Part 4/10:
The last couple of years have seen rapid adoption of RISC-V-based AI accelerators. Major tech players like Meta and Google have integrated RISC-V cores with their AI accelerators. For instance, Meta is building its AI chips on RISC-V, and Google is integrating RISC-V with its SciFi CPU cores, enabling enhanced tensor processing capabilities.
Chinese companies like Stream Computing are also entering the scene, developing accelerators with RISC-V cores combined with their proprietary tensor processing units (TPUs). These efforts illustrate a broad industry trend: RISC-V serves as a stable foundation for domain-specific AI accelerators across data centers, edge devices, and high-performance computing (HPC).