Silicon, foundries, GPUs — the substrate of everything.
123 stories · page 3 of 6
Despite a 20% contraction in the desktop CPU market, AMD has hit record market share, signaling a fundamental divergence in how Intel and AMD are navigating the current silicon glut.
LG's new maskless Laser Direct Imaging machine aims to solve the critical throughput constraints currently plaguing high-density chip packaging and interconnect scaling.
The conviction of a former Samsung engineer reveals the technical blueprints behind CXMT's push to bypass years of DRAM process development and challenge global memory leaders.
Samsung Foundry has reportedly raised contract prices by up to 15 percent on its mature and advanced nodes, signaling tight capacity and a shifting geopolitical landscape for fabless designers.
Driven by US export controls and the diminishing returns of throttled Western GPUs, Chinese hyperscalers are pivoting to domestic silicon, forcing a rapid maturation of the country's independent semiconductor ecosystem.
With backing from AMD, Renesas, and Boeing, the Xen Project is pursuing safety-critical certifications to enable secure hardware virtualization in automotive and aerospace systems.
Google is reportedly collaborating with AMD to develop a next-generation Tensor Processing Unit that integrates custom CPU cores on-package to handle the complex, sequential logic required for reinforcement learning and agentic AI workloads.
Geopolitical volatility and resource concentration continue to threaten semiconductor production, highlighting the urgent need for supply chain diversification beyond current regional dependencies.
Intel is extending the lifecycle of its LGA 1700 platform, citing a surge in demand for Raptor Lake as high DDR5 memory costs keep cost-conscious builders tethered to DDR4.
Advanced multiphysics simulation tools are becoming indispensable for identifying and mitigating failure points in high-voltage electronics, a critical requirement for electric vehicles, data centers, and renewable energy infrastructure.
As modern AI accelerators push thermal and power limits to the brink, legacy silicon like Nvidia's 2020-era Ampere-based A100 is finding an extended lease on life through long-term cloud contracts.
Intel's successful $19.7 billion capital raise provides the critical financial runway needed to scale its 14A process node and deploy expensive High-NA EUV lithography.
As a global high-bandwidth memory shortage bites, Nvidia is reportedly testing Rubin Ultra configurations with as little as 192 GB of HBM4, forcing a strategic shift in next-generation AI cluster architecture.
Leaked Geekbench results for Nvidia's upcoming ARM-based RTX Spark SoC reveal 20-core and 18-core variants, showcasing massive multi-threaded performance that threatens both x86 mobile processors and Qualcomm's Windows on ARM dominance.
Chinese memory maker CXMT has demonstrated DDR5-8800 speeds on AMD platforms, proving that domestic silicon can match the performance of global giants despite strict export controls.
Researchers at MIT have uncovered TONTOU, a microarchitectural attack that leverages hardware timer interrupts to bypass existing Spectre mitigations on modern Intel and AMD processors.
Anthropic is establishing an in-house chip design team, signaling a strategic shift towards vertically integrated hardware-software co-design to enhance the efficiency and speed of its large language models. This move reflects a growing trend among major AI developers to control
As the semiconductor industry prepares for the post-nanosheet era, leading chipmakers are splitting over whether monolithic or sequential integration is the viable path for complementary FETs.
As vertical high-bandwidth memory stacks hit a critical thermal ceiling, packaging engineers are turning to lateral integration and hybrid bonding to sustain AI accelerator scaling.
As traditional copper interconnects hit physical limits, major foundries are racing to deploy co-packaged optics, each taking a fundamentally different approach to silicon photonics integration.
As physical limits choke copper's conductivity at advanced nodes, researchers are turning to topological materials and exotic quasiparticles to solve the semiconductor interconnect crisis.
A technical deep dive into Nvidia's upcoming Vera CPU reveals a highly customized Arm architecture designed to eliminate memory and interconnect bottlenecks in massive AI clusters.
The semiconductor industry faces a prolonged memory supply crisis as the production of high-bandwidth memory for AI hardware consumes disproportionate wafer allocations.
Intel’s discontinuation of Optane memory, known for ultra-low latency and high endurance, leaves AI and memory-intensive workloads reliant on traditional DRAM amid ongoing supply constraints.