The NimbleAI Horizon Europe project leverages key principles of energy-efficient visual sensing and processing in biological eyes and brains, and harnesses the latest advances in$\mathbf{33D}$stacked silicon integration, to create an integral sensing-processing neuromorphic architecture that efficiently and accurately runs computer vision algorithms in area-constrained endpoint chips. The rationale behind the NimbleAI architecture is: sense data only with high information value and discard data as soon as they are found not to be useful for the application (in a given context). The NimbleAI sensing-processing architecture is to be specialized after-deployment by tunning system-level trade-offs for each particular computer vision algorithm and deployment environment. The objectives of NimbleAI are: (1)$\mathbf{100x}$performance per mW gains compared to state-of-the-practice solutions (i.e., CPU/GPUs processing frame-based video); (2)$\mathbf{50x}$processing latency reduction compared to CPU/GPUs; (3) energy consumption in the order of tens of mWs; and (4) silicon area of approx. 50 mm2.

This article presents an instruction-set extension to the open-source RISC-V ISA (RV32IM) dedicated to ultra-low power (ULP) software-defined wireless IoT transceivers. The custom instructions are tailored to the needs of 8/16/32-bit integer complex arithmetic typically required by quadrature modulations. The proposed extension occupies only two major opcodes and most instructions are designed to come at a near-zero energy cost. Both an instruction accurate (IA) and a cycle accurate (CA) model of the new architecture are used to evaluate six IoT baseband processing test benches including FSK demodulation and LoRa preamble detection. Simulation results show cycle count improvements from 19 to 68 percent. Post synthesis simulations for a target 22nm FD-SOI technology show less than 1 percent power and 28 percent area overheads, respectively, relative to a baseline RV32IM design. Power simulations show a peak power consumption of 380 µW for Bluetooth LE demodulation and 225 µW for LoRa preamble detection (BW = 500 kHz, SF = 11).

The fast and accurate processor simulator is an essential tool for effective design of modern high-performance application-specific instruction set processors. The nowadays trend of ASIP design is focused on automatic simulator generation based on a processor description in an architecture description language. The simulator is used for testing and validation of designed processor or target application. Furthermore, the simulator can produce the profiling information. This information can aid design space exploration and the processor and target application optimization. In this paper, we present the concept of automatically generated just-in-time translated simulator with the profiling capabilities. This simulator is very fast, and it is generated in a short time. It can be even used for simulation of special applications, such as applications with self-modifying code or applications for systems with external memories. The experimental results can be found at the end of the paper.

Application-specific instruction set processors used in embedded systems are highly optimized for a given task. On this type of processors runs a specific application. Therefore, the designer should have a tool which helps him or her in the task of processor and application optimization. One of such tools is profiler. It can discover problematic parts, such as bottleneck points, in the processor and application design. Then, the designer can easily find which parts of the processor or application should be modified, so that performance gets better or power-consumption is reduced. In this paper, a way how to generate cycle-accurate profiler for C language from a processor model described with an architecture description language is proposed.