Is this your business? Claim it to manage your IP and profile
The Calibrator is a precision tool for AI System-on-Chips, utilizing architecture-aware quantization through post-training methods. It maximizes accuracy in fixed-point systems like INT8 by accommodating varying bit-width architectures. The Calibrator also supports multiple data path engines, ensuring its versatility across different hardware. Its advanced workflow leverages entropy calculation policies to maintain high precision without altering the network model topology, thus effectively bridging the gap with quantization-aware training scripts.
The ONNC Compiler is an advanced suite of C++ libraries designed to enhance the efficiency of deep learning accelerators. It facilitates the transformation of neural networks into distinct machine instructions, optimizing across varying architectures from simple single-core systems to complex multicore systems with layered memory hierarchies. Supporting prominent frameworks such as PyTorch and TensorFlow, the compiler ensures seamless integration for a wide array of AI systems, particularly those focusing on heterogeneous multicore designs.
Forest Runtime is designed to execute compiled neural network models seamlessly across diverse hardware platforms. Offering C++ APIs compatible with C and Python bindings, it caters to a wide range of AI applications. This runtime is modular and retargetable, making it adaptable for platforms from data centers to mobile and TinyML devices. Notable for its novel 'hot batching' technology, Forest Runtime flexibly alters batch sizes and input shapes during runtime, without needing compiler transformation, significantly speeding up processing times.
HyperThought accelerates large language models with unparalleled efficiency suitable for multimodal applications requiring intense computational capacity. Featuring innovative compression techniques that diminish memory requisites and boost throughput, this IP offers security-hardened interactions through the integration of LISA v3 architecture. Its scalable multicore design capitalizes on high-speed processing capabilities, accommodating both modest and extensive model requirements across various industry settings.
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!
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.