Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Wireless Networking Intel Hardware

Intel Putting Wi-Fi into Future Chipsets 216

Posted by CmdrTaco from the obvious-inevitabilities dept.
Ridgelift writes "Wired's got the story on Intel's plan to incorporate Wi-Fi into the motherboard chipset. "The chipset, however, will not include an actual Wi-Fi radio, so users will still need a wireless add-on card. Intel has said it eventually intends build a Wi-Fi radio into its microprocessors." This would make setting up a wireless network a lot simpler."
This discussion has been archived. No new comments can be posted.

Intel Putting Wi-Fi into Future Chipsets More

Intel Putting Wi-Fi into Future Chipsets

Comments Filter:

  • Is wifi on CPU a good idea? (Score:5, Insightful)

    by brejc8 ( 223089 ) * on Sunday November 30, 2003 @11:00AM (#7592535) Homepage Journal
    It's all very well putting more peripherals onto processors but with the shrinking feature sizes and an increase in cross talk is placing a powerful 2.4GHz source on die really a good idea? You would have to shorten long lines and slow down the processor or suffer random errors. Doesn't inspire me with confidence.

    • Re:Is wifi on CPU a good idea? (Score:4, Interesting)

      by IAR80 ( 598046 ) on Sunday November 30, 2003 @11:08AM (#7592572) Homepage
      It is not on the cpu, but the chipset and the RF part is not there. It will just be something like the 802.11 asic dealing with the the protocol not with power up/down conventers or anything like that.
    • As others have mentioned, this is just logic, not the transmitter itself. But even if they do roll out integrated wifi, it'll never be "on die" with the processor. It'll most likely be a motherboard component. If it ever did come as part of the same component as the processor, you can bet there would be a wall of the a faraday cage between the xmitter and the processor.

    • Mobo features (Score:3, Informative)

      by t0ny ( 590331 )
      It will probably be easy for mobo makers to make use of this as an optional feature, much like SATA, USB 2.0, etc. Then you just need to have an antenna lead from the mobo, and enable it in the BIOS: it will work just like parallel/serial ports.
    • Putting WiFi technology on the motherboard seems like the next logical leap, but hopefully they keep the RF transmitter hardware outside the box. Otherwise say goodbye to HiFi soundcard quality.
    • No, putting more peripherals onboard is not a good idea. Moreover, putting wifi onboard is an extremely bad idea.

      There are many reasons but I can give you at least one--noise. I operate a WISP and one of our customers dropped off our network and it took me at least a half hour to figure out what happened. He put his brand new Centrino laptop into scan mode and it knocked his CPE offline due to side channel noise.

      Just what do you think will happen to your 2.4 ghz cordless phone when every apartment abov

  • Like Microsoft... (Score:4, Insightful)

    by Jake73 ( 306340 ) on Sunday November 30, 2003 @11:03AM (#7592548) Homepage
    This is a lot like Microsoft's business plan, but at the hardware level. It leaves ISV's out of the loop. "Centrino" was perhaps the most serious example of this that I've seen.

  • Intel linux support sucks (Score:5, Interesting)

    by pe1rxq ( 141710 ) on Sunday November 30, 2003 @11:03AM (#7592549) Homepage Journal
    They will probably promise to provide Linux driver like with the centrino chipset and then not even make specs available.
    You will get all kind of lame excuses:
    - We are working on a driver.... (For half a year already)
    - We can't tell you how to operate it because the FCC won't let us (Complete bullshit but sounds nice: 'linux hackers want to interfere with police radio')
    - They might release some binary only modules... (Redhat version bla.bla, kernel version bla.bla and nothing else)

    Jeroen

    • Not complete bullshit (Score:3, Insightful)

      by mindstrm ( 20013 )
      The FCC DOES require that ISM band consumer devices opearate within specifications. They are certified for a given antenna configuration, among other things. Even the anntenna connectors are non-standard, for this very reason (so the consumer doesn't think you are supposed to hook it up to an amp).

      It's nto that you aren't allowed to modify it legally, of course you are, as long as you operate within spec... but that the company has to make it so.

      It's not a big stretch for them to feel releasing driver co

      • Re:Not complete bullshit (Score:5, Interesting)

        by timeOday ( 582209 ) on Sunday November 30, 2003 @11:25AM (#7592638)
        It's not a big stretch for them to feel releasing driver code is a risk, as anyone who gets it from them could easily use it to operate outside of spec.
        It's no different than selling cars that can exceed the speed limit. My compliance with FCC regulations is not Intel's concern.
        • They might be able to exceed the speed limit, but how many can go over 250 km/h without being modified? :-)
          • That would be a good analogy if these cards were able to produce just about any frequency and any power.... Thats not true.
            The differences are in the specific channels available for various regions and even then it is a difference of only a few channels at most. The power is also limited by the hardware, you can't just push it way over the limits just by software.
            The amount of interference you could produce is comparable with a car that can go slightly faster then the speed limit, not some suppercharged hot

    • Re:Intel linux support sucks (Score:5, Informative)

      by Homology ( 639438 ) on Sunday November 30, 2003 @11:57AM (#7592783)
      Intel are not forthcoming with documentation about NIC and crypto. Here is what OpenBSD has to say about this
      OpenBSD experience with Intel [openbsd.org] :


      Much like Intel does for all their networking division components, and completely unlike most other vendors, Intel steadfastly refuses to provide us with documentation. We have talked to about five technical people who are involved in the development of those products. They all want us to have documentation. They commend us on what we have done. But their hands are tied by management who does not perceive a benefit to themselves for providing documentation. Forget about Intel. (If you want to buy gigabit ethernet hardware, we recommend anything else... for the same reason: most drivers we have for Intel networking hardware were written without documentation).
      • ...

        Why tell Intel the truth?

        Why not borrow documentation from a Windows software engineer who accidentally photocopies it at work or loses his/her backup copy while he/she is at your house for dinner.

        • because that Windows developer will have signed an NDA, and although losing it isn't a legal problem (unless it can be shown he was truly negligent), once you start using it, you will be charged with IP theft.

          I suppose you could stick it on Kazaa and have done with it though! :)
      • Frankly, this is taken out of context and is rather misleading. Sometimes I wish for a moderation option entitled "Wrong."

        First of all, the section above is listed under the header called "Intel Ipsec Cards" and more specifically refers to the Intel Encryption Coprocessor on the card.

        Further, Intel has written and released a free, GPL ethernet driver for their EEPro 100, 1000, and 10000 ethernet cards. I shall transcribe for your benefit the top few lines from linux/drivers/net/e100/e100_main.c:

        Co

    • I wholeheartedl agree with everything you have just said (see my other post [slashdot.org] to this story.) While we're on the subject, does anyone know anywhere where one can purchase retail miniPCI cards?! They aren't 'supposed' to be available to the general public maybe someone here knows someone who knows someone... that's the answer to all our prayers for those of us chained to the shitty Intel pro/2100 WLAN card...

      • Re:Intel linux support sucks (Score:2, Informative)

        by O ( 90420 )
        Um, I know Dell sells them. There was a 802.11b MiniPCI card from Dell on sale for $29 the other week. Check Techbargains [techbargains.com] for it. Also, if you're desperate, you can generally rip apart a consumer AP or "router" and yank out it's MiniPCI card, too.

      • ...and on a similar note, anyone know where I can find a Gemtek micro-USB (internal) wireless card? I've fried the one which came inside my laptop (bad firmware flash), and I'm having real difficulty locating a replacement. Gemtek themselves don't reply to my emails.

      • I know you said retail, but I bought a lucent orinoco minipci card off Ebay a little while ago. $35 for the minipci card attached to a real pci card - removed two screws, popped it out of the socket and replaced my Intel pro/2100 (piece of crap), and have been happy ever since!

        It uses the Orinoco cardbus drivers, since the orinoco minipci card is essentially a PCI card with a TI 1410 PCI/CardBus bridge and a CardBus orinoco card all in a minipci form factor. Cool stuff.
    • It's true that Intel Linux support sux and it's true that the government (which is full of idiots - think Ashcroft) thinks it shoul be "illegal" and that all large established vendors (with the possible exception of IBM) generally despise Linux (BTW they are going to be **seriously** burned - in the same way the U.S. automobile industry was by the Japanese - by Asian manufacturers of Linux devices and computers), but you are wrong on one point:

      Intel hardware and support sucks in general - not just for Linu

  • Really? (Score:5, Insightful)

    by odyrithm ( 461343 ) on Sunday November 30, 2003 @11:03AM (#7592552)
    This would make setting up a wireless network a lot simpler.

    Woah there, whats so hard with the way its setup now? pcmcia is a matter of plug and go, pci is a matter of modprobe if that.. theres nothing hard about wifi... its a nic with a wireless medium.. thats all.

    Now if intel had some new fangled wep replacement then that would make things simple, no more mac rules on my fw would be nice.. which is unlikely.

    • Re:Really? (Score:3, Interesting)

      by Artifex ( 18308 )

      Woah there, whats so hard with the way its setup now? pcmcia is a matter of plug and go, pci is a matter of modprobe if that.. theres nothing hard about wifi... its a nic with a wireless medium.. thats all.

      You guys are not looking at the big picture, here. Adding this on board allows OEMs to specify a motherboard platform with this, and then buy the cards at cheaper prices than full implementations on cards would cost. It's just following the trend of AMR slots and onboard video, and, in the last coupl

      • Re: integration (Score:5, Interesting)

        by King_TJ ( 85913 ) on Sunday November 30, 2003 @11:58AM (#7592789) Journal
        Personally, I'm not really convinced all these integrated parts do us any favors.

        Just last week, for example, I installed a new Pentium 4 motherboard and CPU in a standard ATX case that was formerly running a PII system. This was done for a law firm, and was upgraded on-site, because they couldn't afford to have much downtime.

        Well, as luck would have it, the integrated EIDE controller was faulty. I kept getting "data corrupt" type messages when it tried to boot Win2K on the drive that just worked in the other system. I tried a different hard drive with a fresh format, and had the same issue. Even the secondary channel had problems.

        If it hadn't been intergated, I could have simply swapped a $15 or $20 controller card and gotten everything back up and running for them.

        The more devices Intel can integrate into motherboards using their chipsets, the more often they get to sell people an entire new board when they only need one small part.

        On-board video has been a disaster since day 1, for both PC and Mac users. What seems "high end" when a machine is new turns into "mediocre" within a year or two. Then come all the conflicts trying to get the on-board video disabled when you add a new, add-in video card. (I'm sure many long-time Mac users can remember the dislike for the "Performa" towers like the 6400/6500, largely due to the on-board video only allowing up to 2MB of video RAM.)

        Integrated NICs may work fine when they work, but again - I've seen many a blown NIC card due to power surges/spikes. I'd rather swap a card and have a fully functional machine again than have a dead port permanently soldered onto the back of my computer....

        • but... (Score:3, Informative)

          by sbma44 ( 694130 )
          you probably *could* have swapped in an ide controller board and just turned off the faulty integrated controller in the bios.

          I agree that we should not sacrifice modularity for all-in-one disposability, but for all the applications you list (IDE, NIC, video) you can put in a modular card and override the integrated stuff. Personally, I think ubiquitous integrated mobo NICs are one of the handiest hardware improvements of the last five years.

          • Well, I did think to try adding an IDE controller card - but the card I added didn't allow booting from CD, which I needed to get Windows 2000 or XP loaded on the box. (The existing install of 2000 refused to start up after the motherboard was swapped, since the incorrect HAL was on the drive.)

            My point is, PC's are designed to be a "box of slots". This concept is what initally made them dominate the marketplace. (If you didn't buy a PC or clone, it used to be, your other choices were systems that gave y
            • Well, I did think to try adding an IDE controller card - but the card I added didn't allow booting from CD

              I think you've answered your own question about why mobo manufacturer's supply everything integrated on board. In most cases you can still disable the integrated feature (like sound graphics, lan, usb) and add your own card.

              So really apart from cost, there is no reason why its a problem. and the cost of mobos nowadays is really low...
        • Integrated NICs may work fine when they work, but again - I've seen many a blown NIC card due to power surges/spikes. I'd rather swap a card and have a fully functional machine again than have a dead port permanently soldered onto the back of my computer....

          In a computer lab where I teach, we had 25 Toshiba Equiums with integrated NICs, and in the two years we had them in that lab, we probably sent back 4 or 5 motherboards with fried integrated NICs. Sometimes we could get them to limp along with an ISA

      • Yeah, this sounds exactly like an AMR slot. Has anyone actually used one of those? Or even seen an AMR modem? If this was actually totally integrated and functional, like a built-in NIC, it might actually be useful.
    • I sure hope you don't think you are really secure by simply enabling mac filtering. At least use WEP too. Even the windoze drivers for most cards allow lusers to easily change their card's mac address. It's trivial to capture valid MAC addresses using tools like kismet.

  • What's the point? (Score:5, Insightful)

    by bobthemuse ( 574400 ) on Sunday November 30, 2003 @11:05AM (#7592559)
    The chipset, however, will not include an actual Wi-Fi radio, so users will still need a wireless add-on card. Intel has said it eventually intends build a Wi-Fi radio into its microprocessors.

    Why put in a chipset without a radio? Maybe one could argue an attempt to get market share by making their add-on card cheaper than the others (just radio, no chipset), but this card will have a more limited market, since it wouldn't be compatible with older or non-intel mobos.

    Now if the were to put a software radio on board, *that* would be cool! Think of upgrading to future standards with just a flash rom upgrade...
    • Maybe to avoid more allegations of Intel being a monopoly? This is kind of like Microsoft embedding an HTML rendering engine in Windows instead of the entire browser, or Intel's own onboard graphics set; it's up to the vendors what to do with it. They could ignore it and provide their own (presumably higher spec) system, (like all those onboard Intel boards with an ATI/Nvidia AGP card fitted). Alternatively, they can use it and either provide a cheap radio with short range, or go for it and bundle a radi
    • Initially it will probably be targeted at OEMs of the Dell and HPaq variety, so compatiability with older boards (or future boards for that matter) will be a non-issue. For these guys, saving a few dollars on parts is a fairly big deal, even if it means that users have a hard time getting replacement parts on their own (in fact, this may be seen as a benefit to the OEMs).

      As for the future, I'd imagine that Intel will try to put as much of a software radio on board as they possibly can. Anything that can

  • users will still need a wireless add-on card (Score:5, Insightful)

    by rokzy ( 687636 ) on Sunday November 30, 2003 @11:05AM (#7592560)
    I don't follow tech too closely but this sounds like those worthless AMR slots or whatever they were called - the ones that were like having a built-in sound card or modem (in the sense that the board cost more and had less space available) except didn't actually do anything...?

  • Dumb Question (Score:2, Funny)

    by Anonymous Coward
    Customer:If I can have a 3 Ghz processor why cant I have WiFi at 3 Ghz
    Radio Shack Assistant: Mam the 3Ghz speed refers to CPU speed while Wi Fi works at 2.4 Ghz
    Customer:Dont get technical on me . I know this Ghz speed keeps increasing all the time . It cant be fixed at 2.4 Ghz
    • Customer:If I can have a 3 Ghz processor why cant I have WiFi at 3 Ghz
      Radio Shack Assistant: Oh, i dont know. I bet it does.
      Customer: Ok. good.
  • Building more functionality into the motherboard is an ongoing trend, but adding a radio cannot be a good thing. Due to potential interferance, you cannot go into a hospital or airplane without being told to turn off your cellphone.

    • Re:Hospitals (Score:3, Interesting)

      by Jarrik ( 728375 )
      I work in a fairly large hospital (200+ beds) and I can tell you from alot of experance that this is no longer the case for hospitals with modern equipment. The hospital I work for just had Nextel come in and install signal multiple repeaters on every floor and building of our entire campus and put a cell tower on top of our main building that houses all patent rooms / OR's with no problems what so ever. There are certain areas where we are advised not to have our Cell phones on but that is due more to re
    • This is not a problem because, if you read the blurb above, you will see that the radio is NOT included. If you want to enable the Wi-Fi features you need to add this.

  • Wouldn't this compromise security? (Score:5, Insightful)

    by twiztidlojik ( 522383 ) <dapplemac.mac@com> on Sunday November 30, 2003 @11:07AM (#7592571) Homepage
    I think that if an admin of a corporate network had several thousand wireless clients, a simple airsnort would compromise thousands of computers, without requiring the packet sniffer actually access any sort of ethernet cabling.

    This could have a grave impact on the sales of wireless-based chipsets in the corporate market.
    • It is already a pain in the ass. Many laptops now ship with built in wireless. So far it has been possible to disconnect the antennas. But it has been impossible in most cases to disable or remove the WiFi completely as the wireless card is also the ethernet and modem card.

  • how convenient (Score:5, Insightful)

    by vinsci ( 537958 ) on Sunday November 30, 2003 @11:08AM (#7592574) Journal
    So when Big Brother [cam.ac.uk] wants to know what you're up to, they don't need you to be online to the public internet. How convenient. The "trusted" BIOS can always let them bypass your firewall, as the BIOS is going to handle the net connection too.

  • Unintentional network compromises through bridging (Score:5, Insightful)

    by AgTiger ( 458268 ) on Sunday November 30, 2003 @11:08AM (#7592575) Homepage
    I sincerely hope they're going to have these things configured in one of the following manners:

    1. Disable the Wi-Fi by default
    2. If not disabled, seed the encryption key with a pseudorandom number before the user specifically configures it.

    You don't want new computers forming unintended bridges or access points between the untrusted network/airspace and your trusted internal network between when they're first powered up and when the overworked sysadmin has a chance to configure them properly. So much for your company's firewalls having a chance to do their job.

    • The network name will probably be set to 'intel' and no encryption key set, just like linksys hardware comes up as 'linksys' with no WEP key. Windows configures new interfaces for DHCP by default. Hence if someone sat outside your office with a WAP (or a pc with a wifi nic acting like a WAP) providing dns service, with the network name set to 'intel' (in my example, anyway) then they could assign IPs to your machines and begin hacking.

      The solution is simple, however; Don't buy the systems with wifi in the

  • Huh? (Score:2, Insightful)

    by Aneurysm9 ( 723000 )
    Wi-Fi without a radio? Sounds kinda like a soundcard without a DAC or a videocard without a RAMDAC. What's the point of including functionality while not including necessary pieces except, perhaps, to rachet up the marketingspeek and pressure out other manufacturers?

  • Should wireless really be in the chipset?? (Score:3, Interesting)

    by Anonymous Coward on Sunday November 30, 2003 @11:09AM (#7592578)
    I think it should be integrated into the motherboard (like regular ethernet) but not the chipset.

    Or is intel upset that sales of centrino are so poor? To be an "official" centrino laptop, you have to offer intel's 802.11b wireless. Not surprisingly, many people want faster (802.11b/g or 802.11a/b/g) wireless cards.

    Broadcom has been eating intel's lunch in the oem ethernet (wireless & wired) card market. Sounds like anticompetitive monopolist activity to me.

  • WiFi Security (Score:4, Interesting)

    by Chalybeous ( 728116 ) <chalybeous@yahoo ... k minus math_god> on Sunday November 30, 2003 @11:10AM (#7592583) Homepage Journal

    What about security issues?
    I like my router (at home, I share a cable internet connection between two desktop PCs and, occasionally, my laptop). It has an inbuilt firewall according to the manufacturer, and I know if I ever do have a serious problem which I suspect originates from the internet, I can physically disconnect it. Sure, cables are archaic, but they're cheaper and more secure than wireless networks - especially for the novice (like myself).
    But if you enable a CPU to act as a wireless hub - or, eventually, if WiFi comes as a full onboard feature (rather like many motherboards now have onboard sound and graphics) - would that not open up your PC and network to security issues? My parents would not be best pleased if someone warchalked on the fence, but since they have little idea of technology or computer security, I think if they bought a new machine enabled with this kind of tech, every l33t hax0r in a two mile radius would be camping out to leech their access.

    Any other thoughts, opinions or predictions?

    • I won't leech their access, I promise. :)
    • But if you enable a CPU to act as a wireless hub - or, eventually, if WiFi comes as a full onboard feature (rather like many motherboards now have onboard sound and graphics) - would that not open up your PC and network to security issues? My parents would not be best pleased if someone warchalked on the fence, but since they have little idea of technology or computer security, I think if they bought a new machine enabled with this kind of tech, every l33t hax0r in a two mile radius would be camping out to
  • I think it's silly how Intel is now embracing a trend to include features already provided by another market. This is the same situation that MS caused by intergrating IE into the system. So now people will quit buying wireless AP's and routers becuase it came with the motherboard.

    Intel will probalby do this witout adding more than 20 employees and in turn drive about 10,000 people out of jobs due to their companies going out of business.

    • Re:Why do companies expand into other est business (Score:5, Insightful)

      by Artifex ( 18308 ) on Sunday November 30, 2003 @11:32AM (#7592669) Journal
      I think it's silly how Intel is now embracing a trend to include features already provided by another market. Intel will probalby do this witout adding more than 20 employees and in turn drive about 10,000 people out of jobs due to their companies going out of business.


      Oh, yes, it's Intel's fault if 20 of their employees can make a product better suited to the market than 10,000 other peoples' efforts.

      By the same token, buses and trains and taxis have all taken jobs away from the good hard-working people with horse-drawn taxis. And we really should go back to rooms full of seamstresses making clothing by hand, like before those evil industrial looms were created. Oh, and the cotton for the clothes (synthetics put farmers out of work) should be picked by immigrant laborers.

      It's not society's obligation to prop up inefficient methods of production; quite the contrary. There's a word us old-timers use sometimes. It's called "progress." Might want to study it.

      • I might be wrong, but I don't think the original poster was trying to argue that companies have no business ever developing products superior to what's already on the market!?

        I think his point was, companies are wise to stick with their strengths, and not to venture into areas that over-extend their reach.

        With the flood of complete wi-fi solutions out there, does it really make sense for Intel to start offering a partial wi-fi solution like this, integrated into their chipsets?

        For starters, many people l

  • There are other problems (Score:5, Informative)

    by plusser ( 685253 ) on Sunday November 30, 2003 @11:17AM (#7592616)
    The reason why Intel probably don't want to integrate the Micoprocessor with the actual WiFi transmitter and receiver is quite simple. If they add it inside the IC, they will have to go through radio use approval for every different potential market in the world, before they sell a single component. By letting the motherboard/add on card manufacturer's do this instead, they can concentrate on developing better microprocessors.

    As somebody in the know, I do worry that these new WiFi enabled equipment could be the next mobile phone when it come to interference of avionic systems; especially as many modern microprocessors are prone to soft faults at altitude due to the effects of the upper atmosphere.

  • To clarify: (Score:2, Informative)

    by HoldmyCauls ( 239328 )
    This is to make it so that an average desktop computer can function as a router for WiFi traffic in the home or office. The card is needed NMW, in order to grab that traffic. A poster above mentioned using a software radio, but it seems that that would only be useful if things were reversed: the software radio *interprets* the signal, and *generates* one to return to the WiFi device in question, but ultimately it is a radio device which transmits that signal into the air. The problem Intel will face is e

  • Bluetooth? (Score:5, Insightful)

    by ybmug ( 237378 ) on Sunday November 30, 2003 @11:27AM (#7592650)
    802.11a/b/g on board would be nice.. but, I would really like to see more motherboards coming with bluetooth onboard. This would seem to make sense with things like keyboard, mice, headsets, and cellphones that are bluetooth enabled coming to market.

  • DRM implementation (Score:3, Informative)

    by 4volt ( 611357 ) on Sunday November 30, 2003 @11:32AM (#7592667) Homepage
    I'm all for WiFi everywhere, but it sounds like a pretty big backdoor to me, I don't think I'd want to have a WiFi connection built onto my board that I couldn't disable with anything mroe then software. Next thing you would know Microsooft is using it to send DRM related information or usage stastics without you knowing.

    I realize that it would probably able to be disabled in BIOS, but it wouldn't take much that if M$ wanted to take control they could do it with a few sentences in the EULA.

    Improbable, but possible.
    • I'm all for WiFi everywhere, but it sounds like a pretty big backdoor to me, I don't think I'd want to have a WiFi connection built onto my board that I couldn't disable with anything mroe then software.

      RTFA-- no, wait-- just read the damn /. blurb, it's only a few lines long. There's no radio transceiver. It's just the chipset. If you don't plug in the add-on part with the radio, you've got nothing to worry about.

      Mods, stop marking knee-jerk shrieks of "all intel mobos will have a wireless backdoor!" as

  • Easier? (Score:5, Insightful)

    by toupsie ( 88295 ) on Sunday November 30, 2003 @11:33AM (#7592670) Homepage
    This would make setting up a wireless network a lot simpler."

    How freaking simple can it be now?
    1) Insert Airport Extreme Card into PowerBook 12"
    2) Turn on PowerBook 12" 3) Select Network from Airport menu & Enter WEP if needed
    4) Wirelessly communicat

  • call me paranoid (Score:3, Insightful)

    by ciaran_o_riordan ( 662132 ) on Sunday November 30, 2003 @11:35AM (#7592675) Homepage
    I'm no Wi-Fi expert, but couldn't a wi-fi-enabled cpu transmit data without your permission?

    Unique cpu ids? Treacherous Computing Group data?
    • why wouldn't any wifi enabled anything be able to do it?

      it's just integrating, saving costs in manufacturing(and needing to add new features to be better than the old product and really i don't think this is that big of a deal to incorporate)..
  • If every cpu transmits its presence to the world ( and unique MAC address ) it will make monitoring ones whereabouts that much easier..

    Oh, and make the spectrum a total mess with all that noise....

    ( and just for the record, they were talking about this a year ago, but to discuss it you had to have an NDA )

  • Monopolistic BS (Score:3, Informative)

    by drix ( 4602 ) on Sunday November 30, 2003 @11:51AM (#7592737) Homepage
    This just is another step in Intel's ploy to rule the wireless market through cheap and underhanded business practices. Not many people know this, or at least I didn't till I started shopping for a laptop 2 days ago, but all new laptops carrying the Centrino designation have to come with an Intel miniPCI WLAN card preinstalled [powernotebooks.com] or they cannot be called Centrino. Which is great except that Intel refuses to support Linux [petitiononline.com] on their stinkin' card. (Yes I could go [transmeta.com] elsewhere [amd.com], but for the price, speed, and power consumption, Centrino is far and away the best on the market right now.) If you want to monopolize an entire hardware sector, fine--good luck trying. But don't chain me to a stupid Wintel platform because of it. If Intel had their way they'd be the only supplier of WiFi cards within a few short years--then WTF do we do if we're not on Windows?
    • I totally disagree that Centrino is the 'best' mobile chipset. Transmeta's offerings are way too slow for my taste, AMD CPUs draw gobs of power, and Centrino is 'black boxed' to the Linux community.

      The 'best' mobile CPU in my opinion is available from Apple, as either a PPC750 or a PPC74xx. You really can't beat the iBook line in terms of price/performance/quality. Sure, the clock speed is a bit low, but even the G3 has tons of horsepower. A 900MHz iBook running Linux feels about as fast to me as a 2GHz Ce
      • Your perception is seriously off. First, there is no 2GHz Centrino. The fastest one is 1.7Ghz. Second, a Centrino is *extremely* fast clock-for-clock. Its even faster per-clock than a PIII. A 1.7GHz Centrino is about equal to a 2.6 GHz P4, or a hypothetical 1.9 GHz PIII.

        "tons" is not a precise metric of computing power. Relatively, a G3 has very little horsepower. Its got a pretty crappy FPU, and isn't that much faster clock-for-clock than a PIII anyway. A 900Mhz iBook is probably comparable to a 900MHz or
        • I think a G3 has more horsepower per-cycle than a PIII, I seem to compile things in less time on my G3/450 than on my PIII/600. Of course this is pretty subjective, but most other things seem faster on the G3 too.

          The G3 has much better integer performance than a PIII of similar clocking, and FPU scores are neck-and-neck. I don't know about you, but I tend to make use of the integer units about 100 times more often than the FPU units.

          I think our 'feel' for instructions-per-clock has been whacked by the P4,
          • The CPU in Centrino machines is a Pentium M, which is basically a souped-up Pentium III (includes SSE2 instructions, for example). It is not in the same family as the Pentium 4. I got a Centrino laptop for my dad (Dell Inspiron 500m), and while I was setting it up for him, netbooted NetBSD and ran some benchmarks and compiles to get a feel for the performance. I was impressed by the speed and battery life; it's a bit faster than the PIII per clock, but has much lower power consumption. I had expected it to

    • I agree with your sentiment. My question, and I'm not an experty by any means on WiFi, is this. If a new standard comes out, say 802.11musthave, do I suddenly have to buy a new motherboard (+ CPU/Memory/etc) to use that? It'd be awfully convenient for Intel now, wouldn't it.....

      I think that WiFi has gone through more upgrades lately that processor architectures, and perhaps Intel is looking for another upgrade gravy train.

  • Ack (Score:4, Interesting)

    by FrostedWheat ( 172733 ) on Sunday November 30, 2003 @11:52AM (#7592746)
    But why!

    Onboard soundcards (chips?) are rubbish, onboard NICs are quite often crap (not always), onboard modems are a joke and onboard video is nasty. Apart from some specific cases (VIA's mini-itx stuff) I think manufacturers should be moving away from this onboard-everything obsession.

    PCI was invented for a reason! Customisability is what set the PC apart from the Amiga or similar machines!

    • Re:Ack (Score:3, Informative)

      by antiMStroll ( 664213 )
      We use on-board ACL 650 sound for commercial use and it's fine, much better than being forced to load Creative's nasty software (circa Audigy 1, haven't tried the 2.) The motherboards we select use Intel integrated NICs, vastly superior to the typical Realtek-level junk found in so many machines. Modems, you have a point, integrated video however has its uses. The desktop demands in a corporate environment are typically meagre. Integrated handles the task fine and has the added benefits of reduced noise, he
      • That realtek "junk" has a more versatile driver than pretty much anything else I've seen...Case in point:

        At a school (TSTC) we were given a single ethernet drop. Problem: It's keyed to a Widows 98 machine's MAC address that doesn't exist with us. We know the MAC that needs to be cloned.

        My laptop's Broadcom chipset can't do it.
        Another laptop's Intel chipset can't do it.
        A third laptop's 3Com can't do it. Neither could his Linksys.

        A fourth person's crappy Realtek chipset NIC was the only one that could do M
    • Actually, that's not really true. The audio and NIC chips on NVIDIA's nForce is pretty good, and its not like I need anything more than the GeForce-class graphics processor on the nForce to run my xTerms :)
    • Think of this as of IrDA interface. Most of the chipset is on-board, and what is missing is one small chip, one resistor, some wire and optionally a box and a couple of capacitors. Nobody seems to be complaining though.

  • security (Score:4, Interesting)

    by spoonist ( 32012 ) on Sunday November 30, 2003 @12:07PM (#7592831) Journal

    I, for one, do not like this trend of integrating wireless into everything.

    As a security conscious individual, I want to be able to physically choose whether or not I want wireless when I want wireless.

    I like to be able to physically pull out the wireless card in my laptop because then I know I can't be h4x0r3d via my WLAN card.

    Fine. Call me paranoid. I don't mind.

    (Yeah, I know they said the RF part would still be an add-on... I'm just talking in general that I want add-ons and not fully integrated wireless stuff that I can't pull out without desoldering chips.)

  • Dear God, would you people please think? (Score:3, Interesting)

    by gmhowell ( 26755 ) <gmhowell@gmail.com> on Sunday November 30, 2003 @12:17PM (#7592885) Homepage Journal
    Would you people please think before posting?

    This includes a chipset, not a radio. Therefore, it won't be sending out your world control schemes to everyone in existence. Yes, Intel will at some point in the nebulous future include a radio. As will many manufacturers. At that point, we go to the next paragraph:

    Every integrated soundcard/videocard/ethernet controller/serial port/etc. I've ever seen has a setting in the BIOS. If you don't want the location of your laser embedded sharks known to the black helicopter people, switch it off.

    Finally, when is the last time your built in ethernet card just randomly spewed data out the port to the CIA? Oh, last week? Then you have more problems than just a wireless AP built into your motherboard.

  • This would make setting up a wireless network a lot simpler."

    As in "The cops are at the doors! This is a set-up!" ?

    Say, I want a box that's remotely uncrackable. Nothing simpler, remove all network cards. But Intel is known from claiming its CPU 'features' can't be re-enabled without reboot, while they can. So I have a potentially harmful piece of hardware in my box...
  • just add the radio card to any os9/X than can (slot loading iMacs on up, iBooks on up) and press 'share'
    do all the protective stuff of course...
  • I don't get it.

    I'm happy with my hardwired LAN. Why do I need to think about wireless hardware being in my setup by default? I'm assuming that in this situation I'd have to do something to disable it so I can plug in a good old fashined NIC card. Right?

    This is troubling. OK, maybe putting it on the MOBO isn't so bad, but on the CPU? Why? This is good in what way?

    wbs.
  • You can talk to someone in Bangladesh but can't get the football game commentary on WIP? What's up with that!

  • The real problem is routing (Score:3, Informative)

    by Baldrson ( 78598 ) on Sunday November 30, 2003 @01:05PM (#7593096) Homepage Journal
    Having set up a municipal mesh network it is pretty obvious the real problem to solve for WiFi isn't higher levels of integration so much as routing.

    Internet protocols were designed around wires and it shows when you go to wifi meshes. Meshes are critical due to the fact that meshes scale. If you are going to have a wifi node in every consumer device, as seems potentially viable, then you need to continually discover new routes and do so on nearly every packet. Route-flap is what you get, even with damping protocols, with current internet standards. You can end up waiting minutes for a route to stablize.

    Here's an algorithm for a mesh node that seems to work simply:

    Just keep a table of destination IP addresses in memory with a counter that decays exponentially with time.

    When the counter decays below some threshold, clip its IP address from the list. An IP address with no counter is considered to have a value of 0.

    Every time a packet acknowledgement comes through for a given destination IP address, add one to the counter for that IP address.

    Whenever a packet (not already awaiting acknowledgement) is 'heard' destined for an IP address, queue it for rebroadcast according to a priority established by the IP address's counter.

    Let packets that fall off the end of the queue due to low priority do so without further consideration.

    More complex algorithms are required for transmission power optimization, but even this simple algorithm shows how far off-base current internet protocols are for wifi.

    • In IPV4, you can't allocate space for WiFi meshes anymore. ACK.

      But this is not true for IPV6. In IPV6 you can allocate big classless spaces and use these for your mesh networks. A 10 million city with one IP per PC, washing mashine, toaster... no problem.

      What you describe at the most important problem in todays IP world is not the basic protocol, i.e. IP. You are criticizing the routing protocols, they do not match well with fluctuating mesh networks. Indeed. They are designed for wired networks. But most
      • But this is not true for IPV6. In IPV6 you can allocate big classless spaces and use these for your mesh networks. A 10 million city with one IP per PC, washing mashine, toaster... no problem.

        Well, yeah, and I did try to get Jon Anderson to include IPv6 in his first major release of LocustWorld's MeshAP for precisely that reason. To no avail however. He thinks the ability to set up private subnets, local to the mesh, is adequate to the problem at hand. He may be right if the problem at hand is simply

        • Then... if you know what your problems are (i.e. no IPV6 in MeshAP) and you need it, why don't you fix it?
          I mean... it's certainly harder to get a new routing algorithm (i.e. a considerable piece of rather complex software) working than change the packet type from IPV4 to IPV6...
          • ? I don't follow.

            I said I thought IPv6 was a good idea, not that it was an alternate solution to the routing problem. The problems I'm running into now aren't lack of IP addresses, it is route flap.

  • Excuse me, but I thought this is what Centrino was. Can someone explain this to me?
    • First and foremost, "Centrino" is a 100% pure marketing term, nothing at all new or technical behind it. All Centrino means is that you havea laptop that has an Intel Pentium-M processor with an Intel motherboard chipset and an Intel wireless chip. If you use a Pentium-M processor and Intel motherboard chipset but a non-Intel wireless chipset (eg one of the Broadcom chips that include 802.11g support which is lacking in the Intel chip), you can't call your laptop a "Centrino" laptop and more importantly,

  • Why is this on Slashdot? (Score:2, Interesting)

    by jwr ( 20994 )
    Why do I read about this on Slashdot, instead of an article about Intel declining to provide drivers or specifications for their wireless part of Centrino?
  • To respond to those who are making the slippery slope contention, I respond, so what. The slippery slope argument seems to be...

    "what if in the near future, are new computers boot with an active wifi ap built in!"

    That's obviously not dealing with the issue at hand, which is a wifi chipset, but without the necessary hardware built in, but it seems to be popular, so to address that...

    There seems to be at least one fairly simple solution to this. So incredibly simple that I feel silly proposing it, but,

    W

Slashdot Top Deals

Life is a whim of several billion cells to be you for a while.

Close