Intel’s Programmable Solutions Division today announced that Intel Agilex 7 with R-Tile small chips is shipping in volume. This will bring customers the first FPGA with PCIe 5.0 and CXL capabilities, and the only FPGA product with hard intellectual property supporting these interfaces.
“Customers need cutting-edge technology that offers scalability and customization to not only efficiently manage current workloads but also to adapt capabilities and functionality as their needs evolve. Our Agilex products can provide customers with the speed, power and functional innovation they need while providing flexibility and resiliency for the future. For example, customers are leveraging R-Tile with PCIe Generation 5 and CXL to accelerate software and data analytics, reducing processing time from hours to minutes,” said Shannon Poulin, vice president and general manager of Intel’s Programmable Solutions Group.
Faced with time, budget and power constraints, organizations across industries, including data centers, telecommunications and financial services, are turning to FPGAs as a flexible, programmable and efficient solution. With Agilex 7 and R-Tile, customers can seamlessly connect their FPGAs to processors such as fourth-generation Intel Xeon Scalable processors. the configurable and scalable architecture of Agilex 7 enables customers to rapidly deploy custom technology for their specific needs, reduce overall design costs and development processes with massive hardware speed, and accelerate execution for optimal data center performance.
The Agilex 7 FPGA with R-Tile chip is a technological leader compared to other FPGA competitors, with 2x faster PCIe 5.0 bandwidth and 4x higher per-port CXL bandwidth.
According to a white paper by Meta and the University of Michigan, adding FPGAs with CXL memory to fourth-generation Xeon-based servers, along with efficient page placement using transparent page placement (TPP), can improve Linux performance by 18%.
In addition, UnifabriX demonstrated its CXL-enabled smart memory nodes in multiple performance benchmarks, one of which showed a 28% improvement in HPCG (High-Performance Conjugate Gradient) benchmark scores while leveraging more than 2x the fourth-generation Xeon cores to handle HPC workloads.
