A Scalable Massively Parallel Processor for Real-Time Image Processing

Abstract

This paper describes a high performance scalable massively parallel single-instruction multiple-data (SIMD) processor and power/area efficient real-time image processing. The SIMD processor combines 4-bit processing elements (PEs) with SRAM on a small area and thus enables at the same time a high performance of 191 GOPS, a high power efficiency of 310 GOPS/W, and a high area efficiency of 31.6 GOPS/mm . The applied pipeline architecture is optimized to reduce the number of controller overhead cycles so that the SIMD parallel processing unit can be utilized during up to 99% of the operating time of typical application programs. The processor can be also optimized for low cost, low power, and high performance multimedia system-on-a-chip (SoC) solutions. A combination of custom and automated implementation techniques enables scalability in the number of PEs. The processor has two operating modes, a normal frequency (NF) mode for higher power efficiency and a double frequency (DF) mode for higher performance. The combination of high area efficiency, high power efficiency, high performance, and the flexibility of the SIMD processor described in this paper expands the application of real-time image processing technology to a variety of electronic devices.Index Terms-Image processor, power efficiency, area efficiency, SIMD, scalable architecture, fine grained processing element.

Low Power Multicore Processors for Embedded Systems

“…As the latest SH processor core, the SH -X4 extended its ISA and address space effi ciently for this purpose. The SH -X4 was integrated on the RP -X heterogeneous multicore chip as two 4 -core clusters with four Flexible Engine/Generic ALU Arrays (FE -GAs) [47,48] , two MX -2 matrix processors [49] , a Video Processing Unit 5 ( VPU5 ) [50,51] , and various peripheral modules.…”

A 45-nm 37.3 GOPS/W Heterogeneous Multi-Core SOC with 16/32 Bit Instruction-Set General-Purpose Core

“…This chips integrates two MX-2 cores [10], four FE ("flexible engine") cores [11] and one VPU5. Here, we provide an overview of the processing of each type of core.…”

A Programmable Difference-of-Gaussian Analog Complementary Metal Oxide Semiconductor Image Sensor Operating in the Subthreshold Regime

“…In particular, the high power consumption of microprocessors when running image processing algorithms seriously restricts the usage of software solutions in mobile devices. In this regard, computational CMOS image sensors in which the primitive pixellevel operations are embedded [1][2][3] would provide us with a lot of opportunities to solve this difficulty in much smarter ways.…”