ESL – Present day scenario and possibilities in India

Picture of CMR Guest Contributor

CMR Guest Contributor

The field of using automation and tools for design of hardware (and associated software) is broadly categorized into EDA (Electronic Design Automation) and ESL (Electronic System Level) domains.

EDA mainly deals with development and verification of hardware relating to its behavioral and physical aspects starting with design as its input and making the hardware a reality as a chip or SoC as it ultimate output. The EDA flow has well defined stages, which the design goes through and there is a wide choice of tools, which helps in implementing each of these stages.

ESL is a domain that is intended to help architects and designers to get a feel of the hardware design before development starts. This is done by visualizing the different components that make up the design ‘toplevel’ and getting functional blocks of these models so that the architectural complexity could be understood as early as possible in the project lifecycle. The ESL domain is not as well structured as the EDA and often companies resort to proprietary tools and methodologies to achieve their ESL goals. In recent years an attempt to standardize has been made with SystemC as a language for modeling components that go into an ESL flow and TLM as a methodology for communication between these components.

While EDA field has been adopted, innovated and developed exhaustively in India, the same cannot be said to be true for ESL, which has been now present for more than a decade at least. Most of the ESL usage, development and contributions come from a handful of companies, the majority of whom are the Indian development centers for MNCs who also invariably are involved in the EDA flow. Even in these cases the maximum usage seems to be limited to select internal teams with small, dedicated teams, which cater to the needs of a business unit or sometimes the entire company itself.

In India ESL methodology and tools are mainly used for the following areas:

  • Virtual Platform Development of a SoC for early software development and hardware architecture exploration
  • Synthesizable models for High Level Synthesis (HLS) which helps functional models to be integrated into the EDA flow
  • Power estimation and speed analysis of a Hardware system

Having realized and understood the key areas where ESL is beneficial, the India ESL market has evolved in and around those areas. ESL is employed by major design and product companies using tools provided by ESL tool vendors to achieve their system level goals. Many of them are aided in these tasks by a select group of companies who provide third party level verification and consulting services either on offshore or onsite based models.  Product companies which use ESL and have a presence in India are Intel, AMD, Freescale, NXP, Infineon, Texas Instruments amongst others. ESL tool vendors prominently are Synopsys, Cadence, Mentor Graphics, Bluspec, etc. Many product companies also have their in-house set of tools, which they use for simulation and software development. Companies who have been providing third party solutions in the development and verification aspects of ESL in the India market are Kasura, Tata Elxsi, Sasken, Wipro and others. Public sector organizations involved in defense and electronic design use ESL on a very limited basis.

Key innovations and contributions to ESL domain from the Indian market have been limited but this can be traced more to the slow adoption of ESL by the semiconductor industry as a whole than any other reason. Forums like the Indian chapter of OSCI, Indian SystemC Users Group and events like Synopsys Users Group have helped to promote awareness and benefits of ESL but the actual impact is yet to be seen as major ESL developments.

The IEEE standardization of SystemC , the standardization of TLM2.0 by a working group comprising of eminent professionals across companies , the proposal of merging the OSCI group with Accellera (which is responsible for providing and maintaining the SystemVerilog standard) are all the steps in the right direction for ESL. A co-methodology involving the already popular SystemVerilog (and associated verification strategies like OVM, UVM) will provide a significant boost to ESL industry as a whole. AMS extensions and synthesizable subsets of SystemC language are still in nascent stages of adoption and looks to be a good bet for the future of ESL. These are niche fields which are still waiting to be explored and the India ESL market could do well to focus on them to create a distinct and unique market for their clients. Companies already providing services in the key areas of ESL can also look at these emerging areas within ESL to expand their field of expertise and business. Contribution through ideas and patents will ensure that they get noticed beyond existing horizons.

The ESL market in India is still in a growing phase and has miles to go before it can be considered as a major ESL hub. Fortunately, there is no strong contender for that title as yet and is open to both the developed and emerging markets. With the appropriate investment in the right areas and complimentary R&D efforts, Indian firms could well become key players in ESL, which is a domain predicted by experts to become a necessity in the near future.

(The above are the views of the author in his personal capacity as an industry expert in this area for more than a decade.)