Slashdot Mirror
Intel Sees Communications As Company's Next Frontier
WSJdpatton writes "Intel is mounting a long-term campaign to turn personal computers into more reliable tools for calling and conferencing. Intel business-client architecture director Steve Grobman argues that instead of exploiting the Internet to lower communications costs, the next phase is about adding new features. Among the benefits for business: broader access to online meetings with advanced features such as TiVo-style playback, instant captioning of conversations — or even translation into multiple languages. 'That technology could be a foundation for companies to add improvements such as the ability to identify the current speaker during a conference call ... He eventually expects advanced features -- such as automatic transcription or translation of conferences. Intel has used deals to advance its plans. A February 2006 partnership with Skype included joint development to tailor the service for Intel's dual-core chips, and free PC-based conferencing for as many as 10 participants.'"
So, if the newfangled teleconferencing cost $10,000 a minute, people would be all over it, right?
Perhaps it about *both* the features *and* the cost of communications, hmmm?
Insisting on "correct" English is like saying that there is only one, definitive recipe for chili.
I dont understand why it is a new idea - The whole internet is about information exchange - read "Communication". Of course people have been looking at means to improve the way in which we communicate and make it easier to communicate between different people from different regions of the earth and also to make it easer to communicate with larger number of people. People have been trying to do it from the point Internet was started.
"Be the change you wish to see in the world" - M. K. Gandhi
This is a pretty good move on Intel's part. VoIP is really coming into its own, with Asterisk and Skype and others proving it's a truly viable alternative to the traditional telco. You just have to ensure you have an infrastructure in place that can guarantee availability. And it seems Intel want to make creating that infrastructure a whole lot easier. We recently moved to an Asterisk based phone system at work, and true, there were a few teething problems, but it was definitely worth it. We get all the features of the expensive PBXs without the price tag that comes with it. There's also nothing that helps you get to know your phone system more intimately than hand-crafting your configuration from scratch.
http://www.freeswitch.org/ it can do most if not all the features they say are advanced in the conference.
The 'videophone' has a part of the future since 1927 (Metropolis) and has come up in countless visions of the future (ATT exhibit at 1964 World Fair) but for one problem: mass customers just don't seem to want or need it. We had videoconferencing at my last workplace - so you get to see a funky image of the big boss as he speaks, big deal. Might as well let us tele-smell his cologne for all it added to the conference.
try { do() || do_not(); } catch (JediException err) { yoda(err); }
Now tell us, how do you really feel?
moutains and mountains of data to retain. Accessing, querying, storing and getting this data will be the real challenge. Bandwidth is the issue, but will the long pole be transfering bit accross the 'net (or whatever is next) or getting bits two and from the massive data stores that TiVo like capabilities would demand?
The cancel button is your friend. Do not hesitate to use it.
It's always been a huge pain to find Intel drivers, you are about 10,000 times more likely to find the specs for the hardware you're looking for on the site, rather than the drivers. I just hope that their communications isn't the same way, trying to find the manual or some kind of support and all you find is what your hardware (or software) has in it.
The purpose of a network, in whatever form, is simply communication. Anything more detailed than that is losing sight of the purposes of networks and networking.
The purpose of Computers is data manipulation. Anything more detailed than that is losing sight of the purposes of Computers and Computing.
I'm reminded of the failure of the Railroad companies in the dawn of motor vehicles 120 years ago, and again during the dawn of Aircraft 80 years ago, to realize what business they were in. The long running idea of "we're in the railroad business" was extremely short sighted, because they became focused upon the niche of the greater business; transportation.
When Computers connected to the Network, it created a hybrid business, that of Computing Networks. I believe that INTEL has forgotten what business they are in (Computing) because they've lost sight because of the Hybridization of the Computing Network.
In order to accomplish what they have outlined, which is quite admirable, the computing power driving their visiion has to be greatly increased. The Network side is easily expanded, but the computing side is suffering from the constraints of current technology. Intel (and AMD) ought to pay attention and know what market they are really in; computing.
Agent K: A *person* is smart. People are dumb, stupid, panicky animals, and you know it.
Now this may just be me but I remember a few years ago Intel sold off huge amounts of it's communcations divisions because they wern't very profitable.
I bet they're worth a hell of a lot more now, nice move Intel.
"Oh boy"
I will freely admit that I don't use Skype at all. But the internet radio stations I have heard that have interviewed celebrities over the Internet convince me that this way of communicating is not even ready for voice only, much less voice and video. At times it was nigh impossible to understand what the interviewee was saying. And this was people who are supposed to know how Internet communication works (and who had technicians on both ends of the conversation).
Also, considering that the best speech recognition software still can only reach about 95% accuracy (which means that every 20th word would be wrong, 98% is considered the lowest limit for practical use) and the state of automatic translation technology today, we're looking at many years of development before software will be ready to make on the fly translations of conversations over the Internet.
I hate to break it to these "visionaries" but the stuff that humans find most easy to do is in general what computers finds most difficult to do, and vice versa. That's why we get so much use out of them. :)
Badgers, we don't need no stinking badgers! - UHF
This aught to be good:
CEO: Sales are up, things are looking good!
Caption: Dear aunt, let's set so double the killer delete select all
Computerized Russian Voice: , let's delete
Russian repeats back message.
English computerized voice: The dear aunt, us the duel assassin of to erase has left therefore establishes chooses everything
The Cold War 2.0 breaks out.
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
Apparently /. doesn't like Russian :(
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
With VoIP, a lot of resources are dedicated to make two or more endpoints (usually VoIP phones) talk to each other, especially when each device is talking a different codec. A lot of codecs exist - G.729, G.726, GSM, WAV, Speex et al. And so there's stuff in the middle that's required to translate from one codec to another (this is called transcoding), and at the same time take care of other audio quality issues such as echo cancellation, comfort noise generation, DTMF etc. Usually some sort of PABX takes care of this, but at the expense of CPU processing power.
What I'd like to see is for Intel to come up with a specialized chip that is good at the computation and bit-moving required to do these kinds of transcoding and DSP-type functionality. I've heard from someone in the know that when these things are done in software on generic Intel Pentium/Xeon/whatever type chips, they're not that good at doing it (how accurate that is I don't know, maybe its hearsay).
Now if Intel can throw their resources into creating something like this, that would be very nice for the VoIP space.
The Wknd Sessions - Malaysian and South East Asia independent music
As much as I'd love to see One-laptop-per-child get off the ground... they've had their chance. It's been too freaking long! Face it, it's vaporware. Intel's offering is also likely vaporware. It's a fight between one vaporware vs another vaporware!
What makes one vaporware more `noble' than another vaporware? Just 'cause one has `child' in the name... I guess.
I personally would be happy to get a cheap-durable-light-small laptop from any manufacturer. All I need is bash (& all unix utils), vim, perl, gcc. It doesn't even have to have a GUI (just a gnu screen in text mode).
"If anything can go wrong, it will." - Murphy
Steve Grobman argues that instead of exploiting the Internet to lower communications costs, the next phase is about adding new features. No... I would rather lower communications costs. It sounds like exploiting consumers instead of "exploiting the internet."
Are there any programs out there that actually do a reasonable job at translation? I think I've seen some pages on Microsoft's website that claimed to have been translated by software, and those were pretty good translations, but does it take long to translate something in good quality? Do you think it's possible that Microsoft's lying, and their translations are actually human-made, with a few mistakes snuck in? Is such software very patent-encumbered? (A quick Google patents search reveals a fair few patents on machine translation.)
OLPC isn't vaporware, they have been manufactured and some children have them already. Just need to roll out more of them.
Looking back at things like Intel NICs, webcams, and codecs like Indeo, Intel has been fiddling with this stuff for ages now. If they had to fall back on something not based on making fast high-tech chips, this sounds like exactly the kind of thing they would naturally have on the backburner. Back in the day I loved Indeo - it was the only codec that could play a very clear video at 800x600 on my PIII 450.
.. in gaming with Ventrillo and Teamspeak. It boggles the mind that companies are still using horribly painful teleconferencing services and phones to do teleconferencing. If companies just setup well organized and managed Ventrillo services, with an associated chat channel like guild chat in WOW, it would probably revolutionize business communication with almost no new R&D. About the only thing you need is a virtual white board that works integrated with ventrillo or teamspeak for engineering discussions and those are available. For weekly meetings I would take a point and click on a Ventrillo channel any day over punching numbers in to a phone teleconferencing system. Would also take a Ventrillo list of people in the channel over trying to sort out whose on and whose not in a teleconference.
In my opinion the benefits of videoconferencing are overrated. Sure sometimes visual cues help you understand subtleties of meaning but most of the time it really adds very little to the substance of the communication, and instead results in people focusing more on appearance and dress of their coworkers instead of the substance of whats being said.
Ventrillo and Teamspeak are light years ahead in ease of use, ease of management and cost. If you talk the WoW guild paradigm and translated it in to a channel for geographically scattered sales or engineering teams to shoot the shit, brainstorm and generally stay in touch it would probably help most businesses. Only challenge is to keep it from becoming an out of control distraction like a lot of WoW guild chats. It would also be a pretty good performance evaluation tool, its pretty easy to spot idiots in WoW guild chat.
@de_machina
What is old is new again.
s _dialogic/
n tel-sells-dialogic-to-eicon.html
They bought dialogic for USD $750 million in 1999. http://www.theregister.co.uk/1999/06/01/intel_buy
Then they sold in 2006 for an undisclosed sum. The simple fact they sold it suggests they couldn't make it work for them. Otherwise, they would spin it off differently so the operation shows up on Intel's balance sheet: http://blog.tmcnet.com/blog/greg-galitzine/voip/i
The DSP is still where the action is if you are doing infrastructure. I know the modern CPU can handle telco DSP, but as the asterisk docs show, it's very cpu intensive at medium-sized numbers of users.
I give this one a low probability of success and is probably just something to keep the investors convinced they are growning.
http://www.maxineudall.com/2010/02/should-economists-be-sued-for-malpractice.html
When a hardware company preaches to me how my future involves me spending money on their hardware to bring me yet more 'features' I don't want and no one asked me. And the 'features' I do want, specifically, lower power consumption, lower cost are not in their plans at all.
Guess what? I don't want streaming video/TV on my phone. I don't want a massive hard drive to store a directly replacement of my MP3 player. And I REALLY don't want Intel or anyone else 'partnering' with carriers to nickel and dime me to death for every new gewgaw and function on my phone either.
Here's what I want - a solid PDA phone I can synch with Lotus Notes and a few other groupware alternatives. I want a direct interface to the PC w/o having to pay for the carriers picture mail service in order to move jpegs from my phone to my PC. If you're going to cause me switch from a digital camera then make the camera a real replacement with swappable media and a full suite of camera functions. I want a solidly compatible browser for real web browsing at 3G speeds.
To repeat. I don't want a multimedia player. I don't want a sound system. I want a PDA phone. And guess what, Intel? I really do want it cheaper.
Oh man, because those Norwegians really know how to stick it to /.!
They bought Dialogic (a communications company), then sold it off. They bought StoringARM (a communications/mobile chiop), extended that to the PXAxxx XScale architecture, then sold that off.
What Intel needs is a bit of long term thinking.
Engineering is the art of compromise.
Why are they even bothering......they couldn't make decent graphics hardware, so I guess they thought they'd go fail at something else?
Yeah, it doesn't make any difference if you can or can't see the CEO's face during the big teleconference but let me tell you one example were it would make a world of difference:
One month ago my father-in-law died leaving his wife to live alone for the first time in her life. She lives in a somewhat rural part of Japan North of Tokyo in an area with few neighbours. My wife is her only daughter and we live in Canada, her only son is an engineer for Fujitsu and spends many days on the road. We know she misses our kids and we try to get over to Japan as much as possible however more than once every year or two is a stretch.
If she could have a simple device that hooks up to a broadband connection with a small webcam and directional microphone on top of the TV and which can be operated with a couple of buttons on a remote she would feel so much more in contact with us.
Videophone technologies up to now have required knowledge of computer operation and Instant Messaging software or having to go through the complexities of setting up the traditional video conference. Here we are talking about a 76-year old Japanese granny who has never and will never touch anything more complex than the phone or the TV remote.
I'm looking for someone who can build such a device which can be administered remotely, has 6-8 large "quick-dial" buttons and an emergency button which will try to connect through a list of contacts if required.
If you don't want to repeat the past, stop living in it.