Gyrfalcon launches AI-X — full-stack solution for edge-AI development
Dr. Manouchehr Rafie of Gyrfalcon
Gyrfalcon Technology (GTI) announced AI-X™, a new full-stack solution for AI development ideal for small and medium enterprises (SMEs). Popular applications for AI-X include edge-AI vision devices, consumer electronics, AI-camera sensors and automotive.
AI-X benefits SMEs by accelerating AI product turnaround time, maximising on-chip resource utilisation, minimising CPU runtime, boosting performance efficiency, and reducing power consumption by leveraging GTI Networks (GNetâ„¢) and holistic HW-SW optimisation.
Powered by Gyrfalcon high-performing accelerator chips, AI-X delivers chip-level optimisation solutions for realising deep neural network development on edge devices. AI-X provides an efficient HW-SW design framework for developing CNN-based models on GTI’s chips.
“For smart devices such as IoTs, smartphones, autonomous vehicles and many other platforms and end-point devices to reach their maximum potential edge AI co-processing is imperative,” says Dr. Manouchehr Rafie, vice president of advanced technologies at Gyrfalcon.
“AI-X combines an end-to-end HW and SW solution for quickly deploying your data, tasks, AI custom model, algorithm design and chip implementation challenges on the AI devices equipped with the GTI high-performing chips. These AI edge devices benefit from the AI-X higher performance, improved privacy, reduced bandwidth and latency, less CPU computational load, efficient energy use and less BOM cost.”
AI-X Benefits:
- Vision-based solution for AI chip design challenges
- Customised model conversion (MDK)
- Minimising CPU utilisation/runtime (Over 3x CPU Saving)
- Maximising on-chip resource utilisation (>90%)
- Enhanced performance efficiency (by 5x)
- Reduced latency during inferencing
- Chip-level optimisation & model pruning
- Quantisation-aware training
- Per-layer quantisation and parameter optimisation
Powered by GTI Accelerator Chips:
- Unmatched TOPS & power efficiency
- Integrated on-chip memory
- Scalable multi-chip architecture
- Executes a large number of calculations in Tensor operations
The AI-X product information is available here.
Comment on this article below or via Twitter @IoTGN