Breaking News
Asynchronous Techniques Push Moore's Law to the Physical Limit
2/13/2014

Image 1 of 3      Next >

Image 1 of 3      Next >

Return to Article

View Comments: Threaded | Newest First | Oldest First
Max The Magnificent
User Rank
Blogger
I love asynchronous logic
Max The Magnificent   2/13/2014 11:34:59 AM
NO RATINGS
I don't know why, but I love the concept of asynchronous logic -- I know, I know, synchronous logic was good enough for my father and for his father before him, so it should be good enough for me ... but still, I love the concept.

Max The Magnificent
User Rank
Blogger
Re: I love asynchronous logic
Max The Magnificent   2/13/2014 11:37:13 AM
NO RATINGS
Actually... did you know that, in the early days, asynchronous logic was the norm. This was for several reasons:

a) Synchronous techniques were not well understood.

b) Synchronous logic used more transistors, which were expensive

c) Everything was crafted by hand -- no automatic tools.

 

TechColab
User Rank
Rookie
Re: I love asynchronous logic
TechColab   2/13/2014 1:50:32 PM
NO RATINGS
Surprised the ARM asynchronous processor didn't get mentioned.
http://en.wikipedia.org/wiki/AMULET_microprocessor

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/13/2014 5:55:49 PM
@TechColab: "Surprised the ARM asynchronous processor didn't get mentioned."

Well, this blog is not supposed to be an historical introduction to asynchronous logic, but a discussion about the REAL problems that synchronous logic is facing today. Thus, the main purpose of this text is setting the scene for introducing the Globally-Asynchronous, Locally-Synchronous approach, A.K.A. GALS, a tradeoff between hand-made async communication and synchronous island synthesized with COTS EDA tools.

The point is that the same Manchester University team that built the Amulet, are now one of the most active advocates for the GALS approach and the asynchronous-NoC -- it's worthy to note that Steve Furber, one of the main architects of the original ARM, is one of the leaders of this Manchester async research group... and he lauched the Silistix Spin-Off too, a startup based in async and backed by Intel. but I'm digressing so much...

The real purpose of this blog is starting a logical line of thinking in which the different components  and main ideas for building async circuits over COTS FPGAs are introduced and justified. FPGA technology is intended in this framework as an educative platform for learning and playing with REAL async circuits. Next step will be understanding the basic  Ivan Sutherland's micropipelines; after that, I'll introduce you an open project full of resources that should be in the logic designer toolbox... and Finally, I promise a mayor and exciting upgrade from one of the greatest gurus on Async: CalTech's Alain J. Martin.

 

I you want learn more about this marvelous topic, check the conference I'm going to present at ESC / EELive!: Technology Born @ CERN, a tutorial

 

PS. Sorry if I digress too much, but I really loves this topic. I've bet & lost for Async so many times, that now I'm completely aware the next big async thing is really near -- too much lessons learned!!

tb100
User Rank
CEO
Re: I love asynchronous logic
tb100   3/12/2014 4:30:12 PM
NO RATINGS
Intel has been involved with a number of asynchronous logic companies. They bought Fulcrum, which has a number of chips, including Ethernet switches, that use asynchronous logic inside. And they partner with Achronix, which makes FPGAs that utilize asynchronous techniques for internal data pipelining. 

You would almost think that Intel was up to something!

 

 

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   3/13/2014 7:39:24 PM
NO RATINGS
@tb100: "You would almost think that Intel was up to something!"

Very impressive!! You really manage to stay on top of things!!

But I would add one more deal: Intel funded Silistix, a Manchester Unversity start-up leaded by Steve Furber (one of the original ARM architects) and which focused in the development of an asynchronous On-Chip core interconnect.

You can find more info in the next Silistix Limited website -- where references to Intel has been convenently suppressed, of course ;-)

betajet
User Rank
CEO
Re: I love asynchronous logic
betajet   2/13/2014 1:51:01 PM
NO RATINGS
And don't forget...

d) Back in those days, you had to keep winding the clock :-)

Max The Magnificent
User Rank
Blogger
Re: I love asynchronous logic
Max The Magnificent   2/13/2014 1:53:29 PM
NO RATINGS
@betajet: Back in those days, you had to keep winding the clock :-)

It was the only exercise I got :-)

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/13/2014 5:21:02 PM
NO RATINGS
@Max: "Actually... did you know that, in the early days, asynchronous logic was the norm"

In fact, most of the computers that were built in the 50ths and 60ths had asynchronous logic in their inner gears, e.g. ORDVAC, BRLESC II.

As it's explained in this blog, the rise of Moore's law enabled by CMOS technology empowered the synchronous paradigm adoption, but the fall of Moore's law while hitting the nano scale is signaling the beginning of the new Async era!! -- just ask to the Intel guys what is behind most of the papers are presenting this year at ISSCC ;-)

 

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/13/2014 5:05:10 PM
NO RATINGS
@Max: "I don't know why, but I love the concept of asynchronous logic"

The reason for this is quite simple: this is the way in that Nature works!! Just a couple of examples:
  1. BIOLOGY: asynchronous logic  behaves just as a neuronal network, i.e. data Synchronization is handled by local handshaking and feedback.
  2. PHYSICS: Thes synchronous logic paradigm assumes that you can build a clock that ticks at every point in the space --the registers-- at the same time... but this is just impossible, as you cannot build such a physical system due to the limitations imposed by Einstein's Relativity.

And, by the way, I'm really glad of seeing you and me putting the Async Logic into the headlines again together: this is how we get acquainted just a year ago!!

Max The Magnificent
User Rank
Blogger
Re: I love asynchronous logic
Max The Magnificent   2/13/2014 5:24:10 PM
NO RATINGS
@Garcia: ...this is how we get acquainted just a year ago!!

Is it only a year? It seems like I've known you for much longer (and I mean that in a good way LOL)

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/13/2014 6:00:29 PM
NO RATINGS
@Max "is it only a year? It seems like I've known you for much longer (and I mean that in a good way LOL)"

I tell this to my wife at least twice a week -- the good thing is that Rebeca understands how fuzzy I am someties and that there is not pun intented in my words.

Max The Magnificent
User Rank
Blogger
Re: I love asynchronous logic
Max The Magnificent   2/13/2014 6:19:25 PM
NO RATINGS
@Gracia: I tell this to my wife at least twice a week ... the good thing is that Rebeca understands...

Ah, but is Rebeca the name of your wife or your girlfriend?

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/13/2014 6:24:12 PM
NO RATINGS
@Max: "Ah, but is Rebeca the name of your wife or your girlfriend?"

She used to be my girlfriend, now she is my daughter's mom and, as a side effect, my wife LOL

Max The Magnificent
User Rank
Blogger
Re: I love asynchronous logic
Max The Magnificent   2/13/2014 6:30:02 PM
NO RATINGS
@Garcia: She used to be my girlfriend, now she is my daughter's mom and, as a side effect, my wife LOL.

Also she is your mother-in-law's daughter. Now my head hurts. Have you ever heard that song: I'm My Own Grandpa?

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/13/2014 6:43:40 PM
NO RATINGS
@Max: "Also she is your mother-in-law's daughter. Now my head hurts. Have you ever heard that song: I'm My Own Grandpa?"

Now, I'm comlpetely lost. I'm glad to see there is a flow diagram indicating the relations... but I've gotten crazy after the first minute -- I need mor colors in the map LOL

AZskibum
User Rank
CEO
Re: I love asynchronous logic
AZskibum   3/30/2014 1:50:16 PM
NO RATINGS
I'm curious what tools all you "I love asynchronous logic" guys are using to implement and verify your designs.

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   5/15/2014 5:20:20 PM
NO RATINGS
@AZskibum: "I'm curious what tools all you "I love asynchronous logic" guys are using to implement and verify your designs."

Well, last week I used standard VHDL in order to describe a 4000 stages-depth asynchronous micropipeline in the Xilinx Zynq device that powers my Zedboard. No strange circuitry, tools.... just PlanAhead+Xise+plain HDL description.


P.S.: Sorry for my late reply, I missed this post!!

Paul A. Clayton
User Rank
CEO
Re: I love asynchronous logic
Paul A. Clayton   2/13/2014 9:29:55 PM
NO RATINGS
Asynchronous design is interesting at various scales, but I also find wave pipelining interesting (which is sort of related to timing-dependent asynchronous design). It seems neat that one can avoid a pipeline latch by timing more pulse-like signals (waves) such that they do not overlap. Sadly, wave pipelining is a victim of fast clocks, process variation, and other modern factors.

Garcia-Lasheras
User Rank
Blogger
Re: I love asynchronous logic
Garcia-Lasheras   2/14/2014 9:21:50 AM
NO RATINGS
@Paul: "Sadly, wave pipelining is a victim of fast clocks, process variation, and other modern factors."

This issues are common for every design in which you must do deterministic timing asumptions. This is one of the reasons because a well known approach to asynchronous design is considered a glitter alternative for the future: the delay insensitive logic.

This kind of circuits are "correct-by-design", as you don't need to make any extra assumption about process variation, timing, etc. In addition, this circuits automatically adapt their performance to environment variations, such as temperature derives or a swinging power supply voltage.

Plus, delay insensitive design can be used in flexible substrates. An interesting example is the 8 bits MCU that Seiko developed some years ago:



 

 

 

Max The Magnificent
User Rank
Blogger
Re: I love asynchronous logic
Max The Magnificent   2/14/2014 9:35:45 AM
NO RATINGS
@Paul A. Clayton ...I also find wave pipelining interesting...

Wow -- I've not heard that term for more than 20 years I think -- I'd forgotten all about it -- when I first heard about that technique, I was really excited ... but then it seemed to fade away...

tpfj
User Rank
CEO
Not what Moore's Law says
tpfj   2/13/2014 5:28:53 PM
NO RATINGS
Moore's Law says nothing about speed and everything about circuit density. Speed WAS a side effect, but that has waned with advanced nodes. The title should read something like "Asynchronous Techniques Push Circuit Speeds to the Physical Limit". A rather obvious statement if you ask me. Asynchronisity (is that even a word?) has nothing to do with and is completely orthogonal to circuit density.

Max The Magnificent
User Rank
Blogger
Re: Not what Moore's Law says
Max The Magnificent   2/13/2014 5:37:50 PM
NO RATINGS
@tpfj: Suddenly a rather obvious statement if you ask me.

Suddenly? Did this "rather obvious statement" take you by surprise? LOL

 

tpfj
User Rank
CEO
Re: Not what Moore's Law says
tpfj   2/13/2014 5:42:49 PM
NO RATINGS
Fair comment, just corrected. So I take it you disagree with my statement then?

Max The Magnificent
User Rank
Blogger
Re: Not what Moore's Law says
Max The Magnificent   2/13/2014 5:46:05 PM
NO RATINGS
@tpfj: So I take it you disagree with my statement then?

Not at all -- in fact I agree with you -- it was just that your use of "surprised" took me by surprise LOL

Garcia-Lasheras
User Rank
Blogger
Re: Not what Moore's Law says
Garcia-Lasheras   2/13/2014 6:13:15 PM
@tpfj: "joore's Law says nothing about speed and everything about circuit density"

Moore law is all about jumping from one process node to the next and the overal exponential advantage. Density, Speed and Power consumption are all related with the transistor shrinking process.

Moore's law is meaningless beyond CMOS or derivatives, as it badly collapses when data communication replaces data processing as the most expensive process in terms of power consumption, propagation delay, occupied layout area and, of course, monetary cost.

Remeber that by just adding cores/logic, you cannot increase the overall performance in General Purpose Programming.

Back in the nineties, CPU were sold by highlighting their clock speed, as this was really the parameter that supposed an advantage in day to day application. e.g. office suite, graphics, boot-up time.

tpfj
User Rank
CEO
Re: Not what Moore's Law says
tpfj   2/13/2014 6:21:14 PM
NO RATINGS
http://en.wikipedia.org/wiki/Moore's_law

"Several measures of digital technology are improving at exponential rates related to Moore's law, including the size, cost, density and speed of components. Moore himself wrote only about the density of components (or transistors) at minimum cost."

Garcia-Lasheras
User Rank
Blogger
Re: Not what Moore's Law says
Garcia-Lasheras   2/13/2014 6:36:19 PM
NO RATINGS
OK, Mr. Moore is the most reliable voice in order to define his own law.

Now, lets define what costs means nowadays when building a new IC:

There are:
  • costs in delay, that translates in speed
  • costs in power consumption
  • costs in the silicon real-state area a transistor occupies in a chip
  • costs in the monetary expenses for building a transistor at a given process node.

Less call it Moore's law, Javi's law or tpfj's law: the point is that such a profitable trend has started to collapse, no matter how do we want to call it ;-)

tpfj
User Rank
CEO
Re: Not what Moore's Law says
tpfj   2/13/2014 6:58:41 PM
NO RATINGS
I agree, the trend is collapsing. Surprisingly, Moore's component, ie. circuit density, is the one that is holding out the longest. We lost speed a long time ago, we lost power fairly recently, we are now rapidly losing silicon $ cost (just look to current process node fab and related mask costs). I don't believe that asynchronous circuits will solve this either. Sure they will scoop out all those currently wasted picoseconds on the non-critical synchronous paths as well as any delay savings through flops that will be absent from an asynchronous equivalent design. But realistically, how much of an advantage do you think this will give you against a well balanced (ie many paths near critical speed) sychronous equivalent. I doubt it is double? Also, remember, a synchronous circuit gives you inherent parallism through pipe-lining. How do you achieve the same parallism asychronously? Really what I'm saying is the laws of physics that are currently kicking our asses will unfortunately still apply to asynchronous circuits, and hence should then suffer in similar ways.

Garcia-Lasheras
User Rank
Blogger
Re: Not what Moore's Law says
Garcia-Lasheras   2/13/2014 7:22:12 PM
NO RATINGS
@tpfj: "Really what I'm saying is the laws of physics that are currently kicking our asses will unfortunately still apply to asynchronous circuits, and hence should then suffer in similar ways."

I completely agree with you in this point. Indeed, the really worthy advantages comes from other side effects derived from this situation, not from a direct speed advantage.

The main message I pretended to explain, is that our live has been easy as we had not needed to pay attention to datapath internal geometry for a long while by using synchronous design, but now the paradigm has changed and clock starts to be a problem by itself.

The huge clock network in modern ICs sucks as many as 50% of the total power being wasted. For this reason, there are most practical applications of async circuits in RFID cards, in which async MCUs are heavily used.

In addition, the lack of a global clock with a fixed frequency reduces the EMI by orders of magnitude in async parts. For this reason, Async is used too in low power wireless communication modems.

And the list goes on... ;-)

 

betajet
User Rank
CEO
I never metastable I liked
betajet   2/13/2014 7:52:31 PM
NO RATINGS
I'm concerned about the GALS (Globally Asynchronous, Locally Synchronous) approach.  Aren't you going to have metastability problems with multiple clock islands?  Wouldn't one be better off with all islands having the same clock frequency, but with relaxed timing between islands?

David Ashton
User Rank
Blogger
Re: I never metastable I liked
David Ashton   2/13/2014 8:00:46 PM
NO RATINGS
@betajet I must admit that occurred to me as well.  Back in the day I used to deal with strings of modems used to poll terminals.  You HAD to have one clock through the whole system or you'd get errors.  If anything puts data out faster than something else can accept it, you'll start losing it.  You can use buffers to take up phase differences and slight clock slippage but eventually the buffer will over- or under-flow.

If your buffer is big enough and you're only sending bursts of data you can get away with differences, but we're talking big amounts of data here I think?

jackOfManyTrades
User Rank
Manager
Re: I never metastable I liked
jackOfManyTrades   2/14/2014 4:15:37 AM
NO RATINGS
This GALS idea seems to be just using a similar approach on a chip that is tradionally used on a PCB. The chips on a PCB will probably all be synchronous designs, but they certainly won't all be sharing the same clock tree!

betajet
User Rank
CEO
Re: I never metastable I liked
betajet   2/14/2014 9:01:40 AM
NO RATINGS
Jack wrote: The chips on a PCB will probably all be synchronous designs, but they certainly won't all be sharing the same clock tree!

Actually, they may be sharing a single PLL-based clock generator that derives all the clocks from a single reference.  In that case they are sharing the same clock tree, with the trunks implemented on the PCB and the branches inside different ICs.  For example, one of my designs has a 33.333 MHz master clock with matched clock lines so that the clock arrives at the various ICs and I/O modules at the same time.  Very clean, plenty of timing margin.

jackOfManyTrades
User Rank
Manager
Re: I never metastable I liked
jackOfManyTrades   2/14/2014 9:08:03 AM
NO RATINGS
Interesting - did you design all the ICs as well as the PCB?

I'm coming at this as an ex-digital IC designer (65nm was considered sexy when I stopped). I was designing standard components, which customers could take and use as they wished. From that point of view, once signals left the IC, all bets were off: not only did I not have any control over the customer's PCB layout, I didn't know what other ICs they might be using.

betajet
User Rank
CEO
Re: I never metastable I liked
betajet   2/14/2014 9:35:12 AM
NO RATINGS
Jack wrote: Interesting - did you design all the ICs as well as the PCB?  ...once signals left the IC, all bets were off.

I designed the contents of the FPGAs.  Actually, as an IC designer you do have some control over the PCB since you provide a "data sheet" that specifies the external timing of the chip: combinational delays, clock to output delays, and setup/hold constraints.  As the chip designer, you promise the PCB designer that if the timing constraints are met the IC will function properly -- no bets about it.

jackOfManyTrades
User Rank
Manager
Re: I never metastable I liked
jackOfManyTrades   2/14/2014 10:28:54 AM
NO RATINGS
Well, yes, I can specify stuff on a datasheet, but I want to make it as easy as possible for a customer to use my IC. For a start, my ICs would interface to an IC with processor cores on it, which need software. Consequently, the customer isn't going to change that at the drop of a hat because they've invested a lot of money in that software. So, if my IC is not compatible with the processor IC, my IC isn't going to get selected. If I had specified some interface which required the processor IC to be using some sort of balenced, shared clock, sales revenue would have been $0.


Was your PCB full of FGPAs really fully synchronous? 33MHz only gives you 30ns and that doesn't seem very long to get off one IC, across a PCB and on to another. Another path longer than 30ns would then become a Multicycle Path, which IMHO makes the design not fully synchronous (there will be some combination of process corner, voltage, temperature where the path delay is exactly a multiple of 30ns, which is no good unless you have some circuitry to mitigate the resulting metastability). However, I've been designing at the Matlab level for the last 6 years, so perhaps 30ns isn't so short now?

betajet
User Rank
CEO
Re: I never metastable I liked
betajet   2/14/2014 6:41:19 PM
NO RATINGS
Jack wrote: 33MHz only gives you 30ns and that doesn't seem very long to get off one IC, across a PCB and on to another.

30 ns is plenty of time if you register inputs and outputs at the I/O pads and have point-to-point connections that are less than 12 inches: clock to out is 5-10 ns, propagation delay is another 5-10 ns including reflections, and setup is maybe 5ns, so plenty of time.  A 33 MHz PCI bus has that same 30 ns clock period, and since it's a bus it has slower propagation than point-to-point if there are multiple cards, and requires that some wait state logic be done in the same cycle.  People do 66 MHz PCI, though you may need custom logic to meet the timing.

jackOfManyTrades
User Rank
Manager
Re: I never metastable I liked
jackOfManyTrades   2/17/2014 3:17:01 AM
NO RATINGS
30 ns is plenty of time if you register inputs and outputs at the I/O pads and have point-to-point connections that are less than 12 inches

Ah...interesting.

Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/14/2014 9:41:19 AM
NO RATINGS
@jackOfManyTrades: "This GALS idea seems to be just using a similar approach on a chip that is tradionally used on a PCB"

This is a very clever intuition. In some way, today ICs resembles a big system which has been collapsed inside the chip package. For this reason, terms such as "System-on-Chip", "Sytem-in-Package" or "Network-on-Chip" are a common topic in state-of-the-art VLSI design -- and the thing promises to get even more interesting in the future ;-)

Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/14/2014 8:54:52 AM
NO RATINGS
@Betajet: "Aren't you going to have metastability problems with multiple clock islands?"

Metastability is a real issue in all asynchronous logic designs, just as it is in synchronous designs dealing with asynchronous external inputs.

About the synchronous islands, let me clarify that having such designs blocks doesn't always mean you are using a conventional periodic clock. The synchronous island implies that you have a local clock distribution network that acts as an "isochronic fork". This is, by limiting the clock network to a local boundary, you can assure that you can insert a clock/trigger signal in the clock network input and this signal will reach the local registers/flip-flops with a controlled skew inside the synchronous island. By this way, you can use conventional synchronous EDA tools in order to design, synthesize and lay down the digital logic inside the island.

But the point is that the "clock" you are injecting into such a local synchronous island may be a locally generated signal, being aligned by this way with the asynchronous handshaking control circuitry. This is what is called a "pausable clock", and it supposes an advantage in power consumption as this is only tiriggering when the synchronous logic block has an actual work to do.

betajet
User Rank
CEO
Re: I never metastable I liked
betajet   2/14/2014 9:18:42 AM
NO RATINGS
Garcia-Lasheras wrote: This is what is called a "pausable clock", and it supposes an advantage in power consumption as this is only triggering when the synchronous logic block has an actual work to do.

I've always called those "gated clocks", and I used them decades ago with TTL.  It was easy to do in TTL, because you could make a clock tree from NAND gates instead of inverters and the NAND gates provided a good place to put in a clock enable.  You could also use the output enable of a tri-state driver as a clock gate.  You can do this as a fully synchronous design: when clocks do toggle, they do so at the same time.

The FPGAs I'm familiar with won't let you do this, because the global clocks don't have clock enables [that are visible to the designer].  So if you want to gate clocks, you have to do it through logic that adds skew and I suspect screws up your hold timing so that it's impractical.

Update: See Brian_D's comment below -- Xilinx does have this capability.

Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/14/2014 9:34:42 AM
NO RATINGS
@Betajet: "I've always called those "gated clocks", and I used them decades ago with TTL"

A pausable clock is a different concept. As you point out, clock gating implies that you have a clock signal that is always running and waiting for being injected into the clock distribution network.

A pausable clock is based in locally generated clock bursts. This is, there is an specific digital circuitry that is in charge of generating "trigger" signals only when required. In this way, you don't have a "real clock", but a kind of "pulse" generator -- I understand that the "pausable clock" term may lead to some confussion, but I didn't invent it ;-)

zeeglen
User Rank
Blogger
Re: I never metastable I liked
zeeglen   2/14/2014 9:54:55 AM
NO RATINGS
>A pausable clock is based in locally generated clock bursts.

Are you describing an independent clock oscillator that starts up on a trigger so there is always a defined and repeatable time between trigger and first clock edge?


Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/14/2014 10:27:17 AM
NO RATINGS
@Zeeglen: "Are you describing an independent clock oscillator that starts up on a trigger so there is always a defined and repeatable time between trigger and first clock edge?"

Well, what I'm describing is a circuit made from logic gates and that includes some "delay" feedback loops -- similar to a ring oscillator, but there are a lot of alternatives for implementing this.

About the defined and repeatable time, this is a very good question. A very clever alternative is having different delay loops, in such a way you can choose the clock period in real time. By this way, you can change the operating speed, but also implement EMI mitigation techniques -- similar to spread spectrum techniques.

zeeglen
User Rank
Blogger
Re: I never metastable I liked
zeeglen   2/14/2014 10:45:36 AM
NO RATINGS
>Well, what I'm describing is a circuit made from logic gates and that includes some "delay" feedback loops -- similar to a ring oscillator, but there are a lot of alternatives for implementing this.

Sounds similar to a triggered oscillator using an active digital LC delay line; it had an interesting bug. "Ornery oscillator finds problem with IEEE 802.3 Ethernet spec"

http://www.eetimes.com/author.asp?section_id=30&doc_id=1284625

Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/14/2014 11:24:27 AM
NO RATINGS
@Zeeglen: "Sounds similar to a triggered oscillator using an active digital LC delay line"

Yes, the working mechanism is very similar, but in the case of integrated asynchronous circuits, LC delay lines are not an option -- they are just to big and "expensive".

Instead, you must use the intrinsic delay of the logic gates. I order to make accurate logic delay calculations, the best option is using the logical effort theory, developed by Ivan Sutherland and Bob Sproull in the early nineties.

Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/14/2014 10:01:31 AM
NO RATINGS
@Betajet: "The FPGAs I'm familiar with won't let you do this, because the global clocks don't have clock enables"

I should clarify that I'm not trying to endorse the general use of async logic over COTS FPGAs -- but I did it in the past ;-). The point is that FPGAs are a really affordable platform for learning asynchronous techniques, just as they are for learning conventional synchronous logic.

I truly believe that async methodologies are being to be icreasingly used in the future as a way of dealing with the issues associated to deep submicron nodes and new nanoscale technologies.

Now, let me point to a paper I wrote some years ago about the efficient implementation of asynchronous logic over COTS FPGAs. Here, I expose some interesting experimental results on the old and good Xilinx SpartanIII and VirtexIV. Maybe the most interesting ones are the speed reached for both data communication [Mega Data Items per second] and pausable clock generation [MHz] -- I needed to perform some live demos in order to convince some researchers these were true:
  • Spartan III, -4 speed grade: >175 MDI/s, >300 MHz
  • Virtex IV, -4 speed grade: >550 MDI/s, >700 MHz


Brian_D
User Rank
Freelancer
Re: I never metastable I liked
Brian_D   2/15/2014 12:53:20 PM
NO RATINGS
@betajet "The FPGAs I'm familiar with won't let you do this, because the global clocks don't have clock enables [that are visible to the designer]"

Xilinx introduced a synchronously gated global buffer with the Virtex-2 (IIRC) family, which is called BUFGCE in the library primitives guides.

The underlying global buffer also supports sync/async clock muxing capabilities, exposed under a different library primitive name.

Their newer families also support gated regional clock buffers.

-Brian

betajet
User Rank
CEO
Re: I never metastable I liked
betajet   2/15/2014 2:22:33 PM
NO RATINGS
Thanks!  I didn't realize they had that capability.  I may need that feature some day.

 

Garcia-Lasheras
User Rank
Blogger
Re: I never metastable I liked
Garcia-Lasheras   2/15/2014 3:25:24 PM
NO RATINGS
Clock gating is also supported in most of the clock buffers in Xilinx 7 Series devices, e.g. in global clock buffers (BUFG) and horizontal clock buffers -- a regional clock distribution network -- (BUFH).

You can also select a lot of different clock inputs for each different clock distribution network, including internally generated clocks.

Ther is more info about this in the next user guide:

Xilinx UG472 7 Series Clocking

David Ashton
User Rank
Blogger
Moore's law.....
David Ashton   2/13/2014 7:52:49 PM
NO RATINGS
tpfj wrote: "Moore himself wrote only about the density of components (or transistors)".

Based on that, and the impending proliferation of 3D ICs, are we not extending Moore's law vertically, so to speak?  True, the transistors you can fit in one layer are getting limited, but we're still putting more transistors on a chip by putting in more layers.

Max - you're an expert on this, what do you think?

alex_m1
User Rank
CEO
Re: Moore's law.....
alex_m1   2/13/2014 8:48:54 PM
NO RATINGS
Most likely 3d chips will buy us 1-2 or maybe 3 generations of moore's law(both in density and partially in price). After that prices won't go down(according to zvi orbach from monolitic 3d), and it would be hard to put more layers due to thermal limits.

David Ashton
User Rank
Blogger
Re: Moore's law.....
David Ashton   2/13/2014 8:59:29 PM
NO RATINGS
@Alex-m1.....by the time we run out of that, someone will have thought of something else....  :-)

Garcia-Lasheras
User Rank
Blogger
Re: Moore's law.....
Garcia-Lasheras   2/14/2014 9:09:28 AM
NO RATINGS
@David: "True, the transistors you can fit in one layer are getting limited, but we're still putting more transistors on a chip by putting in more layers"

Yes, this is true. But this increases the issues related to clock distribution problems too!!

In state-of-the-art digital ICs, as much as the 50% of power consumption is drawn by the clock distribution circuitry in order to assure that the signal will be "skew-free" across the whole layout. This is because we need to include lots of "H" forks and signal buffers for getting a clear signal.

When we jump to a extra 3rd dimension, this problem gets worse, as illustrated in the next image -- shared from the Ecole Polythecnique Federale de Lausanne (link):



 

 

 

Max The Magnificent
User Rank
Blogger
Re: Moore's law.....
Max The Magnificent   2/14/2014 9:50:48 AM
NO RATINGS
@David @Garcia: Did you see my column on Monolithic 3D IC technologies?

Garcia-Lasheras
User Rank
Blogger
Re: Moore's law.....
Garcia-Lasheras   2/14/2014 10:08:25 AM
NO RATINGS
@Max: "Did you see my column on Monolithic 3D IC technologies?"

Arghhhh... I missed that one!!! Graphene, 3D ICs... Very, very interesting -- I'm reading it just now!!

David Ashton
User Rank
Blogger
Re: Moore's law.....
David Ashton   2/14/2014 6:25:19 PM
NO RATINGS
Thanks @Max & @Garcia for those explanations.  Max, I did not see your 3d ic blog either - a serious problem on EET these days is that if stuff makes it to the home page it is gone within days or less.    Fascinating stuff though.  How do they make the Vias in the layers?

Max The Magnificent
User Rank
Blogger
Re: Moore's law.....
Max The Magnificent   2/14/2014 6:27:36 PM
NO RATINGS
@David: How do they make the Vias in the layers?

Very carefully LOL

David Ashton
User Rank
Blogger
Re: Moore's law.....
David Ashton   2/14/2014 6:36:43 PM
NO RATINGS
@Max thanks for that very lucid explanation.   Maybe you have another forgotten article that goes into this?  

Max The Magnificent
User Rank
Blogger
Re: Moore's law.....
Max The Magnificent   2/14/2014 6:42:29 PM
NO RATINGS
@David: Maybe you have another forgotten article that goes into this?

If so, I've forgotten LOL

But... did you see this one by Zvi and this one and this one by yours truly?

Garcia-Lasheras
User Rank
Blogger
Re: Moore's law.....
Garcia-Lasheras   2/14/2014 7:04:08 PM
NO RATINGS
Max, what a cool collection of 3D ICs related blogs. They suppose a pretty nice reading before going to sleep for me now ;-)

By the way, back in 2007, I was trying to launch a project with a big company in order to build a new asynchronous 3D FPGA architecture. The project was really interesting and very smart planned. Unfortunately, those were bad times for the company, and there were rumours about IBM going to buy it -- finally, Oracle was the one which absorbed the company I was dealing with: SUN Microsystems :-( 

Max The Magnificent
User Rank
Blogger
Re: Moore's law.....
Max The Magnificent   2/14/2014 7:06:25 PM
NO RATINGS
@Garcia: ...finally, Oracle was the one which absorbed the company I was dealing with: SUN Microsystems

I remember a time when I thought Sun Microsystems would be around for ever ... it's like the old saying goes: "The bigger they are, the harder they fall" -- it was a sad day when they disappeared :-(

Garcia-Lasheras
User Rank
Blogger
Re: Moore's law.....
Garcia-Lasheras   2/14/2014 7:12:32 PM
NO RATINGS
@Max: "I remember a time when I thought Sun Microsystems would be around for ever ... it's like the old saying goes: "The bigger they are, the harder they fall" -- it was a sad day when they disappeared :-("

Yes, a very sad day. I really loved SUN Microsystem: its technology, both hardware and software, its openness, its advanced research... Some people think that they were too smart, commited and professional for staying alive in these weird days we are living now.

Radio
NEXT UPCOMING BROADCAST
EE Times Senior Technical Editor Martin Rowe will interview EMC engineer Kenneth Wyatt.
Top Comments of the Week
Like Us on Facebook

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
EE Times on Twitter
EE Times Twitter Feed
Flash Poll