Home    Bloggers    Messages    Webinars    Resources   
Tw  |  Fb  |  In  |  Rss
Jan Decaluwe

MyHDL: The Case for a Better HDL

Jan Decaluwe
Page 1 / 2 Next >
Page 1 / 8   >   >>
Max Maxfield
Max Maxfield
10/10/2012 10:37:13 AM
User Rank
Blogger
Re: RTL vs higher-level description
@Josyb: This is wonderful news -- very interesting -- if you email the files to me at max@CliveMaxfield.com I will push them up on a server and post a link to therm here (can you put the .qar into a .zip file also, because .zip files have an easier time going thru the email ...)

50%
50%
josyb
josyb
10/10/2012 10:27:59 AM
User Rank
Guru
Re: RTL vs higher-level description
@Max: here are the results ...

I drew up a 'contrived' (pun intended) example in 4 coding styles, actually 3 and a half ...

The "TWOPROCESS" follows the textbooks and uses 44 LEs but achieves 416MHz.

The "SYNCHRONOUS" actually uses 46 LEs and makes it to 346 MHz. You can optimize this one to use less resources, but only by adding a combinatorial process.

THe "HYBRID"  uses 36 LEs and climbs to 360 MHz

The fourth one, "HYBRIDJOSYB",  combines the "HYBRID" style with an external component and uses 35 LEs and is a bit quicker at 382 MHz.

There is also a testbench that shows that the RTL simulates equal for all 4 variants.

I attach the .qar (for the Altera gang) and a .zip for the others :) It is all pure VHDL, so it should be portable.

But how do I attach a file?

50%
50%
Myplanet
Myplanet
10/4/2012 7:21:30 AM
User Rank
Guru
Re: RTL vs higher-level description
Cfelton, for us weekends are for enjoying with family and kids. They won't allow me to go to work or even to open any office files. In most of the time they want to spent time with me for outing. I think that's the only time am spending with my family as a family man.

100%
0%
Max Maxfield
Max Maxfield
9/28/2012 9:56:32 AM
User Rank
Blogger
Re: RTL vs higher-level description
@cfelton: I have a bookshelf here in my office that is groaning under the weight of books and DVDs that are clamoring for my attention -- plus I now have a 50+ "wish list" on Amazon -- there's just so much fun stuff to do and so little time to do it all :-)

 

 

50%
50%
cfelton
cfelton
9/27/2012 2:32:46 PM
User Rank
Guru
Re: RTL vs higher-level description
@max "I'm really trying to make the effort to save weekends for stuff I want to do like reading ..."

 

Yes, my to read pile is growing not shrinking, ugh.  The pile also act like a LIFO, poor neglected books at the bottom.  I like the mega mosaic.

50%
50%
Max Maxfield
Max Maxfield
9/27/2012 11:49:48 AM
User Rank
Blogger
Re: RTL vs higher-level description
@cfelton: "I have been looking for one of those :)"

 

I know what you mean -- I have spend far too many weekends working -- life is too short -- I'm really trying to make the effort to save weekends for stuff I want to do like reading, watching science fiction films, and my hobby projects (like my mega mosaic)

50%
50%
Max Maxfield
Max Maxfield
9/27/2012 11:47:27 AM
User Rank
Blogger
Re: RTL vs higher-level description
@josyb: Seriously though, I would like to see this...

50%
50%
Max Maxfield
Max Maxfield
9/27/2012 11:46:25 AM
User Rank
Blogger
Re: RTL vs higher-level description
@josyb: "...is a bit of work though."

But it would be you doing the work, so I'm OK with that (grin)

50%
50%
cfelton
cfelton
9/27/2012 10:29:05 AM
User Rank
Guru
Re: RTL vs higher-level description
@joysb "one lazy sunday coming ..."

I have been looking for one of those :)

 

50%
50%
thrakkor
thrakkor
9/27/2012 9:56:28 AM
User Rank
Blogger
Re: RTL vs higher-level description
I agree a synthesis tool should be able to recognize and optimize any FSM style.  it really comes down to preference.  can you hand-optimize a FSM?  sure, but it seems like a waste of time to me, unless timing closure dictates we add extra pipelines or states to improve timing.

if you need combinatorial signals based upon the FSM, just create them...

I keep everything in a single process unless I know timing closure is going to be a bi#$h.  then I start moving registered outputs into their own process.

100%
0%
Page 1 / 8   >   >>
More Blogs from Jan Decaluwe
Here is an example of solving a hardware design problem by first developing an algorithm, and then using HDL language features to write code that resembles this algorithm as closely as possible.
Thanks to its VHDL and Verilog conversion functions, MyHDL development can be seamlessly integrated with standard HDL development flows.
Until a few weeks ago, I knew next to nothing about cellular automata. Therefore, my first step was to try to discover as much as possible as quickly as possible about this application domain.
Why don't we make a little modeling contest out of this? Given a width 'n,' write a reference model in Verilog or VHDL that returns the corresponding Gray code in a representation of your choice...
VHDL and Verilog have served us exceptionally well, but they also have significant issues. In order to realize the full potential of HDL design, we need a better HDL. In this blog, the creator of MyHDL -- Jan Decaluwe -- considers the concept of signal assignments.
flash poll
follow us on twitter
follow Xilinx on twitter
like us on facebook
like Xilinx on facebook
All Programmable Planet     About Us     Contact Us     Help     Register     Twitter     Facebook     RSS