TECHNOLOGY
April 17, 2026 • 6 min read

From record-breaking chip profits to quantum computing roadmaps and the battle for AI infrastructure dominance, this week in technology has been defined by the growing tension between AI’s insatiable demand for compute and the real-world consequences rippling through consumer markets, public opinion, and enterprise strategy.

TSMC Posts Record Profits as AI Chip Demand Surges

HARDWARE

Taiwan Semiconductor Manufacturing Company reported a 58% jump in first-quarter net profit, marking its fourth consecutive quarterly record. Revenue beat analyst forecasts as advanced chips accounted for approximately 75% of total wafer revenue. The company also raised its annual revenue forecast and announced plans to increase capital spending this year to meet what it described as “extremely robust” demand for AI-capable processors.

The results underscore how the AI boom is delivering tangible, realised profits in the semiconductor supply chain. TSMC’s performance came despite ongoing geopolitical tensions in the Middle East, suggesting that AI infrastructure spending has become largely decoupled from broader macroeconomic uncertainty. The company’s confidence in raising forward guidance signals that demand from hyperscale data centres shows no sign of slowing.

OpenAI Locks In $20 Billion Cerebras Chip Deal

BIG TECH

OpenAI has agreed to spend more than $20 billion over the next three years on Cerebras-powered server capacity, according to reports from Reuters and The Information. The arrangement reportedly includes warrants for a minority equity stake in Cerebras, roughly doubling the size of OpenAI’s previously disclosed agreement with the chipmaker.

The deal highlights a fundamental shift in the AI competitive landscape: model leadership is now inseparable from access to compute infrastructure. By securing long-term capacity with Cerebras, one of the few credible alternatives to Nvidia’s dominant GPU ecosystem, OpenAI is positioning itself to control more of the hardware stack that underpins its models. The move also signals that the battle for AI supremacy is moving from research labs to data centre floors and chip supply contracts.

Meta Raises Quest VR Prices as AI Infrastructure Squeezes Consumer Tech

CONSUMER TECH

Meta Platforms will raise US prices for its Quest VR headsets starting April 19, with the entry-level Quest 3S jumping from $299.99 to $349.99 and the Quest 3 increasing by $100 to $599.99. The company cited surging memory chip costs driven by massive demand from AI data-centre buildouts by OpenAI, Google, and Microsoft.

The price hike illustrates a growing collateral effect of the AI infrastructure boom: consumer hardware is becoming more expensive as data centre operators outbid gadget makers for component supply. Similar pressures have already prompted price increases from Dell, HP, Microsoft, and Sony on its PlayStation 5. Meta’s move comes alongside layoffs of roughly 10% of its Reality Labs team and a broader pivot from metaverse ambitions toward AI priorities.

Google Gemini Adds Personalised AI Image Generation via Google Photos

SOFTWARE

Google has rolled out an opt-in feature connecting Google Photos to Gemini and its Nano Banana image generation model, allowing paid subscribers to create AI-generated images featuring their real-life people and scenes without manual uploads. Users can prompt Gemini with requests such as “create a claymation version of me and my family hiking” and the system pulls context from labelled photos in their library.

The feature represents a significant step toward deeply personalised AI assistants that draw on a user’s existing data ecosystem. By integrating personal photo libraries directly into the image generation pipeline, Google is differentiating Gemini from competitors through its vast repository of user data. The system reportedly pulls limited context such as labels and prompts rather than raw image data, though the privacy implications of linking personal photo libraries to generative AI remain a topic of debate.

French Startup C12 Unveils Quantum Computing Roadmap Targeting 2033

QUANTUM COMPUTING

French quantum computing startup C12 has released a multi-generation product roadmap aiming to deliver utility-scale, fault-tolerant quantum computers by 2033. The roadmap progresses from early logical qubits in its Aïdôs system planned for 2027 to a large-scale system called Panopeia, designed to scale to over 100,000 physical qubits and hundreds of logical qubits.

C12’s approach centres on a unique architecture using purified carbon-12 nanotubes to host spin qubits, leveraging the near-ideal one-dimensional pathway of nanotubes to address the twin challenges of qubit coherence and scalability. The Paris-based company is positioning itself as a European contender in a field currently dominated by IBM, Google, and startups like IonQ and PsiQuantum. If successful, the roadmap would place fault-tolerant quantum computing within reach for practical applications in materials science, chemistry, and financial modelling.

ZY Media Productions

IT • Music • Technology