Renesas Develops Massively Parallel Processor Based on Matrix Architecture

Feb 09, 2006
Renesas Develops Massively Parallel Processor Based on Matrix Architecture
Chip micrograph of Matrix Architecture Processor

Renesas Technology Corp. today announced the development of a massively parallel processor based on a matrix architecture suitable for image and audio multimedia data processing.

This innovatively configured processor is a massively parallel programmable device featuring tight coupling of 2,048 processing elements and 1Mbit SRAM, and has been confirmed to achieve 40 GOPS (giga operations per second) performance at a 200 MHz clock frequency.

Renesas Technology researchers unveiled details at the 2006 IEEE International Solid-State Circuits Conference (ISSCC) being held in San Francisco from February 5.

The image and audio multimedia data processing capability is essential for digital home appliances and other electronics, and involves a combination of complex operations such as fast Fourier transform, convolution, and sum of absolute difference operations. Up to now, processing of these operations has generally used hard-wired logic circuits or a DSP (digital signal processor) specialized for digital signal processing. However, recent dramatic advances in multimedia applications such as the rapid increase in pixel counts in image applications have increased demands for major improvements in multimedia data processing performance. At the same time, there is a growing demand for such processing to be implemented by means of programmable devices in order to simplify support for various multimedia data standards.

One way of improving processing performance is to increase the operating frequency through the use of finer semiconductor processes. However, it will be difficult to continue to gain major improvements in performance while maintaining lower power consumption, and to achieve the required levels of performance with conventional DSP and similar architectures. Meanwhile, a coarse-grained MIMD (multiple instruction multiple data) processor has been announced as an architecture that increases processing performance, but this also has issues with reducing power consumption.

To solve these issues, Renesas Technology has developed a matrix type processor based on a different memory technology from that of a DSP or MIMD type processor.

This new processor is a fine-grained SIMD (single instruction multiple data) type massively parallel programmable device, featuring the following structural characteristics.

1. Basic configuration : 2-bit processing elements (PE) and 512-bit SRAM assigned as data registers
2. 2,048 PEs and a total of 1 Mbit SRAM, together with tight coupling between Pes .

The key to the increased performance of this processor lies in how efficiently the individual processing elements are operated. Also, the layout and connection of the processing elements and data registers are important factors in achieving reductions in area and power consumption.

A prototype processor using the new technology was implemented in 90 nm CMOS with a core area of 3.1 mm2, and achieved processing performance of 40 GOPS at a 200 MHz clock frequency and 250 mW power dissipation. These metrics show approximately 70 and 13 times better energy efficiency in terms of unit area ratio and unit power ratio, respectively, compared to a conventional in-house DSP.

Source: Renesas Technology

Explore further: Canada's CBC websites hacked by pro-Syrian group

add to favorites email to friend print save as pdf

Related Stories

Oak Ridge to acquire next generation supercomputer

Nov 14, 2014

The U.S. Department of Energy's Oak Ridge Leadership Computing Facility (OLCF) has signed a contract with IBM to bring a next-generation supercomputer to Oak Ridge National Laboratory (ORNL). The OLCF's new ...

A GPS from the chemistry set

Oct 27, 2014

Swiss scientists teamed up with colleagues from Hungary, Japan and Scotland to develop a chemical 'processor' that reliably shows the fastest way through a City maze. As the method is basically faster than ...

TSMC, ARM see impressive results with FinFET process

Oct 05, 2014

Technology from a Taiwanese semiconductor foundry is to bring considerable benefits in performance and power efficiency to big.LITTLE implementations, in the name of FinFET. Hsinchu, Taiwan-based TSMC announced ...

Recommended for you

A green data center with an autonomous power supply

5 hours ago

A new data center in the United States is generating electricity for its servers entirely from renewable sources, converting biogas from a sewage treatment plant into electricity and water. Siemens implemented ...

After a data breach, it's consumers left holding the bag

5 hours ago

Shoppers have launched into the holiday buying season and retailers are looking forward to year-end sales that make up almost 20% of their annual receipts. But as you check out at a store or click "purchase" on your online shopping cart ...

Can we create an energy efficient Internet?

6 hours ago

With the number of Internet connected devices rapidly increasing, researchers from Melbourne are starting a new research program to reduce energy consumption of such devices.

Brain inspired data engineering

6 hours ago

What if next-generation ICT systems could be based on the brain's structure and its cognitive and adaptive processes? A groundbreaking paradigm of brain-inspired intelligent ICT architectures is being born.

User comments : 0

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.