As we expected, the concerns that emerged in the final months of 2025 and weighed on the AI trade are rapidly fading. Early earnings reports this season have shifted the focus back to what matters most: impressive growth and expanding margins. And worth noting, the equity rally is broadening. While our core AI portfolio is…
In recent weeks, Nvidia has announced a series of investments with a clear common thread: AI inference, i.e. the running of AI applications. The company signed a $20 billion technology licensing deal with Groq, a specialist in inference chips; took a stake in Baseten, a fast-growing startup providing inference infrastructure for AI applications; and, more surprisingly, acquired a 2.9%…
Bloomberg reported yesterday that Meta and EssilorLuxottica are considering doubling annual production capacity of their Meta Ray-Ban smart glasses to 20 million units by year-end, up from 10 million currently, with the option to scale further to 30 million units if demand warrants. This move, which points to strong demand for AI-powered wearables, follows Meta’s…
For the first time in years, memory has become interesting again (see here and here). Once the most predictable component in a computer’s bill of materials, it has quietly turned into a global bottleneck and artificial intelligence sits at the center of the shift. The disruption begins with high-bandwidth memory, or HBM, a premium version…
Marvell’s $3.25 billion buyout of Celestial AI isn’t just another semiconductor acquisition. It’s a wager—big, loud, and physics-driven—on the idea that the future of artificial intelligence won’t run on copper wires at all. To bankroll the bet, Marvell unloaded its automotive Ethernet division to Infineon for $2.5 billion, then pivoted hard into the fast-forming world…
Meta’s reported spending cuts in its Metaverse division—reportedly as high as 30%—combined with the high-profile hire of Apple’s former head of design, Alan Dye, and its recent push into AI glasses, suggest that the company is rapidly reallocating resources away from VR (Quest headsets) and virtual worlds (Horizon Worlds) toward AI-powered wearables. This pivot should…
The recent correction in AI equities increasingly looks like the two major sell-offs of the past 18 months (February–March this year and July-August 2024), suggesting that we are more likely in a healthy profit-taking phase than in the early stages of a bear cycle. The severity, speed, and magnitude are strikingly similar, with AI and…
Recent capex guidance from hyperscalers has once again been impressive, underscoring the sustained momentum in AI data center buildouts. However, the next phase of this investment cycle will hinge increasingly on monetization — turning AI capabilities into scalable, revenue-generating products. In the past few days, both OpenAI and Microsoft have taken significant steps in that…
Amazon is racing toward a future shaped by automation, but it has not forgotten the humans still on its payroll. Its latest experiment, a pair of smart glasses for delivery drivers, shows how the company aims to boost efficiency without replacing every worker with a machine. The glasses automatically scan packages and provide turn by…
Data centers are evolving at breakneck speed. Artificial intelligence (AI) and high-performance computing (HPC) are pushing power demands to new heights, with next generation AI clusters requiring nearly a megawatt per rack. Traditional alternating current (AC) distribution systems, built for lower loads, are buckling under these demands. Multiple power conversions, bulky cabling, and energy losses…