Circuit Level Soft
Error Analysis
Assess Soft Error Risk During
Design
Memories are not the only sensitive parts contributing to SER, but logics are playing a significant role at 65nm and beyond. The SER is becoming a part of system reliability with a need to define clear metrics and to identify the causality between a SER in a component and its effect on the whole system.
A systematic approach for identifying these connections requires the usage of dedicated tools to link raw cell SER to circuit SER, to System SER and eventually to user SER.
A systematic approach for identifying these connections requires the usage of dedicated tools to link raw cell SER to circuit SER, to System SER and eventually to user SER.The SoCFIT tool platform is a prediction and analysis tool linking raw cell SER to circuit SER and System SER.
For ASIC designers who need to accurately assess the Soft Error Rate performance of their pre-silicon design, SoCFIT is a comprehensive simulation tool platform that accounts for the different types of de-ratings in an explicit way and with good accuracy.
This offer is available both as a stand alone EDA tool and as a service run by iRoC experts.
SoCFIT ARCHITECTURE
SoCFIT presents a modular architecture according to the need of the user. The standard module is the SEU module (sequential logic module)paired with the reporting module. the Memory Bloc Mmodule and the Combinational Logic Module can be added later on.
KEY ANALYSIS TO DEFINE THE BEST PROTECTION OF ASIC
SoCFIT is an EDA tool that computes the FIT of a circuit by taking into account the intrinsic sensitivity (cross-section) of each feature of the circuit (cell instances, memory blocks, IPs, black boxes etc..) multiplied by de-rating factors. The intrinsic sensitivity of each feature is retrieved from a database (called the Soft Error Rate Database).
Depending on the soft error sensitivity shown by the target block, user can elect to modify the design in order to reduce the effect of soft error if needed for the target application. The tool can be used to run fast trade-offs to determine the best mitigation technique.