Null Convention Logic
October 11, 2007It is extremely rare in our industry that totally new approaches for Logic circuit design are taken. I don’t know the exact reasons and I really don’t want to get into the “fight” between tool vendors and engineers.
Null Convention Logic, is a totally different approach to circuit design. It is asynchronous in its heart (I guess half of the readers of this post just dropped now).
It is not new and being currently pushed by its developers in Theseus Research.
They published a book, which I really recommend reading. It is not very practical with the current mainstream tools and flows but it is a very interesting reading that will open your eyes to new approaches in logic design.
You can get a good introduction to the book’s content by reading this paper. It is fairly technical and would need a few good hours to digest and grasp the meaning behind, especially given the fact that it is so much different than what we are used to – forget about AND, OR and NOT gates…
Book link here.
Adventures in ASIC Digital Design 
Very interesting blog. I am watching!!
It is more or like forcing a paradigm shift :-). The paper is intersting.
It is a paradigm shift. The problem is that as I understood, after reading some papers, that it is not all bright.
They (Theseus research) claim area saving. It is also claimed that throughput is better, which I believe is an issue which is still under heavy debate.
I personally find it hard to believe this saves area, and once we scale down, leakage is getting more and more dominant. this means that even if you got less dynamic power spent in your design but you are significantly larger, leakage will kill ya.
I experimented with some minor blocks and never found an area save. There is the possibility I am dumb and did things wrong of course…
maybe this will become more interesting if we will find some way to create these special threshold gates in transistor level in a more optimized fashion.
Hi, this blog brought up memories. One way to describe NCL is looking at the soldiers.
A few walking in lock-step compared to 1 million as in synchronous logic? Which group will react faster and with less idling?
I had to understand NCL at the time and you might find it useful to look
http://www.exemark.com/asynch/downloads.html
and read the articles or try the examples.
http://www.exemark.com/publications/status.html
Look for Theseus and Fulcrum.
As soon as it clicked I wondered why such a technology is not used more, but it is, like using such a construct as a wrapper to link synchronous IP blocks.
Hi, I do not believe that saving gates with NCL is really the target. From my point of view saving gates is nice but not really important. The money is is in time to market and design time of engineers.
Scaling and moving to other technologies is a no brainer as the function will just work faster or slower.
In most of the cases the cost is NOT in having a few gates more or less, but Verification taking some months more or a block can be reused without thinking about any synchronous issues.
Or a new respin of the ASIC.
The main problem with NCL IS the shift to a new way of thinking, but did we not have this when VHDL and Verilog arrived??
Juergen,
I agree and disagree…
There are many factors which determine a successful product.
That is true that theoretically time to market would be faster with fully asynch. logic design, but you have to comply to the world around you. Some systems are synchronous by nature and will stay that way forever.
I also have to disagree when you say area is not the target – this again depends heavily on the product. In the DRAM industry for example, area is a key factor and one fights on every square micrometer.
I also didn’t see any clear cut advantage in performance when using NCL to do the job and comparing it to old synch. design.
I do believe that it was (and is) not given enough thought and consideration by engineers.
There is also the problem of CAD tools and the migration costs to asynch. logic.
when RTL languages came they significantly improved productivity and the move was mostly well accepted in the industry. I wish there would be a GOOD asynch synthesis tool that would give better or even comparable results with synch design tools.
Are you aware of any commercial products that use NCL and the reasons it was preferred over synch. design? I probably just do not know enough about it.
Have you had a look at the website and the articles, SW?
At the time I had quite a good connection to some of the people.
There is always a best theoretical way to design a product assuming you have all the tools and people that know how to use them. And a realistic one …
The tools issue is correct, this is why for the moment it is done more “under the surface” and where the pain is worst.
Look at companies like SUN, ntel, Infineon, Handshake Solutions, Fulcrum, Silistix just to name a few.
To debate which approach is better does not help.
Fire up your Excel, add in all of the costs and times people and tools and tools knowledge, and other issues related to your project, do the sums and decide then.
Asynchronous can be looked at as well as the
“GREEN – GREEN ROUTE for ELECTRONICS and EMBEDDED design with a NEW approach”, burning the least energy (as soon as the tools are there) to get a product into production.
And the Re-USE of IP is easier as well and technology independant.
The usual chicken and egg situation. Have a look at Theseus.com. They just merged, Most of the projects seem to be in the military area, so a solution to aspects they cannot solve otherwise. Actually I think VHDL developed the same way …
It is basically a similar issue as C to Gates into FPGA and ASIC. Here we are a few steps further already.
Readers of this post might be interested to know that Null Convention Logic has now been acquired by start-up Wave Semiconductor Inc and is being commercially deployed.
Refer http://www.camgian.com/2010_01_05.html
There are many increasingly important advantages to NCL, some of them have been touched on above. If you want to know more feel free to contact me via the company website http://www.wavesemi.com, or post questions here.