Recent #High-Performance Computing news in the semiconductor industry
➀ High-performance CPU design is transitioning from traditional Out-of-Order (OOO) architectures to Time-Based OOO, leveraging RISC-V's open ecosystem to improve power efficiency and scalability.
➁ Condor Computing's Cuzco processor uses a slice-based microarchitecture and predictive scheduling via a Time Resource Matrix, enabling flexible configurations for datacenter, mobile, and automotive applications.
➂ Key advantages include superior performance-per-watt, simplified verification, and ISA extensibility, positioning RISC-V as a competitive alternative to legacy architectures like x86 and ARM.
➀ AMD processors have helped achieve a new world record in a recent Ansys Fluent computational fluid dynamics (CFD) simulation on the Frontier supercomputer.
➁ The simulation, which tested Baker Hughes' next-generation gas turbines, was completed in 1.5 hours using 1,024 AMD Instinct MI250X accelerators and AMD EPYC CPUs.
➂ This marks a 25x speed increase compared to the previous 38.5-hour completion time on 3,700 CPU cores.
FMD has launched the Chiplet Application Hub, a central platform for chiplet technology development and application. The hub aims to bridge the gap between research and industry, accelerate the development of German-made chiplets, and reinforce Germany's technological resilience. It complements FMD's role in the Chips for Europe Initiative and builds on the APECS pilot line infrastructure.
The hub is designed to foster collaboration with industry partners, develop new chiplet solutions, and drive innovation in chiplet technologies. It is expected to enhance energy efficiency, performance, and the reusability of high-cost design components, particularly in the automotive and high-performance computing sectors.
➀ TSMC reports a strong demand for 2nm nodes over 3nm, with A16 attracting AI server clients.
➁ TSMC's 3nm shipments accounted for 20% of total wafer revenue in Q3 2024.
➃ TSMC's 2nm is expected to enter volume production in 2025.
➄ TSMC plans to expand its 2nm capacity to meet strong demand.
➅ TSMC's A16 process is highly attractive for AI server applications.
➀ The need for RDMA in high-performance computing, big data storage, and machine learning domains;
➁ Detailed introduction of RDMA network architecture, key components, and protocols;
➂ Explanation of the RDMA working mechanism, including software configuration and hardware design;
➃ Overview of RDMA technology protocols such as InfiniBand, iWARP, and RoCE;
➄ Future technological innovations and challenges for RDMA, including scalable RDMA controllers, efficient QP management, congestion control, network topology optimization, and security.