By Daya Nedamuni, Courtesy of Embedded Systems Programming
Feb 10 2005 (13:33 PM)
URL: http://www.embedded.com/showArticle.jhtml?articleID=60300161 Electronic system level tools, the next big change in design-automation tools, will enable concurrent hardware and software design.
Every 10 to 12 years, a new wave sweeps across the electronic product design arena. Historically this causes a revolutionary but gradual change in design methods and tools, which leads to a new design style at the next higher level of abstraction.
While that change is underway, the industry undergoes a lot of uncertainty. Tool vendors seek to protect their existing markets while strategizing and positioning themselves for the new era. Users deal with shorter design times, rising hardware costs, and software complexities while also strategizing on how to incorporate these emerging tools and methods to gain greater productivity from their design processes. From automated gate-level design, to RTL, to the new generation of electronic system level (ESL) tools, it's been a long and tumultuous journey but one marked by significant changes in the way design is done.
ESL design means the design of hardware and software are done concurrently. The notion of ESL design is not a new one. Many attempts have been made to develop tools that would handle the concurrent design of hardware and software. These tools, however, tended to be domain-specific and couldn't provide optimized, cross-domain effectiveness. Starting in 2000, a new generation of ESL tools started appearing in the market. On the hardware side, these tools were helped along by the adoption of SystemC as the language of choice to describe systems at higher levels of abstraction than was possible with hardware-description languages such as Verilog and VHDL. The increasing use and reuse of IP (licensed intellectual property) and the use of customizable devices (such as FPGAs) to drive new architectures and applications are also helping to increase the use of ESL tools, especially on the verification side of the process.
ESL efforts are being driven primarily from the hardware side. On the software side, a similar change is happening with the slow but steady adoption of the Unified Modeling Language (UML) to describe the system, though currently UML doesn't have the necessary hardware extensions to be a full ESL tool.
Areas of specialization within ESL
ESL design and simulation
Design at the conceptual level, including hardware/software co-design, design partitioning and specification; includes neither RTL nor logic-level descriptions; non-timing-based behavioral simulation, primarily transactional simulation Behavioral synthesis or ESL synthesis
Synthesis of an ESL-design description to the RTL ESL test and verification
Tools that introduce the system test requirements into the design flow, to be passed down during the design process; tools for developing the intelligent test bench |
Gartner Dataquest predicts that the move to ESL is the next big change for the design-tools-automation market. ESL deals with both hardware and software. Therefore the emerging design methods and tools will affect developers in both domains, forcing the pace of concurrent development of hardware and software to increase and demanding closer interactions between software-centric groups and hardware-centric groups at a much earlier stage in the design process.
Users of design-automation tools can be divided into three broad and sometimes overlapping categories:
- The SoC design style. Those groups whose products include a lot of customized hardware development, and where the system has significant software content use the SoC design style. The competitive advantage of this style is in the silicon.
- The embedded system design style. Those who use standard products with a lot of customized software content use the embedded systems design style, where the competitive advantage is in the software.
- The component design style. This design style is used by system designers or component designers who are putting together systems with off-the-shelf parts and whose products are typically targeted at high-volume, low-margin markets. The competitive advantage is not so much in the design; rather, the advantages are the cost of design and the ability to do special designs.
The SoC design groups are likely to be the first to see the greatest impact of the new tools and techniques and, indeed, are already starting to do so. Embedded systems designers are evaluating these tools and are slowly starting to use them. Component style system level designers usually tend to be late adopters unless the design specifically requires usage of the new tools and methodologies. As with any emerging technology, the field is wide open right now. Verification and architectural tools startups have been leading the way in the ESL market. We predict the ESL market will be one of the fastest growing segments of the design-automation market.
Table 1 forecasts the market over the five-year period 2003-2008.
Table 1: ESL market forecast 2003 to 2008 (millions of dollars U.S.)
| Year | 2003 | 2004 | 2005 | 2006 | 2007 | 2008 |
| ESL Revenues | 118.6 | 163.3 | 225 | 356.6 | 474.5 | 639.7 |
Source: Gartner Dataquest (January 2004)
Public sources for more information on ESL
ITRS Design working group
The ITRS Design working group has been engaged in creating a framework within which to describe these tools and methodologies. More information about their activities can be found at www.itrs.net/Common/2004Update/2004Update.htm Industry consorTia
Several industry consortia are working to create appropriate standards so that vendors can develop standards-compliant/conformant products that can fit into the overall ESL design and verification flow.
www.systemc.org
www.spiritconsortium.org |
Daya Nadamuni is an analyst with Gartner Dataquest. She can be contacted via email at daya.nadamuni@gartner.com.
Copyright 2005 © CMP Media LLC
English
