25.6 Terabytes in 1 second!?

New data transfer speed record!!!

On 3/28 the record for data transfer speed was broken by Alcatel/Lucent, with a whopping 25.6 Terabytes transferred over a fiber line in 1 second!

That is eqivalent of sending 600 DVD's to someone in one second.

How long, if ever, do you think it will take before this kind of ridiculous speed is offered to the public, and where does this lead us for the future?

Interesting development.

I can't see this ever coming to mainstream use, but it shall be especially useful for industry.

Now all we need is storage devices that can keep up. Such a line will be hopelessly bottlenecked with even the fastest writing hard drives.

Well, when we have mass producable multi terrabyte storage devices,
and create files that large, than I see it becoming mainstream.

I think though this will stay in the confines of science and business though.

I do think though this could be a key advancement in the future of neural
upload transference.

Yup. I recall when the fastest link you could get was an OC12 (620Mbps), and at the time the best drives on the market were Ultra Wide SCSI-3, half the speed (320Mbps).

Storage rate will always lag behind data transfer.

That doesn't mean it won't make a killer backbone, though.

>
btw, it's 25.6 teraBITS per second. big difference, but still damn fast.



[edit on 30-3-2007 by apc]

wow thats fast!

and I believe the conversion comes into around 3.2 Terabytes...and if each dvd movie is 6 gigs...that makes it 533 movies.

It will be a long times before people get this kinda speed at their home...a long time.

currently, the fastest I can download things is 1MB/per sec....which (i think) is till pretty good.

Never say never, I remember when my modem was capable on 28k and that was a wast improvement over my 14k one Now my home cable gets 15mbs, it will take time but its coming.

In two years this will be on the market, but faster: crystal bio chips are the future, lights fast and add a few pipelines. Well PC games will be like reality soon

This is precisely the development needed for the next step in the evolution of the web.

I predict in the not too distant future all CPU processing will be handled externally within huge processing plants.

You will literally pipe in processing power.

No longer will it be a case of this application won't run efficiently or we can't do this calculation because we don't have the processing ability.

Processing power will become a huge commodity in the not too distant future.

People will come to rely on processor power being piped to them and will consider it as important a utility as electricity.

If anyone wants to make a long term investment and make a mint on the stocks and shares would do well to invest in the large data centres that exist around the globe.

For a truly embedded comms future I think a hybrid of this outstanding tech and WiMax will be where it's at in 2020...

All the best,

NeoN HaZe.

These test are all great and fine and so on but far from practical implementation.

I don't know the specifics of this one but often it's a direct line, random data and no computers actually sending and receiving.. it's basicly testing the limits of the fibre optic.

a line like that to your home pc is overkill as your computer can't handle that (the harddisk) you'd be no where close to that speed. e.g. ata133 has a max capacity of 133mb/s, a harddisk can't even fill that up completely.

interesting tests though and surely an indicator of future computing technology


this isn't practical for the line coming from a switchbox to your house anyway.. technology like this is for backbones.

500.000 people downing video's and what not and reading websites, combined need backbones like this.

it's like the major highways of cable-land.. and noone has a house next to one with direct access.


I can't agree on this.. it sounds like a really slow system.

every bit has to be sent to and from the "cpu-center".
I know a lot of peopel that would not want every bit of information about them to be handled that way and possibly snooped at.

Despite how outragious the transfer speeds sound they aren't that outragious if you take into account how many people send and receive data over the backbones. More and more multimedia will go through these same backbones demanding vastly bigger amounts of bandwidth.

HD tv information sent along the same lines instead of the good old cable tv system, people want more quality and speed from webvideo's instead of those 320x240 youtube video's and so on and so on.

As the capacity increases so will demand for quality, which will suck up the additional bandwidth..

that's not to say someday we'll find even better and novel ways of data transmission. but the centralized processing.. not me anyway. I don't want the potential privacy and even security risks that come with that idea.

[edit on 2-4-2007 by David2012]

Nailed it on the head. I don't believe this will, in the near future, become the sort of internet connection offered to anyone BUT I can see it being used in WAN/Internet back bone applications where you have thousands of users. The contention rations will drop so radically that your internet connection would give you a true 1mb/s rate etc.

I would assume that even in the next 5 years it would be way too expensive for anything other than WAN related applications like internet etc.

i recently heard of adding water to nanowires could hold upto 10,000 terrabytes of information heres the link

www.newscientisttech.com...

I can get One terrabyte storage devices? Need more storage? Get 5 of them? Does that count?

This was actually already tried a couple of years ago and was a huge market failure. The failure wasn't due to technology restrictions, either. Instead, people hated the fact they had to pay a subscription to use their own computer, and then additional subscriptions for the applications you wanted to use.

Ive read that this is Google's ultimate goal for this business. Their dream is to have massive computer banks all over the country, and just sell the consumer "dummy terminals" that have the sole purpose of connecting to their mainframes around the area.

Actually, Google is on public record saying they plan to use GoogleOS for a centralized mega computer that people could subscribe to for under $10 USD a month, and get the power of a 2.8 P4, 512MB Ram and 100GB hard drive as performance specs to start. It was brought up in the latest google conference I was forced to attend due to my work-industry. Not that I have any problem with Google mind you, I am simply wishing they would finally reply to my request for 1 million dollars for the Get Adam Retired Foundation.

Just found this website for the first time tonight while bored out of my skull, and obviously suffering from insomnia due to timestamp of msg. I must say, what an awesome website! So much entertainment at my fingertips from a single website is extremely rare to find now a days, with every teenager blogging up the engines with "my life is so rough" garbage. I think I have found my new late night hobby!

Also, I am a Information Security Specialist for an IT Company that primarily deals with enterprise businesses and government sectors. So I rarely wear my trusted life saving fin foil hat. It is really not that unbelievable that or a stretch of the creative mind to say that tera speeds will be out to public sectors within the next 2-3 years. Carrier uplink speeds are already shifting to the 10GigE uplink as the norm. Also, with such technology as 5 Layer multi direction writing hard drives already being out now at 1Tera storage, the technology for speed and massive scale speed hopping is already proven to exist and happen very frequently.

Wish they would just hurry up and bring out the 3TB single drives which are basically already in existence, just not in production yet

Back to roaming for me,
Adam

Yep, that is almost exact I believe haha.


I disagree. Well, naturally it will take a considerably long period between private section to public resident sector. However, look at the speed increases by % over the last 10 years, seems to be exponential almost. Look at the hard drive capacity standards for example. Took about 15 years to break the 1GB, then 2 or so years to break TB. As the need for tech grows, naturally investors dump more money, making quicker research. Plus as well all know, all modern computing technology and infrastructure comes from downed alien crafts.

By the way, you should grab Rogers Elite when they release it if you are able too. If not, grab the equivalent for your area when it comes out. 25mbit up, 10 down, only for $89 ? That's only 20 dollars more per month than their "elite" garbage which is only 8 up, 0.5 down. Even if you get half of the rated speeds (which my own tests have proved will in no way be the case) you would still be getting over 1mb/s



10 true mbit average? If you are paying under 100 a month, that is EXTREMELY good. I believe the average for Canada and US is still under 1.5mbit on the down duplex rating. Though *cough* my internet uplink to work kicks yours in the bum.

Cheers,
Adam

Perhaps I am crazy, but I do not find 25.6 Terabytes in a second impressive AT ALL. I know people with multi-terrabyte computers and even one person with a 9 terrabyte computer (4x4 with 12x Seagate 750)

I would of thought that a world record would be MUCH higher...


Just buy two of these:
www.newegg.com...

However the speeds? I doubt this will become mainstream for another 10 years, and I doubt anyone would really even need the capability then.

[edit on 22-4-2007 by PisTonZOR]

Good call cancerman.

I can see this kind of technology surging to the forefront of things really quick.

Granted there are still many questions about the viability of nanotechnology in society, but I believe that we will overcome these obstacles and push forward with using this nanotechnology for all kinds of things.

Without nano then how are we going to get our space-elevator, right?