Alphawave Semi is making significant strides in the semiconductor industry with its development of a compute chiplet based on the ARM Neoverse N3 Compute SubSystem (CSS). This marks a strategic shift from their primary focus on intellectual property (IP) to producing high-performance custom silicon solutions. The company’s acquisition of RISC-V developer SiFive further positions Alphawave Semi at the forefront of advanced AI computing technology.
The move to chiplet-based design represents a significant evolution for Alphawave Semi, transitioning from licensing IP to manufacturing physical chiplets. This technology allows for modular and efficient chip design, leading to faster innovation and better performance. Specifically designed for AI computing, the chiplet addresses the rapidly growing needs of this sector. Using pre-designed chiplets like those with ARM Neoverse N3 cores and high-speed connectivity enables faster and potentially lower-cost development of new custom AI chips compared to traditional methods.
The benefits extend across the tech industry. Chip designers can integrate Alphawave Semi’s chiplet into custom AI chips, saving development time and reducing costs. AI hardware manufacturers can create high-performance AI hardware for various applications, including data centres, autonomous vehicles, and research. Businesses and researchers can leverage faster development of powerful AI hardware for advancements in fields like healthcare, finance, and self-driving cars.
Alphawave Semi, already notable in the semiconductor IP market for high-speed connectivity solutions, underscores its commitment to technological expansion with the acquisition of SiFive. This strategy leverages emerging technologies to develop cohesive, high-performance computing solutions.
The ARM Neoverse N3 Compute SubSystem provides an advanced platform meeting the demands of modern high-performance computing (HPC) and artificial intelligence (AI) applications. Renowned for its efficiency, scalability, and performance, the ARM Neoverse architecture is ideal for developing next-generation compute solutions.
Alphawave Semi’s new chiplet, designed for high-performance AI computing, utilizes the ARM Neoverse N3 cores. The chiplet includes the ARM Coherent Mesh Network (CMN) for efficient performance and is manufactured on leading-edge process nodes for optimal efficiency. This adaptable design is suited for various applications, including HPC, data centres, and AI/ML.
This development is a significant milestone for Alphawave Semi, marking a shift from traditional IP licensing to custom silicon design. The chiplet is a key component of their new platform, enhancing performance and efficiency in AI computing.
The chiplet-based custom silicon design platform integrates various IPs, such as I/O extension chiplets, memory chiplets, and compute chiplets. Leveraging Alphawave Semi’s expertise in PCI Express 6.0 and 7.0 connectivity and advanced packaging capabilities, the new compute chiplet promises robust performance and flexibility.
Alphawave Semi’s chiplet offers several advantages for AI computing. Its modular design allows seamless integration with other chiplets, enabling a flexible custom silicon design platform. Advanced packaging techniques combine the compute chiplet with high-speed connectivity solutions, offering superior data transfer rates. The design, featuring ARM Neoverse N3 cores and the ARM Coherent Mesh Network, allows for efficient performance scaling. Companies can also benefit from faster time-to-market by using pre-designed chiplets.
Overall, Alphawave Semi’s chiplet is a versatile, high-performance solution for demanding workloads in AI, data centres, and HPC environments. This strategic partnership with ARM positions Alphawave Semi as a strong contender in the growing chiplet-based chip design space.
Alphawave IP Group plc (LON:AWE) is a semiconductor IP company focused on providing DSP based, multi-standard connectivity Silicon IP solutions targeting both data processing in the Datacenter and data generation by IoT end devices.
