Synopsys Synphony Synopsis

sheet_music.jpgI was contacted a few weeks ago by Synopsys’ PR agency to see if I’d be interested in covering an upcoming product announcement. I usually ignore these “opportunities” since the information provided is usually carefully wordsmithed marketing gobbledygook and not enough for me to really form an opinion. However, it turned out that this announcement was on a subject I know a little bit about, so I took them up on their offer.

The announcement was “embargoed“, that is, I was not to make it public until today. Embargoes are a vestige of the days when traditional journalism ruled the roost and when PR departments thought they could control the timing of their message. I don’t think embargoes benefit companies anymore since news is reported at light speed (literally) and people will write what they want when they want. Still, I consider it a sort of gentleman’s agreement so I’m not writing about it until today.

I also waited a little bit until the “mainstream press” wrote their articles. That let’s me point you to the best of them and conserve the space here for my own views, rather that regurgitating the press release and nuts and bolts.

(Update: Here is a very good description of the Synphony flow from Ron Wilson).

Today, Synopsys announced a new product called Synphony High Level Synthesis. You can read about this here. Basically, Synopsys is introducing a high level synthesis (aka behavioral synthesis) product that takes as its input Matlab M-Code and produces RTL Code, a cycle accurate C-model, and a testbench for simulation. Since I have not used the tool, I cannot comment on the capabilities or the quality of results or compare it to other tools on the market. However, I have had some past experience with tools like Matlab (specifically SPW) and Synphony (specifically Behavioral Compiler). So, here are my thoughts, observations, opinions that come to mind.

  1. Synopsys, once the leader in behavioral synthesis, is now the follower - When Synopsys introduced Behavioral Compiler over a decade ago they were the first to preach the gospel of high-level synthesis and all the associated benefits. Architectural optimization. Faster simulation. Bridging the gap between system design and ASIC design. Smaller and easier to understand code. Dogs and cats living together. The promises never fully materialized and Synopsys seemingly moved out of the market. Meanwhile, Mentor introduced Catapult C, Cadence introduced C-to-Silicon, and several others including Forte, Agility, Bluespec, Synfora, ChipVision, and AutoESL introduced their own high-level synthesis tools. Now, after acquiring Synplify DSP through Synplicity, Synopsys is finally re-entering the market (at least for ASIC design) with Synphony. The hunted have become the hunters.
  2. Synphony takes M-code from Matlab as its only source - Whereas most (but not all) other high-level synthesis tools input C like languages, Synopsys has chosen to input M-code only, at least for now. According to Chris Eddington, who is Director of Product Marketing for System-Level Products at Synopsys (according to his LinkedIn profile), approximately 60% of those who say they do “high-level design” are using M-code or some form of C (ANSI C, C++, System-C) for some portion of their design activities. Of those, slightly more use the C variants than M-code, which means that somewhere close to 25% of all ASIC designers could be a possible market for this tool.
  3. Synopsys can try to leverage the Matlab installed base - As mentioned above, Synopsys estimates that 25% of high-level designers could use the Synphony tool which is a pretty big market. By targeting mainly algorithmic design, not control logic, Synopsys can try to serve the Matlab installed base with a more narrowly targeted offering which should make it easier to support. It also allows Synopsys to avoid a bloody battle over C dominance and to pursue a blue ocean strategy with Matlab’s installed base. Interestingly though, there is no partnership with MathWorks implied by this announcement.
  4. Synphony leverages existing IP libraries - Libraries already exist for many common functions that were available for the Synplify DSP tool. The library elements are available as well for Synphony, allowing the designer to specify his functionality using this library or using M-code as the source.
  5. An FPGA tool is being adapted for ASIC - This is probably one of the first times that a tool initially developed for FPGAs (Synplify DSP) is being adapted for ASICs. It’s usually the other way around (e.g. FPGA Compiler grew out of Design Compiler). It should be interesting to see if the FPGA tool can “cut-it” in the ASIC world.
  6. Ties to implementation are seemingly tenuous - A tool that can take M-code as its input and produce RTL and C and do all the other things is all fine and good. But for Synphony to become more than an experimentation tool, it has to produce results (speed, area, power) as good or better than hand-coded RTL. However, the ties to the implementation tool (Design Compiler) are not as direct as even Behavioral Compiler was a decade ago. It seems that Synphony takes an approach where it pre-compiles and estimates timing for various blocks (kind of like building DesignWare libraries), but it assembles the design outside of DesignCompiler without all the associated timing views and engines necessary for true design and timing closure. It’s hard to understand how this can reliably produce results that consistently meet timing, but perhaps there is something that I am not aware of?
  7. Focus on “algorithmic design”, not control - As mentioned above, Synopsys is going after the folks using Matlab. And those designers are developing algorithms, not state machines. In essence, Synphony can focus on the fairly straightforward problem of scheduling mathematical operations to hit throughput and latency goals and not deal with more complex control logic. Much simpler.
  8. Conversion from Floating Point to Fixed Point - Anyone who has designed a filter or any DSP function knows that the devil is in the details, specifically the details of fixed point bit width. One choice of bit width affects downstream choices. You have to decide whether to round or truncate and these decisions can introduce unexpected artifacts into your signal. Synphony converts the floating point Matlab model into a fixed point implementation. Supposedly, it then allows you to easily fiddle with the bit widths to tweak the performance. Some earlier Synopsys products did this (Cossap, System Studio) and it’s a nice feature that can save time. We’ll see how useful it really is over time.
  9. Synphony produces real RTL, as well as C-code and a testbench - One of the drawbacks of Behavioral Compiler is that it never produced a human readable form of RTL code. This made it hard to simulate and debug the RTL. Synphony supplies readable RTL (or so I am told) as well as cycle accurate C-code for system simulation and a testbench for block simulation. This should help facilitate full chip simulations for chip integration, since Synphony will probably only be used on blocks, not entire chips.
  10. Couldn’t Synopsys come up with a better reference than Toyon Research Corporation - No offense to Toyon, but they are hardly a household name. It makes me wonder how many partners Synopsys has engaged in this development and how well tested this flow is. Not saying it isn’t well tested, just that Synopsys is making me wonder. Gimme a name I’ve heard of, please.

Only time will tell if Synphony is truly music to our ears, or if it is just SYNthesis that is PHONY.

harry the ASIC guy

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)
Loading ... Loading ...

Tags: , , , , , , , ,

13 Responses to “Synopsys Synphony Synopsis”

  1. Ry Schwark Says:

    In defense of my colleagues at Synopsys, an embargo is about being fair and giving everyone equal access to the new information.

    If you got the news on tuesday, and wrote it up, and somebody else on thursday, they’d feel like a chump. If you give it to everybody on tuesday, then they can all write it or not whenever they want.

    Since typically a company would want to talk to someone about it, not just give them a release, you embargo the news so that the logistics of who the company talks to first or second doesn’t give an advantage to one or the other.

    There is, in fact, some logic to old school PR and journalism.

    Anybody in PR who thinks they can control what people write hasn’t been in PR very long!

  2. harry Says:

    Hi Ry,

    I was involved in a similar process when I was with Synopsys when we announced the Pilot Design Environment. We briefed the press and Cooley (this was before bloggers) and also embargoed the materials. The main concern of the PR people was controlling the timing of the message. They wanted to avoid it coming out when there were other hot news stories and industry events because it would have gotten buried. They wanted it on a certain day of the week. They wanted it to coincide with a speaking opportunity for one of the execs at a conference. None of it had to do with being fair to the journalists. You might be the only one.

    Harry

  3. John Blyler Says:

    Hi Harry. First, nice coverage!

    Second, I’ve got to agree with Ry. Most PR folks and especially journalists are not that sinister. From the journalist side, pre-embargoed briefings are as much a matter of work flow timing (allocation of resources) as anything else. There are so many things that deserve coverage, but not enought journalists to cover them.

    In the past, journalists used to have editorial planning meetings with all the major companies to make sure all the important technical topics were covered in a year’s time. Naturally, vendors used this to their advantage, hoping to time technology coverage with product releases. But back when journalists covered the entire industry, such vendor specific advantage was minimal since all of technology was being covered anyway.

    Just be to clear, I’m talking about cover that consists of actual content development (such as your blog) - which is what journalists used to do (and a few still do it). Not merely removing the adjectives from a press release and pitching it as content.

    BTW: Major companies seldom want their names mentioned as reference in press release. Legal issues and whatnot.

    You’re learning the ropes. Before long, you’ll be a journalist - whether you want to or not. :)

  4. JB’s Circuit » Bloggers Turning into Journalist?! Says:

    […] you don’t read Harry the ASIC Guy’s blog, you should. Hes just a regular engineer writing about technology trends/updates and […]

  5. Lou Covey Says:

    Harry, the practice of embargoing news releases came out of the days when publications were given exclusive coverage of certain news items. It used to be that daily publications had advantages in timing over weekly, weekly over monthly, etc. (see http://commbasics.typepad.com/my_weblog/2005/08/the_law_of_dimi.html)
    PR people were trained to be sensitive to the timing of news. Electronic Design actually used to have contracts that stated they would give front page coverage of announcements as long as they were the only ones who covered it for a week after the announcement.
    Social Media and the internet completely killed that reasoning, but PR people in electronics still live by that rule. They put different resaonings on it now, none of which make sense, and journalists still try to “get the scoop” even though that no longer makes any sense.
    What makes it all more ridiculous is that the media that might care about embargoes and such is pretty much gone, hence by Tweet about marketing impotence.

  6. Steve Meier Says:

    Harry: Really good coverage and analysis.

    You did not talk much about the ghost of Behavioral Compiler and the view from within the application consultants on how they view this product? are they ready and excited to promote it ? I would predict that the Systems group has a hard time getting cycles from account teams, thus you have Innovators Dilemma, potentially good technology but no way to foster and grow it.

    Also in terms of verification flow, you did not mention the need for formal equivalence between the user input model and the synthesized result. Calypto offers C-to-RTL equivalence, but no one offers M-to-RTL equivalence. Output of a gate level C model is good, but suffers the classic problem of dynamic stimulus verification being inherently incomplete. Bad logic bugs are a problem which can undermine confidence, particularly when they cannot be discovered. For any synthesis methodology to work there should be 1 golden source for verification, and the rest needs to be automatically statically verified through complete equivalence checking.

    On the embargo topic, you are in good company as TechCrunch has established a no embargo rule after getting burned with their competitors jumping the gun and breaking embargo rules. I guess you are not large enough to set the rules yet, but others are making the same point. The good EDA editors like Richard Goering would actually conduct research, talk to competitors and contacts and have a balanced treatment with multiple points of view, all before a print deadline. Nowadays that balanced point of view can come about real-time through blogs like yours.

  7. Brian Fuller Says:

    Harry, great post, and Ry, John and Lou, nice conversation.
    Embargoes are dead. They died in November 1994 when Netscape and the commercial Internet was born. They just haven’t, like a chicken with its head cut off, realized they’re dead yet.

    I have hated embargoes forever. I have hated them since I began my career at UPI where my job was to get my story on the wire faster and more accurately than the AP. To Ry’s point about fairness, at EE Times, the embargo dragged me down to the same gutter that my competitors wallowed.

    The situation needs to be this: When a company announces something, that’s the first strike, the jump ball. Its press release hits the wire and then everybody takes their shot at building a story around it.

    Nobody gathers in the mailroom anymore to pick up their weekly trade journals. They gather their information in burst mode online and so a Monday release date makes no sense.

    In addition, the whole notion of a press release is anachronistic. We live in a info-streaming world; we should evolve our communications strategies and tactics accordingly.

    Lastly, added kudos to Harry for covering an announcement in a Web-friendly/sensitive manner. We are bombarded by so much information that we generally know the story by the time we read the headline. The traditional press tells us what we already know (even online). To read a text-bloated story with no breakheds that parrots a company’s press release is a ticket to perdition.

    Harry gives us an angle and insight we probably won’t get anywhere else and does so using some nice Web-useability techniques.

  8. SKMurphy » Growing the Pie in EDA, Part 3: Add MATLAB Says:

    […] as an ESL language and Mathworks as an EDA player. Harry Gries provides a good overview in “Synopsys Synphony Synopsis” (say that three time fast) Basically, Synopsys is introducing a high level synthesis (aka […]

  9. Richard Goering Says:

    I don’t entirely agree that “embargoes are dead.” Anyone announcing a significant product will set a date for the announcement. There may be a good reason to brief editors before that date, assuming somebody is going to do some research and post more than a rehashed press release.

    I do think, however, that being “first” with a story has become less important than being insightful. I doubt whether readers of Harry’s blog really cared whether it came out the same day Synphony was announced. What matters is his unique perspective and analysis. If Harry hadn’t received a pre-briefing under embargo, the blog would be no less valuable.

  10. Ry Schwark Says:

    I’m with Richard.

    There have been publications throughout my career that didn’t want to honor embargoes, and that was fine. You just didn’t pre-brief them. No big deal, and hardly new.

    If publications feel the world has changed enough that they want to ‘jump for it’, fine. I’ve certainly worked both ways.

    But I think much of what our world cares about is not who breaks the news, but who ‘gets’ it. The stuff we talk about is technical and complicated. It isn’t something that you can pick up off a news release on the wire and deliver useful analysis to engineering readers. You want to talk to the company in detail and ask your tough questions. Telling people that somebody delivered a new product first doesn’t help if you can’t tell them what that tool does and why it’s relevant.

    That’s where the differentiation comes in. Not in who does the story first, but in who does it _right_.

  11. Mike Santarini Says:

    Harry, don’t listen to these guys. You are right, embargoes are an evil collusion between the trade journalists and the vendors who kept the journalists in Ferrari’s and furs. Come-on Fuller, Goering & Schwark– time to fess up!

    But seriously, there is one somewhat divisive part to an embargo. If you take a vendors release embargo, it means you can’t go to their competitors and get their views on the product. This in most cases doesn’t turn out to be a big deal because in the times I did actually get to talk to competitors for a new product announcement, they of course bashed their competitor’s product. In my many years of reporting, I never heard a vendor say of his competitor’s product, “Wow, that sounds way better than what we have” or “our customers love their product so much and they are cancelling their licensing with us like crazy.” Embargoes made the job manageable especially in the month or so leading up to DAC where we not only covered a ton of products but breaking news as well—and tried to do so as comprehensively as possible. Pop open the EET archives circa 2001 and see how many product writeups Rich and I did a week.

    JB, I’ve never been to an editorial planning meeting with vendors to plan the timing of their product announcements. If they ever invited me to something like that, I would have had one hell of a cover story in the following Monday’s issue or even later that day on the web.

  12. A Talk With An EDA Consultant | Chris Gammell's Analog Life Says:

    […] abducted by aliens for a fortnight. I then spent 7 years as a digital designer at TRW, 14 years at Synopsys as an AE, consultant, consulting and program manager. Synopsys and I parted ways and I have been […]

  13. Blog Review: Oct. 14 | System-Level Design Says:

    […] also a hot topic outside of Synopsys. For a reality check on high-level synthesis, check out Harry Gries’ look at what he calls Synopsys Synphony Synopsis. While there’s lots of tongue-tying […]

Leave a Reply