Is this your business? Claim it to manage your IP and profile
The Avispado core by Semidynamics is a highly efficient, in-order RISC-V processor optimized for energy-conscious applications, such as AI edge computing and IoT devices. Its 64-bit architecture supports in-order execution and is particularly notable for its energy efficiency, making it suitable for power-sensitive environments. Avispado is designed to manage two instruction widths, which enhances its flexibility and scalability for embedded systems and AI workload implementations. Avispado features the integration of Semidynamics’ Gazzillion Misses™ technology, ensuring high-bandwidth access by efficiently managing up to 64 simultaneous memory requests. This high data throughput capability makes it ideal for edge AI applications where rapid data processing is essential. Additionally, Avispado is vector-ready, supporting RISC-V’s Vector Specification 1.0, which facilitates AI acceleration in applications such as machine learning and embedded computing. This core is also built to support multiprocessing, allowing scalability to multicore implementations. With compatibility for Linux environments, Avispado expands its usability scope into more traditional computing environments that leverage open-source platforms for development and deployment. Its branch predictor and various customizable options make it a versatile choice for integrators seeking a robust solution within a highly adaptable processor design.
Atrevido is a powerful out-of-order RISC-V core from Semidynamics, designed to support high-performance processing requirements in AI and HPC domains. This 64-bit processor features an advanced out-of-order execution system that is configurable for 2, 3, or 4-wide instruction handling, ensuring peak computing efficiency. Its architecture incorporates both vector and tensor processing capabilities, providing zero-latency integration which is critical for AI inferencing and high-speed data processing. One of Atrevido's standout features is the integration of Semidynamics’ proprietary Gazzillion Misses™ technology, which enhances data throughput by overcoming memory bottlenecks, allowing up to 128 concurrent memory requests. This makes the processor ideally suited for demanding tasks such as machine learning inferencing, sparse data management, and complex analytics. Atrevido’s multiprocessor readiness, with support for both AXI and CHI interfaces, further enhances its adaptability to high-bandwidth infrastructure requirements. The processor supports a wide range of extensions, making it adaptable to various use-cases including bit manipulation and cryptographic operations. This makes Atrevido particularly attractive for users needing a robust and configurable core that can be tailored to meet specific performance and energy efficiency needs. Its architectural flexibility and high processing power make Atrevido a quintessential element for building next-gen AI-driven solutions.
Gazzillion Misses™ technology is a memory bandwidth optimization solution developed by Semidynamics, designed to significantly improve the throughput of high-performance computing and AI systems. This technology addresses the common bottleneck of memory latency, ensuring that systems can handle simultaneous memory requests efficiently. By doing so, Gazzillion Misses™ enhances the performance of systems that require rapid access to large data volumes, such as AI inference engines or HPC environments. This technology is integrated into Semidynamics' RISC-V cores, complementing their processing architecture by maximizing the available bandwidth. It allows for an unprecedented number of concurrent memory accesses, which reduces wait times and expediting computational processes. This feature is particularly beneficial in scenarios where data-intensive applications necessitate continuous and fast read-write operations. The incorporation of Gazzillion Misses™ into AI processors and high-performance cores provides a significant acceleration in data processing capabilities. This means AI applications can achieve zero latency in memory access, which is crucial for maintaining efficient throughput and processing speed in big data applications. As such, this technology is a critical component in Semidynamics’ suite of solutions aimed at advancing machine learning and AI capabilities by overcoming memory-related constraints.
The Cervell™ NPU by Semidynamics is a groundbreaking neural processing unit designed to cater to the requirements of future-ready machine learning applications. This scalable NPU core provides a high level of efficiency by delivering performance capabilities ranging from 8 to 64 TOPS at INT8, and up to 256 TOPS at INT4, making it an ideal choice for AI tasks that fluctuate between various demanding scenarios. Its architecture supports all-in-one processing by integrating CPU, vector, and tensor capabilities, thus offering a comprehensive AI acceleration platform. Optimized for deep learning, recommendation systems, and large language models, the Cervell™ NPU ensures that latency is minimized while maintaining a high-performance output. This is achieved through fully programmable resources, allowing developers unfettered creativity without vendor lock-in. The NPU is built to work effectively across a wide array of applications, from edge AI to datacenter solutions, providing flexibility and power as required. Designed for efficiency, the Cervell™ NPU boasts features such as an open, all-in-one architecture that combines diverse processing capabilities within a unified framework. This makes it exceptionally versatile, enabling seamless execution of various AI workloads without the need for multiple or discrete components. With such rich capabilities, Cervell™ stands out as a leader in next-generation AI compute solutions, capable of adapting to specific processing needs with ease and precision.
Discover Semidynamics' Cervell NPU, which tackles AI's memory bottleneck with a scalable RISC-V design, perfect for diverse deployment from edge to datacenter. Read more
AI is transforming chip design with generative techniques, enhancing productivity, and integrating disciplines, according to industry experts. Read more
Join the world's most advanced semiconductor IP marketplace!
It's free, and you'll get all the tools you need to discover IP, meet vendors and manage your IP workflow!
No credit card or payment details required.
Join the world's most advanced AI-powered semiconductor IP marketplace!
It's free, and you'll get all the tools you need to advertise and discover semiconductor IP, keep up-to-date with the latest semiconductor news and more!
Plus we'll send you our free weekly report on the semiconductor industry and the latest IP launches!
To evaluate IP you need to be logged into a buyer profile. Select a profile below, or create a new buyer profile for your company.
