Category
Micron has opposed Apple's proposal to source memory chips made in China. A gap exists between near-term price relief and supply improvements expected from 2028 onward, with the Trump administration's additional regulations set to determine adoption conditions.
Exploration in Idaho, USA has confirmed a xenotime deposit containing high concentrations of heavy rare earth elements. As China's export restrictions heighten supply risk, this project—capable of recovering high-grade yttrium and other elements from near-surface deposits—is drawing significant attention as a foundation for North American self-sufficiency in strategic materials.
Intel's Q2 2026 earnings showed 25% revenue growth and a return to operating profit, yet the company posted a net loss of $11.033 billion. The main driver was a $12.529 billion mark-to-market loss on escrow shares, while the narrowing Foundry loss and the fate of 14A external customers also drew attention.
Reports say Anduril is in talks for additional funding at a valuation of roughly $100 billion. We examine whether actual government orders and production capacity can support a price proposal that emerged just 72 days after its last round.
Qualcomm has reportedly notified customers of a double-digit price increase for shipments starting September 1. Product-specific terms have not been disclosed, and whether the increase will be passed on to device prices or lead to spec changes remains unconfirmed.
China's DRAM giant CXMT has set prices above Samsung's for 64GB DDR5 server memory and rejected even Huawei's demands for a discount. As AI-driven HBM production shifts squeeze general-purpose DRAM supply, this move signals a shift toward whoever holds pricing power.
NVIDIA CEO Jensen Huang made his first-ever X post to back a letter opposing regulation of open-weight AI. Behind the wariness over China's Kimi K3 lies an alignment with NVIDIA's own business interests in expanding GPU demand. OpenAI, Anthropic, and Google did not join.
A technique using "Lossless Scaling" to split rendering and frame generation across separate GPUs is gaining attention. Unlike legacy technologies such as SLI, this method allows GPUs from different generations to be combined to distribute rendering load, offering a promising new approach to affordable performance gains by repurposing older hardware.
Etched announced a $300M raise that included participation from SK hynix, and has begun initial production of inference clusters combining low-voltage compute with shared memory. Supply agreements and comparable performance figures have not yet been disclosed.
Claude Opus 5 matches Opus 4.8's API base pricing while approaching Fable 5's practical performance. Migration requires attention to the now-default thinking feature and unsupported APIs.
