member banks has been strong, with hundreds expected to introduce crypto trading over time.

Supporters say integrating digital assets into existing banking platforms could bring cryptocurrencies to a broader audience by allowing customers to trade through institutions they already know and trust. 🚀

As German cooperative banks begin offering crypto access through familiar banking apps, digital assets could become a more common part of everyday investment services across Europe. 🇪🇺

@QSIMedia