Slashdot Mirror


Network Card for Gamers - Uses Linux to Reduce Lag

Cujo writes "The folks at GDHardare have an interview with Bigfoot Networks discussing the pending release of their Killer Network Card which is said to greatly reduce in-game latency. According to the Interview, this card uses a Linux-based subsystem to do its magic."

410 comments

  1. Is it credible? by Whiney+Mac+Fanboy · · Score: 4, Funny
    From TFA:

    All the founders met at the University of Texas while getting their MBAs.

    Oh - and it runs FNapps, so as well as being good for games, its suitable for FNapping.
    --
    There are shills on slashdot. Apparently, I'm one of them.
    1. Re:Is it credible? by __aaclcg7560 · · Score: 4, Funny

      You mean they went to the University of Texas to find Linux instead of God?! I guess that answers the root question then.

    2. Re:Is it credible? by AuMatar · · Score: 1

      THe root questions? That would be "Password:" right?

      --
      I still have more fans than freaks. WTF is wrong with you people?
    3. Re:Is it credible? by snowgirl · · Score: 5, Insightful

      It's obvious that they're all about the business here.

      "Powered by Lag and Latency Reduction (LLR) Technology"
      "Future-Proof: Field Upgradeable"
      "UltimatePing(tm)"
      "MaxFPS(tm)"
      "FNA(tm)"
      "GameFirst(tm)"
      "PingThrottle(tm)"

      Seriously, who else but a marketting department would think that it's a good idea to trademark a name describing everything "new" that your product does? And the page is so full of TLAs (three letter acronyms) that you need a glossary to read it.

      So, yes, I'd have to weigh in with everyone else, it's snakeoil. Basically, any product designed entirely by a marketting group is going to be snakeoil, and this definitely was.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    4. Re:Is it credible? by Irish_Samurai · · Score: 1

      Seriously, who else but a marketting department would think that it's a good idea to trademark a name describing everything "new" that your product does?

      Three MBA's who only took a basic marketing class.

    5. Re:Is it credible? by DoofusOfDeath · · Score: 1

      Greasy marketing doesn't necessarily imply an inferior product. It could be that three college comp. eng. buddies developed the card, and the only marketing talent they could afford was that guy Jeff who used to hit on everyone while drunk at a frat party.

      The nice thing about network card performance claims is that they're pretty straight-forward to emperically test, once the hardware is out. I expect Toms Hardware will do a good job testing its claims, like they did with the recently released physics accelerator cards.

    6. Re:Is it credible? by lullabud · · Score: 2, Interesting

      I like the past tense tone of your post. "This definitely was." That's has a nice finality to it. Screw hype.

    7. Re:Is it credible? by tigga · · Score: 2, Insightful
      Greasy marketing doesn't necessarily imply an inferior product. It could be that three college comp. eng. buddies developed the card, and the only marketing talent they could afford was that guy Jeff who used to hit on everyone while drunk at a frat party. The nice thing about network card performance claims is that they're pretty straight-forward to emperically test, once the hardware is out. I expect Toms Hardware will do a good job testing its claims, like they did with the recently released physics accelerator cards.

      The problem is they describe NIC in snake-oil terms. They try to bullshit people.

      The funny part majority of gamers should not care about NICs. Just anything should work if Internet connection at 1-6 Mb/s.

    8. Re:Is it credible? by Godji · · Score: 1

      Well, it sorta sounds like a Razer product page...

    9. Re:Is it credible? by SoloFlyer2 · · Score: 5, Insightful

      While the network card Might improve the network connectivity for that one person, it more than likely does it by hurting the performance of the the network as a whole...

      IE see the ALOHA protocol...
      http://en.wikipedia.org/wiki/Aloha_protocol#The_AL OHA_protocol

      Changing the way that the network card responded so that rather than using a random backoff time it just retried staight away would decrease the network latency but if there were several of the cards on the network all trying to do the same thing the network would grind to a halt...

      There are alot of protocols with backoffs etc which could be changed to improve the individuals connection but making them faster would be a BAD THING for the network as a whole!

      but, there are also other things which could be done to improve network connectivity without hurting the network as a whole... just setting up some kind of QoS on the network card so that small packets like ping response would always get pushed to the front of the queue would improve the ping time signifigantly under non-idle situations...

      --
      "I reject your reality, and substitute my own" - Adam Savage
    10. Re:Is it credible? by Anonymous Coward · · Score: 0

      Hang on... ...isn't it a TCP/IP offload engine with marketing gumff added?

      Thus, it will actually "add" to your processing power by offloading the decoding of the IP stack to the card and probably DMA it to the CPU?

      Ok, so it's marketing bollox on top, but there is some substance here.

      (Or is it hip to diss it 'cause it's got too much marketing on top - as if most gamers actually understand the technologies they use?)

    11. Re:Is it credible? by snowgirl · · Score: 1

      (Or is it hip to diss it 'cause it's got too much marketing on top - as if most gamers actually understand the technologies they use?)

      I wouldn't call it "hip" to "diss" it just because it has so much marketing. Rather, I prefer to specifically diss the marketing itself. Who knows, you might see an increase in performance with this thing, everyone said the web accelerator of "up to 5x faster" was a bunch of snakeoil, and well... well, actually that was kind of icky, just using a Python script to gzip things as they pass over the phoneline using a proxy server (yes, that is exactly how all of the ISPs do it, someone would have to have a patent on it, so only their implemention is legal).

      So, yes, it's possible there might be a performance gain, but as people said before, modern CPUs aren't really spending that much time on TCP/IP anyways, and they said they weren't even targetting TCP, but rather UDP in TFA.

      So, could it be real? Maybe, but then snakeoil remedies did help some people with at least depression... well, at least until they got addicted to the cocaine in it. Just as it's wrong to assume a conclusion is wrong just because there's a fallacy in the argument, it's wrong to assume that a product won't work because it's covered so high in market-speak, that you can't see even an inkling of what the product is supposed to actually do.

      I will personally remain suspect of this though; although you my certainly go ahead--if you want--and spend the $250+ to get one; just let me know how well it works... preferably with emperically testable data, not personal testamony.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    12. Re:Is it credible? by Andy+Dodd · · Score: 2, Informative

      Most such "abuses" become useless/irrelevant with any situation these cards would be put into. i.e. in most cases there will be one of the following on the other end of the cable:

      1) A cable/DSL modem with an ethernet bridge. i.e. the network is only being used as a point-to-point link
      2) A switch - no one uses hubs any more. Again, since CSMA/CD is effectively never used in this situation, tweaking its behavior does nothing.

      Unrelated to your post
      3) Most games use UDP - Almost every "network accelerator card" I've seen was designed to offload the complexity of TCP. UDP and IP are incredibly simple and there is little to no benefit whatsoever to trying to accelerate them.

      --
      retrorocket.o not found, launch anyway?
    13. Re:Is it credible? by __michikal · · Score: 0

      "Effin' A, Cot, effin' a."

    14. Re:Is it credible? by Anonymous Coward · · Score: 0

      I suspect the card behaves normally on the network side. It sounds to me like it's acting as a network coprocessor, so it handles breaking long packets into shorter pieces, reassembly of fragments, etc. The game only has to toss data at the card and pick up big data packets, so the CPU running the game can focus on the game and not have to be interrupted every few milliseconds to do trivial network tasks.

    15. Re:Is it credible? by arth1 · · Score: 1
      I suspect the card behaves normally on the network side. It sounds to me like it's acting as a network coprocessor, so it handles breaking long packets into shorter pieces, reassembly of fragments, etc. The game only has to toss data at the card and pick up big data packets, so the CPU running the game can focus on the game and not have to be interrupted every few milliseconds to do trivial network tasks.


      And how is this different from existing cards that do the same, like the good old 3c990 series (Typhoon) from 3com? Except that the 3com cards also works with Linux and Novell, and cost a lot less, that is?

      Regards,
      --
      *Art
    16. Re:Is it credible? by tehcyder · · Score: 1
      TLAs (three letter acronyms)
      Thanks for the explanation.
      --
      To have a right to do a thing is not at all the same as to be right in doing it
  2. Yes. by Spazntwich · · Score: 5, Insightful

    I'm sure another layer of abstraction to the network is exactly what gamers need to reduce lag.

    Overloaded and slow routers will say, "Whoah, his network card RUNS LINUX. I'll shuffle these packets through more quickly."

    I'd believe their hype more if we already had an openly tiered internet and these guys gave you a free year's subcription to the top tier with purchase of the card.

    1. Re:Yes. by Beuno · · Score: 4, Informative
      I agree, although, also from TFA:

      Many network products today claim to 'offload' network calculations (like checksum, tcp segmentation, etc.). Those technologies are usually only for TCP/IP networking (which most games that Hardcore Gamers play don't use). Those technologies are also incomplete as they still go through multiple layers of the gaming network stack to eventually get data to the game. With Killer, we completely bypass your gaming PC's operating system and go directly from our card to the game. Our card automatically handles things like IP Reassembly, UDP/IP checksum, UDP and IP header verification and stripping, etc, etc, etc. By bypassing your gaming PC's operating system and allowing Killer to handle everything, Killer can achieve levels of gaming network performance well beyond the offloading features claimed by other consumer networking products (NICs or onboard chipsets).
    2. Re:Yes. by jandrese · · Score: 2, Insightful

      Woo! It saves that 0.1% of my modern CPU that is going for UDP checksum calculations and uh, well UDP doesn't exactly require a lot of processing...

      --

      I read the internet for the articles.
    3. Re:Yes. by tomstdenis · · Score: 4, Insightful

      TCP isn't avoided because it's slow but because it's totally useless for streaming applications (e.g. games). Missing packets is much more easy to deal with than halting waiting for missing packets.

      Maybe that shows the founders don't know that much about networking?

      Tom

      --
      Someday, I'll have a real sig.
    4. Re:Yes. by CyberBill · · Score: 5, Insightful

      A lot of games are TCP based... World of Warcraft for example. Even games that arent exactly TCP are typically a reliable messaging system on top of UDP that pretty much mimics TCP.

      With that said, I cant see how this network card could reduce your latency by more than 1ms or 2ms round trip. Latency isnt introduced because your PC is stupid, its introduced because you're waiting the time it takes for packets to travel to your ISP, to its ISP, to its ISP, down to its child, down to its child, and back to some other PC, and having to interact with the 20 routers, gateways, and switches along the way. Most switches use something called Hold and Forward (I might have the name wrong...) which listens for the whole packet, reads the header information, and then passes it along, rather than writing the bits as they come in like a hub does... (Please dont read into this and think hubs are better :P )

      --
      -Bill
    5. Re:Yes. by Zebra_X · · Score: 4, Insightful

      WOW. Just wow. I think that I have seen it all. This fellow has actually posted a smart, witty and insightful comment about a totally bunk product and got modded troll as a result.

      And yet others defend this weak, limp wristed marketing gimmick and have been modded up.
      Is there no justice on slashdot!? Have the Mod gods forsaken us for the last time!?

      We pray to you mod gods, remove the blight from the parent post and restore the balance of good and newb on slashdot!

    6. Re:Yes. by russ1337 · · Score: 5, Interesting

      ....and probably bypasses any software firewall on your machine at the same time... how long till there is an exploit to get the network card to trigger 'gaming mode' for a worm... my bet is 2 days from release...

    7. Re:Yes. by Jherek+Carnelian · · Score: 5, Informative
      Even games that arent exactly TCP are typically a reliable messaging system on top of UDP that pretty much mimics TCP.

      With that said, I cant see how this network card could reduce your latency by more than 1ms or 2ms round trip.
      Given the constraints - TCP and various homegrown reliable protocols on top of UDP, it isn't too hard to come up with some options to improve latency. But they all involve violating the RFCs.

      First you have to wrap head around one important factor that can absolutely kill latency for any transport with guaranteed delivery -- packet loss. Packet loss means you have to discover what packets were lost and then retransmit them - those two steps can easily introduce delays on the order of seconds.

      So one trick would be to pre-send the retransmits. Send duplicate packets spaced apart by a few miliseconds. If the other end receives multiple copies of the same packet, it will silently discard any extras - but if one copy gets lost en route, the other packet might still make it through, thus eliminating the whole timeout/retransmit cycle. It should be possible to do this for both TCP and UDP.

      However, doing something like that is very unfriendly because it wastes resources. The primary reason packets get lost en route is because of bandwidth saturation. So, if you double or triple your traffic you are just making the problem worse. If you are the only one out of thousands who "breaks the rules" you will probably get away with it and probably even benefit from it since packet loss will be a somewhat even distribution among all traffic, so chances are if one of your packets gets dropped the copy won't get dropped - instead someone else's packet gets dropped.

      But if a significant minority of users were to do the same thing, it would probably result in a complete collapse of any usuable bandwidth. Which is exactly the kind of thing I would expect a bunch of MBA's to come up with.
    8. Re:Yes. by snowgirl · · Score: 1

      However, doing something like that is very unfriendly because it wastes resources. The primary reason packets get lost en route is because of bandwidth saturation. So, if you double or triple your traffic you are just making the problem worse. If you are the only one out of thousands who "breaks the rules" you will probably get away with it and probably even benefit from it since packet loss will be a somewhat even distribution among all traffic, so chances are if one of your packets gets dropped the copy won't get dropped - instead someone else's packet gets dropped.

      Redundant Transmission of All Packets? RTAP(tm), it's mine, they can't use it!

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    9. Re:Yes. by vadim_t · · Score: 1

      You don't have to retransmit. Many things can coded in such a way that any lost data can be ignored without ill effects.

      For example, suppose a game where you control a character. The user is running forward. Packets are sent with the content of "Player Alice moving to (x,y)", at intervals, whether the user is actually moving or not. You have these packets:

      #1: Alice moving to (0,10)
      #2: Alice moving to (0,20)
      #3: Alice moving to (0,30)
      #4: Alice moving to (0,30)
      #5: Alice moving to (0,30) ...

      The receiving side knows at which speed Alice runs, so it interpolates to find what happens in the middle. If packet #2 is lost, the program assumes Alice is still running in the same direction. #3 arrives, and it's quite possible the user won't notice anything went wrong. If enough packets get lost, Alice will be seen running forward for a while, then run back to the correct position, or suddenly teleport to where she should have stopped.

    10. Re:Yes. by jank1887 · · Score: 1
      hmmmm... does anything out there do anything like this? Would make for an interesting LAN protocol. Especially in a lossy wireless network that doesn't use the full bandwidth. When I'm connected in some 'poor coverage' areas of my home, it still connects at 6-11(?)Mbps, but a ping to the wireless router can have anywhere from 5-20% packet loss (intermittant). My DSL can't fully utilize 6-11Mbps, but avoiding retransmission (I think) by pre-empting would reduce disconnects in some online games, which have stated that their implementation handles packet loss very badly. (for the sake of preserving my already tenuous reputation, I won't name the game here :) )

      obviously from the way I've stated the above, this could all be crap. Please point out why if that's the case.

    11. Re:Yes. by mrbcs · · Score: 5, Insightful
      IMNSHO, most latency is caused by "CRAPPY $5 onboard NETWORK CARDS!"

      Get a 3com NIC and be done with it. I've seen a machine getting pings of 200 to 300 and when I switched the crap card to a real 3com the ping times went down to 50.There is a reason that 3com charges $35 for their cards... they work well. I use them exclusively in my network.

      Shitty hardware always causes problems.

      --
      I'm not anti-social, I'm anti-idiot.
    12. Re:Yes. by Lord+Kano · · Score: 1

      Games aren't really streaming media. Straming infers one-way traffic. Gaming is most certainly two way. Your client needs to tell the server what it's doing and the server needs to tell your client what has happened with everything other in-game entity.

      LK

      --
      "Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
    13. Re:Yes. by dyamkovoy · · Score: 1

      This sounds exactly like what good ol' Diablo did. Players and NPCs jumped around like that all the time.

    14. Re:Yes. by D3viL · · Score: 2, Funny

      Unfortunatly in addition to packets not getting there (packet loss) you also get packets getting there out of order so what you'd get is:

      #1: Alice moving to (0,30)
      #2: Alice moving to (0,20)
      #3: Alice moving to (0,30)
      #4: Alice moving to (0,10)
      #5: Alice moving to (0,30) ...
      "Look Alice is haveing a seizure"

    15. Re:Yes. by tomstdenis · · Score: 1

      Yes, but it's still just streaming. It isn't like you are downloading the data, working it, then uploading it.

      You're sending your telemetry and you're receiving theirs.

      Tom

      --
      Someday, I'll have a real sig.
    16. Re:Yes. by Lord+Kano · · Score: 1

      It isn't like you are downloading the data, working it, then uploading it.

      I respectfully disagree. The data that you send to the server is influenced by the data that you receive. If you get killed in an online game, there is no need for you to send data that you've taken a shot.

      LK

      --
      "Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
    17. Re:Yes. by flithm · · Score: 1

      As already pointed out many, many... many, many, games use TCP.

      Transmission control protocol is itself designed specifically for stream based transmissions. If you want a stream transmission, generally speaking, you use TCP, not UDP.

      But it sounds like you probably meant streaming, as in "streaming" video, that kind of thing? It all depends on the situation.

      Also missing packets isn't easier to deal with than waiting for missed packets at all. In fact it's a real PITA. But it does cause delay.

      And that being said, you're right... the founders likely don't know much about networking, and are counting on the fact that most other people don't either.

      It's pretty obvious that the only reason it would run Linux is if it does QoS on the packets... so maybe this card will let you play games and leave your downloads running at the same time. If so that might actually make it worth while.

      But it's not really possible for the device to lower ping times on an unsaturated connection.

    18. Re:Yes. by AnyoneEB · · Score: 1

      Out of order packets are easily noticed by either (1) numbering packets in sequence (TCP does this) or (2) timestamping packets at a higher level (if you got packets out of order, then you must have been using UDP).

      --
      Centralization breaks the internet.
    19. Re:Yes. by Anonymous Coward · · Score: 0

      Totally useless? Then why does YouTube, Google Video, Microsoft Streaming Media, and the majority of games use TCP for streaming then?
      The notion that TCP is unsuitable for streaming is an outdated notion.

    20. Re:Yes. by tomstdenis · · Score: 1

      Yes, but that's not how the games work in practice. Suppose we have

      Machine A [you]: Shoot gun
      Machine B [me]: Shoot and kill you

      And both happen at the same time. What will happen?

      Video games that host many clients use a post/commit process for actions. You posted that you shot but because the games master state (and the state others have) doesn't agree (you're dead afterall) the action is not commited to the game state. This is how, for instance, you can seemingly "duck" but still get plucked off [happens to me all the time]. In the games master state your "duck" action has not been comitted yet. even though in your copy of the game you have. Therefore, when someone shoots you, as far as others are aware you're not ducking yet.

      But it isn't as if the client updates it's telemetry based on what it receives other than limiting your actions (e.g. hitting walls, dying, etc). So in reality you're dealing with two one-way channels [think dual-mono] of relatively independent data.

      Tom

      --
      Someday, I'll have a real sig.
    21. Re:Yes. by tomstdenis · · Score: 1

      Most of those are over TCP because they use HTTP which is defined as using TCP in general.

      Video streaming [and audio] is a task better left to UDP as it avoids all of the retransmissions and other crap. Unless you need a copy of the media, UDP is the proper way to go since lost packets are tolerable.

      Tom

      --
      Someday, I'll have a real sig.
    22. Re:Yes. by Jherek+Carnelian · · Score: 1

      You don't have to retransmit. Many things can coded in such a way that any lost data can be ignored without ill effects.

      Yes, you are correct for some types of games. However, you have ignored the predicated constraint of a reliable protocol.

      A network card can't do anything about the way the software is designed, but it can fart around with the network protocols itself.

    23. Re:Yes. by Schraegstrichpunkt · · Score: 3, Insightful

      My guess is that the drivers are broken. I can't imagine anyone shipping an Ethernet NIC that, by design, adds 75-125 ms of latency to each frame.

    24. Re:Yes. by Lord+Ender · · Score: 1

      A guess:

      This card has a smaller packet/frame (IP? ethernet? both?) size. It also tripple-sends every packet, so that if a packet is lost, its replacement gets to the game server without having to wait for the "where the hell is packet X?" packet to get back to your gaming rig.

      Also, becuase it is upgradable, perhaps they intend to release certain optimizations for the most popular games?

      --
      A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
    25. Re:Yes. by Mia'cova · · Score: 1

      I think that is essentially what is happening at the hardware level. It's slowing down so it's easier to pick up the signal on the other end. Longer pulses are easier to pick up with a weak signal and noise. But at the higher levels (TCP, IP, etc), it works the same regardless of the underlying transport. We don't want to change TCP so much of the work going on for wireless focuses on the protocol that carries the packets, or the smaller frames of whatever size the 802.11_ go by. I'm not very familiar with any of the specifics there but if you drop by a computer science reading room and leaf through some wireless journals/conference articles, you're going to see a whole ton of research playing around with the variables.

      But also, keep in mind that if you decide repeating packets is cool and use more bandwidth, that's less that's going to be available for your neighbor's wifi network. We all share the same spectrum so I'd have to imagine that any agreed upon protocols will be very minimalist and nice to other traffic. You need to leave gaps for other machines to speak up since it's system of random delays (for everything else, I'm assuming wifi too).

      If you're interested in this stuff, take a computer science networking class that digs into the protocols. It's actually pretty interesting how the delays, retransmits, etc work. Or just get your fill looking around online or from a text, just might be a little dry that way.

    26. Re:Yes. by Anonymous Coward · · Score: 0

      Store and forward does a CRC check on the frame. Cut through will read enough about the frame to properly direct it and pass the payload without any additional delay. Some switches can be configured to do either.

    27. Re:Yes. by Feyr · · Score: 3, Funny

      in other news, there is such a thing as a "gaming network stack". forget ipv6! we're going full on gaming network!

    28. Re:Yes. by CaptKeen · · Score: 1
      Latency isnt introduced because your PC is stupid, its introduced because you're waiting the time it takes for packets to travel to your ISP, to its ISP, to its ISP, down to its child, down to its child, and back to some other PC, and having to interact with the 20 routers, gateways, and switches along the way.

      Close... Routers don't take very long forwarding packets - < 1 ms. What takes so long is driving the data down the pipes. Once your signal gets to the Internet backbone - and hell, even before then - you start getting limited by things like the amount of time it takes the light to travel down the fiber. Remember that you can't just connect a router in San Jose to a router in New York; you have to connect it to some form of long haul optical transport gear. That gear will take the signal (be it OC3 up to OC192) and take it cross country. Depending on the gear, every 60-150km you need to go through amplification, and every few amplifiers you have to go through signal regeneration where the optical signal is converted back into an electrical signal and then generated anew as an optical signal.

      It takes 28-29 ms to drive from DC to Dallas, for example, 4 ms to drive from Dallas to Houston, 23 ms to drive from Houston to Phoenix, etc. So if you are in, say, San Jose and you're talking to a server in DC, you're looking at about 72ms of delay right there. And thats backbone router to backbone router, not factoring in any local optics (backbone to your ISP, your ISP to you, etc).

      Thats not to say that there aren't other issues involved, but you can never factor out actual transit time to push those packets around. If you have to hop through a few routers in a single market, any delay seen isn't necesarrily from routing delays unless the router is severely overloaded but is instead from the time taken to drive the signal router to router.

      --
      --
    29. Re:Yes. by Suidae · · Score: 2, Interesting

      Or better yet, how long until there is an exploit to root the OS in the card itself?

    30. Re:Yes. by Anonymous Coward · · Score: 0

      Most well designed games only use PARTIALLY reliable messaging systems - ie, only critical messages are re-sent. Positional updates, etc, when done properly, can be discarded as the next update will correct for the missing packet.

      Critical messages are things like, "You just died" or "Bob: u r a haxx0r!!!11" or such. Well, maybe Bob's message wasn't so critical in your eyes, but if it went missing, the conversation might seem a little off.

      Some games, the worst kind for behind a firewall, open one or more TCP ports for these "critical" messages, and use UDP for non-critical, resulting in a big tangle of rules in the firewall to handle it.

      ---

      I think the expression you're looking for is "Store and Forward". And yes, when only one host is transmitting, a hub is better than a Store-and-Forward switch, latency wise. Like all things in life, there's no black and white - while a switch is powerful, it is not _universally_ better. You can get a myriad of other styles of packet handling mechanisms, but they're always sacrificing something - hubs can't do multiple client communication efficiently, cut through switches can propagate corrupt packets, store and forward switches add latency.

      By the way - 20 routers and switches is a very short trip. Those hops you see on the traceroute may represent ten devices each - MPLS and other types (X.25, ATM, etc) of underlying protocols that IP can ride over hide many, many hops.

      Also, one big huge factor in cross-North America links, and overseas links, is that every 300KM (that's what, 200 miles?) adds another full millisecond to each direction of travel - thus the total ping round trip time increases by 2ms every 300KM. For light. Travelling in a vacuum. Usually your light and electronic signals are travelling through fiber or copper - at most 70% of light speed.

      Satellites totally ruin the picture - geosyncronous orbit (satellites that stay put) is something like 50,000KM above the surface of the planet, and may have to bounce off other satellites also at 50MM altitude, making a complete circuit around the planet like 300,000KM - a full second at light speed.

      ---

      Despite these corrections, your message is fundamentally sound - unless this card is telepathic, and/or hooks you into an imaginary FTL circuit that connects directly into the server... it's not going to help much with anything.

    31. Re:Yes. by sketerpot · · Score: 3, Interesting

      One of the cooler features of this card is that it can be programmed to run things like firewalls on its own processor.

    32. Re:Yes. by sketerpot · · Score: 2, Informative

      Bah, I created it first and called it MegaPacketBlast X-Treem!

    33. Re:Yes. by a_ghostwheel · · Score: 1

      I don't need another firewall on it's own processor. I already have cheap Linksys router between PC and cable modem doing just that (in addition to being switch and bunch of other things. And it Linux based too).

    34. Re:Yes. by arth1 · · Score: 1
      Games aren't really streaming media. Straming infers one-way traffic. Gaming is most certainly two way. Your client needs to tell the server what it's doing and the server needs to tell your client what has happened with everything other in-game entity.

      In one word: No.
      Your games and the game servers don't stop to analyse what was last received before sending out packets. That would make the game work like a turn-based game, which would be dead slow.
      Instead, the game server immediately tells your PC about any changes on the server side, while the PC immediately tells the game server about any changes on the PC side. Traffic one way isn't replies to traffic going the other way.

      If you look at your network connections with e.g. "netstat -an", you'll see multiple connections between yourself and the game server. There's usually at least two connections, and more often three -- one for data from the game server to you, one for data from you to the game server, and one for priority messages. All but the priority channel are almost certainly working as streams, and not as two-way traffic (except for an occasional "please resend" or "you're sending too fast, please quench").

      Regards,
      --
      *Art
    35. Re:Yes. by Anonymous Coward · · Score: 1, Interesting

      Drivers or Card? Who gives a rat's ass?

      I am not a driver engineer nor hardware engineer.

      There have been several times throughout history in which ATI had (theoretically) better hardware. But, as we all know, their drivers sucked some sludge into the intake valve. Only a nimrod stands around bragging about the superiority of their hardware while playing BF2 without textures.

      Spruce Goose anyone? Better drivers and it might have REALLY flown a second time. Fix the drivers after delivery when they are pretty much completely busted? Well, we'll just plain cancel ya!

    36. Re:Yes. by Vihai · · Score: 1

      You are simply confusing a stream-based protocol (TCP) with a protocol suitable for streaming (which would not be TCP).

      ...and you are wrong. Most real-time games use UDP over which they stream time sensitive data.

    37. Re:Yes. by Anonymous Coward · · Score: 0

      It's still a pretty cool feature, whether you need it or not.

    38. Re:Yes. by Rogerborg · · Score: 1

      See the classic X Wing vs TIE Fighter analysis on Gamasutra for an example of developers learning the slow, painful way about this. They settled on resending duplicates.

      --
      If you were blocking sigs, you wouldn't have to read this.
    39. Re:Yes. by Rogerborg · · Score: 1

      ... and shortly after you implement that, you realise that all you really need are the packets that tell you that Alice has started and stopped moving, with occasional absolute positional updates to keep everything in sync. Don't worry, you'll get there eventually.

      --
      If you were blocking sigs, you wouldn't have to read this.
    40. Re:Yes. by Rogerborg · · Score: 1

      I do not think that word means what you think it means.

      --
      If you were blocking sigs, you wouldn't have to read this.
    41. Re:Yes. by Rogerborg · · Score: 1

      So cynical! If you read the technical specs, you'll find that it really does work; the packets slip through the cable faster because they've been liberally coated in snake oil.

      --
      If you were blocking sigs, you wouldn't have to read this.
    42. Re:Yes. by Eivind · · Score: 1
      It doesn't need to be that bad. You can do various much-smarter things.

      For example, you can interpolate. If you are sending a stream of position-data for a player, for example, a single lost packet in the middle can be interpolated around. This will in most cases work better than waiting for a retransmit.

      A player that moves (0,10) - (5,15) - (x,y) - (15,25) can be assumed to have been on (10,20) at the time of the missing packet. That is a very fair guess, and will likely work out fine. There are details. (there always is) like making sure all clients agree if the player was hit at that time or not. (perhaps in reality he was at (8,22) and narrowly escaped a bullet according to his local copy of the game)

      You can also add redundancy much more intelligently than simply sending everything double. To take a trivial example, you can send a single checksum-packet after a series of packets that enables the reconstruction of any lost packet. So for example send 5 packets, then a single checksum-packet, and aslong as max 1 packet is lost, you lost nothing.

    43. Re:Yes. by Anonymous Coward · · Score: 0

      Most switches use something called Hold and Forward (I might have the name wrong...)

      Store and forward is the correct term (not standard on switches because it increases latency, most switches use cut through witch just reads the the first few bytes of the packet (until it knows the destination address) and then forwards it.

    44. Re:Yes. by Darren+Winsper · · Score: 1

      I've gamed on 3Com, RealTek, Intel and NVIDIA network cards. I have never ever seen any of them deviate in performance by more than 10ms on a LAN. If you are seeing a 150-250ms ping time between your computer and your local router, you have seriously fucked up somewhere.

    45. Re:Yes. by Iamthefallen · · Score: 1

      Except the packets are numbered...

      The server builds say 20 frames per second, caluclating each players position based on movement from the previous frame. From this the server builds an update to the state of the world, gives it a number and sends it to the clients. When the client receives an out of order frame, it's quietly discarded.

      Normally, the people who design game network code, aren't complete idiots.

      There are exceptions.

      --
      Wax-Museum Fire Results In Hundreds Of New Danny DeVito Statues
    46. Re:Yes. by TheRaven64 · · Score: 1

      This is why you encode a timestamp in each message. Typically, I would include time, position, and velocity (and apologise to Heisenberg later). From here, you can calculate the correct position by applying the velocity times the current time minus the timestamp to the position. There will be some error in here, of course, particularly if the object is changing velocity regularly (although sometimes you can improve this by encoding the acceleration as well). The advantage of this is that you have relatively accurate data in low latency conditions, and it degrades gradually with increased latency and packet loss. Any message with a timestamp earlier than the last one you processed is simply discarded; no retransmission required.

      --
      I am TheRaven on Soylent News
    47. Re:Yes. by TheRaven64 · · Score: 1
      Sounds suspicious to me. I get <1ms pings to machines on the LAN with Realtek 8139 cards. These, according to the FreeBSD driver source 'redefine low end' and are a horribly broken design. The only reason they are popular is that they are so cheap (easily <$5).

      I don't know what was causing your issues, but I would be very surprised if a network card would add anything like this latency. One thing I could suggest is poor IRQ handling. I came across an issues with a Linux 2.6 kernel that used ACPI for IRQs. It added about 200ms to the RTT time due to poor ACPI implementation (either in the kernel or the BIOS). Turning off ACPI support in the kernel made it go away. Moving the network card to a different IRQ might have done the same thing, and if you swap the card then it may well be assigned a different interrupt.

      --
      I am TheRaven on Soylent News
    48. Re:Yes. by 6031769 · · Score: 1

      OTOH, I can well imagine loads of folks shipping an Ethernet NIC that, by lack of design, adds 75-125 ms of latency to each frame.

      --
      Burns: We're building a casino!
      McAllister: Arrr. Give me 5 minutes.
    49. Re:Yes. by CRC'99 · · Score: 1
      But if a significant minority of users were to do the same thing, it would probably result in a complete collapse of any usuable bandwidth. Which is exactly the kind of thing I would expect a bunch of MBA's to come up with.


      Ahhh - but what they didn't include in this press release is that the card is so advanced that it opens up a sub-space window to tunnel data through. This means that your data travels faster than normal light because it's in a sub-space window. The reason for this is obvious - because there's less space in sub-space.

      The idea was used under license from the creators of the Stargate and incorporated into a network card for everyone to use. The only reason linux is on the card is that a DHD is required.

      The really bad news however is that because of the way this card works, to get any possible chances of having any improvement, you need to have one of these at each end - otherwise the auto-negotiation of the card falls back to Ethernet from Sub-Ethernet.

      It's just like magic really.
      --
      Sendmail is like emacs: A nice operating system, but missing an editor and a MTA.
    50. Re:Yes. by SwissCheese · · Score: 1

      This is why packets have sequence numbers. Out of order UDP packets are simply ignored.

    51. Re:Yes. by TheBogBrushZone · · Score: 1
      A lot of games are TCP based... World of Warcraft for example.
      Interesting then that they give World of Warcraft as the example in the 'Sneak Peek' performance statistics when compared to an onboard Marvell GigE chip known to have relatively high CPU usage. But then they also describe the piddly 10-20ms reduction in latency as 'stunning'. I don't know about others but that wouldn't make much of a dent in my 300ms ping times.
      --
      And behold, a command prompt and he who sat upon it, his name was shutdown and -h 3:11 followed with him
    52. Re:Yes. by vadim_t · · Score: 1

      My system is better. In most games you can click on the place where you want to move, so sending movements in terms of "Player is moving to (x,y)" will produce better results than "Player started moving in X direction at Y speed; Player stopped moving". While the later is less bandwidth intensive, just try to get that to work well with variable latency.

    53. Re:Yes. by mrbcs · · Score: 1
      What was even more fun was that the ping issue was intermittant. Do a ping -n 100 and about every twenty pings it would timeout.

      Switched the card, no more calls, no more problems.

      And btw, it wasn't me that fucked up, this was the customers shit hooked up to Motorola wireless running on the supernet fibre. Every other customer in the area was fine... playing the same game on the same servers. It WAS his shitty nic.

      --
      I'm not anti-social, I'm anti-idiot.
    54. Re:Yes. by heson · · Score: 1

      It might even be able to run a spam relay or a ddos zombie. Oh joy.

    55. Re:Yes. by Darren+Winsper · · Score: 1

      As I said, I've used NICs from all sorts of vendors, more than I care to remember thanks to living in a house full of geeks. At least three of us are heavily into gaming and none of us had these sorts of issues that couldn't be attributed to other things. Perhaps his NIC was shitty, but it could have just been a duff card, hardly a reason to write off the entire brand.

  3. Pricey by HeWhoRoams · · Score: 5, Informative

    Pre order cost is $280. You'll see a better FPS increase spending that on a graphics card, RAM, or some groceries for 6 months.

    1. Re:Pricey by Anonymous Coward · · Score: 0

      There's some sort of truism that every engineer slash MBA thinks this is a good idea and tries to implement it.

    2. Re:Pricey by wiggles · · Score: 1

      I totally agree. $280 is way too much to spend for what would be a very small FPS increase. On the other hand, if they added some functionality and repackaged it as an SSL acceleration card, they would make big bucks from secure web sites looking to offload the SSL calculations from the cpu without buying a more expensive external accelerator.

    3. Re:Pricey by bcmm · · Score: 4, Funny

      Yes, because we thought we'd get a better FPS increase from lower network latency.

      --
      # cat /dev/mem | strings | grep -i llama
      Damn, my RAM is full of llamas.
    4. Re:Pricey by SatanicPuppy · · Score: 1

      Wow, yea, that's a good idea! I've got a handful of secure proxys set up that could use a little boost. Reminds me of something I heard about people trying to take advantage of all the processing power in the newer graphics cards, to speed up server-side number crunching.

      --
      ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
    5. Re:Pricey by Clazzy · · Score: 1

      Looks like I'll be spending a few days buying some food, I have FPS to gain!

      --
      If we can hit that bull's-eye, the rest of the dominoes will fall like a house of cards... Checkmate.
    6. Re:Pricey by Anonymous Coward · · Score: 0

      "groceries for 6 months" - 46.666666 dollars a month on groceries...what do you eat? more like 1.5 months of groceries

    7. Re:Pricey by Anonymous Coward · · Score: 0

      Interesting you should say that, considering the article has nothing to do with system performance, and frames per second have no bearing on network latency.

    8. Re:Pricey by strstrep · · Score: 1

      I bet that the old Intel e100's that I got off eBay have lower latencies than that card. Seems a bit too complex to actually improve performance.

    9. Re:Pricey by NUBlackshirts · · Score: 1

      Food Per Sitting?

    10. Re:Pricey by Anonymous Coward · · Score: 0

      Ramen noodles = $0.08 per pack at my local Wal-Mart. $46.67 will get you about 583 packs of noodles.

    11. Re:Pricey by cswiger2005 · · Score: 2

      Yeah, only you can get a decent NIC from Intel, Broadcom, maybe 3com, which offloads most of the IP stack into hardware *AND* get a HiFN 79xx-based SSL/RSA/DSA/AES cryptoaccelerator card from Soerkis for about $100. I guess the extra $200 or so these people want covers the heatsink and a couple of pretty LEDs.

      I'd mod this product +5 snakeoil.

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
    12. Re:Pricey by Achromatic1978 · · Score: 2, Funny
      Ahh, ramen noodles. High in carbohydrates, low in fibre, minerals and vitamins. High in saturated and trans fats. High in MSG and sodium.

      The cornerstone of any nutritious diet.

    13. Re:Pricey by tapehands · · Score: 1

      I'm going out on a limb here, but I bet they include a window applique with the NIC. The bonus of this total package? 50 EXTRA HORSEPOWER!!!11!

      Ok...I'm exaggerating, but it IS a bleeding waste. I could almost see paying $100 for this card, but $280? It should do a hell of a lot more than give me a 1 or 2ms bonus to my ping times, and free my computer from that horribly arduious task of communicating a few more bytes of data to my network card...maybe they'll have some useful devices to plug in to that USB port, -winkwinknudgenudge- know what I mean?

      And to offer constructive criticism, maybe they could team up with Ageia? That would probably give them a good reason to move to PCI Express (where it seems everything else is moving to), and would definitely add more value to the board.

    14. Re:Pricey by e2d2 · · Score: 1

      If he meant ramen he would have said $280 can buy you food for 6 decades, not months.

      Seriously though, at about 10 for a dollar, then maybe throw in the bulk rate from costco, you can get an good 3000 packs of ramen! even at 3 packs a day you could survive for 1000 days, "live" is too good a word for a life eating ramen.

    15. Re:Pricey by Anonymous Coward · · Score: 0

      On the other hand, if they added some functionality and repackaged it as an SSL acceleration card, they would make big bucks from secure web sites looking to offload the SSL calculations from the cpu without buying a more expensive external accelerator.

      you mean like this?

      (sorry about the store link, only thing i could find on google quickly - not affiliated, never ordered, never even heard of them until this search)

    16. Re:Pricey by quantum+bit · · Score: 1

      Yeah, only you can get a decent NIC from Intel, Broadcom, maybe 3com, which offloads most of the IP stack into hardware *AND* get a HiFN 79xx-based SSL/RSA/DSA/AES cryptoaccelerator card from Soerkis for about $100. I guess the extra $200 or so these people want covers the heatsink and a couple of pretty LEDs.

      Yeah, but from their recent dealing with the OpenBSD folks it sounds like HiFN may not be an option in the future as they're refusing to provide specs on any of their newer models. I know the FreeBSD drivers for the only model that supports AES have had accelerator lockup problems since forever; and it doesn't look like we'll be able to get enough information to fix them. Google the soekris lists for people having Hifn troubles.

      Intel did make the Pro/100S series with 3DES offloading, but as usual they refused to give any information on the crypto part of the card. So you can use them as dumb NICs in Linux/*BSD, but none of the fancy features...

    17. Re:Pricey by Anonymous Coward · · Score: 0

      Don't forget all that yummy starch and sodium.

    18. Re:Pricey by sjames · · Score: 1

      TCP offload SOUNDS like it might help, but in the real world, it doesn't do much good. TCP computations just aren't all that CPU intensive, especially since a packet being sent out already tends to be in the cache (so checksumming doesn't cause a cache miss).

    19. Re:Pricey by Cid+Highwind · · Score: 1

      even at 3 packs a day you could survive for 1000 days

      Probably not. Top Ramen has 190 calories per package, so 3 of them per day comes in at 570 C per day, less than a third of the 1800 C needed to keep up an average man's base metabolic rate. There's no way anyone could sustain that much caloric deficit for 3 years, even if they were an epic lard ass starting out.

      --
      0 1 - just my two bits
    20. Re:Pricey by Superfarstucker · · Score: 1

      You're right that FPS has nothing to do with "network" latency but it does affect apparent latency quite a bit. Figure you're getting a mean of 60 fps and a 3 second low & high mean of 30 and 80 respectively with abs instantaneous mins and maxes of 20 and 100 (10 to 50 ms between frames). Take into account double or triple buffering (frame delay & buffering come out to: 20-150 ms). That's pretty reasonable for a modern game, do the math, that's a bit of latency. Add in the fact that most panels are synced @ 60 Hz (20 ms) and the delay they add when actually drawing what happened. Now granted it is likely many of these delays will collide or overlap, but as you can see there are many components to "apparent" latency. Certainly you can't draw a frame containing information that is still being transmitted down the pipe. Network transit is only 1 component, and seeing as this product is clearly shenanigans I think it is altogether reasonable to suggest a system performance upgrade as a latency reduction. It will certainly do more than this, which frankly, is a joke.

    21. Re:Pricey by cswiger2005 · · Score: 1

      Under low traffic volumes, sure, the packets being sent out are likely to still be in cache, but as your traffic load goes up, especially with jumbo frames, that's less likely to be the case-- and obviously, incoming packets aren't going to be resident in the CPU cache. Just like interrupt mitigation, which is not needed for low-to-moderate traffic volumes, but becomes quite valuable under greater loads...

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
    22. Re:Pricey by sjames · · Score: 1

      In high performance apps, the packet likely will still be in cache when it is submitted to the stack (having just been created).

      While an incoming packet won't be in the cache, the cost of bringing it in when checksumming is offset by the cache being warm when userspace recieves the data. The key is to process incoming packets in the timeslice of the app that wants them.

      If the app is not designed for performance, then it hardly matters.

      On the other hand, interrupt mitigation is quite helpful for throughput (potentially at the cost of latency) because a high interrupt rate tends to trash the cache.

    23. Re:Pricey by Builder · · Score: 1

      You can get groceries for 6 months for $280? I want to hire you to do my shopping for me!

    24. Re:Pricey by cswiger2005 · · Score: 1

      Remember that just getting data to or from userland requires a copyin()/copyout() or equivalent, and a context switch between user and kernel mode-- just because a program in userspace sends data (whether via read()/write() or whatever), that doesn't mean that the system actually immediately goes into the kernel and dispatches the packet; because of that, it's quite possible to send many packets worth of data from userland and bust the cache before the system does a context-switch and runs the kernel's side of the networking code.

      In high performance apps, you might even use things like sendfile() to push your data out, which means that you're using DMA between RAM and the disk controller to fetch the data into mbuf chains, and then you have the NIC driver adding these chains to the NIC's list of Tx descriptors; when ready, the NIC performs DMA to move the data in the mbufs to it for checksumming and then transmission. The CPU will be used in generating the packet's protocol headers, but the contents of the packet are never seen by the CPU and thus never reside in cache.

      That doesn't mean that your point is completely wrong-- yes, doing TCP/IP on the CPU is relatively efficient and yes, it is relatively common for the packet data to still be resident in the CPU's cache, but as the load goes up, or if you have an SMP system where the CPU running the stack and the CPU running the userland program may not be the same and may not share an L2 or L3 cache, having the NIC around to offload the gruntwork of checksums matters more.

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
    25. Re:Pricey by Jamu · · Score: 1

      Certainly you can't draw a frame containing information that is still being transmitted down the pipe.

      You usually can't do this reliably and often making guesses can lead to confusion. But I'd like to add that it is possible to draw a frame containing data that is still being transmitted down the pipe. Consider the drawing of that last sentence: After "that is still being" it's possible (and somewhat trivial in this example) to predict and draw "transmitted down the pipe" without waiting for the data to arrive. Prediction like this though is often more difficult with better data compression.

      --
      Who ordered that?
    26. Re:Pricey by sjames · · Score: 1

      I'll grant that checksumming in hardware is not a total loss, it will help in the case of sendfile. The case of many small writes isn't that important since if the app is doing that rather than aggregating the writes it has already lost due to the system call overhead. Marketing would have us believe that the checksumming is a tremendous burden on the CPU but in practice checksum offloading is just not the huge boost some imagine.

      I do maintain that the rest of TCP offload IS a net loss, in part because the typical card offloads just enough to claim it should cost more, but not enough that the kernel can 'fire and forget'. For that, the card would have to implement firewalling internally and have plenty of buffer memory to handle retransmits. It would have to present the BSD socket API and a firewalling API to the kernel. Even then, it would probably be a barrier to evolution as much as anything.

    27. Re:Pricey by tehcyder · · Score: 1
      Ahh, ramen noodles. High in carbohydrates, low in fibre, minerals and vitamins. High in saturated and trans fats. High in MSG and sodium.
      The cornerstone of any nutritious diet
      Not quite, genius. You forgot coffee and doughnuts.
      --
      To have a right to do a thing is not at all the same as to be right in doing it
  4. Huh? by SatanicPuppy · · Score: 4, Insightful

    It's always been my understanding that the bigger bottlenecks are upstream of your NIC. I mean, my home network set up goes gigabit from my desktop to my hardware router, gigabit from my router to my gateway firewall, then gigabit (minus a few MTU) to my DSL modem, and after that the speed gets massively reduced and there's nothing I can do about it. My lan latency is practically non-existant.

    Can you really reprioritize your packets coming from your desktop in such a way that you make a significant gain after it hits your ISP? Or is this just cyberpenis enlargement? Seems to me that, unless you're hosting a bunch of internet spyware or network-heavy background processes, you're not going to be making much of a gain.

    --
    ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
    1. Re:Huh? by MustardMan · · Score: 2, Funny

      Seems to me that, unless you're hosting a bunch of internet spyware [...], you're not going to be making much of a gain.

      Uh, this is for gamers, right? Don't most gamers run WINDOWS?

    2. Re:Huh? by grasshoppa · · Score: 1

      Can you really reprioritize your packets coming from your desktop in such a way that you make a significant gain after it hits your ISP? Or is this just cyberpenis enlargement? Seems to me that, unless you're hosting a bunch of internet spyware or network-heavy background processes, you're not going to be making much of a gain.

      In a single device network, this is mostly true. However, most people have at least two devices on their network ( at the very least, two seperate systems ). And it's not about saturating the line, it's about saturating the line for brief moments of time.

      Case in point; VoIP. Plenty of people have vonage. You always want time sensitive packets like VoIP packets to go out ahead of all other traffic. In a theoretical situation, it's possible that even random web traffic could delay voip getting out in a timely manner, which you'd hear as pops and crackles.

      Not that this is relevant to the card; It's a marked up POS looking for a sucker to buy it.

      --
      Mod me down with all of your hatred and your journey towards the dark side will be complete!
    3. Re:Huh? by nine-times · · Score: 1

      I didn't RTFA, but maybe there's some sort of QoS that makes sure Windows Update or something like that doesn't slow down your games? I think I'd rather have manual control, though, and just turn off things that I'm downloading in the background.

      I'm not sure what else they could do, though, and I guess that wouldn't reduce latency as much as increase throughput for games. If you're halfway around the world from your destination, I don't know how a better network card is going to get you a faster ping.

      But hey, someone tell me if I'm just not getting it.

    4. Re:Huh? by fm6 · · Score: 2, Informative

      If you go to the web site, and read the white paper, you'll see that they're mainly thinking in terms of LAN usage.

    5. Re:Huh? by smbarbour · · Score: 1

      Can you really reprioritize your packets coming from your desktop in such a way that you make a significant gain after it hits your ISP?

      Yeah, technically if the packet is flagged with a higher priority CoS, and ALL of the equipment between hither and yon support CoS, it is possible.

      After reading through the whitepaper, it seems this card is also able to flag INCOMING packets as well. If this were possible, it would CAUSE incredible amounts of lag for everything else waiting for packets (not to mention requiring a cache for all of the suspended packets it would have to set off to the side while waiting for the "preferred" packets to arrive). If I'm not mistaken packets still arrive one by one from the network.

    6. Re:Huh? by forkazoo · · Score: 5, Insightful
      It's always been my understanding that the bigger bottlenecks are upstream of your NIC. I mean, my home network set up goes gigabit from my desktop to my hardware router, gigabit from my router to my gateway firewall, then gigabit (minus a few MTU) to my DSL modem, and after that the speed gets massively reduced and there's nothing I can do about it. My lan latency is practically non-existant.


      Now, maybe I'm completely misunderstanding teh point of this NIC, but...

      You are correct. The NIC isn't an appreciable source of latency. Right now, I ping'd a server on another subnet, and I averaged 0.3 ms latency. This is bog standard 100 Mb. Nothing the least bit fancy. That server might have a nice NIC of some sort, but this desktop certainly doesn't. And, that's hopping between subnets. Crossing between buildings over a T-1, with a few routers involved in about 5 ms. Pinging my home machine over the internet is abou 150 ms. So, assuming that of the .3 ms latency I have inside this building, none of it is due to the switch, and none of it is due to actual wire delay, then about half of the latency is my system, and half is from the server. So, my NIC is responsible for abou 0.15 ms of latency.

      Now, assuming that I was playing a game with my home computer, moving to a NIC that cut the latency of my PC down by 2/3 (from .15 ms to .05 ms), I'd be shaving my total latency for the connection to 149.9 ms (from 150ms).

      Which would improve my lag by .06%

      No, dammit. You won't see a noticeable improvement from a lower latency NIC. There are probably a few microbenchmarks where you will get a phenomenal speedup. Gaming isn't one of those cases.
    7. Re:Huh? by SatanicPuppy · · Score: 1

      Yea, packets only ever arrive/depart one at a time, unless you've got multiple independant interfaces enabled. That's just tcp/ip. If this thing could change that well that would be something. Maybe some kind of quantum dynamics? Or magic packets where an outgoing packet could travel through an inbound packet. Trippy.

      --
      ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
    8. Re:Huh? by Anonymous Coward · · Score: 0

      then gigabit (minus a few MTU) to my DSL modem,

      You have a GB DSL modem? Why on earth would anyone make one of those?

    9. Re:Huh? by malejko · · Score: 1

      Total marketing, and definitely some cyberpenis enlargement. No other way to explain this product.

      The only other way I can see it being viable is so that you can download your P2P pr0n, keep being a spambot and have the NIC do some gaming QoS for 'ya. Either that or you have a built-in NIC that's hogging your bus. There are some good built-in NIC's, but there's a lot of shitty ones too.

      Still not worth even close to $280. $100 tops, and even then I'll still stick to just 'turning off' the other network stuff on my machine. D-Link's gaming router can help if you're not a Linux geek and want to stick something in to keep your sister from hogging the bandwidth. But to save yourself from hogging your own computer's bandwidth? Ugh.. I can't believe I'm even commenting on this story now that I think of it.

      --
      -Adam
    10. Re:Huh? by PitaBred · · Score: 1

      What about playing on a LAN with other people, though? You'll be moving from 0.3ms to 0.1ms. Or something. I don't actually know, but I'd guarantee you that's where this card is targeted. Not at the people playing online games, but those playing networked, LAN games.

    11. Re:Huh? by Not_Wiggins · · Score: 1

      Parent is spot-on about latency when playing over the Internet.
      I haven't RTFA, but I suspect that the claims of increased performance will be on a LAN-only basis.

      And not to try and legitimize what looks to be a sucker-magnet product (they seriously need to add neon to it!), but there could be optimizations related to having a local network of these things talking to one another. Or, the optimizations could be related to a cross-over cable scenario (not fun for gaming, but could be enough to "qualify" the marketing claims).

      As others have pointed out, one still needs to optimize all the pieces that actually make-up the network. What about the switch that all these local machines would plug into? Wouldn't that need to be optimized, as well?

      Indeed, this smells like MonsterCable selling GOLD PLATED Optical Fibre... only suckers need apply. 8/

      --
      Diplomacy is the art of saying, "Nice doggie!" until you can find a rock.
    12. Re:Huh? by smbarbour · · Score: 1

      No, magic packets are something else entirely. Magic packets are used to activate the Wake-on-LAN feature of some network cards.

      Oh wait... You're talking about some OTHER type of magic packet. My bad.

      BTW, is there actually a practical use for Wake-on-LAN (i.e. a real-world scenario for it)

    13. Re:Huh? by Jett · · Score: 1

      I've yet to see this happen and I've tried with my cable modem. I saturated my connection with torrents and then tried making and receiving a vonage call. I didn't notice any difference except my torrents slowed down a little and web pages loaded very slowly. Actual phone service had no noticeable difference.

    14. Re:Huh? by Anonymous Coward · · Score: 0

      So exactly the sort of dickheads who spent $$$ tripping out their machines with 733t blue LEDs? It's all starting to make sense now.

    15. Re:Huh? by gallwapa · · Score: 2, Informative

      BTW, is there actually a practical use for Wake-on-LAN (i.e. a real-world scenario for it)

      I work for a school district with a sub-standard WAN (READ: 1 T1 link for a school with 500 computers with internet access) and no WSUS or patch management distribution servers. Wake on LAN, have 45 sites with 100-500 computers connect back to our ...*shudder* singular *shudder* SUS server...shut down.

      Or, wake on lan, re-image machines with Novell Zenworks, shut down
      Or, wake on lan, to unfreeze machines from DeepFreeze
      or, wake on lan to virus scan instead of when users are at work

      The possibilities are endless :)

    16. Re:Huh? by chrisv · · Score: 1

      Perhaps it's a DSL modem with a built-in switch? I don't know... I know the DSL modem / firewall that I've got sitting here has a built-in switch and such on it, but it still only runs 100baseTX-FD, though the way I've figured it's not like it makes much difference unless I'm communicating with the machine on the floor next to me. The rest of the house is behind an ancient 10baseT hub, not that, for internet access, anyone would ever notice. The hub never gets saturated in the first place... 3Mbit in from the DSL line isn't enough to do that.

      --

      Dogma: Dead (mostly because your Karma ran it over)

    17. Re:Huh? by mute47 · · Score: 1

      I haven't seen any recently, but WOModem functions were well-used in the days of BBSs.

      There was some discussion on it regarding a campaign to conserve energy in the IT business, run from the PC Pro magazine, called Switch IT Off, but the result was that it wasn't usable without some serious thought and planning.

      --
      Don't mind me, I'm just carping the diem...
    18. Re:Huh? by fimbulvetr · · Score: 1

      Then something is wrong with your experiment, because lack of latency is of paramount importance to a VoIP call. Maybe the router vonage sent you does auto QoS? Maybe your computer cannot generate enough bandwidth to saturate the connection?

    19. Re:Huh? by EvilNecro · · Score: 1

      Once upon a time when I did network backups overnight, users would turn off their machines when they left (even though they weren't supposed to). In the backup script, you wake up the machines with the 'magic packet', then back them up. =]

    20. Re:Huh? by grasshoppa · · Score: 1

      I've yet to see this happen and I've tried with my cable modem. I saturated my connection with torrents and then tried making and receiving a vonage call. I didn't notice any difference except my torrents slowed down a little and web pages loaded very slowly. Actual phone service had no noticeable difference.

      The router vonage sends you does QoS. I was merely using vonage in the hopes more people would know that.

      I suppose I could have used Skype as well and had the same effect.

      --
      Mod me down with all of your hatred and your journey towards the dark side will be complete!
    21. Re:Huh? by cswiger2005 · · Score: 1

      > Can you really reprioritize your packets coming from your desktop in such a way that you make a significant gain after it hits your ISP?

      Doubtful. It's almost never the case that you get out-of-order packets on your LAN; that happens upstream between ISPs when routers flap routes as a result of an outage or topology/config change.

      If you can't afford to wait a few seconds for reassembly or a retransmit, the game or application is better off using UDP and timestamps or a sequence # to provide ordering without the guaranteed in-order delivery provided by TCP.

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
    22. Re:Huh? by Jett · · Score: 1

      that was the point of my story.

    23. Re:Huh? by nuzak · · Score: 1

      > I didn't RTFA, but maybe there's some sort of QoS that makes sure Windows Update or something like that doesn't slow down your games?

      Windows Update uses the Background Intelligent Transfer Service (aka BITS), which uses only idle bandwidth. It'd be neat if BITS could be used for things like bittorrent or TOR or freenode, but it's really only designed as a pull-only file transfer mechanism. Limited as it is, it still works pretty well, and it's something Linux still hasn't copied in any mainstream distro.

      --
      Done with slashdot, done with nerds, getting a life.
    24. Re:Huh? by mdboyd · · Score: 1
      I'm pretty sure that the kind of people that would purchase this don't have many friends to play on their LAN with them.

      From the website:
      PingThrottle- When other games complain that your ping is too low, adjust it a little higher until they stop whining. Then, dial it back down and go in for the kill!
      So is this just an expensive tool to gain an unfair advantage (dare I say, "cheat")?

      I'd really love to see this thing in action because if it does everything it claims it does, it would appear to break a lot of the laws of networking...like, you could only be as fast as your router and other connected networking devices can support?
    25. Re:Huh? by Anonymous Coward · · Score: 0

      The latency from your computer to your router is practically non-existent, but the router doesn't have unlimited memory. If your network card bombarded the router with packets, the router would have to drop any packets that didn't fit into it's buffer. To prevent lost packets, the network card throttles outgoing packets and keeps a buffer of it's own where it queues packets to be sent. Instead of a simple FIFO, this network card most likely prioritizes gaming packets over other non-gaming packets within it's own buffer.

      I've seen demonstrations of a similar technology (NVIDIA's First Packet) and it does actually make a difference - but only if you have some other program (such as Bittorrent) saturating the link. Anandtech ( http://www.anandtech.com/cpuchipsets/showdoc.aspx? i=2764&p=3 ) saw ping times improve between 15% to 40% with NVIDIA's stuff.

    26. Re:Huh? by Anonymous Coward · · Score: 0

      You misread the quote. It allows you to increase your ping, meaning give you a disadvantage.

    27. Re:Huh? by quantum+bit · · Score: 1

      Unrelated to this NIC, but the problem with BITS is that it only knows how much bandwidth your local machine is using. If you have multiple computers on a network (sharing an uplink), one machine may think it's idle and start transferring a hundred meg service pack; completely killing the MMORPG or FPS you're playing on one of the others.

    28. Re:Huh? by quantum+bit · · Score: 1

      Hehe, one of our engineers has a mega overkill machine (read: Dual Dual-Core CPUs, gigs of RAM, half a terabyte worth of mirrored disks). So we decided to appropriate some of his storage space to act as a cached repository for software install CDs. DFS server is set to direct requests for the software repo coming from his subnet to his machine instead of the main server.

      The machine doesn't usually get powered off as he leaves number crunching jobs running overnight. Just in case, we have a cron job that sends a magic packet his way every 5 minutes or so ;-)

    29. Re:Huh? by Drgnkght · · Score: 1

      Certainly, it allows you to turn off your PC and still have it be available for the next internet worm to infect.

      Other than that, I can't think of anything except maybe LAN file/printer sharing or administration purposes. You certainly wouldn't want a real server to be napping and have to wake up to reply to a request.

    30. Re:Huh? by hcob$ · · Score: 1

      Any chance it could also be a pre-fetch, lookahead, cached dns/best route algorithm that could actually speed the connection by directing the packets through certain links?

      --
      Cliff Claven
      K.E.G. Party Chairman
      Founding Leader of: Koncerned for Egalitarin Governance
    31. Re:Huh? by Anonymous Coward · · Score: 0

      geee whizz dickhead.. think you're king with your gigadick network

    32. Re:Huh? by runningduck · · Score: 1

      0.3ms of latency across a T1 circuit? I would be impressed it you were getting 3ms across a T1 much less 0.3ms.

      --
      -rd
    33. Re:Huh? by mdboyd · · Score: 1

      No, I didn't misread it. It clearly says that when gamers whine that your ping is too LOW, you can turn it up (more lag), then turn it back down again. I'm just trying to say that multiplayer competition should be about skill and not hardware and if too low of a ping is deemed as unfair, then knowingly turning it up and then back down to fool players is ethically cheating.

    34. Re:Huh? by Anonymous Coward · · Score: 0

      The largest gains that this kind of card would make is actually on the WAN, where the overhead of the transport protocol is one of the most significant bottlenecks. You can actually make some pretty huge gains in both reliability and latency if you don't mind shitting all over the RFCs and choose to be the bastard in the prisoner's dilemma that is cooperative networking. For example, sending each outgoing packet multiple times can lower the average latency of reliable messaging protocols (by automatically addressing dropped packets most of the time), and increase the reliability of streaming protocols like UDP, all at the expense of everyone else on the network.

    35. Re:Huh? by demon · · Score: 1

      Given that TCP/IP doesn't work that way, no. IP routing uses a naive next-hop-only routing arrangement; all a single station knows is the next station to pass the packet on to. It doesn't know where it goes beyond that, and has no control over it.

      --

      Sam: "That was needlessly cryptic."
      Max: "I'd be peeing my pants if I wore any!"
    36. Re:Huh? by Anonymous Coward · · Score: 0

      Whilst I haven't heard a problem with my Vonage phone when running torrents, the person on the other end has said they have, this would be because my upload is only a eighth that of my download speed. If you have an asynchronous connection it may only be the person on the other end that hears the problem. Other possiblities are your router does a good job on QoS for the Vonage packets (although mine should do that if I set it correctly) or you have a better connection which can handle both without problem.

  5. For Gamers? by aquowf · · Score: 1

    Aw, nice marketing. Yea, thats all I have to say.

  6. Klingons on the starboard bow! by cannonfodda · · Score: 4, Funny

    Wow the first network card with built in Bat'leth!

    --
    Hmmmmmm
  7. Oooo... Killer by snowgirl · · Score: 5, Funny

    OMG, they named it the "KillerNIC"? Like, does this kind of advertising actually work?

    "This NIC is so hardcore it KILLED SOMEONE!"

    I can just imagine their second version coming with a muzzle a la Silence of the Lambs.

    --
    WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    1. Re:Oooo... Killer by 0racle · · Score: 1

      Since the majority of gamers see their computers as the mysterious and magical box that their parents do, my guess is yes, it works perfectly.

      --
      "I use a Mac because I'm just better than you are."
    2. Re:Oooo... Killer by Tumbleweed · · Score: 4, Funny

      I'd buy it if had a better name. Something like Xtreme NIC Ultra or something.

      And some VTEC stickers and a big-ass wing on it.

      Yeah, that'd be phat, yo!

      Pimp my NIC, bitches!

    3. Re:Oooo... Killer by utopianfiat · · Score: 1

      yup.
      12-year-old: "Mom, I want the killerNIC because it will make me more hardcore and I will be able to pwn more newbs."
      mother: "I don't know what you're talking about, but if it will get you to shut up so I can get back to my whiskey, I'll buy it."
      The kid will go on to tell all his friends "I've got a KillerNIC!" and they'll say "oooo, I don't know what that means but it sounds so hardcore" and one snot-nosed brat will say "Oh yeah? We'll I've got a 10 megabit super gigaflop bytemaster 200001 with PBX capabilities!" and the kid will go home, defeated, at least until he asks his mom for that nonexistent coil of dogshit.

      --
      +5, Truth
    4. Re:Oooo... Killer by Anonymous Coward · · Score: 0

      10 megabit super gigaflop bytemaster 200001 with PBX capabilities!

      Link please.

    5. Re:Oooo... Killer by kaddeh · · Score: 1

      it is called 'KillerNIC' because the people who buy it will KILL themselves after they find out it isn't worth shit in a bag.

    6. Re:Oooo... Killer by cheese-cube · · Score: 1

      Dude yout forgot the spinners!

    7. Re:Oooo... Killer by Xzisted · · Score: 1

      It rubs the lotion on its chip or else it gets no packet round trip.

      --

      Honesty may be the best policy, but apparently by elimination, dishonesty is the second best policy.
    8. Re:Oooo... Killer by BiggerIsBetter · · Score: 1

      I'd buy it if it came with the muzzle.

      Anyone from The GIMP project taking notes?

      --
      Forget thrust, drag, lift and weight. Airplanes fly because of money.
    9. Re:Oooo... Killer by glsunder · · Score: 1

      if it fails, maybe they'll try for the more cheerful Saint NIC next time.

    10. Re:Oooo... Killer by DigitalHammer · · Score: 1

      Plus a free copy of an MC Hawking album with every purchase. NICcas 4 life, homie! :P

    11. Re:Oooo... Killer by Matt_R · · Score: 1
      OMG, they named it the "KillerNIC"? Like, does this kind of advertising actually work?

      &nbsp

      I wonder if it's compatible with an Etherkiller :)

    12. Re:Oooo... Killer by Spasmodeus · · Score: 1

      It's Marketing 101-level knowledge that if you want to sell stuff to hardcore gamers, you have to give it a name that implies violence and/or evil.

      I've already trademarked my upcoming gamer-oriented product, "Remorseless Genocidal Slaughterettes: Crimes Against Humanity Edition"

      They're pre-moistened screen wipes. X-TREME pre-moistened screen wipes.

  8. Oh, Geeky! by Enoxice · · Score: 1

    It's a nice idea, but I'm skeptical about whether or not the promised performance boost will be worth the price. But does it run li...oh, right. So now you can say "you got pwned by my linux-powered nic!".

    --
    Anyone else think the comments just weren't rendering right before they turned off ABP and saw ads?
    1. Re:Oh, Geeky! by harrkev · · Score: 3, Informative

      But... if you read the specs, you have to run Windows in order to use it. It uses Linux, but no Linux drivers...

      --
      "-1 Troll" is the apparently the same as "-1 I disagree with you."
  9. drivers on cards? by LiquidMind · · Score: 1

    kinda OT, but something that came to mind recently....

    why can't manufacturers make hardware (esp NICs) with basic 9x/xp/*nx drivers on a ROM chip? It would be so much more convenient than having to 1) (if you're lucky) have the original disk / CD, insert, and install drivers or 2) download the drivers on another machine, burn to cd / copy to disk, walk it to the other computer, etc etc...

    too much to ask for?

    --
    This sig contains repetition and redundancy.
    1. Re:drivers on cards? by Amouth · · Score: 1

      I would like this too but only if you can flash it and stick what you want on it.. (within reason) personaly for alot of cards i get i stip the drivers down and out of the software they All want to install and only set up what is needed.. If i didn't have a choice because they had it in the card i wouldn't buy it..

      but the better way to look at it is to make your cards ne2000 compatiable.. so that even if you don't have the drivers you get basic networking... or does that just make too much sence?

      --
      '...if only "Jumping to a Conclusion" was an event in the Olympics.'
    2. Re:drivers on cards? by faragon · · Score: 2, Interesting

      Manufacturers don't do it because of:

      a) ROM code implies adopting some sort of code execution (ISA dependant, p.e. x86/PPC/MIPS/etc), CPU related.
      b) There are also dependencies related to the system BOOT process (p.e. IBM-PC / EFI BIOS / Other), i.e. related to the boot "protocol", CPU unrelated.
      b) Ignoring (a) and (b) problems, having 9x/xp/*nx drivers built-in in ROM just as backup for your media, note that the BIOS chip is nowdays quite more expensive than the 0.20$ that costs the driver CD, or the ~0$ that costs the driver update download.

      In the other hand, what you point could be possible -and interesting- if:

      1) Make/adopt an industry standard CPU emulation for booting CPU-independant BIOSes, p.e. using some kind of Java-like CPU emulation (like the way the PPC comunity uses PCI boards with x86 BIOSes).
      2) Make/adopt an industry standard BIOS boot protocol.
      3) Wait until some PC manufacturer put an 64Mbit BIOS, for loading a 8MB Linux.

      BIOS, drivers, and such are so inconvenient given the multiple OS available (which is a good thing). I hope some day someone bring the solution for that nonsense, may be there is necessary some kind of *clean* abstraction layer above these things (and, why not other Linux operating on a secondary CPU I/O dedicated subsystem?).

    3. Re:drivers on cards? by snowgirl · · Score: 2, Informative

      1) Make/adopt an industry standard CPU emulation for booting CPU-independant BIOSes, p.e. using some kind of Java-like CPU emulation (like the way the PPC comunity uses PCI boards with x86 BIOSes).
      2) Make/adopt an industry standard BIOS boot protocol.


      It exists, it's called OpenFirmware by Apple, and both Apple and Sun used it. Of course, since PCs didn't use it, the industry standard died and was replaced by a de facto standard, which is arch dependent.

      Also, a) ROM code implies adopting some sort of code execution (ISA dependant, p.e. x86/PPC/MIPS/etc), CPU related.

      Devices already have a ROM code in them, which contains execution stuff to get the card into a working order. All cards have this. On an x86 PC, this code is in x86 machine code, while for PowerPC CHRP and Sun Solaris systems it was in FCode (a Forth bytecode, which is extremely simple to write a virtual machine for).

      But again, as always, the existing legacy support of the PC destroys yet another better idea.

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    4. Re:drivers on cards? by cswiger2005 · · Score: 1
      why can't manufacturers make hardware (esp NICs) with basic 9x/xp/*nx drivers on a ROM chip?

      Oh, they could, only it's exceptionally rare to find a NIC which actually has on-board ROM, and even then it's more likely to be used for PXE booting then for providing a driver when Windows already ships with generic drivers for the common chipsets. However, there were NICs made for OpenFirmware-based systems (Suns, Macs) which contained basic drivers and hardware diagnostic code.

      --
      "The human race's favorite method for being in control of the facts is to ignore them." -Celia Green
    5. Re:drivers on cards? by dotgain · · Score: 2, Informative

      Just a couple of things to add to your post:
      Sun developed it and trademarked it as OpenBoot, which is the name for their impelementation of it. OpenFirmware is the name given to any IEEE 1275 conforming implementation.
      The ROM code is FCode, or Forth Bytecode. This is machine independent, so you can take an IEEE1275 ethernet card from a Sun Sparc, put in a PCI Machintosh, and boot off the LAN right out of the box.
      Scrambling to keep up, BIOS manufacturers added bling like being able to change the boot priority of different devices, and for the system to just carry on booting in the face of a keyboard error. Truly revolutionary.

  10. Aimbot! by GlobalEcho · · Score: 1

    This sounds like a great place to house an aimbot, if those still exist.

    1. Re:Aimbot! by Rupan · · Score: 1

      You would think so, but that isn't the case. This hardware is designed to be separated from the PC hardware, and as such will probably not have access to system memory. If you can't read memory allocated to the game, you can't figure out where the enemy is running around...

      --
      Ads? What ads?
    2. Re:Aimbot! by quantum+bit · · Score: 1

      Sure you can, just decode the packet stream and you'll have access to all of the game state that the client has. If you don't have to reverse engineer the protocol (say something well known like Quake [1-3]), it might even be easier than mucking around in system memory.

      I vaguely recall mentions of aimbots that were designed to run on a separate machine -- either by acting as a proxy, or by sniffing in promiscuous mode and injecting UDP packets.

  11. network card lag? by Fullaxx · · Score: 3, Insightful

    since when is lag caused by your local NIC? So what if you get an extra .001 ms to your router? Never once have I seen my cpu above 5% b/c of network usage, even full network usage. No way is this legit

    1. Re:network card lag? by Anonymous Coward · · Score: 0
      Look if you're not extreme enough for this card just say so. Aerodynamic ethernet packets are a well established technique for avoiding bottlenecks in internet tubes so if you're not going to support your bashing with actual evidence why don't you just sit down and shut up? You're just embarressing yourself.

      (seriously, It's the Monster Cable of NICs,,, wait 'til they start using emotional language to describe packets... "warmer, smoother packets... not cold harsh brittle packets like your run-of-the-mill NIC")

    2. Re:network card lag? by grolschie · · Score: 1
      "warmer, smoother packets... not cold harsh brittle packets like your run-of-the-mill NIC"
      All of us internetphiles know that the best internet experience is good old tubes (youtube pun intended) and not that modern solidstate crap. :-)
    3. Re:network card lag? by MerlynEmrys67 · · Score: 1
      Obviously you haven't gone fast enough...

      Watch what happens when you start pushing a 5-8 Gbit/sec stream through your dual socket server. That will take a rather large bite out of CPU utilization. I have measured CPU utilization on a real OS above 50% running 9 1 Gbit/sec streams through a 10 gig NIC (not TOE) and watched it drop to about 25% with TOE. Those damned copies get you every time

      --
      I have mod points and I am not afraid to use them
    4. Re:network card lag? by Lord+Crc · · Score: 1

      Obviously you haven't gone fast enough... Watch what happens when you start pushing a 5-8 Gbit/sec stream through your dual socket server.

      Yeah coz like, games are so demanding these days... ;)

      For "real" work, what you say is true (or so I guess), but for gaming, who usually doesn't go above ISDN speeds...

    5. Re:network card lag? by Fullaxx · · Score: 1

      thank you
      exactly what i was getting at. Games do not require a NIC CPU.
      They should be marketing this product toward high-performance network servers.

  12. Viral Marketin 101 by Anonymous Coward · · Score: 0

    81) Make sure to include "Linux" to viral marketing ads posted as 'newsposts' on Slashdot to gain support with the Linux Community.

    82) Hardcore gamers will do anything for the promise of better ping, even adding another hop to their upstream

  13. Two reasons. by khasim · · Score: 4, Insightful

    #1. It's more difficult to issue updated software in firmware.

    #2. It's another chip. Software is far cheaper than hardware for OEM's.

    1. Re:Two reasons. by Anonymous Coward · · Score: 0

      Also recovering from a fucked up flash attempt is way harder than recovering from a fucked up driver install. In the latter case, you can always uninstall the driver, reinstall it again, etc. In the former, however, chances are your nice piece of hardware just became a paperweight.

  14. MBA's, huh? by Anonymous Coward · · Score: 2, Funny

    They got Masters degrees in Business Administration, and yet their typing and conversational skills are on the level of 14-year-olds.

    That's just sad.

    Well, that and the fact that their "product" is clearly incapable of giving anything near the boosts they claim it gives...

    1. Re:MBA's, huh? by Anonymous Coward · · Score: 0

      and yet their typing and conversational skills are on the level of 14-year-olds

      That's because their potential customers are unstable 14 yo AMD- and NVidia-fanbois.

    2. Re:MBA's, huh? by Anonymous Coward · · Score: 0

      We prefer the term NVidiots

    3. Re:MBA's, huh? by flight_master · · Score: 1

      Exactly. That's why All the real geeks use ATI and Intel products :D
      Chris

      --
      "Free software" is a matter of liberty, not price.
    4. Re:MBA's, huh? by Ruff_ilb · · Score: 1

      Idk about ATI, but all the real geeks use Intel now. Conroe has completely wasted AMD.

      --
      http://www.TheGamerNation.com/Forums
    5. Re:MBA's, huh? by jack_call · · Score: 1

      Have you already forgotten the FedEx/MBA commercial?
      http://youtube.com/watch?v=8Vx1BTBhg4c
      It's funny because it's true

      --
      This is my sig. There are many like it but this one is mine. My sig is my best friend. It is my life.
    6. Re:MBA's, huh? by MBGMorden · · Score: 1

      Just becuase it's doing better in benchmarks doesn't mean everybody will buy one. Heck a lot of people (even geeks) will buy one or the other just out of preference. Out of my last 8 processors purchased, 6 have been AMD (2 Intels - Celerons overclocked to 550 and 850 mhz). I don't see myself switching any time soon.

      --
      "People who think they know everything are very annoying to those of us who do."-Mark Twain
    7. Re:MBA's, huh? by Gli7ch · · Score: 1

      Why?

      If you were going to buy a new computer now, you're telling me you would buy a new AMD-based system even though you know it's not as fast as an Intel machine in the same price range?

      So basically, you like wasting your money.

      I was a staunch supporter of the AMD64 right up until Conroe. From the best bang for buck and the budget end all the way up to FX60 kicking the crap out of the too-hot-to-handle P4 EE, but I only chose AMD because it was better value for money.

      Now, Conroes offer more performance for cheaper. Why the hell would you buy second best?

    8. Re:MBA's, huh? by MBGMorden · · Score: 1

      I like AMD more as a company. IMHO they compete more fairly than Intel and as such I'd prefer to give them my business.

      There's also the issue that I don't buy new computers - I build or (more likely) upgrade. My motherboard is already for an AMD processor, so switching to the Intel platform would take more $$$.

      --
      "People who think they know everything are very annoying to those of us who do."-Mark Twain
  15. Is this really needed? by adamwright · · Score: 5, Insightful

    As a small test, I ran up Quake 3 on it's highest settings, and had it play back a reasonably heavy demo. Now, Quake3 isn't the most modern of games, but it can still peg a CPU at 100%. Then, I found the latency to my router.

    Pinging 192.168.0.1 with 32 bytes of data:

    Reply from 192.168.0.1: bytes=32 time1ms TTL=255
    Reply from 192.168.0.1: bytes=32 time1ms TTL=255
    Reply from 192.168.0.1: bytes=32 time1ms TTL=255

    Assuming this product entirely eliminates all latency on the first hop (impossible), that's a net gain of 1ms.

    The entire concept of these FNApps also strikes me as a route to evil; I heard a subtext of "Now, even the most clueless Windows gamer with too much money can run packet scanning cheating tools with no chance of detection!".

    I'm placing this one firmly in the "Snake oil" bin, based on this interview.

    1. Re:Is this really needed? by snowgirl · · Score: 1

      The entire concept of these FNApps also strikes me as a route to evil; I heard a subtext of "Now, even the most clueless Windows gamer with too much money can run packet scanning cheating tools with no chance of detection!".

      Ooo... something even bigger of an issue also. Now black hats don't even have to own your real computer to use you as a zombie, they can just own your NIC.

      person who actually fell for buying one of these: "Why is my network so laggy? I thought this $250 NIC was supposed to help with that?"
      person running packet sniffer: "Oh, here's your problem, you're part of a DDoS on killernic.com

      --
      WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
    2. Re:Is this really needed? by dorath · · Score: 2, Informative

      They also claim to increase your FPS by offloading other apps onto the NIC, namely a music player but apparently also file sharing. With a mic jack one would think it could also run voice communication software.

      Being at work, I'm not in a position to check FPS while running just the game vs the game, music, and chat. :-(

      I'm placing this one in close proximity to the "Snake oil" bin.

    3. Re:Is this really needed? by asuffield · · Score: 2, Interesting

      ping doesn't put much load on your networking stack. Try that with large UDP (or better yet, TCP) frames, at a rate which actually saturates your network, on a gigabit network - and you'll start to see what network card latency is all about (hint: you'll never get the last 300Mbit out of that gigabit network without doing something about the latency of your network card and kernel's networking stack).

      Not sure why it's relevant to games, but low-latency high-intelligence network cards are very important for heavy duty servers.

  16. Retarded, by u16084 · · Score: 5, Insightful

    If your ISP sucks ass, a $250 lan card is not going to help.

    --
    -- I Dont Deserve A Sig I Have Bad Karma
  17. Linux by agentdunken · · Score: 0

    Linux owns in everything hardware/software based. Everything today has Linux in it even if you don't know it. PDAs, Cell Phones, TV's, DVD players, Cars, Airplanes, even the U.S. Army robots use Linux. iRobot is famous for using Linux as the OS in their robots.

    --
    Linux, because a PC is a terrible thing to waste.
    1. Re:Linux by bbingham · · Score: 1

      I am ignorant - why does this card need to host a version of Linux? Wouldn't a much less expensive chipset running embedded Forth or some other such environment do the same job faster and better? Is the upside of writing card-specific applications really that great? Why would UDP processing require a 400Mhz cpu and consume 5-10W?

    2. Re:Linux by Miniluv · · Score: 2, Insightful

      Because this way they had 7 bullets of bullshit instead of 1.

      With just 1 they would've had to only charge $199.99.

    3. Re:Linux by dotgain · · Score: 1

      Mmm, I can see all the devs lick their lips while they relish writing a TCP/IP stack complete with QOS in Forth!

  18. Linux by Anonymous Coward · · Score: 0

    Linux is the best

  19. So it's a QoS Network Card? by tlhIngan · · Score: 5, Insightful

    The only thing I can guess it needs Linux for is to do the routing and QoS services (see lartc.org)...

    Then again, considering I get sub-1ms latencies across my network (only 100Mbps...), and this is with some rather pathetic equipment (Celeron system running Win2k), I fail to see how I can improve my 80ms ping with a better network card.

    It seems that hardcore gamers are starting to become the computing equivalent of the "audiophile". From CRT displays that do 120hz refresh (do they notice the difference between 100 and 120, I wonder?) since LCDs that do 6ms are "too slow". Gaming mice that do 10k-dpi for ultra-precise positioning, videocards that cost the better part of a grand. And now, network cards that cut down microseconds or give you that extra frame per second. There's also keyboards, the gaming mousepad (though, some are nice for general use), and god knows what other accessories, doodads and other monster-cable-type things.

    1. Re:So it's a QoS Network Card? by bcmm · · Score: 4, Interesting

      This is offtopic, but I have to mention this while we're talking about audiophiles. About a month ago, I saw in a shop a device even more blatantly pointless than this NIC. It was an "A/V USB cable". Gold-plated. That's right, ordinary USB cables are not good enough for running a projector, presumably because those cheap stainless steel USB connectors introduce too much noise into the (digital) signal.

      (For anyone who doesn't frequent the same shops as crazy people, it is common to gold-plate the connectors of analogue audio connectors to improve the quality of the signal. Presumably the untarnishable gold reduces the resistance of the connection. This gets taken to rather silly extremes when gold-plated 3.5mm connectors are marketed for use with low-quality stuff like MP3 players.)

      --
      # cat /dev/mem | strings | grep -i llama
      Damn, my RAM is full of llamas.
    2. Re:So it's a QoS Network Card? by ScrewMaster · · Score: 2, Interesting

      The word is "elitism". There will always be people willing to lay down serious money to maintain their egos. Cars, stereos, guns, computers, home theaters, women ... you name it, somebody will pay too much for it just to get that special "God, I'm just so much better than everyone else" feeling. Of course, most elitists are in reality fools (and if male, typically equipped with miniscule sexual apparatus) but if you tell them how idiotic they look they'll just go spend even more money to prove you wrong. That can be entertaining, actually ... just keep pissing them off by saying things like "yeah yeah, that's cool, I guess, but you know my friend Bob has twenty-seven terabytes running on his home network" and watch their credit-card balances soar.

      --
      The higher the technology, the sharper that two-edged sword.
    3. Re:So it's a QoS Network Card? by brainnolo · · Score: 1

      There are lots of this kind of cables. Actually with high-speed device (mostly firewire) they could make *some* sense, because the digital data, while strong, is not immune to interferences. Error correction would fix the few transmission errors caused by a bad cable, but this would slow transmission down. Now, the gold plating itself makes no difference, but they usually gold plate the connector as a way to highlight the good quality of the cable. This said, if a gold plated cable costs 10x a standard cable, then is a scam targeted that "audiophile"-alike people.
      This is NIC is also mostly useless. They were probably trying to develop a way to run p2p/server applications that would run directly on the NIC (the FNA thing), noticed too late it would be of no use and tried to sell it to gamers, who already buy crazy things for no reason at all.

    4. Re:So it's a QoS Network Card? by Anonymous Coward · · Score: 0
      From CRT displays that do 120hz refresh (do they notice the difference between 100 and 120, I wonder?)
       

      FWIW, I can tell the difference between 100Hz and 120Hz. It's not a huge difference, unlike the difference between 75Hz and 85Hz, and certainly not worth paying $100+ for, but there is a noticable difference for me. I'm probably in the minority though.

      Interestingly, as sensitive as I am to CRT refresh rates, LCD response times as low as 25ms don't bother me at all; I can't even tell the difference between an 8ms panel and a 16ms one.

    5. Re:So it's a QoS Network Card? by Sycraft-fu · · Score: 1

      It actually makes good sense. You either gold or silver plate depending on how it will be used. Pro gear is generally silver, home gear is gold. You use gold because it doesn't tarnish and is somewhat soft. That makes it good for a connection that's going to be hooked up and not messed with, it won't corrode and so on. Silver is good in that it's quite conductive and won't rub off with repeated pulgging and unplugging and that action will keep it fairly untarnished.

      I've actually had problems with pro gear because of that. I use pro gear, but in a consumer fashion (never moving it around). So my sound was a little off one day, couldn't say what, jsut wasn't sounding right. I messed with the cable going to the amp, got a nice crackling sound. Disconnected it, cleaned the contacts, reconnected it, everything worked fine.

      For a USB cable, ya that's pretty stupid.

    6. Re:So it's a QoS Network Card? by Anonymous Coward · · Score: 0

      'It seems that hardcore gamers are starting to become the computing equivalent of the "audiophile".'

      Got to take issue with your choice of words there. Hardcore gamer doesn't necessarily mean rich - a stereotypical hardcore gamer is someone at school/college with too much time and not enough money. I think the term you probably want is "hardcore gaming spender" - we need a good slang term for it, can you think of one?

    7. Re:So it's a QoS Network Card? by slackmaster2000 · · Score: 5, Interesting

      Yeah audiophiles are very easily fooled. I laugh every time I hear about digital technology being described in analog terms, which is terribly common in the audiophile realm. It's sort of understandable though because "digital" is so variable whenever we interface with the analog world. Not all AD/DA converters are created equal. Audio CD playback is very similar to analog technology in that it's designed play right through and mask errors which can result in high frequency distortion and various other artifacts that one might not expect from "digital" and how that term is marketed. It can be difficult to understand, so why wouldn't the quality of a USB cable be as important as the quality of a microphone cable or the weight of a speaker cable? Well, most of us would say "duh", but the second big problems is that these people can "hear" the difference! Ah the power of suggestion. You won't find a lot of double blind studies published by audiophile gear manufacturers.

      Aside from $30,000 speaker cables (I shit you not), my most favorite audiophile product is a wooden knob that costs $500. "What does the wooden knob do, though?" Well, nothing on its own. It is in fact, a wooden knob. A knob made from wood. It sounds great when attached to your volume pot, though. Five hundred dollars please.

      So today we learn about a network card that somehow reduces lag by implementing hardcore quality of service on an endpoint that is for all intents and purposes dedicated to a single application. Are gamers analogous to audiophiles? "Quake III is a lot more responsive now. My ping is about the same, but I can feel the difference." It must work though. I mean, who would make something and charge so much for it if it didn't actually work? :)

    8. Re:So it's a QoS Network Card? by Anonymous Coward · · Score: 0

      Oh, the best is the gold-plated SPDIF cable, and *not* of the coax kind.

    9. Re:So it's a QoS Network Card? by slashflood · · Score: 1
      It was an "A/V USB cable". Gold-plated.
      You mean this one?
    10. Re:So it's a QoS Network Card? by slashflood · · Score: 2, Informative
      my most favorite audiophile product is a wooden knob that costs $500
      Here it is! You can also spend 7000 bucks for a volume control - a must have.
    11. Re:So it's a QoS Network Card? by slashflood · · Score: 1
    12. Re:So it's a QoS Network Card? by asuffield · · Score: 1
      About a month ago, I saw in a shop a device even more blatantly pointless than this NIC. It was an "A/V USB cable". Gold-plated.
      As I understand it, the primary reason for this is because a tiny layer of gold plating is cheaper than making the connector out of stainless steel (and you don't want to have any contact surface - analog or digital - made out of a metal that corrodes just from humidity, since that causes ignorant people to return the product to the store).
    13. Re:So it's a QoS Network Card? by Hamoohead · · Score: 2, Insightful

      Some of you may remember the CD pen? A green magic marker that you use to color the edges of your CDs so the laser beam wouldn't leak out the edges? Check out http://www.halfbakery.com/idea/Audiophile_20Hearin g_20Test for details. Or how about the CD demagnitizer (http://www.gcaudio.com/cgi-bin/store/showProduct. cgi?id=190) to remove that annoying light-bending magnetic field from your precious collection. I could go on, but not without deviating more from the thread topic.

      Notwithstanding that the average person sees/hears no difference between these "tweaks" and the normal off-the-shelf fair, companies like this are lucrative enough in that small niche market to make a profit. People with "ophile syndrome" will never be convinced that their perceptions are somehow flawed. As a former manager of an audiophile store and a technician of 30 years, I can attest to that. The more you argue, the more they will be convinced that you are unable to appreciate the finer points of whatever they are into.

      Is this net card snake oil? Probably. Is it stupid to pay $279 for a network card to gain 1-2 FPS advantage? I won't buy one, but value is a matter of perception (however flawed). Yes *ophiles are a curious bunch with deep pockets. The extension to the computer market was inevitable. I'm just surprised no one thought of this marketing angle sooner.

      --
      "If your parents never had children, chances are you wonât either." -Dick Cavett
    14. Re:So it's a QoS Network Card? by dbIII · · Score: 1
      The only thing I can guess it needs Linux for is to do the routing and QoS services
      For at least the last two years you could get a very good firewall/router on a network card for a lot less that can do this - based on uClinux and with a good web based configuration interface. I don't really know how much good it would do for a single user - it's the sort of thing you put in a gateway when space is tight or if you want to let the card handle a virtual private network connection instead of doing it in software.
    15. Re:So it's a QoS Network Card? by dotgain · · Score: 1

      Thanks for the links! Best laugh I've had all week!

    16. Re:So it's a QoS Network Card? by sjames · · Score: 1

      About a month ago, I saw in a shop a device even more blatantly pointless than this NIC. It was an "A/V USB cable". Gold-plated.

      At Best Buy, those are 'Geek Squad Approved' (but the regular steel ones aren't) and cost $35

    17. Re:So it's a QoS Network Card? by Oldav · · Score: 0

      I concour with the last poster, that was the funniest thing I have seen in ages, surely its a joke site, I mean no-one is stupid enough to actually believe that stuff are they, tell me it ain't so!

    18. Re:So it's a QoS Network Card? by Klintus+Fang · · Score: 1

      I like the gold plated optical cables the best. :-)

      --
      In a minute there is time For decisions and revisions which a minute will reverse. -T.S. Eliot
    19. Re:So it's a QoS Network Card? by Shanep · · Score: 1

      It must work though. I mean, who would make something and charge so much for it if it didn't actually work?

      Years ago, I bought one of those audio magazines which are released once a year or so which tabulates the specs of all the audio products currently on the market, complete with prices.

      I was astounded to see that there was a small CD player, perhaps as large as an alarm-clock-radio. The short quote from a previous review had raves for it, but stated that the build quality was also like a cheapo alarm-clock-radio. This CD player (a Meridian) was $50,000 US, yet it had worse specifications, in every respect, than a $100 US Marrantz CD player (which got a ho-hum review). The Marrantz had better THD, SNR, Dyn Range and very flat response and fine build quality, yet got a crappy review compared with the $50,000 piece of trash.

      I could only conclude that the reviewer falls into the category of, "If it is expensive, it is good. If it is cheap, it is bad".

      I wanted the Yamaha CDX-1060. Because it had pretty much the best specs of all of the players, yet was only a few hundred bucks. Unfortunately, I missed out and they stopped making it before I could buy one. I might be able to get one second hand, since the CDX-1060 is still traded as a legend. But for me, the convenience of my iRiver H340 mp3 player is just too good to go back to CD players.

      --
      War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
    20. Re:So it's a QoS Network Card? by proxima · · Score: 2, Interesting
      I like the gold plated optical cables the best. :-)

      Ah, yes, and leave it to Radio Shack to sell them.
      --
      "The universe seems neither benign nor hostile, merely indifferent." --Carl Sagan
    21. Re:So it's a QoS Network Card? by Suddenly_Dead · · Score: 1

      I have never had my stainless steel speaker connectors tarnish or corrode, let alone do so enough to reduce audio quality. It's all a load of bullshit. Any problems that there ever are can be fixed by merely cleaning the thing, as you said.

    22. Re:So it's a QoS Network Card? by TenMinJoe · · Score: 1

      I'll see your gold-plated USB cable and raise you a gold-plated OPTICAL cable. Because gold stops... the light... falling out. Or something.

    23. Re:So it's a QoS Network Card? by im_thatoneguy · · Score: 1

      "Video cards which cost the better part of a grand..."

      Speaking for all of those who own video cards which cost more than a grand: it's worth every penny. Just because a game has been optimized to run on hardware available now doesn't mean there aren't 3d and openGL applications out there which are far more taxing than any game available today. Also high resolution mice provide another level of skip prevention, especially on uneven surfaces. Are they worth the cost? Maybe, but that's up to the user.

      I also own a $100 component video cable... and yes it did make an appreciable difference over the cable that comes with your average big box DVD player.

    24. Re:So it's a QoS Network Card? by slashflood · · Score: 1
      I concour with the last poster, that was the funniest thing I have seen in ages, surely its a joke site, I mean no-one is stupid enough to actually believe that stuff are they, tell me it ain't so!
      I'm sorry, but it's all real. The knob is made in switzerland by the same one-man-company that builds the potentiometer. You can also buy the "C37 laquer" for 215 bucks/2oz from the inventor. Whatever you paint with this laquer sounds better. From a wooden volume knob to metal cd player cases.
    25. Re:So it's a QoS Network Card? by Anonymous Coward · · Score: 0

      I don't think Meridian have sold a CD player the size of an alarm clock in 20 years, and even then it was a cheap Philips with a custom logic board. Also, the only player Meridian have sold even approaching $50,000 is 6U rack-mount, so hardly the size of an alarm clock.

      I suspect if what you say is true, that the rag you read was mad and full of nonsense. Also, don't trust most magazines to actually do the type of test on Hi Fi kit where you actually get numbers for results - most are hopelessly wrong and badly done.

    26. Re:So it's a QoS Network Card? by bcmm · · Score: 1
      No no. The volume knobs are REAL. The site even explains how they work!
      The point here is the micro vibrations created by the volume pots and knobs find their way into the delicate signal path and cause degradation (Bad vibrations equal bad sound).
      ROFL.
      --
      # cat /dev/mem | strings | grep -i llama
      Damn, my RAM is full of llamas.
    27. Re:So it's a QoS Network Card? by bcmm · · Score: 1
      I dont think it was the same one... There must be a lot of them about. However, that site has probably the best quote out of all of these:
      "minimizes noise and interference."
      Such a total misunderstanding of digital data transfer.
      --
      # cat /dev/mem | strings | grep -i llama
      Damn, my RAM is full of llamas.
    28. Re:So it's a QoS Network Card? by Anonymous Coward · · Score: 1, Funny
      No no. The volume knobs are REAL. The site even explains how they work!

             
      The point here is the micro vibrations created by the volume pots and knobs find their way into the delicate signal path and cause degradation (Bad vibrations equal bad sound).


      ROFL.


      Yeah they're made from old Stradivaris.


  20. Makes me wonder what the prices by Moryath · · Score: 1

    of a few other things are these days.

    Like say, snake oil. Or those magnetic gewgaws that are supposed to give you 500% better fuel efficiency in your car.

    Or, other crap that doesn't work.

    1. Re:Makes me wonder what the prices by AuMatar · · Score: 1

      I think its more in line with gold plated speaker cables and $300 power supplies.

      --
      I still have more fans than freaks. WTF is wrong with you people?
  21. It must be good !! by Midnight+Thunder · · Score: 4, Funny

    It must be good! Have you seen the size of the fan on that thing ;)

    --
    Jumpstart the tartan drive.
    1. Re:It must be good !! by phase_9 · · Score: 1

      I love the way there is a serios lack of actual PCB on the card - hence the massive fan / headsink / plastik-crap (tm) on it.

  22. $280 pre order? by Anonymous Coward · · Score: 0

    I guess this gives rise to a new saying: A n00b and his money are soon parted.

    1. Re:$280 pre order? by DarthStrydre · · Score: 1

      The axiom that we have had around for a couple years is...

      "A n00b and his AWP are soon pwn3d."

      I can't believe I wasted the time to post that... :-P

  23. The telling comment at the end by grasshoppa · · Score: 4, Insightful

    All the founders met at the University of Texas while getting their MBAs.

    That says all that needs to be said for the article.

    --
    Mod me down with all of your hatred and your journey towards the dark side will be complete!
  24. I've said it before... by BertieBaggio · · Score: 4, Insightful

    This, of course, was covered earlier. And I still agree with the tag - I think it is snake oil.

    Let's try and remember a few fundamentals. As per RFC 1925, "The 12 Networking Truths":

    [2] No matter how hard you push and no matter what the priority, you can't increase the speed of light.

    (Déja vu? Yes!)

    Right on. This card might process incoming data quicker, or perhaps even send the data to the CPU faster, but it won't reduce latency. The high price ($280? TFA is not responding) does not justify the alleged 'improvements' in lag this card offers. Games communicating over UDP like BF2 have fairly low lag anyway (when they stay connected...). As others have said: spend the money on RAM or some other upgrade. The 'lag' improvement will be much more cost-effective.
    --
    If all you have is a grenade, pretty soon every problem looks like a foxhole -- MightyYar
    1. Re:I've said it before... by Suddenly_Dead · · Score: 1
      Games communicating over UDP like BF2 have fairly low lag anyway


      Are you playing the same BF2 as I'm playing?
    2. Re:I've said it before... by yalla · · Score: 1

      I haven't looked into that "Killer-NIC" (not interested anyway) but i know a bit about latency; people building cluster are investing a great deal of time reducing the latency the operating system's TCP/IP stack introduces. This can make up a difference of ~100-microseconds. So chosing the right network technology to deliver your data from one host to another one takes a great deal of time and lot's of effort is put into the research of new networking-technologies.

      See Infiniband, or the various TCP-offloading features, take Level 5's Etherfabric or projects like MPI/GAMMA who circumvent the whole IP-shebang directly and let the programm build up it's own Ethernet-frames.

      So i wouldn't necessarily denouce the whole "Killer-NOC" thing as bullshit; but on the other i must admit: The network in between (speak: Internet) is the problem for gaming, NOT necessarily the IP-stack or NIC.

      I'm not a gamer so i can't tell if a difference in latency of, let's say, 100 us really brings an advantage (he, folks, that's 0.01 milliseconds!) - but well, let's see what kinda specs the company is going to release.
      Alex.

      P.S.: The Cluster Monkey had quite an extensive comparison of different low-latency high-bandwdith network-technologies: http://www.clustermonkey.net/content/view/124/34/

      --
      You look like a million dollars. All green and wrinkled.
    3. Re:I've said it before... by yalla · · Score: 1

      I'm saying over and over again: Never post before you had your first coffee:
      > he, folks, that's 0.01 milliseconds!

      It's 0.1 ms, yeah, i suck, i know. ;)
      Alex.

      --
      You look like a million dollars. All green and wrinkled.
    4. Re:I've said it before... by BertieBaggio · · Score: 1

      Are you playing the same BF2 as I'm playing?
      Yes. Battlefield 2 has quite low pings for me (reported as low as 14ms on some servers), although I have a feeling this is a misrepresentation. Even still, the point I made is that this card will almost certainly not help any 'lag' you are suffering from. A big component of 'lag' in BF2 (and other such games) is graphics processing (GPU/CPU/memory-bound), not network latency. Depsite what TFA(s) want you to believe with their OTT marketing, spending $280 on this card will not help! Upgrade or buy a new, different component instead.
      --
      If all you have is a grenade, pretty soon every problem looks like a foxhole -- MightyYar
    5. Re:I've said it before... by Suddenly_Dead · · Score: 1

      I get low pings in the game as well, but I still see network-related lag, which was the point of my comment. It's fairly obvious when it's network lag, as your player will suddenly jump a few metres ahead or opponents will do the same. No fancy card can make up for bad netcode.

  25. Look who's in charge of marketing by sammydee · · Score: 1

    " Bob Grim, Vice President of Marketing, has a wealth of sales and marketing experience from his time in Marketing at AMD"

    Well this product is practically guaranteed tobe a complete commercial success then...

    1. Re:Look who's in charge of marketing by BrainRam · · Score: 1

      Ah, so this is the latest in Grim's Fairy Tales... Oh, wait, sorry, he's short an 'm'.

  26. Big laffs! by Ancient_Hacker · · Score: 2, Funny
    The best lie is the boldest.

    Of all the things a spiffy network card could do, reducing latency is just about the least likely.

    They could have put 20mb of buffers on it.

    They could buy glow-in-the-dark pc board material.

    They could have put a handful of bright blinky led's on it.

    They could even put on a 12AX7 vacuum tube to do something useful.

    They could put built-in auto ping.

    But what do they do? Put another layer of OS glop in the way. Big laffs!!

  27. Hardcore gamers use IPX by Jimmy+King · · Score: 0
    From TFA:
    Many network products today claim to 'offload' network calculations (like checksum, tcp segmentation, etc.). Those technologies are usually only for TCP/IP networking (which most games that Hardcore Gamers play don't use).
    What? This may be a neat idea, although has some obvious issues as others have pointed out, but I can't trust a word these guys say after reading that.
    1. Re:Hardcore gamers use IPX by Anonymous Coward · · Score: 0

      I assume you have heard of UDP/IP, and are fully aware that this is the most prevalent protcol for real-time data?

    2. Re:Hardcore gamers use IPX by Jimmy+King · · Score: 1

      If that's what they meant, that's an ass backwards, unclear as hell way to state it. While not necessarily technically correct, most people refer to the IP suite as TCP/IP, and so would include apps that are using UDP as using TCP/IP. Looks like they sort of clear it up a few lines down, but I stopped reading after that first line.

  28. This is such BS by McNihil · · Score: 0

    I have never seen such a blatant premature optimization. Why have the IP stack in hardware and "pre-cache" packets for the host OS? Much cheaper to make good network drivers in the first place (and if you have latency problems you have them because of CPU depletion and no pre-caching will for ever hide that.) These MBA students should get an A+ if this crap gets anywhere above ground level.

  29. What a waste of money by spun · · Score: 2, Insightful

    What kind of geek wannabe would waste money on this? Nowhere on their site do they show benchmarks or even vague references to how much this will speed up your networking or FPS. Oooh! it offloads network processing, leaving your CPU free to PLAY THE GAME!!! That's probably going to speed things up by like .5%. AWESOME!!! TO THE MAX!!!

    This is the tech equivalent of herbal viagra.

    --
    - None can love freedom heartily, but good men; the rest love not freedom, but license. -- John Milton
    1. Re:What a waste of money by Raxxon · · Score: 1

      Gee.... TOE Offloading is an option on several of the Gig Nics I've seen. What else really needs to be moved from the CPU/OS to the NIC?

  30. 400 MHz? by Anonymous Coward · · Score: 1, Interesting

    Somehow it seems unlikely that a 400 MHz "32 bit" CPU (probably an ARM or a Geode) running Linux can increase performance relative to a 3 GHz or so CPU running Windows. In fact, it seems most likely that this would actually slow down networking performance, especially in heavy traffic situations.

  31. Will game soundtracks sound warmer, too? by brainnolo · · Score: 2, Funny

    That's right! The card has high quality vacuum tubes and a special magnetic stone that will make the sound much warmer. Another great feature is that this NIC is so powerful only your packets will get to the server, so nobody can shoot at you! Seriously now, what were these guys trying to do? Probably the card was created for the FNA thing, then when they found it had no application at all, they tried to find a market for it.

  32. Wow. Just wow. by MBCook · · Score: 2, Interesting

    OK, I thought gamers were suckers (paying $600+ for graphics cards) but really.

    A $280 network card.

    But wait, there's more!

    It's also... a Linux box. And not just a Linux box, an "Open Source Linux" box.

    Plus it has USB so you can connect a hard drive or headset???

    OK, the basic idea is interesting. Offload all the TCP/UDP/IP processing. I have to wonder how much impact that would really have. But how does the data get onto the host computer? If it's via a driver that shows up as a NIC, then it still has to go through the network layers of the OS. If it shows up as some kind of memory, then the host applications must be written to use it. The idea of offloading a few other features too (like voice chat) is nice too, but again, you'd have to write special software or drivers or something on the host OS to use that.

    And you can use it for a hard drive. If they open it, background bittorrent anyone?

    Or you could just let your NIC have a hard drive for fun that you can't access. Genius!

    Look, if they had a little ARM processor and it did the network stuff only, that would be cute. But I think they over built it, it's over priced, and I seriously doubt it has much impact.

    I wonder if they'll make Linux drivers available *smirk*

    --
    Comment forecast: Bits of genius surrounded by a sea of mediocrity.
    1. Re:Wow. Just wow. by Jimmy+King · · Score: 1

      I "think" the point of the USB is so that you can use the nic as a seperate computer. It's kind of a neat idea if it didn't cost so much. My first thought was using it as some sort of poor man's blade server... get 3 or 4 and stick em in your pci slots. But not at $280 each, I could just buy a single computer that would outperform all of them combined for less. For a much more reasonable price it has some potential and niche uses. Of course, I don't trust these guys one bit. Their comments, such as (paraphrased) "most games today don't use tcp/ip" show that they don't know shit and are horrible liars. I'm sure there was more, but I stopped reading after that.

    2. Re:Wow. Just wow. by MadMidnightBomber · · Score: 1
      OK, the basic idea is interesting. Offload all the TCP/UDP/IP processing. I have to wonder how much impact that would really have.

      Endace make some decent offload cards, suitable for snorting hundreds of Mb/s and above, if that's what you're looking for.

      --
      "It doesn't cost enough, and it makes too much sense."
    3. Re:Wow. Just wow. by Anonymous Coward · · Score: 0

      Bittorrent traffic will be cached on the NIC for faster download.

  33. No wonder he was so concerned about lag... by Jonah+Hex · · Score: 1
    As a side note, Harlan met his wife in Ultima Online, married her in the game, and then eventually married her in real life.
    As a former Chesapeake player (DogMeat[MoO] of Oberon Pass) I can just see the scenario...

    Do you take this woodelf to be your...
    PKs!!1! Recall1111
    Corp Por Corp Por Corp Por
    Oooo ooooo oo oo ooooooooo ooooo ooooo Translation: I'm going to do something about this!

    Jonah HEX
  34. kinda cool by digitalsushi · · Score: 1, Interesting

    This sort of looks like a crappy TOE - TCP Offload Engine. You can get a TOE NIC from Chelsio for a grand that'll do gig rates. Anyone ever try to get gigabit speeds out of their NIC? It's not so easy. It takes a lot of overhead to encapsulate data inside ethernet frames. Offloading that job to your ethernet card is a nice way to keep your CPU doing the stuff you want it focused on.

    It's sort of clever, I think. If your CPU is pegged calculating physics for a video game, or however you kids crunch math, having the NIC doing the actual packetization of your location info is a small step towards getting better response times. Honestly, I could see this being like those riceboys -- adding so much "bling" to your car that it actually slows it down in an attempt to make it look faster. *Shrug* Either way, I think this company will sell a few of these, and by a few, I literally mean few.

    The fastest we've ever gotten a machine to spit data out the line was with a 10 gig ethernet card, with a TOE in it, and that rate was 1.2 gigabit/second. The bus could handle 7.6 gigabit max, but we got nowhere near that due to the framing involved. We were just piping /dev/random over a netcat connection and into /dev/null, too. Couldn't figure out how to do anything faster. Any ideas? Cause we were stumped.

    --
    slashdot: where everyone yells sarcastic metaphors to themselves to understand the issue
    1. Re:kinda cool by slashjunkie · · Score: 1

      The fastest we've ever gotten a machine to spit data out the line was with a 10 gig ethernet card, with a TOE in it, and that rate was 1.2 gigabit/second. The bus could handle 7.6 gigabit max, but we got nowhere near that due to the framing involved. We were just piping /dev/random over a netcat connection and into /dev/null, too. Couldn't figure out how to do anything faster. Any ideas? Cause we were stumped.

      It's possible you only had enough entropy to generate 1.2Gbps of data from /dev/random. You could try sourcing data from /dev/zero. Although it will be a very repetitive pattern of bits, I don't think that should affect the outcome. Or just source data from a bigass file, but this will be tainted by the speed of your hard disks.

      Also, think about whether you're interested in measuring TCP throughput, UDP throughput (which I'd expect to be a litte higher), or raw Ethernet throughput (which will be as close to the theoretical limit of the card as possible).

    2. Re:kinda cool by KingMotley · · Score: 1

      Use windows and enable the advanced TCP options.

      Linux doesn't yet support some of the more advanced features of TCP/IP. At that rate, I'm guessing that you are having a problem with the ACK window being too small (Forcing the sender to stop sending). The original 64k ack window is woefully small over either high latency links, or over high bandwidth links (Your situation being the later). I think they are working on adding it to the 2.6.17 or 2.6.18 linux kernel, but it's still very buggy. You may need a TOE capable card that can also do the ack window scaling options of TCP. No sure if the one you mentioned does or not.

      google "ACK window scaling" for more information.

    3. Re:kinda cool by Percy_Blakeney · · Score: 2, Informative

      Linux has had TCP window scaling for a long time. The recent "bugginess" with it came from a change to its default settings that caused issues with a non-compliant device on the Internet. Read about it at KernelTrap.

    4. Re:kinda cool by Anonymous Coward · · Score: 0

      I am assuming you are attached to Ethernet. If not look to the more specialized vendors, such as Myrianet. Also feel free to look at InfiniBand (IB) devices. These solutions get pricey.

      Try using Jumbo frames (needs network configuration, too) and doing a cat of /dev/zero instead. Make sure you are sending packets to a machine on the local network as routing requires a Layer 3 decision that will kill the performance of the stream. Make sure that you have a "wirespeed" switches, also. This will help to ensure that the switch in not being a bottleneck. Better yet, use a cross-over cable.

      I am certain that I can improve my network performance on World of Warcraft if I were colocated with their server.

    5. Re:kinda cool by pc486 · · Score: 1

      Like another poster said, /dev/random is probably your problem. Over here we're running Linux with a Myricom 10G card on a PCIe bus. 9.9Gb/sec is easy enough to do so long as you have the right equipment and enough processing power (you'll see about 30,000 interrupts/sec).

    6. Re:kinda cool by StuartFreeman · · Score: 1

      First, I'd use iperf since it is designed for this kind of testing. Additionally, I'd adjust my TCP buffer limits by appending the following lines to /etc/sysctl.conf:
      # increase Linux TCP buffer limits
      net.core.rmem_max = 314572800
      net.core.wmem_max = 314572800
      net.core.rmem_default = 65536
      net.core.wmem_default = 65536

      # increase Linux autotuning TCP buffer limits
      # min, default, and max number of bytes to use
      net.ipv4.tcp_rmem = 4096 87380 314572800
      net.ipv4.tcp_wmem = 4096 65536 314572800
      # number of pages, not bytes
      net.ipv4.tcp_mem = 195584 196096 314572800


      Then run 'sysctl -p' to read in the changes. (sysctl and sysctl.conf may not be available on some distros, in this case you'd have to echo the values into the correct places in proc; google is your friend).

      --
      This is my sig, there are many like it, but this one is mine...
    7. Re:kinda cool by longword · · Score: 1

      /dev/random you say? Those random bits aren't free you know...

      What you need is a random accelerator card.

    8. Re:kinda cool by chris_eineke · · Score: 1
      Offloading that job to your ethernet card is a nice way to keep your CPU doing the stuff you want it focused on.
      Sounds like the wheel of reincarnation.

      We were just piping /dev/random over a netcat connection and into /dev/null, too. Couldn't figure out how to do anything faster. Any ideas? Cause we were stumped.
      Sure! Read from /dev/null instead of /dev/random. What's happens in /dev/null, stays in /dev/null. :P
      --
      "All you have to do is be fragile and grateful. So stay the underdog." Chuck Palahniuk, Choke
    9. Re:kinda cool by belg4mit · · Score: 1

      /dev/zero. /dev/random is also blocking on entropy which is bad... /dev/urandom is better.

      --
      Were that I say, pancakes?
    10. Re:kinda cool by KingMotley · · Score: 1

      It has had a limited version of window scaling for a long time. It wasn't until version 2.6.17 that they actually started to implement a more robust solution, which caused the above problems as posted on KernelTrap. The problem is in the way they "cap" the window scaling in prior versions. I believe even in 2.6.18 and beyond, the window scaling option is still capped, but now it's capped at 6. Of course this all stems from the fact that the TCP stack limits the amount of memory it'll allocate to a maximum of 1/128th of the available memory. For a 512MB system, that means 4MB maximum will be allocated (which is the default maximum), and since the window scale must be no more than 1/2 of the receive buffers, 2MB is the maximum window size (2^6*(64k-1)), which corresponds to a window scale of 6.

      Ok, just to be fair. You can also download the latest linux betas, go through the TCP network stack, change the source to remove the memory limitation checks or modify the formulas, recompile your kernel (and hope nothing else breaks), and you'll have the same effect as windows by default. Then after that, you can make a few changes to a configuration file (Pretty much the same numbers you need for the windows registry), and you're good to go.

  35. Snake oil by MetricT · · Score: 1

    This card is meant for the "audiophile gold cable crackpot" market. It offers no real-world benefit, at a very substantial cost.

    This sort of card *might* (might!) be of use in a server environment where you're trying to transfer gigabits of data at a whack. In fact, they already have it. It's called TCP Offload Engine (TOE). Unless you are rendering Doom 3 on your Beowulf cluster at 1600x1200 and sending the raw uncompressed data over ethernet, you aren't going to need this sort of card for gaming.

    For lower latency, you need to ditch ethernet altogether and use a different switch fabric, like Myrinet or Infiniband. Again, not for video games.

  36. QoS might have an effect by Aelix · · Score: 1

    QoS (i.e. "Gamefirst" I guess?) can have a very significant effect on ping, so it might not be entirely snake oil. However if there's more than one person on your network it's useless to have it on the NIC, and you can get decent, _cheap_ QoS on a custom firmware on a WRT54GL router. You could also get the same effect from just turning off your bloody P2P. While not entirely snake oil, this is pretty useless, people will buy it though because while there's $300++ video and sound cards, there isn't a network card that's that expensive, and you know... err... more money == better?

  37. But... by Shadyman · · Score: 1

    Does it run Linux?

    1. Re:But... by the_humeister · · Score: 1

      Apparently it does. A better question is whether or not you can fully access Linux on the card. It would be an interesting way to get multiprocessing on one computer, just install a bunch of these cards.

    2. Re:But... by Anonymous Coward · · Score: 0

      and once you find out that the onboard CPU is a 75Mhz ARM CPU, you'll be cranking it out, wont you?

  38. qos by trybywrench · · Score: 1

    hah maybe it uses qos to give game traffic higher priority then spy/malware traffic coming out of the os. I bet they did all their testing on an unpatched unfirewalled Windows box.

    btw, latency is not related to bandwidth so all those"well my home network is gigabit and i have no latency" arguments don't apply. My 10mbit network has the same latency as a 10gigabit network.

    --
    I came to the datacenter drunk with a fake ID, don't you want to be just like me?
    1. Re:qos by Anonymous Coward · · Score: 0

      True as far as it goes, but in practice, a household network with more than one user will experience better actual latency with a faster connection than a slower. Minimum latency remains unchanged, but the effects, without QoS set up, of a spouse browsing hefty web pages upon your ping is less with a faster connection. Why? Because the bandwidth is saturated for shorter periods of time.

  39. KillerNic? by idontgno · · Score: 1

    Sounds like a fine BLAZEMONGER product.

    --
    Welcome to the Panopticon. Used to be a prison, now it's your home.
  40. This reminds me... by zarozarozaro · · Score: 1

    of those monster cable products(not a plug.) You take a piece of standard hardware, an RCA cable or a 10/100/1000 nic, put gold contacts on it, and a certain percentage of people will go for it. I don't think that you get much better sound out of a gold plated cable, and I don't think your gaming experience will be any better with this nic, but if it outperforms the other guy's hardware, even just on a bench test...I still won't buy it.

    1. Re:This reminds me... by geekoid · · Score: 1

      the only thing gold plating buys you is no corrosionon the plug at the jack connection.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
  41. A couple of words come to mind... by infosec_spaz · · Score: 1

    Weak Sauce...

    --
    ----- I have bad karma for a reason! -----
  42. Pretty Junk by Anonymous Coward · · Score: 0

    How does the quote by Bill Gates go? "If you can't make it work at least make it look good." or something to that effect.

  43. Sort of old by crabpeople · · Score: 1
    These people have been making these outrageous claims for months. I thought that it must of been on slashdot before, or i would have posted it for the extreme joke level. Firingsquad had a review in july of it, and it seems to be an ongoing joke over there as well.

    Its really quite funny. My favourite "feature" from The manufacturers OWN website PDF:

    Ping Throttle: When other gamers complain that your ping is too low, adjust it a little higher untill they stop whining. Then, dial it back down and go in for the kill!

    The guy who invented it aparently got really fed up with lag and developed special "algorithms" to optimize your packets or something. The original interview is here. The inventor basically comes off as either an amazing exercise in self deception or not terribly bright. If theres even a difference between those two things. I REALLY want to see benchmarks, but it will probably just be the phantom console all over again.

    On March 22nd, we will release some of the details behind our technology. The technology is called LLR, Lag and Latency Reduction. Everyone can read more about LLR at our website on that day. In general, you've said it right, it helps to fight Lag. It does this using a custom designed Gaming Network Processor (the first of its kind) to fight all 3 of the causes of Lag (Client, Network, and Server). If I told you exactly how it worked, I'd have to kill you

    One of the interesting things is that he talks about how the game that drove his lag rage over the edge was UO, which was a really great example of crazy server lag that went unfixed for *years*. You could play on a 28.8 modem and it wasnt even THAT laggy, but everyone would mostly have world lag at the same times. But i guess this LLR tech will upgrade the servers or something. Hes a nutter. Might as well strap some magnets on the side of your nic for all the good that this will do you.

    --
    I'll just use my special getting high powers one more time...
    1. Re:Sort of old by chmod+a+x+mojo · · Score: 1

      Might as well strap some magnets on the side of your nic for all the good that this will do you.

      Ooohhh good idea i'll get right on thklfdashf;gkjdchjads;;vndkjnkad;nvfjnvadk;nfkj;ad z6%^&%$%^

      --
      To err is human; effective mayhem requires the root password!
  44. One cool guy! by dook43 · · Score: 1

    Harlan Beverly is the CEO and Mad Scientist of Bigfoot Networks
    ..snip....
    As a side note, Harlan met his wife in Ultima Online, married her in the game, and then eventually married her in real life


    Nothing more needs to be said.

    --
    This comment was randomly generated by a school of piranhas chewing on the PCB of a Microsoft Natural Keyboard.
    1. Re:One cool guy! by DeathPenguin · · Score: 1

      What's wrong with him meeting his wife on-line in a particularly geeky manner? Did you forget that you're reading Slashdot?

    2. Re:One cool guy! by slashflood · · Score: 1
      Harlan Beverly is the CEO and Mad Scientist of Bigfoot Networks
      ..snip....
      As a side note, Harlan met his wife in Ultima Online, married her in the game, and then eventually married her in real life


      Nothing more needs to be said.
      You think this is cool? Look at this!
  45. Oddly enough... by martinultima · · Score: 1

    Their specifications sheets don't want to pull up in Konqueror / KPDF... they open, but you can't read the text. Guess they're Linux-friendly only if you're running Windows...

    --
    Creative misinterpretation is your friend.
    1. Re:Oddly enough... by makomk · · Score: 1

      Their specifications sheets don't want to pull up in Konqueror / KPDF... they open, but you can't read the text. Guess they're Linux-friendly only if you're running Windows...

      And the same in xpdf too - I had to give up and use GhostScript to view it...

  46. cluster? by brickballs · · Score: 1

    I know theres a beowulf joke in here somewhere...

    --
    "What does slashdotting mean?"
    "You've never heard of slashdot?"
    "I know it makes websites not work."
  47. Well, if it supports PowerPlay, I'll get one by bunions · · Score: 1

    but not until then, because PowerPlay is the Technology of the Future!

    --
    there is no need to sign your posts. this isn't usenet. your username is right there above your post. stop it.
  48. I think you are making some invalid comparisions by Sycraft-fu · · Score: 5, Insightful

    For example take the CRT thing. I own such a CRT, and it's not marketed to gamers, it's marketed to professionals. Why the refersh rate then? Simple function of it's ability to go super high resolution. The monitor is rated to do 2048x1536 @85Hz. To do that, takes some fast electron guns. Well, that ability implies higher refresh rates at lower resolutions. It can do over 200Hz at 800x600 because the resolution is so low. The point is to get extremely high resolutions at usable refresh rates. Also, in general, you want your device spec'd above what it's supposed to actually do. You don't want to run it at it's limits all the time.

    Likewise the mouse thing is a little misinformed. Higher DPI cameras isn't worthless on an optical mouse. It lets it track on more uniform surfaces. No matter how uniform something looks, at some point it's uneven. Well, optical mice need uneveness to track, that's why they don't work on a mirror, or a really smooth surface, they can't track details. One way to make them track better is to up the DPI. The smaller details they see, the more uniform a surface can be. That's also the point behind using a laser. Since it is truly monochromatic light, just one frequency, it shows small details in a starker contrast that is lost with normal LED light.

    Though there's certianly BS targeted at the gamer market, this being some of the BS, there's plenty of products with real legit reasons to be bought. Not everyone wants an experience that is "acceptable" or "works jsut good enough to get the job done." Doesn't mean they are wasting money on the things they buy. Yes a $50 used mountain bike will get me to work and back, but that doesn't mean that I'm wasting money on a deceant $600 street bike. It honestly does work better.

  49. Clearly you missed the point! by XMilkProject · · Score: 1

    Everyone is complaining that the card and marketing are various degrees of bullshit, but who really cares?

    Isn't it obvious that we'll all buy this thing just becuase it *looks* so badass?

    --
    Big ones, small ones, some as big as yer 'ead!
    Give 'em a twist, a flick o' the wrist...
    1. Re:Clearly you missed the point! by Anonymous Coward · · Score: 0

      Looks cool? It's a 3d model, who knows what it actually looks like

    2. Re:Clearly you missed the point! by Anonymous Coward · · Score: 0

      I guess then that I am the only one here with a Intel Quad 1 GBit server adapter in my PC ?

  50. And at this point I would like to announce: by ettlz · · Score: 4, Funny

    Killer IO-APIC!

    • Dispatch those interrupts ten times faster than the enemy so you can pwn!
    • Be on the cutting edge so you can react level when those unexpected events hit!
    • Totally slam your on-line opponents with our 8259-bustin' l33tn355!

    Stay alert, kids, because we'll soon be announcing Killer Keyboard Controller with Bitchin' Gate A20 Technology!

    Pwn!

    w00t!

  51. Looks every bit as credible.... by Anonymous Coward · · Score: 1, Insightful

    ...as the BitBoys Glaze3D video card that was announced about the same time that it was also announced that Duke Nukem Forever was going into development.

    Oy!

  52. Re: your sig by kevlarman · · Score: 0, Offtopic

    macs come with both emacs and vi(m) last time i checked.

    --
    A mouse is a device used to point to the xterm you want to type in
  53. Windows Layer by Doc+Ruby · · Score: 1

    When Windows systems are running on HDs which in turn run on embedded Linux, this is all going to get a lot more interesting.

    --

    --
    make install -not war

  54. Mod article funny by AC-x · · Score: 1

    This is a joke, right?

    It's completely unbelievable, literally. From the poorly rendered mockup to the made up buzz words it's actually quite funny. I'm a bit worried that the online ordering system appears to be functional, hmmm ... Anyone else want to spend $280 to find out ??

  55. What Developers are "Enthuthiastic"? by Akoman · · Score: 2, Interesting

    I've seen two different comments from the founders of this business where they say game developers are "enthuthiastic" (and other exciting, developers-are-pumped-up-words). Who actually cares? It's not like their games are going to be vastly improved with this POS.

  56. Re:Is it credible? Of course not by frovingslosh · · Score: 4, Insightful

    I don't know what's sadder, that some fools would actually hand over money to a bunch of MBA who claim to someow have designed a better network interface than engineers, and who can't understand that these claim are completely bogus, or that Slashdot actually gives them a soapbox to further pitch their snakeoil from (perhaps because of the use of the term Linux in the hype).

    --
    I'm an American. I love this country and the freedoms that we used to have.
  57. software alternative by Intocabile · · Score: 2, Informative

    Cfosspeed is a great little program for Windows XP+ that gives an incredible ammount of control over quality of service. It detects a large number of standard protocols (http,ftp,nntp,dns...) and can set priority on a per program basis for games, voip and p2p. I reccomend it especially if you have an ADSL connection prone to a high ping while downloading. You can setup cfosspeed to prioritize your connection for low ping or high bandwidth. Unfortunately it doesn't work as intended if you have multiple computers on a router because it shapes the traffic independent of other computers, but the developers are working to add multi computer support.

    I've used it for almost 6 months and its given me the highest most stable download speeds I've ever seen on my DSL. My pings while downloading are almost as good as they can be. It's also very lean on CPU overhead.

    here's the developers explanation on how it works
    http://www.cfos.de/traffic_shaping/traffic_shaping _e.htm

    1. Re:software alternative by geekoid · · Score: 1

      What happens if you have multiple computers?
      I would think one the packet leaves one computer, it's got it's priority.

      Apparently I would be wrong, I just don't see where.

      --
      The Kruger Dunning explains most post on /. http://en.wikipedia.org/wiki/Dunning%E2%80%93Kruger_effect
    2. Re:software alternative by Intocabile · · Score: 1

      If you have cfosspeed setup to minimize ping on all computers I think they will try not to get in each others way, but if one computer is trying to max out the connection it just wont work.

  58. $280 ?!?!?! by coaxeus · · Score: 1, Interesting

    A: a non-onboard NIC already offloads "networking tasks" to your NIC from your CPU, not that it matters to gaming. B: upping the QoS priority of ICMP on your PC is very useless. They would be better off making some kind of XXXTreme gaming router with blinky lights and heatsinks all over it, as a router can help a bit more with QoS (although still not very much). C: these guys are idiots, but consumers are idiots too, so they will make money. D: $280 ?? AHAHAHAHAHAHAH

    --
    My name is coaxeus, and I approve this message. In fact, I think it is awesome.
  59. whoa! killer nic! rockin! totally tubular man! by Anonymous Coward · · Score: 0, Insightful

    Totally tubular! tubular, hmmm wait a sec... won't this like killer nic end up blocking the bits in my internet tube? Ah, no got it. Whenever ya squeeze the hose it raises the pressure. No wonder the games will go faster. Freaky rockin' wow! and so cheep.

  60. Wow by Anonymous Coward · · Score: 0

    I guess the mods are basically admitting that the parent is 100% right, since they have now modded *him* troll... Seriously...

  61. Ted Stevens and I believe the card works by User+956 · · Score: 5, Funny

    So, yes, I'd have to weigh in with everyone else, it's snakeoil. Basically, any product designed entirely by a marketting group is going to be snakeoil, and this definitely was.

    Everyone knows the internets is a series of tubes. Well, this card hurls your data through the tubes with such force that it can't possibly get stuck.

    Make sure not to point the jack at anyone. You'll shoot someone's eye out.

    --
    The theory of relativity doesn't work right in Arkansas.
    1. Re:Ted Stevens and I believe the card works by Schemat1c · · Score: 4, Funny

      Everyone knows the internets is a series of tubes.

      Okay that's it, I officially declare the 'internet is a series of tubes' reference not funny any more.

      Thank you.

      --

      "Nobody knows the age of the human race, but everybody agrees that it is old enough to know better." - Unknown
    2. Re:Ted Stevens and I believe the card works by madcow_bg · · Score: 1

      I really cannot understand what is the fun in the internet-as-a-series-of-tubes. Please, will someone explain to me?

    3. Re:Ted Stevens and I believe the card works by MKalus · · Score: 1
      --
      If you want to e-mail me, use my PGP Key.
    4. Re:Ted Stevens and I believe the card works by QuantumFTL · · Score: 3, Insightful

      Okay that's it, I officially declare the 'internet is a series of tubes' reference not funny any more.

      Yeah, to be honest, of all the things said during that horrid speech, that was the most accurage. tubes/pipes, WTF is the difference?

    5. Re:Ted Stevens and I believe the card works by Anonymous Coward · · Score: 0

      Duff man

      Patents

      Accurage (tm)

    6. Re:Ted Stevens and I believe the card works by hmccabe · · Score: 2

      So I shouldn't mention the beowulf cluster?

    7. Re:Ted Stevens and I believe the card works by dr_d_19 · · Score: 3, Funny


      Okay that's it, I officially declare the 'internet is a series of tubes' reference not funny any more.

      Thank you.


      Sure Schemat1c, but not everyone reads Slashdot... could you send out an Internet to everyone about this ?

    8. Re:Ted Stevens and I believe the card works by Anonymous Coward · · Score: 1, Funny

      Yeah, to be honest, of all the things said during that horrid speech, that was the most accurage. tubes/pipes, WTF is the difference?

      Q: How is a tube like a pipe?
      A: Neither of them has anything to do with the Internet.

      Clearly, the Internet is a series of retards.

    9. Re:Ted Stevens and I believe the card works by Anonymous Coward · · Score: 0

      That is the funniest thing I've read for some time, other than the text to Mr Tourette on the UK TV show Modern Toss. It actually made me cry with laughter!!

      If only I had mod points...

    10. Re:Ted Stevens and I believe the card works by Anonymous Coward · · Score: 0

      Slashdot, where cliches and catchphrases crawl home to die.

    11. Re:Ted Stevens and I believe the card works by tehcyder · · Score: 1
      Q: How is a tube like a pipe?

      A: Neither of them has anything to do with the Internet.

      A tube actually is pretty much like a pipe. And it is a well known usage to say "a fat pipe" to describe a high bandwidth connection.

      They're just more or less successful metaphors, not technical drawing plans.

      --
      To have a right to do a thing is not at all the same as to be right in doing it
    12. Re:Ted Stevens and I believe the card works by Pharmboy · · Score: 1

      Slashdot, where cliches and catchphrases crawl home to die.

      That just isn't true. They don't die. On Slashdot, they live forever, which is perhaps the problem.

      --
      Tequila: It's not just for breakfast anymore!
  62. You are correct... by jd · · Score: 1
    ...but missing an important point. Every resend you do has to go through the bus, and PCI is a slug on tranquilizers. The less you go through your bus, the better. You've also got the problem of getting data to and from the card. The PCI bus is 66 MHz, the local LAN will probably be 100 Mb/s. This isn't a problem most of the time. The problem is when you have bursts of data that exceed your PCI bus speed. If your card can store-and-forward these AND deliver all the appropriate responses, then you don't have problems from packet retransmits and timeouts.


    Will this be useful over a WAN? Depends. If the WAN is significantly faster then the bus, then yes. In most cases, people don't have access to a pipe of anything like 100 MB/s. In those cases, you gain nothing at all.


    This device is for LAN parties, as far as I can tell. If it is designed correctly, it'll max out your system under those conditions. You'll get damn-near wire-speeds, even though the internal PC architecture can't handle data that fast.


    The biggest question I have is: Why Linux? I use Linux because I'm a geek and because Linux is damn-near infinitely versatile. It supports protocols that have barely cooled down from being hot off the press. Games don't use, or need, SCTP. They might be able to benefit a little from DCCP or GAMMA, but no game in existance uses those. Network games are almost certain to use TCP/IP, and the fastest TCP/IP stack on the planet is in NetBSD.


    I don't use NetBSD, though I do use some of the other *BSDs, but credit should be given where credit is due - NetBSD holds the record for TCP/IP throughput and so is the sensible system to use for an accelerator of this kind.


    There is one - and only one - exception to this. That is when the hardware (DMA, etc) is designed to be controlled by multiple points. A lot of work has gone into Linux for clustering, and that's essentially what you'd have - a cluster, with the network card acting as the I/O node. You could transfer to/from memory (or other devices) directly without using the CPU at all, in that case. This has limited benefit if used trivially - the memory bandwidth is already so constricted that you've L1, L2 and sometimes L3 caches just to keep things going at all.


    What you'd need to do is - provided the game is multithreaded - move the communications thread onto the network card, with the rest of the game staying on the CPU. The calls to and from the communications thread will generally be much more compact than the data to/from the socket layer, so it'll be much less bus-intensive and therefore substantially faster. This only works if there is a thread that can be firmly identified as handling communications, AND where the assumption of calls vs. sockets is valid.

    --
    It's a small world and it smells funny; I'd buy another if it wasn't for the money; Take back what I paid (SoM)
  63. Bittorrent? by DocSavage64109 · · Score: 1

    Maybe this card reduces your ingame lag that occurs while you are downloading/uploading torrents (or other high-traffice apps) in the background. I don't see why this couldn't be done. Of course, a smarter bittorrent client would have much the same effect and likely be free.

  64. Since I can write code for it... by Xedium · · Score: 2, Funny

    Since it runs Linux and apparently I can write code for it,

    if (packet.sender = "World of Warcraft") {
    packet.contents('health') = "100%";
    }


    I'll never lose at WoW!

    1. Re:Since I can write code for it... by Synic · · Score: 1

      No, but you will *think* you aren't being hurt and be surprised when you die. :)

  65. The patent... by STratoHAKster · · Score: 2, Informative

    ...for this so-called so-called gaming card is readable here. I ain't buying it.

  66. It's really a decellerator by Jthon · · Score: 2, Interesting

    Compared to existing NICs their card actually seems to slow down performance.

    If you read their white paper http://www.killernic.com/KillerNic/PDFs/KillerNic_ LLR_White_Paper.pdf/ you can see that their card can generate 20.15 MegaBytes/s of throughput. From the results in their whitepaper they come out far on top, beating NVIDIA's nForce system by almost 3x the performance.

    Anandtech has an interesting comparison of ethernet performance in one of their mainboard reviews http://www.anandtech.com/showdoc.aspx?i=2696&p=11/ . Anandtech's benchmark paints a totally different story. All the chipsets featured in the whitepaper are included in this review, and as you can see, they perform significantly better than Bigfoot leads you to believe. They are all able to sustain upwards of 118 MB/s of performance (divide the benchmark results by 8 for MegaBytes). While I know it's hard to directly compare benchmarks their results are so far off that I find them very suspect.

    This new item is nothing more than Quantum Speaker Cables for PC Gamers.

  67. Re: your sig by snowgirl · · Score: 1

    :( oh, and here I thought my Mac was safe from all those evil software programs.

    I curse you emacs! Begone and go to the torturous hell from whence you had spawned!

    Hm, it didn't work, emacs is still installed.

    --
    WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
  68. YOU NEED THIS by jarg0n · · Score: 0

    Because everyone needs 400mhz of network card power to save .001 miliseconds of lag!

    --
    Error 2101: all your sig are belong to us
  69. Yes, but does it...? by freehunter · · Score: 0

    I mean, can I... install OSX on it?

  70. OMG!! by Anonymous Coward · · Score: 0

    I can stop being fragged, because ill be going from 25ms pings to 22ms pings...OMG think of how much faster everything is going to be on my $7,500 gaming rig my parents bought me..its going to speed up ordering pizza online too...all i need now is a toilet in the basement, and ill never have to leave...

  71. Re: Most switches don't use store and forward by Jason69 · · Score: 1

    Most switches don't use store and forward. That is fairly old switching technology. Most switches (and routers too) use some form of fast packet switching similar to Cisco Express Forwarding (CEF) that reads the first packet of a stream, learns its path through, and all the following packets of that stream are immediately forwarded 'bit by bit'.

  72. Re: your sig by LordOfTheNoobs · · Score: 1

    You can't curse that many years of elisp of out existance that easily. And for that matter, though he be slightly portly and rather ominously bearded, Richard Stallman hardly passes for even a remote semblance of a torturous hell. Not to mention the difficulties in actually placing the emacs back into the Stallman... :p

    (( for the record, uses emacs and likes it

    --
    They're there affecting their effect.
  73. It gets better by MmmmAqua · · Score: 1
    Did you read the descriptions of their "new" technology?

    MaxFPS(TM) Technology offloads network processing to Killer which frees your CPU to focus on what you need it to - THE GAME!!! Killer's superior performance allows you to experience more Frames Per Second even on max'd out PCs with top of line graphics cards and other peripherals.

    So... they must have invented TCP offload around the same time Al Gore was inventing the Intarweb?
    --
    Arr! The laws of physics be a harsh mistress!
    1. Re:It gets better by Sledgy · · Score: 3, Insightful

      If you read the artical you would see that they are not claiming they invented TCP offload, but adding support for UDP offload (as most games use UDP and not TCP).

      Although in saying that, I can't really see how this card is going to make much difference over the internet. Your connection or the configuration of your ISP is more likely to be where the lag is introduced rather than your NIC.

    2. Re:It gets better by MoxFulder · · Score: 3, Insightful

      UDP offload???? Gimme a break :-)

      The entire processing required to transform a hunk of data into a UDP packet consists of prepending a 6-byte header to the thing, containing the source port, the destination port, and 16 bits of zeroes... not exactly the sort of thing that requires immense processing power. Unlike TCP, UDP doesn't synchronize anything, doesn't reorder anything, and doesn't acknowledge receipt of packets.

      How much of your processing power is ever occupied by the network card when playing a game??? Or when doing anything else for that matter. I can have several hundred bittorrent connections running on my computer, with a total transfer rate of hundreds of KiB/s, all kinds of checksumming and I/O overhead, and it still makes a 1 or 2 percent blip on CPU usage... unless a network card can magically construct a LAN between two computers at a distance, it's not going to affect latency in network gaming.

  74. Re: Most switches don't use store and forward by osbjmg · · Score: 1

    You are mostly right and partly wrong. Store and forward vs. cut through isn't really an issue anymore as it is very rare to see store and forward (unless it's a broadcast frame). These are buffering issues, where CEF is a propriatary forwarding technology whereby the entire routing table (FIB) is put into memory in a quickly accessable hash. Fast switching and CEF are both routing enhancements. You sound as though you are explaining the original multi-layer switching model: netflow switching. The first packet of a stream was inspected at process level and then a netflow entry was built. Subsequent packets in that stream hit the netflow shortcut and required no process lookup. This is specific to multi-layer switches. I would say the majority of latency on the megaweb is not from switches - it's probably an mostly distance and then some is router latency.

  75. it runs linux by kemo_by_the_kilo · · Score: 1

    So would a killerNic "killer" be a nic with firewire it in instead of usb 2.0?
    some one please tell me why i cannot just run a linux box as a router to my normal nic and get the same perf. gains.
    also, even if this card had a sub $100 price tag would anyone buy this?

  76. I know I've been busy recently, but ... by really? · · Score: 1

    ... I hadn't realized time had gone by so fast. Damn, April 1st already. Missed a whole season of skying. :-|

    --

    "Consistency is contrary to nature, contrary to life. The only completely consistent people are the dead." A. Huxley
  77. Re:Mel Gibson joins Hezbollah freedom fighters !! by Anonymous Coward · · Score: 0
    Mel Gibson is an "Australian son" when he's making big movies and being a good guy.

    When he's drunk and loudmouthed, we remember that he was born in the US.

  78. Wow by Godji · · Score: 1

    As a side note, Harlan met his wife in Ultima Online, married her in the game, and then eventually married her in real life.

    'Nough said.

  79. The telling thing ..... by taniwha · · Score: 1
    of course is that it's vaporware enough that they only have a rendered image, not a photo of the real-thing ....

    people have been building off-loading cards for TCP for years (I remember the first round of them back in the mid 80s), as a rule they don't work that well - better to put the smarts closer to the apps

    Now of course if you could 'lower ping times' by rewriting game's packets on the fly to look as if you had faster latency ...... (wait isn't that sort of what they are claiming - the ability to dial in one's ping times ....)

  80. Tytus (Bigfoot Networks) Responds by Tytus_Bigfoot · · Score: 5, Informative

    Hey everyone, I am the inventor of the Killer NIC. I will not try to hijack your thread, so this will be my only post. Thanks for the interest in our Killer Network Card. It has been my personal vision for years. A lot of very good questions have been raised here, and I think a lot of them are answered in our FAQ here: http://www.bigfootnetworks.com/FrequentlyAskedQues tions.aspx . If there are still questions, I would love to try to answer them at our sponsored community site: http://www.endlagnow.org/ELNForums/ Thanks, Tytus

    1. Re:Tytus (Bigfoot Networks) Responds by rufus78 · · Score: 1

      How can you be serious?
            This is funny, someone took way toooooo long to make up the forums on the www.killernic.com website.

            If there is an actual product, where are the pictures? Not graphic images, PICTURES?

            Port: usb? What about the PCI from the image? I could go on all night, but have already waisted tooo much time.

            This is waaayy too funny...

    2. Re:Tytus (Bigfoot Networks) Responds by Mike+Bridge · · Score: 1

      how is it that your product, which according to your documentation, does not accelerate TCP, shows more more performance gains in a game that uses ONLY TCP, then the other game listed, which uses ONLY UDP?

  81. Slashdot should subsidize by maxrate · · Score: 1

    I bet this card is so good it gives you more bandwidth too! In in for two of 'em - I'll setup line bonding and be faster than all the telcos put together!

  82. but does it run... by sjalex · · Score: 2, Funny

    uh... minix?

  83. Nforce 5 by Joe+The+Dragon · · Score: 1

    The Nforce 5 chipset offloads networking form the cpu and it is built in the motherboard. This card is only pci so it can even use gig-e at full speed nforce 570 and 590 for amd or just 590 for intel has 2 ports with teaming.

  84. Correlation by svallarian · · Score: 1

    If *you* suck ass, it's not going to make you any better either.

    (or so says my the rest of my BF2 squad)

    --
    I patented screwing your mom. But it got revoked for "prior art."
  85. Re: your sig by snowgirl · · Score: 1

    Not to mention the difficulties in actually placing the emacs back into the Stallman...

    I don't know... I suppose there are many a VI(M) user that would like like to at least try ;)

    --
    WARNING! This girl exceeds the MAXIMUM SAFE standards established by the FDA for BRATTINESS
  86. Why does it has such large host requirements by gelfling · · Score: 1

    If the goal is to offload as much of the network processing, which btw is going upstream to a link MUCH slower than any NIC, assuming not everyone is in the same LAN, then why does it have such large host processing requirements?

  87. Depends... by Svartalf · · Score: 1

    IF you are driving the card to wireline speed, you'll see a performance jump, under Windows, because the CPU isn't addressing the NIC, it's just handing off the data to be sent to the Transport Offload Engine (TOE) on the card, which then sends it via DMA out to the wire. If you're lan-partying, on a Windows box, it might make a difference on a 100Mbit card. A Gig-E or higher is a slightly different story, as is with running on Linux. With Linux, you're already close to those performance levels with any old NIC at up to Gigabit speeds. With Gig-E or 10Gig-E, your CPU's overhead with all the cache touches, etc., you'll start seeing more and more demand from the machine (And more benefit) from those cards.

    Now, having said this, it's an Offload Engine- this means that you've done an end-run around ANY security measures on-machine you might have
    for networking, etc. (In other words, if you're using a firewalling app on the machine, you're wasting your time if you're using this NIC...)

    --
    I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
  88. 3com cards are so 2001 by Ayanami+Rei · · Score: 2, Insightful

    You want one of the following (in order of desirability):
    1) Intel
    2) Marvel
    3) Via/Rhine
    4) RealTek

    And fuck the rest.
    Yes, fuck Broadcomm and their shitty SMP-deficient drivers.

    --
    THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
    1. Re:3com cards are so 2001 by mrbcs · · Score: 1
      I don't like any of those personally. 3com may be so 2001, but I have yet to see anything outperform them, your list included.

      I've turned off so many onboard cards, realtek and via especially and replaced them with 3com and all of a sudden.. no more network problems.

      --
      I'm not anti-social, I'm anti-idiot.
    2. Re:3com cards are so 2001 by vision864 · · Score: 1

      You are aware that Marvel was an off shoot of 3com? Case in point the asus p4p8x - that board contains a 3com marvel

    3. Re:3com cards are so 2001 by monsted · · Score: 1

      I'll take the intel and you can take the other three and stuff 'em. I haven't seen a VIA/Rhine NIC in a while, but Marvel and Realtek definitely still makes crappy NICs.

      In a pinch i'll use a 3com, but using a Cisco switch with 3com can be painful if you're autodetecting speed/duplex.

    4. Re:3com cards are so 2001 by Anonymous Coward · · Score: 0

      I will also take the intel over anything else. Reason: compatibility. Even their newer 100 MBit Cards worked with old drivers, even the ones included in OPENSTEP! - Different 3com cards required different drivers.

    5. Re:3com cards are so 2001 by Anonymous Coward · · Score: 0

      Since I have a pair of 3com cards sitting on my desk at the moment:

      3com 3C905B Fast Etherlink XL PCI
      3com Parallel Tasking II, 3Com 40-0476-002, 3868 CT0001, BROADCOM 5903

      3com 3C2000-T (gigabit PCI)
      might be a Marvell chip since there's a stylized "M" below the 3com logo on the chip

      The big advantage of 3com in the late 90s up until early 2000s was that they were pretty much guaranteed to work. We rarely (never?) had an issue when we bought 3C905s so it became an automatic purchase option.

    6. Re:3com cards are so 2001 by Ayanami+Rei · · Score: 1

      In a pinch i'll use a 3com, but using a Cisco switch with 3com can be painful if you're autodetecting speed/duplex.

      Seriously. I swore off using 3Com cards once I ran into this and it never improved over a 2 year period. If it's so good how come it can't even manage to autonegotiate properly with the most common infrastructure switches? These chip-level issues that disturbed me, never mind that it was parallel-tasking or whatever. No software can fix broken wire protocol handling. Maybe Cisco were the ones screwing up but it never sat right and I never trusted their physical cards since.

      And on the other side (Via, Realtek, etc.) it comes down to driver quality. I think the RealTek and Marvel hardware works just fine so long as you use the WHQL certified and generic linux drivers. Then it's rock solid and as low latency I could expect before some kinda fancy TCP offload.

      And Intel cards always kick ass. I expect because they control the card manufacture and they source all the drivers. But they are expensive. :-(

      --
      THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
  89. Video Game Hacking Possibilities? by Xzisted · · Score: 1

    You have to wonder that with this thing running a native version of linux on the card itself if you could place linux based game hacking utilities on the card that make use of packet hacks to hack your games. This would allow you to bypass things like Blizzards hack checking utility and things like that. I wonder if anyone has examined those possibilities.

    --

    Honesty may be the best policy, but apparently by elimination, dishonesty is the second best policy.
  90. Not just wireline speed in latency... by Svartalf · · Score: 1

    While I'm not saying this stuff's not snake oil, there is something to be said for something else
    that slows down the framerates that has little to do with latency in and of itself- what's your
    CPU utilization when you're pushing data down the wire?

    With 100Mbit cards, the utilization can approach 50-60% at full wireline speeds.
    With 1Gbit cards, if you don't do clever things to offset the CPU use, it can be as 80%.
    We won't even go into raw 10Gbit speeds...

    If you're burning 20-25% of your CPU handling the NIC at load in a game at a LAN Party, it would
    make some small sense to offload that onto something else- IF it really made that much of a
    difference. It remains to be seen what they've done, but it looks like a TOE that also supports
    UDP packets (Though there's not a LOT to be done there to be offloaded...). At a 100Mbit rate,
    it's a difficult sell to convince me that it's worth >$200 for a card to do transport offload
    just so I can see an extra 5-20fps on an otherwise already maxed out machine. This is being
    sold to enthusiasts with more cash than brains, I suspect. Now, if they offer this as a generic
    offload engine for higher speeds, say 10Gbit Ethernet, for the purposes of a cluster interconnect,
    then they might have my interests...

    --
    I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
    1. Re:Not just wireline speed in latency... by tji · · Score: 1

      If it were really 5-20fps, I could see people buying one of these. But, I'll bet the actually difference would be in the 1fps realm (i.e. somewhere in the noise where you really can't tell if it's having an effect).

      As you point out, there big selling point is that they do UDP offload, but UDP is a much simpler protocol than TCP, so the offload potential is much less. Also - existing cards already do UDP checksum offload (such as the extremely popular Intel Gig-E NICs).

      Then, when you consider that the game companies make the game communications bandwidth efficient, so the games are playable over the WAN, the value becomes even smaller.

      I think the product is a hoax, just playing on those gamers who go to ridiculous extremes.

    2. Re:Not just wireline speed in latency... by Corngood · · Score: 1

      When would you ever use that much bandwidth for a game? If it were a server app using that much, it should certainly be a dedicated machine, right?

    3. Re:Not just wireline speed in latency... by Svartalf · · Score: 1

      CRC calc is only part of the overhead in UDP. Another part is all the copying that typically ends up being done
      to take the header (Which is in it's own chunk of system memory being assembled) and plant it in the packet
      assembly buffer in memory and then copy the entire buffer to the card- this is done for architechtural reasons
      in that it's easier to design a modular system doing this at the slight expense of the normally minimal copy
      overhead... There's also Z-Copy transmission of the packet from it's different places in memory via scatter
      gather DMA operations. But, again, many of the better "dumb" cards do this, so...

      TOE only really sort of makes sense if you're talking something where you're needing to handle 10Gbit Ethernet
      and something like hooking a SAN array into a cluster of server machines. There, you might see
      some advantage.

      --
      I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
  91. Re:I think you are making some invalid comparision by feepness · · Score: 1

    Well, optical mice need uneveness to track, that's why they don't work on a mirror, or a really smooth surface, they can't track details. One way to make them track better is to up the DPI. The smaller details they see, the more uniform a surface can be.

    That's it! I'm off to sandpaper my desk! Who's sniping with the railgun NOW?

  92. I interviewed Harlan by Yert · · Score: 2, Interesting

    ... at the GDC in San Francisco earlier this year, and posted the results - http://www.gamerati.net/index.php?option=com_conte nt&task=view&id=118&Itemid=1 He's a slick little monkey. The second he figured out that I wasn't buying the claims after I kept insisting on a single hard number or physical card, he ended the interview, and then cancelled the rest of the media interviews that day. His "white paper" is hype, and so is his whole company - at the 8th grade level. It's sad, really. I put him up there with Infinium Labs - and I don't expect he'll send us a sample card.

    --
    Truck driver, plumber, Linux systems engineer.
    1. Re:I interviewed Harlan by tehcyder · · Score: 1
      He's a slick little monkey
      Well give him some fucking credit then, how many simians do you know with an MBA? Oh...
      --
      To have a right to do a thing is not at all the same as to be right in doing it
  93. Install a PCI ATM/ADSL card by Colin+Smith · · Score: 1

    Run the ATM stack on your machine, miss out the local area network entirely. In theory that removes a hop and therefore that hop's latency from the route between you and the server. Not impossible, but you'd have to be a hard core gamer and a little dim to bother.

    --
    Deleted
  94. It may offload the networking overhead by Colin+Smith · · Score: 1

    Which is fair enough, some big server cards do this, it's quite neat and frees up the CPU for other tasks. Same idea as SCSI, I2O or the GPU in video cards. in which case a game might well run faster with the card installed.

    --
    Deleted
  95. Ha Ha! by Anonymous Coward · · Score: 0

    We got about 10000 glossy little cards for this to stuff in PAX bags. By the end of the day, all you had to say was "Ping Throttle" and the ENTIRE bagging line burst into laughter. The claims they make are so ridiculous, it's not funny.

  96. I must already be on their beta program by Tim+Browse · · Score: 1

    So, do I already have one of these network cards installed, then?

    C:\>ping -l 2048 media

    Pinging media [192.168.1.10] with 2048 bytes of data:

    Reply from 192.168.1.10: bytes=2048 time<1ms TTL=128
    Reply from 192.168.1.10: bytes=2048 time<1ms TTL=128
    Reply from 192.168.1.10: bytes=2048 time<1ms TTL=128
    Reply from 192.168.1.10: bytes=2048 time<1ms TTL=128

    Ping statistics for 192.168.1.10:
    Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
    Approximate round trip times in milli-seconds:
    Minimum = 0ms, Maximum = 0ms, Average = 0ms

    Or will it make my pings even faster than <1ms?

    And how does it remove latency introduced by my ISP/ADSL connection?

    (Both PCs involved in the ping are just using onboard ethernet interfaces, of course - not even a fancy $30 3Com NIC.)

    Awesome snake oil.

    As for offloading your CPU from all that massive overhead of formulating a packet once per frame, well gosh. I'm thinking of releasing a PCI card that runs Linux that games can use whenever they need to divide floating point numbers by 5 or 7.5. Game developers will be super-excited about that, as well. Won't they? It would have about the same effect. I'm thinking $280 would be a fair price. Who's with me?

    "There are no consumer products available today that are comparable to Killer. In addition to being a Network Interface Card (NIC), Killer is also a computer within a computer. With Killer's Flexible Network Architecture (FNATM) you can literally run a Linux command prompt in a window on your gaming PC."

    They're right. There are literally no other products available that allow me to have a Linux command prompt on my PC.

    WTF would I want that, anyway?

    You are in a maze of twisty passages, all alike.
    quake% move forwards
    You are in a maze of twisty passages, all alike.
    There is a HPB here.
    quake% shoot HPB FTUW!
    You are shot by someone up on a balcony. You are dead.
    quake% echo "camping fagg0t"
    Yu-huh. You're still dead, though.

    And, of course:

    GD: What do you say to the customer who already has great pings to the servers they play on?
    Bigfoot: We designed the Killer for those gamers that care about winning, that care about being the best. With the Killer, your ping will be even lower. So, if you turn the corner and spot your enemy you will have more of an edge with Killer than without.

    Notice how they didn't follow it up with the question "What do you say to the customer who says you are totally full of shit?" Nice job with the interview, guys :-)

  97. Type of Service by Short+Circuit · · Score: 1

    Did it occur to anyone else this card might simply change the Type of Service field in packets to something that gets a higher priority by ISPs and backbones?

  98. Re: your sig by Anonymous Coward · · Score: 0

    The line starts here ;o)

  99. Not just audiophiles... by bokmann · · Score: 1

    The general population is so confused by technology they can easily take one concept they grasp and mis-apply it.

    For instance, my sister-in-law is convinved her cell phone battery dies quicker when she is calling long distance numbers. To the average slashdotter, that is laughable... but think about it from her point of view - batteries die more quickly when they are used for more intensive stuff... her cell phone is calling a number further away. Throw in the 'power of suggestion', and the rest is an obviousl false conclusion. She has no idea how cell phones actually are talking to a local tower, so it is a reasonable but ignorant (in a non-insulting sense) conclusion.

    I still had fun telling her my theory that global warning is caused by the extra hour of sunlight from daylight savings time, and that the ocean levels are rising because the chemicals we are dumping into the oceans are killing all the sponges.

  100. hahah, this is SO a hoax.. by radarsat1 · · Score: 1

    Come on people, this is so definitely a joke. :)
    A good one though.
    Their flash site is practically written like an Onion article.

  101. pfft by Anonymous Coward · · Score: 0

    I've just sent it to the PTO so you are all screwed

  102. Ridiculous. by Shanep · · Score: 1

    How could this reduce latency appreciably?

    1/ Reduce the load on the system CPU by off loading some networking tasks? What will that buy you? 0.1%? 1%?

    2/ Tag all of the packets from that machine as being higher priority for QoS routers which may bump up their priority? You can do that at your own router/gateway. Buy a gaming router or do-it-yourself with BSD/pf, Linux, etc.

    3/ Reduce the latency of the interface itself? Oh yeah, that will provide huge gains!...

    Pinging my OpenBSD firewall, I get an average of 0.165mS. Pinging the gateway at my ISP (1536/256 ADSL), I get an average of 11.3mS. Even if this card performed a miracle on the order of water-into-wine and reduced the latency to my firewall down to 0mS (absolute), the latency to my next hop is still 11.1mS. A whopping 1.8% improvement if Jesus is the man behind this card. Any QoS packet tagging that this card might be doing, would be pointless, because my firewall is dealing with my gaming QoS and other gamers can do the same or use a router which does it for them.

    And they want how much for this card? Is Jesus going to personally come out and install it for me? And while he's here bless me with super lightning fast reflexes, while we break bread?

    --
    War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
  103. PingThrottle(tm) by KarmaMB84 · · Score: 1

    I'm sure the ability to throttle your transfer rate to induce warping in multiplayer games like counter strike when they feel they might be in danger will be extremely attractive to 13 year old FPS fanatics...

  104. Call me sceptical by necrogram · · Score: 1

    Maybe i've been spending way too much around router and switches, but it seems there are few things they forgot to consided. 1) Unlike graphic cards/motherboard/cpu/etc, an IP network is not a closed system. you can have the fasted and best card money can buy, but you're at the mercy of the weakest link. Dodgey router or crapin out switch port? your traffic will slow down just like everyone else's 2) XP Pro's network stack. Just 4 months ago, MS released code for Server 2003 SP1 and R2 to better use TCP Offloading Engines (TOE). While the TOS support (chimney and reverse side scaling) are some really cool shit, i dont recall whitepaper one showing the improvments made to XP's IP stack to kick it up a few notches. 3) QoS. The "Game First" feature is just a glorified name QoS or quality of service. QoS has two basic requirments for it to be worth a damn. a) every rotuer and switch port must respect it. b) no rotuer or switch port rewrites the markings. You can mark your packets CS7 (network control critial data), but if your hitting your $50 dollar linksys home nat-toy, its getting FIFO queued (First in, First Out). And, if the ISP and Carriers have admins worth the papaer the admin's check's printed on, that admin is NOT going to respect an end user's packet and remarking them back to 00 (coach class priorty). I think that is even the default for IOS (dont quote me on it), rewrite the DSCP and COS values back to 00/0. But... If you're still intrerested in the card, i have some prime real estate for sale ;-)

  105. Um... WTF? by SanityInAnarchy · · Score: 2, Interesting

    It's a router... or a bridge... in the network card. WTF?

    Ok, it either breaks RFCs and causes even MORE lag when you get more than one or two on a network, or it basically does what current stateful firewalls do for you anyway -- defragment packets before they get sent through.

    So, if you already have a cheap router on your home network -- and you probably do if you have wireless -- this part is done for you already.

    Now about that ping... Assuming it's not breaking any RFCs, I haven't heard ANYTHING to suggest that this can do ANYTHING to improve traffic, once it's beyond your network. Now, gamers, go ahead, ping your routers. That's almost certainly 192.168.1.1 or 192.168.0.1, but it's not hard to figure out if you know what you're doing. I have NEVER seen a ping to the router of a LAN of more than 2-3 ms, and usually we're talking about .013 ms, so if you're seeing more than that, you have other problems, or you're on a wireless network, in which case your games will always suck.

    As I continue to read through the spec sheet, all I see is either:
    "Gee, any off-the-shelf router already does this!"
    or:
    "Gee, even if that's true, my $100 mobo includes a NIC that already does this. Even if not, the maybe 5% of CPU that you'd save surely costs less than the $280 you're charging for this card."

    Really, go the fuck home. I don't want any of these anywhere near my networks, much less my gaming rig.

    --
    Don't thank God, thank a doctor!
  106. Lies ! by wtarreau · · Score: 1

    From the FAQ :

    Killer shaves off these extra milliseconds while also prioritizing gaming network packets. ICMP Ping that can be run from a DOS prompt (ICMP Ping) does not actually measure the overhead of the operating system's network stack.

    while this is a utterly bogus claim, basically what they say is "don't worry if you don't notice any change, that's expected". Considering that a gamer's machine can easily eat 2-4 Gbps of UDP traffic at 100% CPU, I suspect that 1 Mbps would at most eat 0.025%, which should not be noticeable at all. By the way, I find it funny that this card pretending to increase performance is still limited to a PCI bus, limited to less than 1 Gbps in+out, and with latencies as high as 4 microseconds at 33 MHz.

    Have you seen the design ? It's just a card for hardware modders, it seems to even have leds. I suspect that the card will basically do nothing but the people running it will say "look at my NIC, it's beautiful and blazingly fast, have you seen how faster I run ?"...

    As long as they swindle those stupid people, I don't really mind...

    Willy

  107. Ok, but I didn't love ghost chasing. by cerebis · · Score: 2, Interesting

    Those of us with an embarrassingly large playtime in MMOGs would probably say that filling in the blanks due to lost packates (as seen in games like Everquest and everything today) is certainly better than waiting for retransmits (like in the original Ultima Online), but often enough, you ended up chasing ghosts.

    This was merely annoying as a newbie, chasing orc pawns in overloaded starter zones, but in much later stages the same feature could result in writing off hours of playtime for 40 people in unforgiving raid zones.

  108. Where is the Disadvantage???? by woolio · · Score: 3, Funny

    While the network card Might improve the network connectivity for that one person, it more than likely does it by hurting the performance of the the network as a whole...

    That sounds like killing two birds with one stone....
    1) Decrease my latency
    2) Increase the latency of everyone else, including the snipers in CS

    This could be a gamer's paradise... It will be infinite successful.

    OTOH if it doesn't work, then it makes a great gift for a `friend'.

  109. Even worse... by ImaLamer · · Score: 1

    It's a Linmodem!

  110. tubeless? by freaker_TuC · · Score: 1

    But but, without humor, the Internet will loose all its dry tubes; imagine all that water flowing .. arrr... bedtime zzzzzzzzzzzzz

    --
    --- I am known for the ones who want to find me on the net. Is that a privacy risk or a privilege? One might wonder..
  111. hoax by citizenr · · Score: 0, Troll

    I'v got a question. Have you ever been to 'pound me in the ass' prison? Because thats where you will be going if you really start this fraud.

    --
    Who logs in to gdm? Not I, said the duck.
  112. they charge HOw much? by vision864 · · Score: 1

    I will kick your ass on a $5 Realtek, Mr i got a $230 nic hooked to a 50$ cable modem.

  113. meh l1nux 0wned y00 by gus24 · · Score: 1

    wow, only 235 days till april 1st, or is this post just 130 days late?

  114. Where's the numbers? by Anonymous Coward · · Score: 0

    They claim that game developers are very excited about the improvements in FPS that they've been shown but where are the numbers? Did I miss them somewhere? I mean really if they had really big improvements then they'd be touting them wouldn't they? Hell they'd be comparing using the KillerNIC to other NICs on the same computer at the same location using different online games on public networks. Just wait until any of the gaming sites get their hands on samples and then we'll see if it is just all a load of hype.

  115. AMD prefer it on top.... by Anonymous Coward · · Score: 0

    The tables will turn once more my friend. AMD have come from behind more than once before. Have faith.

    1. Re:AMD prefer it on top.... by heinousjay · · Score: 1

      Having faith in a company is like having faith in pulling out.

      --
      Slashdot - where whining about luck is the new way to make the world you want.
  116. I had something like this.. by jleq · · Score: 2, Informative

    I had a card like this back in 1998. It was a 3Com Etherlink with "Parallel Tasking II"... Except that it actually did make a difference with the PCs of that day, and cost a hell of a lot less.

    Sadly, I know some people who will probably actually buy a network card like this... LOL KILLER! How ridiculous.

  117. Shielded power cables? by samj · · Score: 1

    Can these guys sell me one of those funky shielded power cables for my stereo too? I have a lazy grand that I'd like to blow on stuff that will make absolutely no difference to my life.

  118. Switch Types was Re:Yes. by hauntingthunder · · Score: 1
    yep your right about store and forward though pro switches may use better methods

    There are 3 main types of frame fowarding

    1 Store and forward

    Reads the entire frame checks it then sends

    2 Cut through

    Starts foward as soon as posible ie after it reads the destination MAC address

    Fragment free

    As cut through but it reads a short way into the frame to detect errors (most errors happen very early in the frame)

    --
    You will never get to heaven with an Ak 47... But A Zu 30 is good for Low Flying Cherubim
  119. so 1980 by Spliffster · · Score: 1

    i am not much into hardware, so please forgive me if i misunderstand this. i understand that clients in a lan would benefit of this new technology.

    Nowadays games are played over the internet and a lot more than a couple of NIC's and hubs are involved.

    Plus, most of the game load is usually transferred over udp, tcp chanells usually exist just to transport very few reliable game information (like player infos upon server joins, round start/end info ... you name it).

    So if i get this right, this techmology (sic) would help on a lan, games are not played often on a lan anymore and since we are in the age of affordable gigabit ethernet, where is the point (concering gaming) ?

    -S

  120. Work for UDP by shani · · Score: 2, Informative

    UDP still carries a checksum, which does involve doing calculation over the entire contents of the packet and header. I admit, the calculation is simple, and the amount of data in the packet is small, but it is non-zero.

    Also, IP supports packet fragmentation and reassembly. This is why you can send a 5000 byte UDP packet on an Ethernet network, which sends data in chucks of 1500 bytes. I think the main "win" here is that by handling fragmentation on the network card, you avoid the main CPU having to context switch to collect the state before the entire IP packet has arrived. I also freely admit you probably don't get many packets of this size during gameplay.

    You are right that network processing does not require much processing power, but that's not the point. The point is latency. The checksum calculation can be sped up by doing it on an ASIC (as this board does I think), and the fragmentation/reassembly can be sped up by avoiding extra context switches of the main CPU.

    Personally, I doubt going from (for example) 30 msec to 25 msec ping time is worth it, but I also don't think getting 100 frames per second versus 70 frames per second from your graphics card is worth it, so what do I know? :)

    1. Re:Work for UDP by TheRaven64 · · Score: 1
      Of course, if you are writing a latency-sensitive protocol, then you should ensure that none of your datagrams is longer than the MTU of the network (1500 for a typical Ethernet LAN, a minimum of 576 for the Internet). Doing so ensures that datagrams will never be fragmented, which has two advantages:
      1. There will never be any added latency for reassembly of packets.
      2. If a packet is dropped, then you lose one message, but you don't waste any bandwidth on the bits that are not dropped.
      I would imagine most game designers understand this, and design their protocols accordingly.
      --
      I am TheRaven on Soylent News
    2. Re:Work for UDP by MoxFulder · · Score: 1

      Well, the UDP checksum is optional and isn't even used in most applications (although the IP header does have a checksum).

      I'm don't think what you said about processing power and latency makes sense... if it doesn't take much processing effort to set up a UDP packet, then it won't take much time for the processor to do, and it won't cause much latency!!! The processing required to handle network packets is simply a NEGLIGIBLE source of latency for gaming.

      Look at it this way:
      * With UDP/IP offload, the processor makes up its UDP packet, and sends the raw datagram to the network card, then resumes its business immediately.
      * WITHOUT UDP/IP offload, the processor makes up its UDP packet, then prepends the necessary headers, and sends the complete Ethernet frame to the network card, then finally resumes its business.

      The extra processing time required without offload is probably around a microsecond (a couple thousand clock cycles). Hundreds of packets could be sent and received every second without noticeable slowdown of the game. And the additional latency for the individual packets will be only a microsecond or two apiece.

      Basically, TCP/IP offload only makes sense if the CPU is *overwhelmingly* used to process network packets: it makes sense for a trunk router to have packet-processing ASICs, because ALL it does is punt packets around. It might make sense for a corporate file server with dual 1gb NICs to use offload. But it definitely *does not* make sense to do offload when the CPU is devoting at most 1 or 2% of its time to network processing...

    3. Re:Work for UDP by MoxFulder · · Score: 1

      That's a solid point about fragmentation: any gaming protocol that worries about network processing will avoid using large packets that need to be fragmented, end of story. Basically, any gaming protocol should be designed to use as few whiz-bang network processing features as possible.

      Gaming just doesn't require that much network processing. The Killer NIC may be useful for TCP/IP offload when used by massive servers whose CPUs get overwhelmed trying to saturate a 10gb/s pipe. But there are already less-hyped NICs for that kind of use. Seems like the Killer NIC is a solution in search of a problem...

  121. CONSPIRACY! by Anonymous Coward · · Score: 0

    OMFG, these guys are a shell company funded by the big telecos to put out a product that will destroy the internet! The telecos will then declare that they must go with tiered service levels to avoid these kinds of problems in the future.

    Evil.

  122. The saint of internet?! by bjoeg · · Score: 1

    Ok so this will fight lag?

    Hmm, it will boost the FPS of my system. Well I could believe it, since most SOHO onboard features eat a little CPU.

    But tell me, how will this on fight lag on the internet? I mean, yes the card "might" reduce lag between my PC and router, but the card cant do anything about latencies on the internet......other than fake it.

  123. dunno about faith.... by freakmaster · · Score: 1

    I don't see anything particularly exciting in the current AMD roadmap. They have quad core, but so will intel.
    The only advantage I see left for AMD now is that their memory system scales better, so for 4 & 8 way servers they may still have an advantage.
    That and joining w/ ATI, possible hypertransport gfx/ integrated CPU & GFX core anyone? but that stuff will probably take a _while_ to get to market.

    as far as faith... well that's just plain silly. AMD is a company that want's your dollar just as bad as Intel. This isn't Linux vs Microsoft, where there is a huge philosopical / ethical difference between the two offerings, regardless of how well either one works.

    AMD has come a long way in the last 10 years, from cheap knock of to industry leader w/ x64 and the 32 bit athlon was strong particularly in the early p4 days. Eventually their chips will leapfrog Intel's, but until then the core-duo is hands down the best pc chip out there.

    AMD's major contribution is healthy competition to the market. But buying an inferior product just to artificially support competition is anti-market. It promotes inefficency. Make AMD work for your dollar. That's what intel did. They had to drop netburst & start from scratch. But they now have the better product, not the least of motivation behind it was beating AMD's then superior product.

  124. Re: your sig by Anonymous Coward · · Score: 0

    Sign me up!

  125. Re:Is it credible? Of course not by fitten · · Score: 1

    Yeah... somehow the Linux on the card reaches out across the Internet and makes my wire signals travel faster and makes sure my packets are prioritized across all routers. These guys are betting on advertising that their card has Linux on it being a selling point where the likes of /. readers will buy it simply because it has Linux on it (so they can brag to their similarly clueless friends). Of course, you could probably take the card and get the source and do other neat things with it if you had the time and the interest to do such things.

  126. No but that's good to know. by Ayanami+Rei · · Score: 1

    And that would explain why they are decent.

    --
    THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
  127. Hardware requirements? really??? by aonaran · · Score: 1

    Can someone explain to me why I need a P4 or Althon 64 with a 256MB 3d card to use a NIC with a built-in QOS engine?

  128. People have... by Benanov · · Score: 1

    Look up the Realtek 8139 or some of the older SMC cards. I really do enjoy the part about the small number of registers.

  129. Features of the new Killerizer Network card by TheBogBrushZone · · Score: 1
    Any resemblence to any real (or vaporware snakeoil) network card is purely coincidental.
    • LocalPing - Win more because LocalPing Technology offers the fastest gaming performance possible. No matter how fast your ISP is, you can get a better Ping by having the NIC provide the ping reply instead of all that pesky messing about with the actual network.
    • SneakyFPS - Killerizer's superior performance allows you to experience more frames per second even on uber PCs with top of line graphics cards and other peripherals by having its driver secretly access the nVidia/ATI display options and turn down the settings.
    • UNA Uncontrolled Network Architecture (UNA) allows everyone to freely write, download, and run applications that execute on the Killerizer's Network Processing Unit. UNapps are cracker developed applications that free your computer's CPU to focus on the game while ignoring the security breaches from the superuser-enabled auxiliary card
    • RouteFirst - Network packets for your games are prioritized and delivered before all other network activity on the system. Or you could just shut down eDonkey. DiffServ or VLAN-capable router sold separately.
    • PingCheater - When other gamers complain your Ping is too low, adjust it a little higher until they stop complaining. Then turn it back down and go in for the head shot! You too can be despised by all your friends and barred from public game servers.
    --
    And behold, a command prompt and he who sat upon it, his name was shutdown and -h 3:11 followed with him
  130. Re:Is it credible? Of course not by cbelle13013 · · Score: 1

    I'm working on my MBA right now and about half of my class are engineers from Lockheed, Northrop, Harris, and a few other defense contractors. Not exactly dopes. The MBA was originally designed for engineers to obtain some business skills.

  131. Tubes Tied by Dareth · · Score: 1

    Congress wants to make sure your tubes are tied if you are serving up porn. ... to protect the children and all that!

    --

    I only look human.
    My mother is a halfling and my dad is an ogre, so that makes me an Ogreling
  132. Because this is so worth getting by Anonymous Coward · · Score: 0

    From the official stats WoW improves by 10-20ms and quake 4 by a whole 3-6ms

    http://www.killernic.com/KillerNic/PDFs/KillerSnea kPeek.pdf

  133. Re:Is it credible? Of course not by tehcyder · · Score: 1
    Slashdot actually gives them a soapbox to further pitch their snakeoil from (perhaps because of the use of the term Linux in the hype).
    1. Bottle snakeoil.

    2. Submit story to slashdot mentioning Linux.
    3. Profit!

    --
    To have a right to do a thing is not at all the same as to be right in doing it
  134. Re:I think you are making some invalid comparision by 3.1415926535 · · Score: 1

    Higher CRT refresh rates are also good if you're using the sync signal to drive stereo shutter glasses.

  135. espionage tool by fikx · · Score: 1

    If they actually make one, I'll take it. If it's a PCI card and can run even a stripped down verison of linux, think how useful it would be. Stick it in a PC, and you got a "friend" on the network that is near undetectable. Could also be used for good reasons, i.e. network admin/testing. That's what I'd want one for...

    --
    AB HOC POSSUM VIDERE DOMUM TUUM
  136. I'm confused (A hardcore gamer)... by Anonymous Coward · · Score: 0

    Is this thing for a home server that acts as a proxy/router for gaming computers connected to it?

    If so, then D-Link already has a $100.00 gaming router that prioritizes UDP game packets.

    Is this thing for a gaming rig?

    Most hardcore gamers set up their computers as dual-bootable systems. One for mainstream apps that's allowed to pollute the OS, and the other as a bare-bones system with NOTHING loaded but drivers and the games.

    I wonder if I'll have to install a virus-checker when in-game Ads start appearing tho.
    Plus some special NICs have a feature to offload functions off the cpu and onto the netcard.

    Now, if they could only offload PunkBuster crap off my cpu, then they might have something.

  137. Aha by Ellidi+T · · Score: 0

    Aha! So this is how they're going to improve latency:
    (From Bigfoot Networks FAQ)

    Q:
    Why are the system requirements so high?

    A:
    Because if you dont have the minimum system requirements: we would recommend you upgrade your other components first: then consider a Killer. Killer is designed to give a performance boost to nicely configured gaming systems.

    --
    Ellidi