1 / 6

CRC-32 on Maxeler DFE

CRC-32 on Maxeler DFE. Authors: Klemen Košir, Matevž Žugelj Faculty for Computer and Information Science, Ljubljana. Algorithm. CRC-32 is used for the calculation of checksum for data transfer, compression, mobile networks, etc . T(n) = O(n ).

irisa
Télécharger la présentation

CRC-32 on Maxeler DFE

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. CRC-32 on MaxelerDFE Authors: Klemen Košir, Matevž Žugelj Faculty for Computer and Information Science, Ljubljana

  2. Algorithm CRC-32 is used for the calculation of checksum for data transfer, compression, mobile networks, etc. T(n) = O(n). Our CRC-32 implementation does not use lookup tables and is relatively slow. We wanted to speed-up the algorithm by using a Maxeler machine.

  3. Implementation for(i = 0; i < N; i++) { bit = crc.and(1); crc = 0xFFFFFFFF; mask = 0xEDB88320; crc = crc >> 1; if(bit ^ input[i] == 1) crc = crc ^ mask; } crc = ~crc;

  4. Results

  5. Observations The algorithm does not have enough data reusability, therefore the more bits we send on the DFE at once, the slower the computation is. The execution time is higher compared to the CPU’s, because of the number of ticks we have to wait for out loopback to deliver the previous value. We are not sure if it is possible to implement this algorithm without a loopback, since we have loop data dependency. The algorithm works the fastest, if the stream delivers 1 bit at a time. If we send 1 bit at a time, there are even less arithmetic operations which speeds up the process even more.

  6. Authors Klemen Košir klemen913@gmail.com Matevž Žugelj matevz.zugelj@siol.net

More Related