Empowering SoC Developers with Cutting-Edge AI IP Solutions
Cadence Virtuoso AI IP platform offers a comprehensive solution for developers to design and deliver optimal AI solutions for a wide array of applications and markets. With a range of IP options including Neo NPUs, Tensilica AI Base, and AI Boost, SoC developers can meet the demands for power, performance, and area (PPA) behind the NeuroWeave SDK common software platform. These IP options provide scalable and energy-efficient on-device to edge AI processing, essential for today's AI/ML SoCs in various sectors like automotive, consumer electronics, industrial applications, and mobile devices.
Key Benefits of Cadence Virtuoso AI IP Platform
The platform offers a scalable design to adapt to various AI workloads, efficiently mapping state-of-the-art AI/ML workloads. It provides best-in-class performance for inferences per second, boasting low latency and high throughput, with architectures optimized for high performance within a low-energy profile. Cadence Virtuoso also demonstrates industry-leading performance and power efficiency, delivering high Inferences per second per area (IPS/mm2) and per power (IPS/W). Moreover, the platform offers an end-to-end software toolchain catering to all markets and supporting a variety of frameworks through the NeuroWeave SDK, enabling the compilation of networks across different IPs with flexibility for performance, accuracy, and run-time environments.
Target Markets for Cadence Virtuoso AI IP Platform
The platform answers the needs of a broad spectrum of end applications including IoT, hearables, wearables, true wireless stereo, smart speakers, AR/VR headsets, automotive systems, mobile devices, drones, robots, intelligent cameras, and private on-premises compute environments. By catering to such diverse applications, Cadence Virtuoso AI IP platform demonstrates its versatility and capability to address the evolving requirements of various industries.
Neo NPUs: The Leading Edge of Cadence AI IP Portfolio
Among the offerings, the Neo NPUs stand out as the pinnacle of the Cadence AI IP portfolio. These NPUs provide accelerated AI processing with a wide performance range, supporting up to 80 TOPS in a single core. They cater to a broad spectrum of devices ranging from energy-sensitive IoT and wearables to high-performance AR/VR and automotive systems. The Neo NPUs are designed to efficiently deliver high performance for various neural network types, supporting the transition from classic to generative AI and providing the necessary processing power for cutting-edge AI applications.
Training and Support for Cadence Virtuoso AI IP Platform
Cadence offers extensive training resources to support developers in mastering the Cadence Virtuoso AI IP platform. The Training Learning Maps provide a comprehensive visual overview of learning opportunities, ensuring developers can upskill efficiently. Additionally, the Cadence Online Support (COS) system offers a vast library of accessible materials for self-study and step-by-step guidance. Developers can also engage with the community on technical forums to discuss ideas, troubleshoot challenges, and enhance their design capabilities, making Cadence Virtuoso a go-to platform for cutting-edge AI development.