In most data-centric embedded systems, optimum implementation of a few key applications is required to realize a high costperformance system. However, each application requires operations of a different type and a different performance range. Therefore, a configurable media processor is a suitable solution for such embedded systems. This 4GOPS 3way-VLIW imagerecognition processor for an automobile system is based on a configurable media processor.The configurable media-processor is characterized by template architecture, common components and custom components. The common components are used for every application and their configuration can be optimized for a target application. The template architecture defines internal bus structure and extension interfaces for custom components. The custom components are designed for each target application. All components are assembled together along the template architecture at design-time and implemented by standard-cell design. Figure 9.5.1 shows the template architecture of the configurable media-processor. It consists of a processor core, on-chip memory, user-defined instruction units, a coprocessor, hardware (HW) engines and a DMA controller (DMAC). The processor core and on-chip memory are categorized as common components. The others are the custom components. The processor core has RISC-like architecture with a 5-stage pipeline. It has optional instructions. Extension interfaces necessary for custom components can be implemented.One of the most important features of the media-processor is a VLIW extension. The processor core has VLIW processing mode to support a VLIW coprocessor tightly coupled at pipeline level. In the VLIW processing mode, a fixed-length VLIW instruction is fetched every cycle. The upper portion of an instruction is executed by the processor core and the lower portion is executed simultaneously by the coprocessor. A mode-change between the VLIW processing mode and the normal processing mode is by instructions for subroutine call and return. While the processor core is responsible for instruction execution flow and memory access, the coprocessor executes application-specific operations. By substituting a co-processor customized for each application, application-specific VLIW processors are obtained.The image processor is designed for an automatic automobile rear and side surveillance system. The system detects passing vehicles by processing image data from a rear-view camera to support safe driving [1]. Figure 9.5.2 shows a block diagram of the image processor. The chip integrates parallel I/O (PIO), ROM controller (ROMC), capture unit, andSDRAM controller (SDRAMC). The image processor is controlled by a micro controller through the PIO.The ROM controller supports 4MB E 2 PROM. Image data are taken from the capture unit. The SDRAM controller provides a glueless connection to a 125MHz 8MB(32bx512kwordx4Bank) SDRAM. The I/O pins of the SDRAM interface are located on both edges of the chip so an SDRAM can be mounted on the reverse face of a print circui...
Design of embedded memories for a 64 bit superscaler RISC microprocessor is described. Since the microprocessor issues four instructions per cycle including two memory operations at a time, very wide bandwidth of the primary caches: 2.4 GB/sec is vital. The chip includes 16 KB instruction cache, 2 KB branch cache, 16 KB dual ported data cache and 384 entry dual ported TLB. Unique scheme of TLB hit check greatly reduces critical path. The chip is fabricated in Toshiba's high-speed 0.8 pm CMOS technology utilizing triple metal and triple well. The die size is 17.3 mm x 17.3 mm and contains 2.6 million transistors. The chip achieves 75 MHz at 70 'C 4 16KB
Two peripheral processor LSIs, the FM (Fast Timed Input port) and the FTO (Fast Timed Output port), have been developed for real-time pulse handling. By using the time-wheel scheme, these processors provide a high-level command interface with the host CPU. thus alleviating the CPU load. New features, such as time difference measurement between channels and user reprogrammability during operation have been realized by this novel approach. The prototypes of both F M and FTO were designed and fabricated using a 1 . 5~ CMOS Sea-of-Gates technology, and proved the effectiveness of the time-wheel scheme.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.