InspireSemi has announced the successful completion of the production process of the Thunderbird I accelerated computing chip for manufacturing at TSMC.
This highly differentiated “supercomputer cluster on a chip” features 1,536 custom 64-bit RISC-V CPU cores designed for high-end scientific computing and complex data processing.
Thunderbird I is designed to meet a wide range of compute-intensive applications, from artificial intelligence and machine learning to graph analysis. It leverages the open RISC-V CPU ISA standard, enabling easier development and integration into existing technology frameworks, with access to a robust ecosystem of software, libraries and tools.
PCIe add-on card planned
The chip architecture integrates a high-speed mesh network structure that provides substantial bandwidth and minimal latency communication between cores, important for applications that rely on synchronized operations across multiple threads. This efficient network integration manages interactions within the core chip assembly and memory systems, ensuring optimal performance without common bottlenecks.
The upcoming product release will include a server PCIe add-in card that will house four Thunderbird chips, providing more than 6,000 interconnected 64-bit CPU cores. This configuration is equipped to handle double-precision mathematics, essential for many high-performance computing applications in fields such as climate science, medical research, and complex simulations.
Ron Van Dell, CEO of InspireSemi, said: “We are proud of our engineering and operations team's achievement in completing the design of Thunderbird I and shipping it to our world-class supply chain partners, TSMC, ASE and imec for delivery. production. We hope to begin deliveries to customers in the fourth quarter.” However, there is no information on the price yet.
InspireSemi also emphasizes Thunderbird I's energy efficiency, a holdover from its initial design for energy-sensitive blockchain computing applications. The company says this approach offers a more environmentally friendly alternative to traditional data center GPUs.
