Forget NASA, this is where the physicists are going...
 

Buy neutrinos

Wired.com Article
http://www.ritholtz.com/blog/wp-cont...press-lane.png

That is cool.

I ordered 50/25Mbs internet. Quantam.

Not knowing crap about networking or it's infrastructure, these transatlantic and transartic fiber optic cables seem like a horrible idea to me. You would think there has to be a more cost effective and easily upgradable options taking data through the air rather than running a telephone wire 3500 miles long and 3 miles deep in the ocean.

How would you transmit the data through the air? On land, they use microwave towers (~200' tall IIRC). They actually address one potential application in the full article involving solar powered hovering drones over the Atlantic. :brain:


One thing missed here, is the implication that some of the greatest minds in the world are now working hard (and spending a metric -----ton of money) on moving numbers around in the financial markets. These are people that left scientific research and academic mathematical careers or engineering positions to figure out the best algorithm to make $0.001 profit per trade.

Wait, so gov't provided wifi towers isn't innovation?

I don't understand your question?

All of the technology infrastructure discussed in the article is privately funded. They are spending billions of dollars [about what it cost to put a robot on the surface of Mars, FWIW] in an attempt to shave fractions of milliseconds off of the amount of time it takes a data packet to make a round trip from Chicago to New York or New York to London.

And these are not public pipelines. Annual subscriptions to use them run in the hundreds of thousands of dollars. More than you can afford, pal. Ferrari!

They then devote their genius to creating mathematically-oriented trading algorithms designed to profit fractions of a penny, hundreds or thousands of times per day.

Engineers and physicists are devoting their genius to figuring out the transmission mechanisms to move that data.

Computer and software engineers are devoting their genius to figuring out how to assemble or configure the hard and software necessary to run these systems.

Meanwhile, other groups are creating businesses to assemble, compile and make machine-readable all of the data necessary for the quant traders to create their trading algorithms.


All of this human capital being funneled in to shuffling shares around and ocassionally causing havoc (Knight Capital, BATS) when the machines get ahead of the people.

Impossible.

the current administration doesnt believe that 95% of the populous having access to high speed internet is enough, and they nede to come fill in the "gaps"

just a stupid jab. i do that.

kinda like how my incredibly rich spent tons of money on these bike renting machines to give lower income people any easier way to get around town. problem is you need a CC to rent one, and I've never seen a person of color ever riding one. or anyone below 20 years old or above twenty-five. yet the spent millions to put them all over the city, you know, to help the "poor" (we dont have poor in arlington.)

fuck you Pittsburgh

Sorry...

I'm not sure if you are serious or making a sarcastic remark about the idea that only government can be counted on to provide infrastructure.

The difference here, is that private sector infrastructure has a clear and short-term available return on investment: $250k (or more) per year, per client subscription fees.

This is not pipeline for downloading your torrent faster. This is pipeline to make money by shuffling financial assets around. There is no greater good served and no technology trickle down that I can think of. If anything, there is harm to broader society in general as the HFT and quant traders disrupt trust in the equity markets, ocassionally bring down electronic trading markets and generally front-run individual investors.

That's to say nothing of the posibility of algorithms designed to drive down stock prices actually succeeding better than their authors may have intended and nearly wiping out the market capitalization of companies (see the BATS example where the stock was driven to pennies before trading was halted).


And I'm going to fire up my time-space distortionary transport vehicle to run through some parallel universes, because I am pretty sure some of these current and future geniuses that are pouring in to the financialization of the US economy would have been working on cures for disease or some other great technological breakthrough that would benefit society - or even general consumers - in ways that HFT never will.

The only present-day technology which is capable of moving large quantities of information across the ocean over a "through the air" path is satellite radio.

One problem with satellites is that they're in space. 22,236 miles above sea level, to be specific. So every packet of data sent has to travel 22,236 miles up, and then 22,236 down to the receiver- 44,472 miles total. If this is a packet which requires a response (eg, a ping) then add another 44,472 miles for a total of 88,944.

Radio waves in space travel at roughly the speed of light, which is appx. 186,000 miles per second. So that ping would spend nearly half a second (478 millisec) simply travelling through space.

By way of comparison, a fiber path between NYC and London would be only around 3,500 miles. And even correcting for the reduced speed of light in fiber (around 0.625c) the total path length off 7,000 miles would still yield a 60 millisecond ping, a MASSIVE improvement.

Since this pipe is being put in place specifically to service people who are willing to pay a premium for low-latency communication, that's pretty significant.

It was (almost) entirely sarcasm, but if I wanted to actually pursue that point, I'd have to reject your conclusion. As I was told repeatedly in the other thread, it doesn't matter if we can't predict or anticipate the possible benefits and uses of the technology. It only matters that it's new and fancy.

Oh boy here we go again.

It's nice to have lower communication latency for trading through use of better communication technologies but it's not the only thing that is important. There are many other things that will add latency to the end-to-end communication channel (whatever it may be - copper wire, optical cable, waves, etc).

Take a look at 0mq for example. Project is developed by same guys who originally designed AMQP protocol. 0mq was specifically designed to lower latency in trading applications but ended up being just a great library all around.

That said, I think we should focus more on redesigning all the outdated protocols that we use for data transfer (TCP, HTTP, FTP, etc). Even IP protocols are crappy, and no - IPv6 is not the solution.

About that liquidity that high frequency trading provides...

http://www.ritholtz.com/blog/wp-cont....ES_.Z12.2.png

Click for full-size infographic

NYSE gets hand slapped for rigging trades in favor of certain clients:

https://www.miataturbo.net/attachmen...ine=1347901663