Accelerate Intelligence at the Edge
High-performance APUs and logic processors built for massive data throughput, vector search, and real-time AI inference
The APU Advantage:
Compute-in-Memory
Overcoming the "AI bottleneck" by processing data directly where it resides. Delivering millisecond latency for vector search and reducing LLM hallucinations at a fraction of the power cost of traditional GPUs.
3-Second TTFT
Delivering industry-leading Time-To-First-Token (TTFT) for edge multimodal LLM inference. Ensuring real-time responsiveness for critical interactive AI applications without relying on cloud latency.
30W Sub-system Power
Leveraging revolutionary Compute-in-Memory architecture to eliminate data movement bottlenecks. Achieving massive AI acceleration at a fraction of the power cost of traditional GPU clusters.
Billion-scale Vector Search
Optimized for high-throughput similarity search and Retrieval-Augmented Generation (RAG). Instantly querying billions of data points to provide accurate context and significantly reduce LLM hallucinations.
Leda-E PCIe Accelerator Card
For edge inference and workstation integration.
The Leda-E PCIe accelerator seamlessly integrates into existing workstations, bringing the massive parallel processing power of the Gemini-II APU to your local environment. It is the ultimate solution for running billion-scale vector databases without the cloud latency.
KEY SPECIFICATIONS
- Form Factor: PCIe Gen4 x16
- Compute: 1x Gemini-II APU
- Power: < 50W (Sub-system)
- Memory: High-bandwidth 3D SRAM
TARGET APPLICATIONS
- Edge AI inference servers
- High-frequency algorithmic trading
- Localized LLM deployment for secure environments
- Real-time video analytics
Leda-S 1U Edge Server
High-density compute node for enterprise data centers.
Engineered for maximum efficiency in constrained spaces, the Leda-S 1U server packs up to 16 APUs into a standard rackmount chassis. It delivers unprecedented throughput for retrieval-augmented generation (RAG) pipelines directly at the network edge.
KEY SPECIFICATIONS
- Form Factor: 1U Rackmount
- Compute: Up to 16x APUs
- Host Processor: Dual AMD EPYC™
- Networking: Dual 100GbE QSFP28
TARGET APPLICATIONS
- Enterprise vector database acceleration
- Retrieval-Augmented Generation (RAG) pipelines
- Smart city and intelligent traffic management
- High-density compute nodes for telco edge
Leda-E 2U Enterprise Server
Maximum APU density for the most demanding workloads.
The flagship of our AI compute lineup. The Leda-E 2U server is a powerhouse designed to tackle the most complex generative AI challenges, offering massive scalability and redundant reliability for mission-critical data centers.
KEY SPECIFICATIONS
- Form Factor: 2U Rackmount
- Compute: 1-8x Leda-E PCIe Cards
- Host Processor: Dual Intel® Xeon® Scalable
- Power Supply: Redundant Titanium (1+1)
TARGET APPLICATIONS
- Large-scale LLM fine-tuning and training
- Billion-scale similarity search engines
- Bioinformatics and molecular search
- Synthetic Aperture Radar (SAR) processing
Secure Your Supply Chain Today
Due to market fluctuations, our chip agency services are priced per project.