RX v2: Renesas's New-Generation MCU Processor

Sugako OTANI  Hiroyuki KONDO  

IEICE TRANSACTIONS on Electronics   Vol.E98-C   No.7   pp.544-549
Publication Date: 2015/07/01
Online ISSN: 1745-1353
DOI: 10.1587/transele.E98.C.544
Type of Manuscript: Special Section PAPER (Special Section on Low-Power and High-Speed Chips)
instruction fetch unit,  instruction set design,  microcontroller,  low-power,  dual-issue core,  

Full Text: PDF>>
Buy this Article

RXv2 is the new generation of Renesas's processor architecture for microcontrollers with high-capacity flash memory. An enhanced instruction set and pipeline structure with an advanced fetch unit (AFU) provide an effective balance between power consumption performance and high processing performance. Enhanced instructions such as DSP function and floating point operation and a five-stage dual-issue pipeline synergistically boost the performance of digital signal applications. The RXv2 processor delivers 1.9 - 3.7 the cycle performance of the RXv1 in these applications. The decrease of the number of Flash memory accesses by AFU is a dominant determiner of reducing power consumption. AFU of RXv2 benefits from adopting branch target cache, which has a comparatively smaller area than that of typical cache systems. High code density delivers low power consumption by reducing instruction memory bandwidth. The implementation of RXv2 delivers up to 46% reduction in static code size, up to 30% reduction in dynamic code size relative to RISC architectures. RXv2 reaches 4.0 Coremark per MHz and operates up to 240MHz. The RXv2 processor delivers approximately more than 2.2 - 5.7x the power efficiency of the RXv1. The RXv2 microprocessor achieves the best possible computing performance in various applications such as building automation, medical, motor control, e-metering, and home appliances which lead to the higher memory capacity, frequency and processing performance.