The AI-based applications include RISC-V cores, HBM2 high-bandwidth memory interfaces, and Interlaken chip-to-chip interconnects, and the TileLink coherence protocol. Data center applications can be configured with a variety of RISC-V management and compute cores, optimized on-chip cache and scratchpad memory systems, customer-specific hardware acceleration blocks, and are supported with a full system software stack.
- 7nm, 12/16nm Process
SiFive’s “Idea-to-Silicon” Methodology
SiFive’s “Idea-to-Silicon” methodology provides the solutions for a market-focused design with silicon proven IP and the ability to incorporate your own IP. The innovative methodology process of SiFive will lower the cost and deliver an SoC for your specific application in a short period of time.
Benefits of using "Idea-to-Silicon" Methodology
Ongoing product development costs on derivatives of existing chip designs are high and make it hard to integrate new IP. SiFive’s “Idea-to-Silicon” methodology can reduce cost and cycle time, while putting the hardware you need under your preferred software stack.
Combine SiFive’s standard cores with our SoC IPs, or IPs from our partners and/or your own IP to get your ideal end product.
SiFive’s “Idea-to-Silicon” methodology provides an ideal starting point to explore and enter into key markets with the advantage of a faster design cycle by integrating required IPs from our wide IP portfolio, thereby providing a competitive advantage.
SiFive’s “Idea-to-Silicon” methodology provides the opportunity to optimize layout for the power, performance and size requirement of your target market, and enables rapid iteration for ongoing product roadmap development for optimal time-to-market.
Explore Related Solutions
With over 350 successful tapeouts and 150 million parts shipped, our team brings 15 years of experience to your custom silicon solution.