Cray To Build Australia's Fastest Supercomputer 54
Bismillah writes: US supercomputer vendor Cray has scored the contract to build the Australian Bureau of Meteorology's new system, said to be capable of 1.6 petaFLOPS and with an upgrade option in three years' time to hit 5 petaFLOPS. From the iTnews story: "The increase in capacity will allow the BoM to deal with growth in the 1TB of data it collects every day, which it expects to increase by 30 percent every 18 months to two years. It will also allow the agency to collect new areas of information it previously lacked the capacity for. 'The new observation platforms that are coming online are bringing quite a lot more data,' supercomputer program director Tim Pugh told iTnews.
Re: (Score:1)
You should get out more...
Re: Australia's Fastest Supercomputer. Hmmmmm (Score:2)
Re: (Score:1)
haha. You like getting backdoored!
Re: (Score:2)
Not very familiar with the country, huh?
At the moment temps are ranging from -3C to about 5C here. In summer it gets up around the late 40s or early 50s, but otherwise, it is subject to change without notice.
Further note, BoM is typically "Bank of Melbourne".
Re: (Score:2)
Small nations usually buy big computer for a few historic reasons, very advanced weapons design or a crypto race.
Australia has given up on its own advance weapon design and fully imports its needs from the USA.
Huge amounts of data collected in Asia for itself and the NSA/GCHQ flows into remote secure collection sites.
That bulk flow might find its way into a dual use civilian military telco system with global suppliers and other nations been very close to internal domesti
Re: (Score:2)
Quantum computer work has been great for funding.
"Quantum storage breakthrough key to 'unbreakable' encryption A new quantum hard drive jointly developed by researchers in Australia and New Zealand could lead the way to an 'unbreakable' worldwide data encryption network."(January 12, 2015)
http://www.zdnet.com/article/q... [zdnet.com]
"Powerful quantum computers move a step closer to reality A research team from Australia has pushed q
Re: (Score:2)
Re: (Score:2)
Re: (Score:1)
But can it run Crysis?
Re: (Score:2)
This would be posted when Slashdot was still news for nerds:
"I wonder what a Beowulf cluster of these could do.."
Why don't you ask Natalie Portman why things have gone down hill? I'm sure you can get her comment before her breakfast of hot grits. But just be prepared for an answer that welcomes the new overlords.
The sad thing is that Netcraft will now finally be able to confirm what is going on, to the point that not even an insensitive clod in Soviet Russia will be spared any feelings. Now if you will excuse me, I need to change the onion on my belt.
Just breaking . . . (Score:5, Funny)
Just breaking: Australian Prime minister Tony About-face reports that, in the interests of national security, the weather will no longer be reported, and the machinery will be turned over to the George Brandis / Australian Federal Police, where it will be used to monitor the internet (just meta-data) for unauthorized wind-mills.
I don't know about Aus weather forecasts (Score:4, Insightful)
... but here in the UK our Met Office sometimes can't even get it right 12 hours in advance. I'm not blaming them for that because the point i'm making is it doesn't matter how powerful the computer, if you don't have enough data and/or the software model isn't good enough then the hardware won't make much difference especially when chaos theory is working against you all the time too.
Re: (Score:2)
Australian weather forecasts are easy: Just report that it will be hot! Most of the time you will be right on the money.
Except when it's -6C (last night) or lower. Canberra, National Capital, of Australia - which is not a small island off the coast of California you geographically challenged fool.
Try camping out overnight in the Death Valley and see how "mostly hot" that is - don't worry about a sleeping bag. Please.
Re: (Score:1)
Chaos theory is working against you insofar as a slight change to the input data might yield a large change to the output data. So, they add noise to the input data and run many simulations, to get a probability distribution over outcomes. OBVIOUSLY if you have faster hardware you can run more simulations in the same time and get a higher quality result. But the higher quality is only a better probability distribution. It still won't tell you what will *definitely* occur. That's impossible and you shou
Re: (Score:1, Interesting)
With all these dynamics converging in one place it makes sense that we're building a 16 petaflops supercomputer [bbc.co.uk] to replace the current one the UK met office uses.
The new system will be housed partly at the Met Office headquarters in Exeter and partly at a new facility in the Exeter Science Park, and will reach its full capacity in 2017.
.
At that point, its processing power will be 16 petaflops - meaning it can perform 16 quadrillion calculations every second
The "Cray XC40" machine will have 480,000 central processing units or CPUs, which is 12 times as many as the current Met Office supercomputer, made by IBM. At 140 tonnes, it will also be three times heavier.
That kind of makes my eyes bleed / head hurt to think
Re: (Score:1)
Re: (Score:2)
if you don't have enough data and/or the software model isn't good enough then the hardware won't make much difference
Agreed. This is why they are also assing hardware to record more data. More data = bigger computer = better predictions makes sense to me.
Re: (Score:2)
here in the UK our Met Office sometimes can't even get it right 12 hours in advance.
Here in the western USA, reading the AFD's (Area Forecast Discussions), whenever model tendencies come up, the Euro and Canadian models tend to be way off, perhaps because, despite being global models, they're optimized for high northern latitudes whereas we're in the Horse latitudes. Only a guess, but it's a constant. Further south, for tropical eastern Pacific hurricanes, the results tend to be hilariously off and are dismissed by the forecasters almost immediately.
Hope they remember the power bill (Score:1)
Re: (Score:2)
So Melbourne uni had a great cluster. Spent a fortune on it. It's switched off because they couldn't afford the power. Let's hope the BOM remembers to factor that into the estimate. I guess we know now why they want to increase the GST.
I didn't know that, thanks - if I had mod points I'd mod that up.
I'm pretty sure the push to increase the GST is so Abbott can lower the tax rates for his coal mining mates - and to pay for more helicopter flights by politicians of all parties to fly more instead of having to drive.
Re: (Score:2)
3) Do the CPU fans spin in the opposite direction?
Well done. I tip my hat to you, sir or madam or whatever it is that lurks behind AC posts.
I agree with this title. (Score:2)
But the main question is ... (Score:2)
Re: (Score:2)
It has a 150 MHz processor, 200 megabytes of RAM, 1500 megabytes of disk storage, a screen resolution of 4096 x 4096 pixels, relies entirely on voice recognition for input, fits in your shirt pocket and costs $300.
Except for the screen resolution, we beat them all several times over. Voice recognition has come a long way but still way short of the bridge of USS Enterprise NCC-1701. And Popular Mechanics has been selling flying car on its cover to gullible teenagers since 1930s. Come on Mechanical Engineers. Step up to the plate and deliver something.
Pity (Score:3)
It seems that the more Australia's BoM relies on computer modelling the worse their predictions become. Honestly the predictions of 15-20 years ago were more accurate than the BoM has been able to produce in the last decade or so.
It is too bad (Score:1)