Weekly Science Headlines

[P3nT4gR4m]

Worth watching closely. I can see this meshing well with all sorts of other emerging tech, probably in ways I can't even conceive. I'd place a reasonable sum on Moores law being FUBAR in the next 5/10 years. The sheer number of people working in the field today combined with things like Kickstarter would make me pretty confident that this won't slow down, if anything, quite the opposite. 

The purists definition Moores law (size/count/cost of transistors on silicon) is actually getting a bit quicker. Was originally 2 years. It's down to about 14-18 months now. However it will break in the next couple of years on account of physics. The broader phenomenon of exponential growth, tho, which is often referred to as Moore's law, in which shit in general gets exponentially faster, smaller and cheaper and more complex or powerful I dont think will ever break. If you think about it in those terms (as Ray Kurtzweil does) It applies to everything from computers to the phenomenon we call biology only the periods of iteration change. By the time we hit Moore's transistor limit, we wont be using transistors to make chips anymore. I've no reason to expect whatever we are using by then wont be getting faster, smaller and cheaper every year or two - same as it's always been.

Neural net boards hit 1,000,000 neurons

"The human brain, with 80,000 times more neurons than Neurogrid, consumes only three times as much power," Boahen et al write in the Stanford paper. "Achieving this level of energy efficiency while offering greater configurability and scale is the ultimate challenge neuromorphic engineers face."

80,000 times. If Moores law doesn't break soon, we're 25 years away from neural nets with twice as many neurons as humans, using 0.00002 times as much juice.

Neuronal connections are just a tiny part of the not-very-well-understood mechanism of cognitive processing, so don't read too much into that.

Correct. However, just plain neuronal connections on their own are incredibly powerful. A million of them, with a billion or so virutal synaptic connections is pretty fucking huge news. You should also remember that the "not-very-well-understood mechanism of cognitive processing" was invented by an incredibly dumb system that relies on throwing paint at a wall to plan it's design spec. Let's assume the human brain is not necessarily the best machine for the job.

[P3nT4gR4m]

Holy fuckballs, they've finally cracked reliable mass production of pure graphene!

http://www.thomas-swan.co.uk/sites/pdf/adv_materials_datasheets/TS_Graphene_datasheet_lo.pdf

"What the fuck is graphene?" you ask.

By bye plastics, you're not needed anymore.

Anyone want a mobile battery that lasts a month and charges in under a minute and weighs less than a sheet of paper?

This is BIG fucking news

[Telarus]

Saw your link on FB, this is awesome!

[Junkenstein]

P3nt, speaking of moore's law...
http://www.bbc.co.uk/news/technology-27282732

Sony has developed a new storage tape that is able to hold up to 185 terabytes (TB) of data per cartridge.

Created with the help of IBM, Sony's technology allows for tapes that can store the equivalent of 3,700 Blu-ray discs.

The tape hold 148 gigabits (Gb) per square inch - beating a record set in 2010 more than five times over.

I'm not quite sure if that places us ahead or behind at this point. Seems like a minor acceleration.

[minuspace]

 

P3nt, speaking of moore's law...
http://www.bbc.co.uk/news/technology-27282732

Sony has developed a new storage tape that is able to hold up to 185 terabytes (TB) of data per cartridge.

Created with the help of IBM, Sony's technology allows for tapes that can store the equivalent of 3,700 Blu-ray discs.

The tape hold 148 gigabits (Gb) per square inch - beating a record set in 2010 more than five times over.

I'm not quite sure if that places us ahead or behind at this point. Seems like a minor acceleration.

That would like totally make the difference between informed and prolapsed opinions
[Ed.  That response of mine is intended as total jest, just decompressing here]

[P3nT4gR4m]

Tape? Yeah trust Sony to improve the capacity of completely redundant and obsolete storage media. 

Intel are and always have been one of the key drivers of Moore's Law, right now. To the point where they see supporting it as one of their key missions. With their 3d transistor fab they've managed to extend the lifetime of the performance to price graph. The number of transistors will hit a wall but the capabilities of a single transitor has taken a massive leap since (afaik) core i3.

[P3nT4gR4m]

Latest news from the makersphere

[minuspace]

Holy fuckballs, they've finally cracked reliable mass production of pure graphene!

http://www.thomas-swan.co.uk/sites/pdf/adv_materials_datasheets/TS_Graphene_datasheet_lo.pdf

"What the fuck is graphene?" you ask.

By bye plastics, you're not needed anymore.

Anyone want a mobile battery that lasts a month and charges in under a minute and weighs less than a sheet of paper?

This is BIG fucking news

Love it.  How long before we hatch a graphene loaf fully exploiting possibilities of QED architecture...  I mean, why don't I have one already

[minuspace]

At the other end of the debate, "White Space/band" debate is heating up.  Probably good looking for the future if we stay on it:
http://whitespaces.spectrumbridge.com/Overview/Home.aspx

[P3nT4gR4m]

Scotch Tape, thumb tacks and picture hooks ovar!

Demo

FTR: Gloves can be made out of this shit!

[Pæs]

Tape? Yeah trust Sony to improve the capacity of completely redundant and obsolete storage media. 

Are you thinking of cassette tapes? Because AFAIK high-density magnetic tape storage is still the standard for archiving ridiculous quantities of data, because it's more reliable and much cheaper than other media. They're also more comfortably transported. Taking entire drives to off-site backup is a risky endeavour. Last I heard (a year or two ago) Google primarily used tape for backup of everything Google.

Tape libraries.


Tape is still pretty perfect if you need to read a lot of data sequentially or don't have a need to search through it regularly, which suits archival uses just fine.

[minuspace]

Tape? Yeah trust Sony to improve the capacity of completely redundant and obsolete storage media. 

Are you thinking of cassette tapes? Because AFAIK high-density magnetic tape storage is still the standard for archiving ridiculous quantities of data, because it's more reliable and much cheaper than other media. They're also more comfortably transported. Taking entire drives to off-site backup is a risky endeavour. Last I heard (a year or two ago) Google primarily used tape for backup of everything Google.

Tape libraries.


Tape is still pretty perfect if you need to read a lot of data sequentially or don't have a need to search through it regularly, which suits archival uses just fine.

I ave no idea why, but I have carried an ADAT with me for the past decade.  Actually, I guess it's built like a tank, takes massive abuse and always works.  Guess I could have it repurposed to read graphene tape

[I really like this thread - also just read the quick release super adhesive thing - awesome!]

[P3nT4gR4m]

http://www.huffingtonpost.com/2014/05/28/richest-man-american-medicine_n_5355277.html A fucking one man revolution. Sounds like he's well on the road to picking up the tricorder xprize!

[P3nT4gR4m]

I don't understand quantums (fuck knows I've tried) so I don't understand quantum computers but I'm reliably informed they'll be much faster than silicon ones. I don't understand silicon either but I've a hundred and one ideas as to what I can do with fast. This seems like it could be big news. Maybe someone who does understand quantums could explain roughly how big

https://news.wsu.edu/2014/06/04/discovery-opens-new-path-to-superfast-quantum-computing/#.U48DD_ldWFV

[minuspace]

I don't understand quantums (fuck knows I've tried) so I don't understand quantum computers but I'm reliably informed they'll be much faster than silicon ones. I don't understand silicon either but I've a hundred and one ideas as to what I can do with fast. This seems like it could be big news. Maybe someone who does understand quantums could explain roughly how big

https://news.wsu.edu/2014/06/04/discovery-opens-new-path-to-superfast-quantum-computing/#.U48DD_ldWFV

Granted it's too complicated for me, still I feel reasonably disconcerted by the possibility of quantum computing solving NP-C problems in ways that would render obsolete most encryption protocols.