Pre-silicon and post-silicon testing of SIWA, a low-power RISC-V microcontroller
Resumen
Making chips is a complicated and costly process that requires specialized knowledge. A
single product of the semiconductor industry often involves many teams of engineers who
work to make the chip as flawless as possible. The VLSI design flow is usually the compass
needed to navigate these tricky seas, as it provides a detailed description of what is required
to successfully build a chip.
With each passing year, chips have evolved ever more complex, drastically increasing research
and development costs. A look into how the industry has changed is enough to understand
the previous sentence. At the moment of this dissertation, there are far more companies
in the semiconductor industry that specialize in specific aspects of the chip-making
process than the decreasing number of Integrated-Device-Manufacturers (IDM) around the
world. This trend is due to the increasing costs of chip manufacturing as designs and processes
become more complex. Companies in this industry have learned to chain and intertwine
their specialized business models to create chips at reduced costs.
However, small design teams and academic groups still struggle to design and create their
chips as the effort required and costs are still considerable for them. As part of a group of
academic research in microelectronics, this doctoral dissertation was built from the experiences
and challenges we faced while developing a low-power RISC-V microcontroller for
implantable medical devices, SIWA. In particular, this thesis focuses on the validation of our
microcontroller, and the strategies discussed in this document were developed considering a
low-budget and the scarce human resources at our disposal.
This thesis covers several aspects of the test process that SIWA went through over several
years. In this document, the functional verification strategy to validate SIWA in the presilicon
phases will be found. Then, a detailed description of the physical testing framework
and how it was complemented with FPGA emulation is shown. Later, it is explained how
software applications are built for SIWA and how I/O hardware emulation through software
was achieved. Finally, this dissertation presents a benchmark proposal for classifying lowpower
RISC-V microcontrollers used in implantable medical devices. It is worth mentioning
that part of the results of this thesis were four scientific publications indexed in the Scopus
database, demonstrating that the effort done in my PhD studies made a contribution to the
knowledge in the field of microelectronics.
Descripción
Proyecto de Graduación (Doctorado Académico en Ingeniería) Instituto Tecnológico de Costa Rica. Área académica de Doctorado en Ingeniería, Universidad de Costa Rica, Facultad de Ingeniería, 2025