ANIC is a suite of flexible FPGA IP blocks that have been optimised to accelerate high-performance networking pipelines for Speedster7t FPGAs and the VectorPath Accelerator Card.
As the demand for high-speed data processing continues to grow exponentially, Achronix is working to deliver cutting-edge solutions to address the evolving needs of the networking industry.
By integrating 400 GbE and PCIe Gen 5.0, Achronix is empowering data centre operators, cloud service providers, and telecommunications companies to create SmartNIC solutions with improved levels of performance, scalability, and flexibility.
The key features and benefits of the ANIC include:
- Unparalleled speed – with 400 GbE support, the ANIC IP enables ultra-fast data transfer rates, allowing organisations to process massive amounts of data in real-time. This accelerated network throughput maximises application performance and significantly reduces latency.
- Accelerated time to market – ANIC modular architecture enables customers to select SmartNIC components necessary for their application. Each optimised IP block is pre-verified with closed timing to speed design. Coupled with partial reconfiguration (the ability to dynamically change the functionality of a block within the IP design), solutions can be seamlessly modified in the field.
- Custom IP support – Building on top of ANIC, customers can deploy custom IP functions, such as key-value stores, intrusion prevention, de-duplication, and other network applications, to provide highly parallelized, value-added network solutions at 400 GbE network speeds.
Commenting Steve Mensor, VP of Marketing for Achronix Semiconductor, said, "This breakthrough achievement underlines our commitment to providing our customers with the most advanced, high-performance networking solutions. With the increased speed, bandwidth, scalability, and data acceleration offered by ANIC, customers can unlock new levels of performance and address the ever-growing demands of modern data centres and communication networks."
ANIC modular IP runs on Speedster7t AC7t1500 FPGAs and VectorPath S7t-VG6 Accelerator Cards offering improved performance for networking and compute acceleration applications.
The Speedster7t architecture includes a revolutionary 2D network on chip (2D NoC) that provides 20 Tbps of total bandwidth. The 2D NoC offers high-speed connectivity between the FPGA fabric and the high-speed interfaces including 400 GbE, PCIe Gen 5.0, GDDR6, and DDR4/5.
Additionally, Speedster7t FPGAs have machine learning processors (MLPs) distributed across the FPGA fabric. Each MLP is a highly configurable, compute-intensive block, with up to 32 multipliers that support integer formats from 4 to 32 bits and various floating-point modes including direct support for TensorFlow's bfloat16 format and block floating-point (BFP) format.
