# Accessing Intel FPGAs for Acceleration

- Institution: CERN
- Department: CERN Computing Seminar
- Speaker: Karl Qi
- Year: 2018
- Language: English
- Subjects: Computer science
- Source: [CERN](https://videos.cern.ch/record/3012804)
- Licence: [CERN](http://copyright.web.cern.ch)
- Who can watch: Tweens profiles and older (sources give no ratings; placed by Invisible Network from the source's own words)

In this presentation, we will discuss the latest tools and products from Intel that enables FPGAs to be deployed as Accelerators. We will first talk about the Acceleration Stack for Intel Xeon CPU with FPGAs which makes it easy to create, verify, and execute functions on the Intel Programmable Acceleration Card in a Data Center. We will then talk about the OpenCL flow which allows parallel software developers to create FPGA systems and deploy them using the OpenCL standard. Next, we will talk about the Intel High-Level Synthesis compiler which can convert C++ code into custom RTL code optimized for Intel FPGAs. Lastly, we will focus on the task of running Machine Learning inference on the FPGA leveraging some of the tools we discussed.

About the speaker

Karl Qi is Sr. Staff Applications Engineer, Technical Training. He has been with the Customer Training department at Altera/Intel for 8 years. Most recently, he is responsible for all training content relating to High-Level Design tools, including the OpenCL SDK, HLS Compiler, and DSP Builder as well as Machine Learning applications. Prior to joining Altera/Intel, he was with Lockheed Martin Space Systems. Overall he has 15 years experience with FPGAs. .

0:00:00 Slide 1

0:03:26 Slide 2

0:04:19 Slide 3

0:07:21 Slide 4

0:08:49 Slide 5

0:09:47 Slide 6

0:10:03 Slide 7

0:10:56 Slide 8

0:12:22 Slide 9

0:12:41 Slide 10

0:13:37 Slide 11

0:16:43 Slide 12

0:17:01 Slide 13

0:18:12 Slide 14

0:19:04 Slide 15

0:19:56 Slide 16

0:20:09 Slide 17

0:20:26 Slide 18

0:20:34 Slide 19

0:21:02 Slide 20

0:21:50 Slide 21

0:22:18 Slide 22

0:23:40 Slide 23

0:25:21 Slide 24

0:25:55 Slide 25

0:26:33 Slide 26

0:27:08 Slide 27

0:27:30 Slide 28

0:28:34 Slide 29

0:29:38 Slide 30

0:30:07 Slide 31

0:31:39 Slide 32

0:32:18 Slide 33

0:33:18 Slide 34

0:33:30 Slide 35

0:34:03 Slide 36

0:34:45 Slide 37

0:35:49 Slide 38

0:36:11 Slide 39

0:37:14 Slide 40

0:38:58 Slide 41

0:42:15 Slide 42

0:43:08 Slide 43

0:43:44…

## Lectures

1. [Accessing Intel FPGAs for Acceleration](https://invisiblenetwork.ai/learn/cern/3012804/dadaa4a8d097), 1 h 22 min (CERN's player)