🤖 AI Summary
This work addresses the limited off-chip memory bandwidth and inefficient data transfer capabilities commonly found in open-source microcontrollers, which hinder their support for domain-specific accelerators in machine learning and signal processing. Building upon the open-source RISC-V Croc SoC, this study presents the first integration of a silicon-proven HyperBus controller and a DMA engine, establishing an MCU platform that enables plug-in domain-specific accelerators with low-pin-count, high-bandwidth access to external DRAM/Flash and CPU-intervention-free high-speed data transfers. Leveraging an open-source 130nm PDK and an end-to-end physical implementation flow, the chip can be fully implemented within one hour on a consumer-grade workstation. The first silicon (MLEM) operates reliably at 72 MHz under 1.2 V, validating both the functional completeness of the platform and the robustness of the implementation methodology.
📝 Abstract
Domain-Specific architectures with accelerators for machine learning and signal processing require efficient bulk data movement and high-bandwidth access to large datasets. Such capabilities are often absent from minimal open-source microcontrollers (MCUs). We present HyperCroc, an extension to the end-to-end open-source RISC-V Croc system-on-chip (SoC) integrating a silicon-proven HyperBus controller for off-chip DRAM and Flash memory access and a DMA engine, providing a practical MCU-class platform with streamlined plug-in support for domain-specific acceleration. HyperBus offers a low-pin-count PSDRAM interface at up to 400 MB/s, enabling bandwidth-scaled dataset access, while the DMA engine enables autonomous, high-throughput transfers without CPU intervention. HyperCroc preserves Croc's open-source synthesis and physical implementation flow targeting IHP's open 130 nm process design kit (PDK); the full chip can be implemented in under one hour on a consumer-grade workstation. We further report first silicon measurements from MLEM, the first Croc tapeout, confirming that the silicon is fully functional at 72 MHz @ 1.2 V and validating the end-to-end flow.