backtop


Print 20 comment(s) - last by Proxicon.. on Apr 5 at 10:59 PM


International Square Kilometre Array  (Source: kurzweilai.net)
The idea is to develop fast, but low-power exascale computer systems for the international Square Kilometre Array

IBM has partnered with ASTRON, the Netherlands Institute for Radio Astronomy, for a five-year, 32.9 million euro ($43.8 million USD) venture to develop efficient exascale computer systems for the international Square Kilometre Array (SKA).

ASTRON is one of the top developers of SKA, which is an international consortium to create the world's largest, ultra-sensitive radio telescope. The new SKA telescope is estimated to be completed in 2024, and will explore dark matter, evolving galaxies and even the origins of the universe. IBM said the amount of processing power needed for the telescope to function will be several millions of today's fastest and most powerful computers.

To reach that amount of computing power, the new SKA telescope will need technology and data transfer links that reaches far beyond what is available today. At the same time, it's important for SKA's exascale computer systems to be fast, yet low-power. To achieve these goals, ASTRON has agreed to collaborate with IBM for five years and 32.9 million euro ($43.8 million USD).

The partnership, which is called DOME to reflect the famous Swiss mountain as well as the cover on telescopes, aims to research technologies capable of handling exascale computing, storage processes and data transport on a large scale, but that is also energy efficient. More specifically, IBM and ASTRON will look into advanced accelerators and 3D stacked chips as well as new optical interconnect technologies and nanophotonics for both efficient computing and large data transfers. As far as storage goes, revolutionary tape systems and phase-change memory tech will be researched.

"If you take the current global daily Internet traffic and multiply it by two, you are in the range of the data set that the Square Kilometre Array radio telescope will be collecting every day," said Ton Engbersen of IBM Research. "This is Big Data Analytics to the extreme. With DOME, we will embark on one of the most dataintensive science projects ever planned, which will eventually have much broader applications beyond radio astronomy research."

Between ASTRON's experience with SKA and IBM's research and methodologies regarding realistic, yet powerful and efficient computer systems, the DOME partnership is expected to make the new SKA telescope a reality. DOME received funding from the Netherlands, the Province of Drenthe, and the Dutch Ministry of Economic Affairs, Agriculture and Innovation.

"Large research infrastructures like the SKA require extremely powerful computer systems to process all the data," said Marco de Vos, Managing Director of ASTRON. "The only acceptable way to build and operate these systems is to dramatically reduce their power consumption. DOME gives us unique opportunities to try out new approaches in Green Supercomputing. This will be beneficial for society at large as well."

Source: IBM



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

RE: The mother of innovation
By geddarkstorm on 4/2/2012 5:49:19 PM , Rating: 2
The new iPad is 2x the pixel pushing performance because they stuck a second dual core GPU next to the original dual core GPU. They doubled the amount of hardware and got double the performance. No mystery, no great advance. And it comes at a near doubling the amount of energy used by the device (hence the massive battery) and much greater heat production.

That is not a fundamental advance or paradigm shift in the base technology like spintronics or phase change memory or any of the other exotics needed to make exocale a reality would be.

Also, you're thinking about the wrong Pentium. The Pentium 4 came out November 20, 2000, that's almost 12 years ago, or the length of time until this technology is supposed to be finished and the array fully operational. And P4s are still sold to this day in applications where you don't need multicore performance. Think about that.

Is there a sigifnicant change in the base technology between a P4 and a Ivy Bridge? Nope. Just interative improvements in design and most critically die shrinks. The only real new technology breakthrough is the 3D gates on the transistors. That is a real change in the base technology, if minor.

Also, don't think that as technology for this array project is built that it won't leak out bit by bit into the consumer space. We won't have to wait till 2024 for IBM and others to already be starting to put to other uses the production technology they have to invent and move from a lab to the real world for this exoscale project--assuming the project is successful to any degree. And hey, it could always fail, that's what makes it a risk, and also why businesses themselves are loath to try such a big fundamental game changing project like this, unlike science.


RE: The mother of innovation
By Reclaimer77 on 4/2/2012 9:56:07 PM , Rating: 1
You need to tone down the rhetoric just a bit here. Are you their new PR guy or something?

To say this project will explode our current rate of technological advancement in the field of computer electronics where the market has failed seems a bit naive.

Let's be honest here. I've seen stories like this multiple times a year for decades now. Very rarely do we see a direct technological trickle down from it. There's always some amazing "breakthrough" right around the corner, then it disappears and we hear nothing else about it.

quote:
And hey, it could always fail, that's what makes it a risk, and also why businesses themselves are loath to try such a big fundamental game changing project like this, unlike science.


Okay I don't know why you keep referring to "science" as if it's some great autonomous force, but anyway, you show a clear lack of understanding in how things work. A scientist might come up with a computer that's 400 times faster and more energy efficient than what we have today, but it's businesses taking risks that figure out how to make it affordable and marketable. Assuming that's even possible in the first place with current technology and manufacturing processes.

Do me a favor, when you find a computer product that has "Made by Science" stamped on it, and not a business, you give me a call.


RE: The mother of innovation
By geddarkstorm on 4/3/2012 2:06:01 AM , Rating: 2
You seriously misunderstand.

The market hasn't failed in any sense. It simply doesn't have the motivation to spur people to take the risk on exploring such advanced concepts for new technologies. How is that anything new or novel to say? Don't you understand how business even works, and what is necessary to make a successful one? A scientific endeavor is not limited by such constraints, and so can do the off the wall stuff that the market later down the road finds a use for.

quote:
Let's be honest here. I've seen stories like this multiple times a year for decades now. Very rarely do we see a direct technological trickle down from it.


What are you talking about? ALL technology is born this way.

quote:
Do me a favor, when you find a computer product that has "Made by Science" stamped on it, and not a business, you give me a call.


Reclaimer, you are rarely this intentionally ignorant. ALL computers are "made by science". Scientists researched and developed computers. Scientists researched and developed the internet. Some were employed by industry, some were employed by the military. But all computer products come straight from scientific R&D.

Don't you realize that industry spends more on scientific R&D than the government and academia combined?

What I've been talking about is motivation. Necessity is the mother of all innovation--and the necessity for completely new foundational computer technology to meet the needs of this array breeds motivation for an innovation speed that cannot be fueled through any other source but War.

The market is slow at innovating compared to the challenges science presents. And that's why if you want to advance our technology rapidly, you invest in science. That's what I'm simply saying; I honestly have no idea where your ideas came from, but they were so out there I doubt they even caught a glimpse of left field.


RE: The mother of innovation
By Paj on 4/3/2012 7:53:19 AM , Rating: 1
The technology that is going to power the next generation of cool shit already exists. In labs everywhere, people are demonstrating examples of awesome, cool shit that works every day. Sometimes these labs are run by governments, sometimes corporations, often partnerships between the two, as in the SKA.

Do you really think the first iphone came out in 2007? They would have had a working model of it well before that, and the various technologies within it would have been working and demonstrated years before 2007. The Wi Fi technology that we consider a 'modern' invention was functioning in labs in the early 1990s.

Sometimes the tech in question isnt quite ready for market yet though. Maybe its too big, runs to hot, costs too much, doesnt have the necessary economies of scale, is too hard to manufacture cheaply with current techniques. But its there.

Sometimes, the reason the tech isn't in products yet is, in many cases, due to commercial reasons. Maybe there's still more money to be made from current generation hardware. Maybe companies made too many units of the previous model that they need to shift. Product life cycles need to run their course. Profit forecasts need to be met. It's often economically prudent not to play all your cards at once, but to slowly let them trickle out over a period of years or even decades to maximise profits.


RE: The mother of innovation
By Proxicon on 4/5/2012 10:59:19 PM , Rating: 2
RAMBUS was gonna be Huuuuuge!


RE: The mother of innovation
By modus2 on 4/3/2012 9:01:58 AM , Rating: 2
I think you overestimate the size of this project ($43.8M), Intel alone spent $6.2B in 2010 on R&D. That being said IBM have a good track record for researching HPC, just that the kind of money involved here sound more like they are going to combine emerging technologies from other sources rather than actually build from scratch themselves.


"If you look at the last five years, if you look at what major innovations have occurred in computing technology, every single one of them came from AMD. Not a single innovation came from Intel." -- AMD CEO Hector Ruiz in 2007














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki