The following is the report of the first review of the Esprit/COSY project which was held in Brussels on the 3 July 1998.

The COSY project will deliver a complete system design methodology and tools supporting it for SOCs based on the use of functional and architectural IPs. A few point solutions exist today, but not a complete solution that integrates a selection of system tools into a seamless system to silicon design environment.

End-Users of this methodology are small design houses who have specialized on the design of very specific IP, semiconductor vendors as well as system suppliers that are facing the challenge to implement complete system solutions but who have to rely on other suppliers to deliver sub-systems in order to meet the cost and time-to-market targets.

In order to meet these targets it is imperative to utilize a well-defined design process starting at the system level and allow the refinement of the system implementation in a very structured manner.

Consumer electronics, computers, telecommunications, and automotive electronics are the potential markets for this methodology.

The COSY project aims at integrating system level tools that are either commercially available or owned by either of the involved parties. There is currently no complete system methodology commercially available, therefore, the COSY deliverables will for the first time ever provide a tools set closing the gap between system design and silicon implementation. It fully supports IP reuse and allows to perform architectural trade-offs leading to a significant reduction in time-to-market.

It will be important to clearly define the COSY objective with respect to the project deliverables. From the first review it was not clear whether a documented methodology or a complete system design framework including industry standard tools will be delivered at the end of the project. The participation of CADENCE should make it rather easy to provide a complete tools set supporting the defined system methodology.

There is various EDA competitors with plans to develop a system design environment similar to the one defined by the COSY project. None of them, however, publicly announced plans on when such a capability would be commercially available. The consortium consisting of one of the world’s leading EDA companies as well as two very strong European semiconductor and system companies, PHILIPS and SIEMENS, has definitely the potential to deliver the COSY methodology ahead of anyone else in industry.

The consortium has to clarify what will be the delivered product at the end of the project and the added-value of COSY as compared to the Felix partnership project undertaken by Cadence with partners like BMW, Ericsson, Nokia, ST and others.

CADENCE is one of the largest suppliers of EDA tools and services world-wide. They recognize the trend to develop SoC solutions which requires the development of a new system methodology supporting architectural exploration and IP reuse. Cadence is already developing a tool/methodology named Felix that will be commercially available by end of this year. Cadence is in a very strong position to penetrate the world-wide market with this new methodology.

PHILIPS is a worldwide leading supplier of ICs for consumer electronic equipment’s. The company intends to exploit the results of the COSY project to develop generic CPU/RISC architectures for future semiconductor devices as SOCs ; to design, manufacture and sell system solutions for applications in set-top boxes and for Digital Video Broadcasting.

SIEMENS AG also intends to exploit the COSY results to develop generic SoC solutions. In the first review it was unclear on how the cooperation between the central R&D organization and the business units (including the semiconductor group) would ensure effective proliferation of the COSY results into the SIEMENS business groups. It was made clear though that the SIEMENS Automation Department has high interest in this methodology for control oriented industrial applications and is one of the partners in this project.

Firstly a clear definition of the project deliverables will be essential to ensure market success. This should be done in the next project phase and the involvement of Cadence should allow to effectively communicate these to the open market. The development of similar capabilities through the Felix Partnership projects and delivery of a complete methodology ahead of the COSY project would limit the marketability of the COSY results. Again, clarification of the added-value of COSY and communication of those to the market will ensure the market success of the COSY results.

The main risk of the COSY project is the non-acceptance by the market through the integration of proprietary tools as used by Phillips and SIEMENS. A newly defined and developed system design methodology is only of interest to the whole market if the tools supporting the methodology are commercially available. Adequate definition of the project deliverables will be essential to understand the impact of this risk regarding the acceptance of the COSY results in the market.

The objectives are clearly stated in the Project Program. The result will be a system design flow and methodology incorporating the tools supporting the defined methodology and models for system design, the re-use of hardware and software IPs.

Complex embedded systems require advanced tools and a system methodology that allows to perform architectural exploration and optimization. The reuse of software and hardware IP is crucial to deal with the complexity of SoC solutions. Therefore, the project is fully in line with real needs of the industry.

The partners appear complementary in the project and no conflicts relative to the objectives could be identified. Each of the partners brings specific expertise (Siemens, Philips - systems know how, Cadence - tools and methodology expertise as well as their productisation, Research Institutes - tools and methodology knowledge) to the project.

As the project aims at developing a product that will be commercially available on the market, all three industry partners will benefit either directly (through tool sales) or indirectly (through enormous cycle time reduction or better product optimizations) from the project.

The progress reports are clear and give a good overview and the progress of each task. The technology assessment for all partners is very clear and concise.

Project tasks were well defined and appropriate project plans were developed to ensure progress towards the defined project goals/objectives. The project is strictly following the defined project plan.

As already outlined, each partner brings state-of-the-art technology and methodology know-how to the project. It will be important that the project partners use their respective technology know-how to influence the development of the final product.

The method followed by the consortium sounds technical and professional.

It will be mandatory that the project partners prepare a project requirements document that clearly outlines the requirements of the COSY methodology. Project deliverables can then easily be assessed as to whether they meet the defined requirements. Before moving into the next phase of the project the requirements document needs to be available and reviewed by all parties.

Secondly, the demonstration of tools efficiency is also important for the success of the project. Therefore it is important to define accurate measurement criteria that allow to monitor the progress being made with the development of the COSY methodology. Progress has to be monitored during each of the project phases. The demonstrators to be built as part of the project play a crucial role in measuring whether the initially defined project requirements will be met.

The report explains that up to 80% of the objectives of the first period have been reached. Some tasks have been delayed and 2 deliverables have been postponed  (D1.2.2  and D1.10.1). These are, however, not on the critical path for the project and the delay can easily be recovered.

Small modifications were made to the plan. Several tasks were started earlier as planned in order to ensure that results that were on the critical path for other projet tasks were obtained earlier.

Project Management took more time than anticipated, but the overall management of the project seems to be excellent, thus that this time seems to be very well spent and will finally help to execute the project in a timely fashion .

Assessments of the Felix technology by Philips and Siemens are described in this document. Philips assesses the use of Felix with respect to the architectural exploration of a DVB/VLD benchmark. Siemens describes the use of Felix to perform the architectural simulation of the Data Transfer Manager example (DTM).

A Felix Question and Answer list prepared by Cadence is attached to this report. The Questions and Answers deal with Modeling, Performance Analysis, Delay Script Language, Mapping, and Simulation.

The presentation during the review meeting followed the hereafter agenda :

The presentations were of high quality and very concise. All the presenters were very well prepared.

No specific comments.

No specific comments.

Synergy’s with other projects are currently not defined. It will be important to define the synergy between the COSY project and the Felix partnership projects to understand the added-value of the COSY project and it’s potential of exploiting new methodology features in future commercial products.

Information dissemination and internal technology transfer has been good so far.

The review meeting has been very helpful to understand the project objective, current progress to workplan as well as the level of cooperation and involvement of the COSY partners.

1- Requirements for COSY

The preparation of a requirements document is mandatory for the project to ensure that project deliverables can be clearly defined and success of the project can be assessed at the end of the project. It is not acceptable to finish the project with a newly defined methodology but not real physical deliverables that can be used by other companies.

The demonstrators play a crucial role in this as they document the effectiveness of the COSY methodology.

Benchmarking criteria and measurements must be defined to monitor the progress made through exploitation of the COSY methodology. This progress must be tracked during the duration of the whole project. A clear list of metrics to monitor the progress must be defined for at least the following 3 dimensions: performance, design effort, silicon cost.

The partners must continue to complete the benchmarks and clearly prove that the use of the COSY methodology provides clear advantages over conventional design methodologies.

3- Siemens demonstrator

The demonstrator presented by Siemens seems to be very simple and it is not clear whether this demonstrator will add significant value to the project. It is also not very clear to what extent the COSY methodology will be applied to develop the demonstrator.

The demonstrator example has to be modified such that it will clearly add value to the project and can be used to prove that the exploitation of the COSY methodology generates a faster and more optimized solution.

4- Felix tool

Cadence has to clarify the specific added value of the COSY project for the Felix product. How will Felix improve through participation in the COSY project and when will those features commercially be available to the market.

5- Synergy with other projects

Other European projects are currently dealing with hardware/software co-design methodologies. Therefore it is important that the COSY project defines its relationship to other projects regarding its synergy’s, complementary efforts and solutions, etc.

The requirements for the COSY methodology and the definition of real project deliverables (methodology features, tools, models, etc) have to be clarified. This is very important for the acceptance of this methodology in the market. The definition of Benchmarking criteria and measurements is mandatory to prove the effectiveness of the new COSY methodology over currently available conventional design methodologies.

No real major problems to report.