By using AWS re:Post, you agree to the AWS re:Post Terms of Use
AWS re:Postre:Post

FPGA Acceleration in AWS [F1 Service]

1

I am trying to do some optimization for network traffic handling by utilizing FGPA acceleration on AWS [F1?]. Per my understanding the F1 service is based on Xilinx Alveo SmartNIC hardware.

So my question is, is it possible to configure the EC2 instance with FPGA acceleration in a way that traffic is passed to the EC2 from the FPGA smartnic and not the ENI?

Topics
Compute
Tags
FPGA Development
Language
English
estro
asked 10 months ago253 views
2 Answers
1

No. The FPGAs do not have direct access to the network. The FPGAs communicate via PCIe to the instance CPU, where the Enhanced Networking (ENA) drivers run. ENA provides a high-performance, low-latency virtualized network interface suitable for data movement to the F1 instance. See the EC2 ENA driver documentation for more details.

Documentation: https://github.com/aws/aws-fpga/blob/master/FAQs.md

profile pictureAWS
ladybug
answered 10 months ago
  • hackrsio
    3 months ago

    Any chance this is on the roadmap? I understand it’s a complicated feature to add, but I think this would add a lot of value to many users. I know there would be a decent number of users of it in AWS Tokyo. I believe you can do it while keeping the network safe by putting a switch in front of the card heavily restricting the activity of those network ports. Although my use-case requires production data I think if AWS was very worried about it messing up the network, there is still value even if it’s only an isolated network back to host ENI for testing designs.

0

Please, accept this answer if it was useful to you.

Thank you.

profile pictureAWS
ladybug
answered 7 months ago

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions

Relevant content