Subscribe to our newsletter
Get the latest updates delivered to your inbox every day, and stay up-to-date for free 🧠📈
Share
Linkedin
Twitter
Facebook
Whatsapp
Copy Link
Meta Platforms erased its 2026 losses with a 15% weekly surge, its strongest performance since early 2024. The rally came after the company unveiled plans to sell AI computing capacity through Meta Compute and launched new AI models including Muse Spark 1.1 and Muse Image. Wall Street's enthusiasm signals a shift in sentiment around Meta's massive AI infrastructure spending, though the company has yet to prove it can compete with established cloud providers.
China announced an immediate temporary ban on helium exports as renewed US-Iran military conflict threatens global supplies of the gas critical for semiconductor production. The country imports around 85% of its helium needs but has become a key intermediary for Russian helium re-exports to Europe, raising concerns about tightening global availability.
Taiwanese chipmaker Nanya Technology announced plans to spend $6.2 billion in 2027, quadrupling its current investment, as it races to expand memory chip production amid an AI-driven shortage. The company reported second-quarter revenue up 684% year-over-year, with gross margins climbing to 79.5% as customers including Nvidia, Qualcomm, and Google drive demand for AI-specific memory products.
AMD confirms its next-generation Epyc 'Venice' processors will debut at the Advancing AI 2026 conference on July 22 in San Francisco. Built on TSMC's 2nm process, the Zen 6-based EPYC server CPU lineup features up to 256 cores and promises up to 1.7 times faster performance for AI workloads compared to previous-generation chips.
Nvidia has shed nearly $1 trillion in market value over two months, driving its forward P/E ratio to its lowest point since early 2019. Despite the steep correction, the company continues posting record revenue growth with Data Center sales reaching $75.2 billion. Wall Street analysts see the selloff as a buying opportunity, with Bank of America projecting $20 billion in new Vera CPU sales and maintaining a $350 price target.
Applied Materials shares surged after CEO Gary Dickerson revealed the company has tremendous visibility into customer demand for the next 24 months. Mizuho raised its price target to $650, projecting wafer fabrication equipment spending to reach $192 billion in 2027 as AI infrastructure growth accelerates across memory and foundry sectors.
Rocket One Inc. shares jumped 21% following its acceptance into the SpaceXAI API program, granting access to multimodal artificial intelligence models. The company plans to integrate these capabilities into its AI-first platform focused on defense technologies and space computing applications.
Arm, Intel, and AMD are redesigning their next-generation CPUs to handle AI workloads directly, moving computation away from cloud servers and power-hungry GPUs. The shift promises faster on-device assistants, better privacy, and significantly improved battery life for phones and laptops running Windows 11 and beyond.
SK hynix and TetraMem have developed an experimental analog in-memory computing chip that performs AI calculations directly within memory arrays, achieving 21.3 TOPS/W energy efficiency. The memristor-based system-on-chip targets edge AI devices by eliminating data movement bottlenecks, though performance remains limited at 2.54 TOPS peak throughput.
AMD's stock has surged over 140% as its CTO Mark Papermaster reveals a critical shift in AI hardware requirements. Agentic AI workloads need roughly four times more CPU resources for orchestration and reasoning, not just GPU power. This insight is reshaping how enterprises design AI infrastructure and driving AMD's evolution from chip maker to system optimizer.
Samsung is sampling its Gaia AI chip with major PC makers HP and Lenovo, marking its return to PC processors after 13 years. Built on 4nm process technology, the dedicated AI accelerator chip targets generative AI workloads and could enter mass production by 2027, potentially challenging Qualcomm and Nvidia in the AI PC market.
Meta plans to begin manufacturing its custom AI chip, code-named Iris, in September 2026 as part of its MTIA program. The chip cleared testing in just six weeks and will be produced by TSMC with Broadcom's design support. Meta aims to double its computing capacity to 14 gigawatts by 2027 while reducing dependence on Nvidia and AMD GPUs.
Nvidia CEO Jensen Huang reveals his software engineers now prefer building AI agents to writing Python code, framing the shift in software engineering as a promotion rather than a threat. He argues AI is creating jobs, not eliminating them, as engineers focus on designing agentic systems, benchmarks, and guardrails instead of mundane coding tasks.
AI chip startup Cerebras accelerates its European expansion with plans to deploy 200MW of AI data center capacity by 2027 across France and the Nordics. The multibillion-dollar investment responds to soaring demand for local AI infrastructure and data sovereignty concerns, positioning the company to challenge Nvidia's dominance in the region.
Taiwan's central bank Governor Yang Chin-long has raised concerns about an AI bubble risk despite acknowledging real growth in the sector. Speaking at a parliamentary hearing, Yang warned that aggressive corporate borrowing and speculative capital expenditures in the tech sector could lead to over-expansion, even as Taiwan's AI-driven economy hits record highs anchored by TSMC and partnerships with Nvidia.
Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.