This Task will establish the principle requirements and the specification of the automotive debug support IP block. Initially an expert review of the technology requirements will take place to further identify the specific needs the automotive sector has beyond those of mainstream markets. The review will look closely at the key area of engine and powertrain control, which will lead to Carbon reduction and will note the main shortcomings of existing solutions. Important factors will include: the extreme temperature range automotive engine control units must work in and the large cost pressures within the market due to its high volume and significant engineering effort. It is clear that the challenging move towards MPSoCs is becoming very pressing as the complexity of automotive systems is rising faster than the performance of single processor core SoCs. The automotive industry expects to shortly face a step-change where in order to make cars more fuel efficient, they will need to replace their current approximation tables that determine the fuel injection cycle with computational models that run within the engine control unit. This will require at least one additional processor core to execute the model but will offer dramatic fuel and CO2 savings in return. Furthermore, the development of more effective hybrid combustion-electric powertrains is also linked to a substantial rise in complexity that will shortly necessitate an MPSoC.
The requirements specification will be based on the input from the project team and its collaboration with leading automotive systems manufacturers and industry veterans in order to ensure relevance for real automotive application scenarios and to closely link in the end users of the technology at an early stage. The end-users we have already engaged see a need for two core capabilities that differ fundamentally from current technologies. Firstly, they need a holistic system-centric debugging capability in order to deliver MP systems; secondly, they see a need for the MP debug to function in their shipped product and aid the system in self-diagnosing failures at run-time – these requirements are key in order to rapidly create low-carbon products using an MPSoC. UltraSoC knows how to realise such capabilities but lacks the financial resources for doing so.
Risks include the potential oversight of important requirements that may affect subsequent stages of the development process such as the architecture design and implementation phases.