Tag
Explore the latest semiconductor news and technology.
Panmnesia and Meta have unveiled a next-generation AI infrastructure design that uses CXL to tightly couple an entire data center as if it were a single chip. The architecture cuts cross-rack communication latency to a few hundred nanoseconds and forms a shared memory space spanning up to 960 accelerators.
imec has demonstrated 3-layer NbTiN superconducting circuits with 30nm wiring. We examine how close this brings the technology to 300mm manufacturing, based on the 3.8 million junction achievement, the gap with the 400 million device vision, and untested issues around cooling and yield.
IonQ's 256-qubit Superion 256 links electronic control with semiconductor fabrication. Beyond QPU manufacturing and ion trapping, questions remain about integrated performance, yield, and customer delivery.
Samsung and TSMC are moving toward High-NA EUV mass production before the larger photomask standard is finalized. Their 2028–2033 roadmaps reveal a two-stage transition: capturing resolution gains first, then boosting throughput later.
We examine what Intel's cumulative 1 million wafers on High-NA EUV reveals about process learning, and unpack the 2030-2033 roadmap ASML and TSMC have set for moving to 12-inch photomasks along with the constraints on mass production.
France's Mistral AI has raised a record €3 billion for Europe, led by Samsung Electronics. The two companies will jointly develop dedicated models for defect detection and yield stabilization, running in air-gapped, on-premise environments inside semiconductor fabs that require strict confidentiality, advancing the practical rollout of sovereign AI.
Arm has unveiled CSS for Mobile 2, its next-generation smartphone compute platform. With a C2 CPU cluster featuring double the SME2 units and a GPU with a dedicated integrated AI accelerator, it aims to deliver both agentic AI and desktop-class graphics within a 1W power envelope.
Arm's Neoverse CSS N4 offers up to 128 cores, but its real significance lies in the pre-integrated RTL that reshapes how much design work customers must still do themselves. This piece examines the conditions behind Arm's performance claims and development-time figures, and what they reveal about Arm's revenue strategy.
Starting from reports of an approximately 10% October price hike on Intel's PC CPUs, this article traces the pattern of incremental price revisions and server-first supply allocation, examining the shift toward prioritizing gross margins and the conditions—such as 18A and 14A mass production—needed to resolve ongoing supply constraints.
Huawei's two-layer stacked Kirin 2026 reportedly delivers major power savings at equivalent performance. But at peak performance, power density exceeds the previous generation, and absolute junction temperature figures remain unpublished.
