A semi truck, barreling down one of the world’s major highways. Imagine the added safety if semis are equipped with road marking recognition technology that can adjust a vehicle’s steering if a tired operator fails to do so.
Imagine a world where image recognition becomes more available and more efficient for organizations needing to harness the technology, helping to catch more bad guys.
At the Embedded World 2019 Exhibition & Conference, February 26-28 in Nuremberg, Western Digital, in collaboration with Mipsology, will demonstrate an impressive, innovative new prototype in Booth #3A-429.
Western Digital’s small form factor (SFF) accelerator offers a highly optimized platform for inference at the edge of the data center, with performance and power optimized for the most common mid-range networks workloads. Specifications include:
1. Seamless Deployment:
- Same software environment as CPU/GPU
- No application software changes required
- No new training required
- U.2 SFF (SFF8639)
2. Broad Network Support:
- Works under Caffe, Caffe2, TensorFlow, MXNET
- Same CPU/GPU-based trained neural network with similar accuracy without any change
3. Power Efficient:
- Scalable and Lower Power than GPUs
- Ideal for Edge or Data Center Applications
Able to quickly and dynamically process multiple in-bound neural networks at once, it has the potential to improve the intelligence and analytics of a wide range of deep learning and neural network-enabled applications, including smart surveillance and image classification and recognition.
Machine Learning Accelerator Highlights
- Speeds of 239 ResNet50 (8-bit) frames per second*
- Power-efficient, utilizing less than 20W
- Optimized for all neural networks and wide range of deep-learning inference workloads, including ResNet50, GoogleNet, Inception V3 and V4, Yolo V2 and V3
- No need to sacrifice a PCIe card slot or redesign systems to accommodate
- Runs in TensorFlow, Caffe, Caffe2 or MXNet
- Utilizes Xilinx Zynq® Ultrascale, MPSoC ZU7EV FPGA, and Zebra from Mipsology
*Based on internal testing.
- Solution Brief: Machine Learning Accelerator
- Find out what the Office of the CTO is up to: Innovation at Western Digital
- Blog post: Top Machine Learning Trends 2018
- Blog post: Machine Learning for Edge Devices
At Embedded World this year, we are showcasing a variety of complete data technology solutions, including advanced in-vehicle systems, high-endurance industrial and IoT designs, our optimized performance portfolio for Smart City surveillance applications, and technology innovations enabled by the open, scalable RISC-V instruction set architecture for purpose-built, data-centric applications. Stop by Booth #3A-429 from February 26 to 28!
Certain blog and other posts on this website may contain forward-looking statements, including statements relating to expectations in the market for our products and applications of our products. These forward-looking statements are subject to risks and uncertainties that could cause actual results to differ materially from those expressed in the forward-looking statements, including development challenges or delays, changes in markets, demand, global economic conditions and other risks and uncertainties listed in Western Digital Corporation’s most recent quarterly and annual reports filed with the Securities and Exchange Commission, to which your attention is directed. Readers are cautioned not to place undue reliance on these forward-looking statements and we undertake no obligation to update these forward-looking statements to reflect subsequent events or circumstances.
Richard New is Vice President of Research at Western Digital, in charge of the Western Digital Research Lab.