FPGA Acceleration of Convolutional Neural Networks (CNNs)
White Paper FPGA Acceleration of Convolutional Neural Networks Overview Convolutional Neural Networks (CNNs) have been shown to be extremely effective at complex image recognition problems.
SNIA defines several components collectively called “computational storage.” For a typical IA-220-U2 deployment, the acceleration tasks are called CSS (computational storage services).
For example, the IA-220-U2’s Agilex FPGA can perform compression much faster than a CPU—even surpassing the transfer rate of the storage system for transparent compression.
CSPs (like the IA-220-U2) work alongside FLASH, providing accelerated computational storage services (CSS) by performing compute, such as compression or encryption. This lets users build out storage using standard SSDs instead of being locked into a single vendor’s FLASH storage.
What features make the IA-220-U2 a powerful CSP?
Built with the latest PCIe Gen4 interface, the IA-220-U2 can transfer up to twice the bandwidth of Gen3 devices.
NoLoad provides FPGA IP and host components. For the FPGA IP, you can see in the diagram the main components. More details are below:
As a complete solution, NoLoad provides host software with a choice of implementation:
The HRG gives you much more detail about the card such as block diagrams, tables and descriptions.
r1 v0
Our technical sales team is ready to provide availability and configuration information, or answer your technical questions.
"*" indicates required fields
White Paper FPGA Acceleration of Convolutional Neural Networks Overview Convolutional Neural Networks (CNNs) have been shown to be extremely effective at complex image recognition problems.
High-speed networking can make timestamping a challenge. Learn about possible solutions including card timing kits and the Atomic Rules IP TimeServo.
BittWare customer OVHcloud built a powerful anti-DDoS solution using FPGA technology, specifically the XUP-P3R card.
Panel Discussion How Today’s FPGAs are Taming the Data Deluge Problem From Gen5 to AI, NOCs to RF at the Edge Watch the recording for