ABR builds time-series AI processors (TSPs) that reduce AI cost, power and latency by using ABR’s AI compression method, called an LMU, which makes time-series AI models optimally small. Larger AI models are more powerful and more accurate AI models. TSPs are 100x more efficient than CPUs & GPUs for time-series processing and can process larger AI’s than small edge processors while using far less power and costing a few dollars.
TSP IP is available for companies looking to enhance their chip by incorporating intelligence.
Applied Brain Research IP Listing
1 Product(s) Listed
Applied Brain Research
101 Jefferson St. Menlo Park CA 94025 USA