Cadence has unveiled a neural network DSP IP core offering 1 TeraMAC (TMAC)/sec computational capacity to run all neural network computational tasks in vision, radar/lidar and fused-sensor applications. The Vision C5 DSP is designed for automotive, surveillance, drone and mobile/wearable markets.
The Vision C5 DSP offers 1 TeraMAC (TMAC)/sec computational capacity in less than 1mm2 area. This is four times the performance of the existing P6 DSP.
As neural networks get deeper and more complex, the computational requirements are increasing rapidly. Meanwhile, neural network architectures are changing regularly, with new networks appearing constantly and new applications and markets continuing to emerge.
The firm notes these trends are driving the need for a high-performance, general-purpose neural network processing solution for embedded systems that not only requires little power, but also is highly programmable for future-proof flexibility and lower risk.
The Vision C5 DSP is architected as a dedicated neural-network-optimized DSP, to accelerate all neural network computational layers (convolution, fully connected, pooling and normalization), not just the convolution functions. This frees up the main vision/imaging DSP to run image enhancement applications independently while the Vision C5 DSP runs inference tasks.
The Vision C5 DSP completely offloads all the processing and minimizes the data movement between the neural network DSP and the main vision/imaging DSP, where much of the power is actually consumed. This makes it a lower power solution than competing neural network accelerators. It also offers a simple, single-processor programming model for neural networks, according to Cadence.
“Many of our customers are in the difficult position of selecting a neural network inference platform today for a product that may not ship for a couple of years or longer,” said Steve Roddy, senior group director, Tensilica marketing at Cadence. “Not only must neural network processors for always-on embedded systems consume low power and be fast on every image, but they should also be flexible and future proof. All of the current alternatives require undesirable tradeoffs, and it was clear a new solution is needed. We architected the Vision C5 DSP as a general-purpose neural network DSP that is easy to integrate and very flexible, while offering better power efficiency than CNN accelerators, GPUs and CPUs.”
“The applications for deep learning in real-world devices are tremendous and diverse, and the computational requirements are challenging,” said Jeff Bier, founder of the Embedded Vision Alliance. “Specialized programmable processors like the Vision C5 DSP enable deployment of deep learning in cost- and power-sensitive devices.”
“Compared to commercially available GPUs, the Vision C5 DSP is said to be up to 6X faster in the well-known AlexNet CNN performance benchmark and up to 9X faster in the Inception V3 CNN performance benchmark”, according to the firm.
The Vision C5 DSP also comes with the Cadence neural network mapper toolset, which will map any neural network trained with tools such as Caffe and TensorFlow into executable and highly optimized code for the Vision C5 DSP, leveraging a comprehensive set of hand-optimized neural network library functions.
For more information, visit www.cadence.com/go/visionc5.
Click here to view the full Press release.