Semiconductors
SK hynix has introduced StreamDQ, a near-memory dequantization architecture designed to accelerate LLM inference. By performing weight dequantization directly in HBM, the system reduces data transfer bottlenecks and improves energy efficiency.
Semiconductors
NVIDIA’s new Vera CPU architecture targets 'agentic AI,' focusing on single-threaded performance at scale. This new category of silicon is designed to handle the reasoning and planning tasks required by physical AI systems.
Semiconductors
Researchers are leveraging LLM-based agents to automate the refactoring of software for High-Level Synthesis (HLS). This 'self-evolving' workflow promises to bridge the gap between high-level code and hardware-optimized semiconductor design.
Semiconductors
The semiconductor industry is facing critical I/O design challenges as AI data centers and HPC clusters grow. Engineers are grappling with bandwidth chokepoints that threaten to stall the progress of large-scale AI training.
Semiconductors
Intel’s 18A process node is gaining significant momentum, with the company showcasing new breakthroughs in backside power delivery and routed designs. These innovations are critical for the next generation of AI and automotive chips that require unprecedented power efficiency.
Semiconductors
NVIDIA's Blackwell architecture has dominated the latest MLPerf Training 6.0 benchmarks, setting new records for AI training speed and efficiency. This hardware serves as the silicon foundation for the next generation of physical and agentic AI.
Semiconductors
Researchers from imec and KAIST are pushing the boundaries of AI hardware. New 3D heterogeneous memories using IGZO FeFETs and reconfigurable hardware for Zero-Knowledge Proofs are set to redefine chip efficiency.
Semiconductors
As AI workloads move to the edge and space, the challenge of securing Terabit Ethernet and preventing hardware-level breaches has become a top priority for chip designers.
Semiconductors
As the industry moves toward the sub-2nm node, engineers face a paradox: shrinking features can lead to diminishing returns without radical new approaches to variation management and workload-specific hardware.
Semiconductors
Researchers from Stanford and Google have developed 'ITHICA,' a new approach to detecting Silent Data Corruptions (SDCs) in CPUs. As chip nodes shrink, these hardware 'flips' are becoming a major threat to the reliability of hyperscale AI data centers.
Semiconductors
As large language models move from training to inference, the underlying hardware architecture is evolving. New research into SRAM-based pipelines suggests a path toward drastically faster and more efficient AI serving.
Semiconductors
The industry is shifting toward heterogeneous integration and chiplets to overcome the physical limits of monolithic silicon. New roadmaps are required to manage the cooling and power delivery challenges of these complex stacks.