Accelerator IP for
An AI accelerator platform that scales
for any application
Expedera's unified compute pipeline architecture enables highly efficient hardware scheduling and advanced memory management to achieve unsurpassed end-to-end low-latency performance. Origin enables system designers to meet a full range of low latency, power-efficient and high-performance requirements. The architecture is mathematically proven to utilize the least amount of memory for neural network (NN) execution. This minimizes die area, reduces DRAM access, improves bandwidth, saves power, and maximizes performance. Since NNs generate a tremendous amount of intermediate data, minimizing the memory allows high-resolution NN processing, such as 4K/8K video, to run real-time on-chip.
The Origin architecture allows designers to run their trained neural network unchanged, without the need for hardware-specific optimizations.
This enables them to achieve greater accuracy and predictable performance. The architecture enables a simplified software environment that reduces complexity and eases the integration effort.
Origin achieves sustained performance and can process more throughput with less power than competitive products. Origin excels at image related tasks like computer vision, image classification, and object detection. It is capable of NLP related tasks like machine translation, sentence classification, and generation. Origin offers deterministic performance, scalable on-chip execution with the smallest memory footprint, and 18 TOPS/W effective performance. It scales from edge solutions with little or no DRAM bandwidth to high-performance applications such as autonomous driving and cloud applications without the software bloat that other solutions require.
Origin Line of Products
Origin E2 is designed for low power applications such as mobile phones. It uses on-chip memory only. By eliminating the need for DRAM access, it saves system power while increasing performance, reducing latency, and cutting BOM costs. Its highly efficient engine uses less than 1W. Origin E2 is tunable for specific workloads to provide an optimal performance profile for unique application requirements.
- Processes without DRAM during runtime
- Tailored for application workloads
- Highly efficient engine using less than 1W
Origin E6 runs the breadth of models for general applications such as smartphones, tablets, and edge servers. It enables L3 cache or DRAM access during runtime. Expedera’s advanced memory management ensures sustained DRAM bandwidth and optimal total system performance.
- Runs NN models for smartphones, tablets and edge servers
- Enables L3 cache or DRAM access during runtime while minimizing data transfers
- Highly efficient engine using less than 2W
Origin E8 provides the high performance needed for ADAS, autonomous, and datacenter applications. It dramatically reduces DRAM requirements, cutting BOM costs. Origin E8 enables multi-job support for better utilization of hardware resources and reductions in system costs. Its highly efficient neural network engine allows designers to develop products using passive cooling, further reducing system costs.
- Supports high-performance applications such as ADAS, autonomous and datacenter
- Drastically reduces DRAM requirements, cutting BOM costs
- Enables multi-job support
- Highly efficient engine scaling up to 100 TOPS at 5W
to our News
Sign up today and receive helpful
resources delivered directly
to your inbox.