Comments

You have commented 339 times on Rantburg.

Your Name

Your e-mail (optional)

Website (optional)

	My Original Nic Pic-a-Nic Sorry. Comments have been closed on this article.
Foto

Submit your comments on this article

Science & Technology

How AI Can Keep Accelerating After Moore's Law

2017-06-02

[MIT] Google CEO Sundar Pichai was obviously excited when he spoke to developers about a blockbuster result from his machine-learning lab earlier this month. Researchers had figured out how to automate some of the work of crafting machine-learning software, something that could make it much easier to deploy the technology in new situations and industries.

But the project had already gained a reputation among AI researchers for another reason: the way it illustrated the vast computing resources needed to compete at the cutting edge of machine learning.

A paper from Google’s researchers says they simultaneously used as many as 800 of the powerful and expensive graphics processors that have been crucial to the recent uptick in the power of machine learning (see "10 Breakthrough Technologies 2013: Deep Learning"). They told MIT Technology Review that the project had tied up hundreds of the chips for two weeks solid--making the technique too resource-intensive to be more than a research project even at Google.

A coder without ready access to a giant collection of GPUs would need deep pockets to replicate the experiment. Renting 800 GPUs from Amazon’s cloud computing service for just a week would cost around $120,000 at the listed prices.

Posted by:Besoeker

#5 Fascinating stuff. No way I can even begin wrap my brain around any of it, but fascinating stuff.

Posted by: Besoeker 2017-06-02 17:15

#4 We are crossing a big paradigm shift now as the industry shifts from Deep Ultraviolet (DUV) to Extreme Ultraviolet (EUV) lasers. One of the challenges to Moore's law was the wavelength of the Deep Ultraviolet beam was too fat and double-patterning the waffers was sloppy. EUV solves that.

Posted by: rjschwarz 2017-06-02 14:48

#3 Two paradigm shifts.

The longer term one is new chip architectures based on neural connections similar to those in mammal brains. Such chips have been demonstrated in early prototypes by IBM and others.

The shorter term shift is already in place in 'deep learning' based on nested neural nets.

Posted by: Angusock Chuting6110 2017-06-02 09:40

#2 Now it doesn’t take much to figure out that this cannot carry on indefinitely. Somewhere along the line there MUST be a stop, a ceiling of some sorts, a saturation point if you will.

And then at some point there will be another paradigm shift.

Posted by: JohnQC 2017-06-02 08:40

The electrical engineer and billionaire businessman, Gordon Moore, founder of “Intel Corporation” is the man who first observed in 1965 that the amount of transistors in a printed circuit board would double roughly every two years and predicted that it would carry on like that for at least another ten years or so. 50 years later his prediction is still happening. Therefore, it is called Moore’s Law

Now it doesn’t take much to figure out that this cannot carry on indefinitely. Somewhere along the line there MUST be a stop, a ceiling of some sorts, a saturation point if you will.

For instance, Moore’s prediction has become a target for miniaturization in the semiconductor industry, but components cannot be made smaller and smaller all the time, because somewhere along the line you cannot make it smaller than the atom.

Over the years lots of people have predicted the death of Moore’s Law, yet it is still alive. It carries on, and on, and advances technology to the benefit of the human race. However, every time someone says, this is now definitely the end of Moore’s Law, it is not.

Some sources say we will have tri-gate transistors, perhaps around 2020. After that there will be "gate all around" transistors and nanowires. The mid-2020s could bring monolithic 3D chips, where a single piece of silicon has multiple layers of components that are built up on a single die, but still…every two years we see Moore’s Law in action. …In other words, Moore’s Law, although being 50 years old already, is still alive as you are reading this and there are no signs that it will be letting up anytime soon despite predictions.

Posted by: Besoeker 2017-06-02 07:57

00:00