As this writer makes clear nano photonics is kicking over the traces ofconventional computer technology and replacing electrons with photons is eventurning Moore
This throws away all inhibitors in achieving the holodeck. I bring that up because all can imagine itand can also comprehend the massive computational power necessary. This makes it all true. We will live to see it.
Actual cost will continue to drop like a stone and that gig of storagewill seem as quaint as the meg you were once proud to own. How about we supply you with the entirecontents of civilization digital efforts?
Exaflop computing: Moore
By JasonPerlow | December 2, 2010,
Summary
Silicon Nanophotonics will usher in a completelynew age of computing and power applications that we’ve only seen on Star Trek.
Silicon Nanophotonics will usher in a completely new age of computingand power applications that we’ve only seen on Star Trek (image: IBM)
Theuniverse has a funny way of playing karma tricks on us writers that follow thetech industry and dare to make sweeping prognostications about future trends.
For example, yesterday, myZDNet Storage Bits blogging colleague Robin Harris wrote that theindustry may have hit thewall in terms of increasing computing performance.
Then on the same day myemployer, IBM, pulled the rug outfrom under him. SorryRobin. It happens to the best of us. Seriously, I feel for you man.
The funny thing about Moore’s Law isthat every single time the industry calls for its inevitable demise, the Godsof Technology come and knock you on your ass. When we think we’ve pushedlithography and compacting transistors to their absolute limit, an advance intechnology allows the trend to continue as it always has been.
Thistime, however, instead of just proving itself consistently correct, Moore’s Lawis going to have to be completely re-written — instead of microprocessortechnology doubling its performance every two years, we’ll be looking forwardto ten to twenty fold increases in computational power, at a bare minimum,every five years.
Thisincrease in performance is so significant that the math itself is mind-bogglingand it becomes difficult to actually relate to it in conventional terms, oreven express it in a quantifiable fashion that makes sense to informationtechnology practitioners outside of very high-end scientific research.
Today, advances inmicroprocessors are built on the premise of cramming as many transistors onto apiece of silicon as possible. Over the last four decades, we’ve continued toadvance processing power by using different lithography techniques that allowsemiconductors to be manufacturedin densities of continually decreasing nanometers in width.
Robin may indeed be correctthat we may have hit the wall with this approach and advances usingconventional microprocessor design may only result in very small incrementalimprovements. Eleven nanometers may bethe practical limit as to how small we can go before we hit the physical limitsof what can be done using current semiconductor technology.
However, what IBM showed tothe public on December 1, 2010 changes the game dramatically, especially for supercomputing applications.Eventually, these advances will filter down to the enterprise systems and evenconsumers.
Artist’s conception of the future application of photonic routingelements onto a silicon wafer (IBM)
This technology — with a namepulled seemingly right out of Star Trek is called CMOS Integrated Silicon Nanophotonics.
Without getting toointergalactic and too technical of a description of how it actually works, itis essentially the fusion of optical technology with semiconductor technology.Instead of using semiconductor pathways to route data and for the processorinterconnects, light pulses are being used instead, using componentscalled SiliconNanoscale Photonic and Electronic Transceivers, or SNIPERS.
To put all of this techno-jargon in the properperspective, one must understand what is currently the benchmark for mostpowerful supercomputer in the world, the Tianhe-1A.
This powerful Chinese supercomputer has achieved 2.67 Petaflop/s or 2.67 quadrillion floatingpoint operations, per second.
This is so amazingly fast that it realizes the kind ofscientific research and advanced simulation that only 10 years ago could onlyexist in a computer scientist’s wet dreams. But the Tianhe-1A uses strictlyconventional computer technology, assembled from over 14,000 Intel Xeon 5670-seriesx86 processors and 7,000 nVidia Tesla GPUs using a custom high-speed interconnectnetwork that operates at 160Gbps per second.
But compared to future nanophotonics-based systems,assessing the legacy performance of the Tianhe-1A will be like comparing thetop speed of a Segway personal transport to that of a Bugatti Veyron.Or an X-15.
The first application of IBM’s nanophotonics technologywill be used in the Blue Waters supercomputer, which will be installedat the National Center forSupercomputing Applications (NCSA) in Urbana , Illinois
However, eventhe Blue Waters supercomputer will become a clunky dinosaur once more and morenanophotonics elements are integrated onto microprocessors and replace many ofthe functions that traditional semiconductors perform today. And as with allsupercomputing technologies, the equivalent of N minus 2 generations ofprocessor performance will quickly find their way into enterprise systems andeven consumer electronics.
Projected technologicalprogression of photonics integration in supercomputing (IBM)
Eventually, perhaps by the year 2020 or even sooner,nanophotonic supercomputers will exist that approach the Exaflop range in terms of performance. What’san Exaflop? Well, try measuring the aggregate performance of about 250 Blue Waters or 1000Tianhe-1Asupercomputers and that’s what you’ll get. Did your head justexplode? No? Because mine just did.
So with Exaflop-level supercomputing and enterprisecomputing, or even Petaflop-level consumer computing, justwhat exactly can you do withall of that processing power?
Well, with Exaflop/s, you can do the sort of things thattake current supercomputers weeks or years todo in only minutes or days. Itwould allow the average citizen to gene sequence their babies the moment theyare born to anticipate future diseases for an entire lifetime, or sequencetheir own DNA and apply corrective action as needed, such as synthesize custommedications.
It would allowfor the real-time simulation of complex systems such as world weather andallow for meteorological science to advance at a level approaching magic orwitchcraft — the ability to accurately model how destructive storms suchas tornadoes and hurricanes actually work and form, and accurately predict howthey will behave.
It would permit our various governmentintelligence agencies, such as the National Security Agency and well as the National ReconnaissanceOffice toperform advanced signal intelligence (SIGINT) and spaceimaging in real time, determinethreats to our national security and respond to it in kind with militarystrikes and with covert operations almost instantaneously. It would completelyre-define what we understand today as rapid response and operational readiness.
The very same technology used for military applications,along with a much larger global array of radio telescopes than what we possesstoday, could also be used to perform much more comprehensive real-time signalspectrum analysis of extraterrestrial radio emissions and actually allow SETI to prove the existence of the E.T’swe’ve been trying to locate for over 40 years, if those signals do indeedexist.
Beyond bruteforce computational applications, Exaflop and commodity Petaflop computing willalmost certainly allow for the creation of intelligent robots and softwareagents, perhaps as brilliant as a trained circus dog, an advanced primate, afive year old child, or even more depending on advancements in computerlearning.
It would allow for the real-time renderingof computer-generated imagery from today’s biggest and most expensive Hollywood blockbuster films (or even better) in virtualreality or virtual worlds for the average citizen. These types of games andfully-immersive artificial realities would make the most advanced XBOX 360 orPlaystation 3 first-person game look about as sophisticated as PONG.
Essentially, we’re talking about the delivery of The Matrix as a commercial software product.
There arecertainly other applications for commodity Petaflop and Exaflop supercomputingthat people haven’t even dreamed of yet. But I think we can say for sure that Moore
What else willwe do with commodity Petaflop and Exaflop computing?Talk Back and LetMe Know.
Disclaimer: The postings andopinions on this blog are my own and don’t necessarily represent IBM’spositions, strategies or opinions.
No comments:
Post a Comment