Choosing a Processor for Machine Learning at the Edge
By Manisha Agrawal, Texas Instruments
EETimes (June 23, 2019)
Not all machine learning models need processing on the order of several TOPS. Understanding the performance, latency and accuracy need of your application is a critical first step to choose a processor for machine learning at the edge.
Machine learning has become popular for solving machine vision and other embedded computing problems. While classical machine learning algorithms need human intervention to extract features from data, machine learning algorithms or network models learn how to extract important features in data and make intelligent predictions about that data.
Below, figure 1 shows a few examples where machine learning technology is adding intelligence to a variety of devices. In smart home appliances like a smart oven, machine learning can be used to classify food inside the oven and set the cooking temperature and time of the oven accordingly. In factories, machine learning can be used for detecting defects in the products or it can be used for predictive maintenance to help predict the remaining useful life of the motor or detecting anomaly in motor operations. In a vehicle, it can be used to detect cars, pedestrians, traffic signs, etc. on the road. It can also be used in devices doing natural language translation.
To read the full article, click here
Related Semiconductor IP
- UALink 200G v1.0 Verification IP
- CCSDS SCCC Demapper and Turbo Decoder
- NPU IP
- JPEG XL Encoder
- I2C Master/Slave Controller Core
Related Articles
- Low-Energy Reduced RISC-V Instruction Subset Processor for Tsetlin Machine Inference at the Edge
- Peregrino: A Full-Hardware Accelerator for the Complete Falcon Post-Quantum Digital Signature Scheme on Resource-Constrained Edge Devices
- MEGATRON: a 28nm Analog PCM CiM/Digital System-on-Chip for Edge GenAI at 57.5 TOPS/W and 1.52 Mparam/mm²
- MIPI in next generation of AI IoT devices at the edge
Latest Articles
- Anechoic: Onsager-Corrected Parallel Annealing for Sub-0.1 ms Dense Max-Cut on an FPGA
- Power, Performance, and Area of HPC-Based Malware Detectors: An FPGA and Open-PDK ASIC Study
- Terracotta: Enabling the Adoption of New DRAM Techniques via a Flexible DRAM Interface and Memory Controller
- A Framework for Accelerating Transformer Inference on RISC-V for Edge AI
- An Interleaved Parallel Dependent Quantization Hardware Architecture for H.266/VVC