Nvidia describes 10 teraflops processor
November 18, 2010 // Rick Merritt
Nvidia's chief scientist gave attendees at Supercomputing 2010 a sneak peak of a future graphics chip that will power an exascale computer. Nvidia is competing with three other teams to build such a system by 2018 in a program funded by the U.S. Department of Defense.
Nvidia's so-called Echelon system is just a paper design backed up by simulations, so it could change radically before it gets built. Elements of its chip designs ultimately are expected to show up across the company's portfolio of handheld to supercomputer graphics products.
"If you can do a really good job computing at one scale you can do it at another," said Bill Dally, Nvidia's chief scientist who is heading up the Echelon project. "Our focus at Nvidia is on performance per watt [across all products], and we are starting to reuse designs across the spectrum from Tegra to Tesla chips," he said.
In his talk, Dally described a graphics core that can process a floating point operation using just 10 picojoules of power, down from 200 picojoules on Nvidia's current Fermi chips. Eight of the cores would be packaged on a single streaming multiprocessor (SM) and 128 of the SMs would be packed into one chip.
The result would be a thousand-core graphics chip with each core capable of handling four double precision floating-point operations per clock cycle—the equivalent of 10 teraflops on a chip. A chip with just eight of the cores would someday power a handset, Dally said.
The Echelon chip packs just twice as many cores as today's high-end Nvidia GPUs. However, today's cores handle just one double precision floating-point operation per cycle, compared to four for the Echelon chip.
Many of the advances in the chip come from its use of memory. The Echelon chip will use 256 Mbytes of SRAM memory that can be dynamically configured to meet the needs of an application.
For example, the SRAM could be broken up into as many as six levels of cache, each of a variable size. At the lowest level each core would have its own private cache.
The goal is to get data as close to processing elements as possible to reduce the need to move data around the chip, wasting energy. Thus SMs would have a hierarchy of processor registers that could be matched to locations in cache levels. In addition, the chip would have broadcast mechanisms so that the results of one task could be shared with any nodes that needed that data.
To ease programming, the design is cache coherent across both graphics and traditional processor cores. Indeed, finding ways to program many-core processors is one of the chief challenges for today's computer scientists.
"We are about to see a sea change in programming models," said Dally. "In high performance computing we went from vectorized Fortran to MPI and now we need a new programming model for the next decade or so," he said.
All the languages use similar ingredients. For example, they try to build into their semantics support for advanced memory sharing mechanisms.
Nvidia's Echelon system will compete with teams from Intel, MIT and Sandia National Labs, each taking different approaches to build power efficient exascale systems.
The Ubiquitous High Performance Computing program is sponsored by the Defense Advanced Research Projects Agency. DARPA tasked the teams to build by 2014 a prototype petaflop-class system into 57 kilowatt rack prototype computer. Such systems could be used as building blocks to create an exascale system to be built by 2018.All news
Automotive industry sees innovation in products, not business models
December 10, 2013
Innovation strategies in the automotive industry are aiming to technology and better customer orientation, states management ...
Big moves in chip vendor top 20 ranking
Extendible processors go head to head backed by EDA giants
Koch Industries acquires Molex for USD7.2 billion
EnOcean joins the OSGi Alliance to define universal open interface for energy harvesting wireless
GaN-on-silicon LEDs to grow market share to 40 percent by 2020
December 09, 2013
The penetration of gallium nitride-on-silicon (GaN-on-Si) wafers into the light-emitting diode (LED) market is forecast to ...
Pebble CEO turns back clock
Tyndall National Institute captures snapshots of moving atoms under bursts of light
Measurement technique allows survey of 3D micro objects
- 3mm × 3mm QFN IC Directly Monitors 0V to 80V Supplies
- UltraCMOS® Semiconductor Technology Platforms: A Rapid Advancement of Process & Manufacturing
- Adaptive Cell Converter Topology Enables Constant Efficiency in PFC Applications
- Isolated 4-Channel, Thermocouple/RTD Temperature Measurement System with 0.5°C Accuracy
InterviewPerformance monitoring solution helps provide intelligent control of high power systems
A performance monitoring solution designed to enable companies to monitor high power IGBT module systems in locomotive, wind turbine, High Voltage DC and industrial drive applications was unveiled this ...
Filter WizardCheck out the Filter Wizard Series of articles by Filter Guru Kendall Castor-Perry which provide invaluable practical Analog Design guidelines.
Linear video channel
READER OFFERRead more
Internet of Things (IoT) manufacturer Ciseco has launched the Raspberry Pi ‘Wireless Inventors Kit’ (RasWIK), featuring 88 pieces to provide everything a Pi owner needs to follow a series of step-by-step projects or to create their own wireless devices, without the need for configuration or even writing code.
RasWIK has been designed to be highly accessible, demystifying the dark art of wireless and enabling anyone with basic computing skills to begin building wireless devices with a Raspberry Pi. You can create anything from a simple traffic light, to a battery monitor, or even a temperature gauge that sends data to the Xively IoT cloud so billions can access the data.This month, Ciseco is giving away twelve Raspberry Pi Wireless Inventors kits, worth £49.99 each for EETimes Europe's readers to win.
And the winners are...
In our previous reader offer, Farsens was giving away five kits for EEtimes Europe readers to evaluate its FenixVortex, Kineo and X1 wireless, battery free sensor tags.
Lucky winners include Mr A. Neil from the UK, Mr. E. Delvaux from Belgium, Mr Lengal from the Czech Republic, Mr H. Bijlsma from the Netherlands, and Mr G. Pfaff from Germany. All should be receiving their packages soon. Lets wish them some interesting findings with their projects.
December 15, 2011 | Texas instruments | 222901974
Unique Ser/Des technology supports encrypted video and audio content with full duplex bi-directional control channel over a single wire interface.