High Priests & Gurus
“Our tool input is in an object-code format - nothing in our tool is processor core specific. It gives an extra dimension that the ARM products don't have. Why doesn't ARM buy us you ask? Well, [with a laugh], we're not thinking of being acquired by anyone. Our technology is, of course, highly complementary to what ARM does and I believe it actually makes their offerings more appealing. We don't need any specific cooperation from ARM, but we do have a good relationship with them and we certainly hope to continue that.”
“Over the years, many people have explored hardware/software co-design, usually by focusing on finding the optimal partition. This breaks the problem into two distinct pieces and just stores up trouble for later in the development flow. Instead, we're recognizing that some pieces of that software need to be migrated into a more hardware-optimized solution within a very short development time. We're offering a way for the users to control the migration process using a different approach from what has been done in the past.”
“What we're doing with our tool is the kind of thing you do inside of a compiler, only the knowledge we extract is used to create an optimized hardware architecture. We're coming at things from an embedded software viewpoint, not one centered about hardware design. We firmly believe that as tool providers, we need to provide a solution that allows customers to migrate their software into appropriate hardware, whether for performance, or to reduce silicon costs/power, or to reduce time to market.”
“We've started raising our profile in the EDA community and now need to do that in the embedded community as well. We will be exhibiting at DATE and DAC, and will be at the Embedded System Conferences. We've started discussions with a number of the major embedded tool providers. The embedded space is a medium term play compared to where we are in the mainstream SoC domain. Our short-term focus is on delivering value within customer embedded-core SoC projects, such as the work we have done with ST Microelectronics. Meanwhile, the next two market opportunities we're seeing are in the FPGA area building accelerators for soft cores, and opportunities to build accelerators for stand-alone processors in the embedded space.”
“Our technology is not language specific. SystemC is fine, but so is C, or C++, or Java, or anything else. Since we work off of the object code, if you can compile it, we can use it. We don't need special languages, special coding styles, or special hardware. That makes it much easier for users to leverage their considerable investments in embedded software and hardware implementation tools.”
“Richard Taylor, our CTO, is a compiler guy and in that sense, it's a little unusual to find him in an EDA start-up. He's been an end user and a manager of both hardware and software people. He took his compiler knowledge, his deep domain knowledge, and his hardware/software experience, and said there must be a better way to do this.”
“The three co-founders of CriticalBlue came together and started looking at this whole area. We saw that there was a need and an opportunity in the end user community as more product functionality moves into the software domain, and I believe fundamentally that these types of innovations only occur in small companies. We may of course look youthful, but we're long enough in the tooth in that we understand how to make it easier for our customers to adopt our technology.”
2 - Sound bytes on the Synopsys acquisition of ADA
I had a chance to speak with Sanjiv Kaul and Matthew Raggett on the 28th about some of the implications of the Synopsys/ADA announcement. Kaul is Senior Vice President and General Manager for New Ventures at Synopsys. Raggett is currently CEO of Analog Design Automation (ADA). They were both in Silicon Valley at the time of our phone call.
Q: When was ADA founded?
Matthew: The company was founded in 1999.
Q: Who are your principle customers?
Matthew: We're a private company, so we're not giving out our customer list. Although Potentia, who was quoted in the announcement, is representative of the class of customer that ADA has [engaged with]. We have customers who are semiconductor companies and analog companies, among others.
Q: When was the decision made to pursue this acquisition?
Sanjiv: We're not giving out those details. However, Synopsys was an investor in ADA, so we have had people involved in ADA from the early inception of the company. As we were looking at what to do in analog, we targeted circuit optimization as an important technology. We concluded that ADA had what we needed. We have invested in other analog companies, as well.
Matthew: It all started with the original investment. It's like any girl friend, the more time you spend together and the more you get to know each other, the more you like each other.
Q: Why now for the ADA acquisition?
Sanjiv: With the Avanti merger, we got HSPICE, and now we have a very strong transistor-level simulator. We're seeing, in talking to our customers, that circuit optimization is a key need, and we're seeing that the ADA technology is ready for deployment. So all of this naturally came together now.
Matthew: Actually, the technology was ready at a certain level a while ago, but now it's hitting mass production and getting multiple seats and attention. The thing you look at [in this situation], is the customer base as a whole. ADA now has examples of customers getting first tape-out and working silicon. Having gone through that process with our customers and seeing the impact we can make, the analog community has started to spread the news about us by word of mouth. People could be very skeptical because they have felt that analog tools haven't really changed in 20 years. But when they discovered that we can provide a big improvement, and that in the last 6 months the semiconductor companies have put in place a budget for analog tools, [we have started to get more notice].
Q: Are there other initiatives underway at Synopsys that dovetail with this purchase?
Sanjiv: I'm sure you're aware that Synopsys is the leader in analog simulation, and also in extraction technology. We also have a complete analog solution in physical implementation. Over the next 12 to 24 months, you'll see a series of announcements from Synopsys about delivering new technology in this space.
Q: Is there any overlap between the Synopsys technology and that of ADA?
Sanjiv: They are completely complementary.
Q: What is the advantage to ADA in being part of Synopsys?
Matthew: It's that pure ability to penetrate the marketplace. Access to the established sales channels of Synopsys is a tremendous boost to ADA. Other CEOs have told me that it can be amazing to see how much volume and product movement can be had by linking with a big company. Also, the synergy between HSPICE and NanoSpice means that a tighter integration will now be available for analog engineers.
Q: How many employees are there at ADA and how many will be moving over to Synopsys?
Matthew: We have 25 employees.
Sanjiv: Around 20 of them will be coming to Synopsys.
Q: Where are they all now?
Sanjiv: Most of them are in Ottawa. They'll be moving our to our R&D facilities in Ottawa, which we're lucky that we already have in place there.
Q: What will Matthew's new title be at Synopsys?
Matthew: I will not be joining Synopsys. I don't know yet what I will be doing when this process is complete. Right now, it's fairly time consuming.
3 - Guido Arnout at PowerEscape
Be the first to review this article