1
\$\begingroup\$

What is the difference between an FPGA Intellectual Property core and an Accelerator Function Unit?

As far as I understand, an AFU is an IP core developed by the end-user (as opposed to IP cores delivered in the hardware by the vendor). Is it correct?

\$\endgroup\$
2
  • \$\begingroup\$ AFU is not a very common term. Can you point to it's definition? It sounds like it is just a specific IP. \$\endgroup\$ – Eugene Sh. Dec 4 '15 at 15:49
  • 1
    \$\begingroup\$ What are these tags? Power supply? Transistors? Programing? These do not relate to the question \$\endgroup\$ – efox29 Dec 4 '15 at 16:09
1
\$\begingroup\$

The term IP core typically refers to a design delivered by a third party.

The accelerator function unit might be part of that IP core or it can be your design.

If you distribute your design for others, then it will be an IP core for them.

\$\endgroup\$
3
\$\begingroup\$

I interpret the question a little differently than the other answer.

An IP core is a general purpose processing unit.

An acceleration function unit is generally a term given to a block of IP that can not do general purpose processing. The most common example of this would be a graphics processing core. In that case, you can't send it generic code like you would with a general processing unit. It can only do certain things really well.

In either of these cases, an AFU or an IP Core can be either in asic form or in FPGA IP blocks. A general direction the industry appears to be going currently is to have hardware IP cores and AFUs as soft ip in an on-board fpga.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.