Key facts
- Microsoft is hinting at new devices that will not host traditional operating systems or apps, instead featuring AI agents.
- Microsoft unveiled its AI-designed Majorana 2 quantum chip, aiming for commercially useful quantum machines by 2029.
- IBM plans to invest over $10 billion in quantum computing over the next five years.
- IBM aims for a large-scale, fault-tolerant quantum computer by 2029.
- Microsoft announced seven new in-house AI models at its Build conference.
- Microsoft reported Q3 earnings with EPS of $4.27 and revenue of $82.89 billion.
- CEO Judson Althoff sold $7.1 million in stock.
- Applied Digital signed a 15-year lease for 210 megawatts of computing capacity with a U.S. hyperscaler.
- The Applied Digital lease is expected to generate approximately $5.2 billion in revenue.
- Microsoft announced Project Solara, a new platform for devices that run AI agents instead of traditional apps.
- Microsoft launched the Surface RTX Spark Dev Box, capable of running a 120 billion parameter AI model on-device.
- Anyscale has launched a public preview of its platform on Microsoft Azure.
Microsoft is signaling a significant shift in personal computing, hinting at new devices that will operate with AI agents instead of traditional operating systems and applications. This vision is embodied in Project Solara, a platform for devices running AI agents, with prototypes demonstrated for healthcare and retail tasks utilizing Qualcomm and MediaTek chips. The company also launched the Surface RTX Spark Dev Box, capable of running a 120 billion parameter AI model on-device. Microsoft is also bolstering its AI capabilities by developing its own in-house AI models, aiming for fine-tuning and cost optimization, thereby reducing its reliance on external providers like OpenAI. This move follows strong Q3 earnings, with EPS of $4.27 and revenue of $82.89 billion, though CEO Judson Althoff sold $7.1 million in stock. Additionally, Microsoft is leveraging AI to enhance its cybersecurity defenses, proactively identifying and neutralizing cyber threats. The company's new AI agent, Scout, is designed for autonomous task completion and aims to foster user 'addiction' through deep workflow integration, built on OpenClaw and WorkIQ.
In the realm of quantum computing, Microsoft has unveiled its AI-designed Majorana 2 quantum chip, with a revised target of commercially useful quantum machines by 2029, shortening its previous timeline. IBM is making a substantial commitment to quantum computing, planning to invest over $10 billion in R&D, manufacturing, and partnerships over the next five years, with the goal of a large-scale, fault-tolerant quantum computer by 2029. The broader quantum computing landscape faces challenges, as qubits are fragile and prone to errors, requiring thousands of physical qubits for reliable logical qubits, suggesting practical applications remain distant. However, breakthroughs are occurring, with Chinese scientists developing the world's first superfast quantum memory to address data-reading bottlenecks, potentially aiding fields like drug discovery and financial fraud detection. The U.S. and Japan are launching a $1 billion partnership, with each nation investing $500 million, to advance AI, quantum computing, and fusion energy. France's Quobly has raised €115 million for its silicon-based quantum computers, aiming to leverage existing chip infrastructure. MicroAlgo Inc. has developed reconfigurable simulation technology for quantum algorithms, and Quantum Secure Encryption Corp. is creating technology to protect AI data from hackers, citing increased government investment in quantum computing as a driver for urgency. The Trump administration has publicly endorsed D-Wave's second quantum computing approach.
Beyond these core developments, the AI sector sees Applied Digital signing a 15-year lease for 210 megawatts of computing capacity at its Delta Forge 2 AI Factory campus with a U.S. hyperscaler, a deal projected to generate approximately $5.2 billion in revenue, potentially reaching $12.7 billion with renewal options. Meta is investing $115 million in a data center jobs training program, America's Workforce Academy, to support its AI infrastructure buildout. Anyscale has launched a public preview of its platform on Microsoft Azure, enabling enterprises to run large-scale AI workloads within their Azure tenancy, potentially saving up to 90% on API costs. Taiwan Semiconductor's (TSMC) advancements in AI are viewed positively for the long term by Wedbush analyst Daniel Ives. The UK plans to purchase AI chips from domestic technology companies to incentivize their presence in Britain.