Slashdot Mirror

← Back to Stories (view on slashdot.org)

NVIDIA nForce 4 SLI Intel Edition Launched

Posted by CmdrTaco on from the coming-to-you-live-from-the-third-dimension dept.

Spinnerbait writes "NVIDIA took the wraps off their nForce 4 SLI chipset platform for Intel Processors today and there's a full review and showcase with benchmarks up at HotHardware. As with NVIDIA's AMD version of this chipset, motherboards based on the technology will support dual PCI Express graphics cards for load sharing in 3D Gaming applications. What's perhaps even more interesting is how the new NVIDIA memory controller actually allows the platform to out-pace Intel's own i925XE in virtually all of the benchmarks."

35 of 133 comments (clear)

  1. Full article mirrored by winkydink · · Score: 3, Informative

    Here

    --

    "I'd rather be a lightning rod than a seismometer." -Ken Kesey

  2. I heard that its isn't supposed to be as good by Anonymous Coward · · Score: 4, Informative

    I remember reading on the Inquirer, that on a one on one comparison, the nForce Intel boards weren't able to keep up to the AMD ones, on more than just a processor basis. Was a few weeks ago though, so possible could have been fixed, i.e. driver probs

    1. Re:I heard that its isn't supposed to be as good by mauriceh · · Score: 2, Insightful

      If anything the test was "fixed" to make the Intel CPUs look better.
      For example the test systems were Intel 3.73GHz versus the AMD 4000+ at 2.4GHz.

      Considering the competitive AMD CPU is the 2.6GHz FX55 model, this is obviously a skewed result.

      They pitted an $1,100 Intel P4 against the $500 AMD Athlon64 4000.

      Even if they had compared against the much faster AMD Athlon64 FX55, the price delta is still huge. The FX55 is an $835 chip versus the $1,100 Intel!

      Even so, on most tests the AMD soundly won.
      Some of the tests would have been so embarrassing they skipped testing the AMD altogether.
      Where are the Sysoft Sandra results for the Athlon64?

      Add to that the other hidden factor: Chipset Cost!

      With the AMD64 on board memory comtroller, the Intel version of this chipset costs nearly double the AMD version.

      So, let's sum up the cost results:
      $1,100 Intel CPU, with $400 RAM on a $250 motherboard. $1,750 total.

      Versus:
      $500 AMD CPU, with $250 RAM, on a $150 motherboard. $900 total.

      The only reason anyone would buy this is if they are an Intel fanboy, or are still drinking the Intel KoolAid.

      --
      Maurice W. Hilarius Voice: (778) 347-9907
  3. Re:Humans in my game by BigWhiteGuy_27 · · Score: 5, Funny

    Imagine you can select characters like Brad Pit or Tom Cruse for the game

    Or CowboyNeal...

  4. EM emissions by Pants75 · · Score: 5, Funny
    When are these things going to start spitting out microwaves? We're already into TV and FM Radio Band emissions at 200-400Mhz. Microwaves really aren't that far off.
    While there are some radar bands from 1,300 to 1,600 MHz, most microwave applications fall in the range 3,000 to 30,000 MHz (3-30 GHz). Current microwave ovens operate at a nominal frequency of 2450 MHz, a band assigned by the FCC.

    You'll be glad you kept your old steel PC case when we get this sort of speed out of MBs

    Pete

    1. Re:EM emissions by Rude+Turnip · · Score: 2

      You do realize that MHz is a measure for any sort of frequency (ie clock cycles) and not just radio waves, right?

      --
      Bill Clinton: Pimp we can believe in. - The Shirt!!!
    2. Re:EM emissions by tehcrazybob · · Score: 2, Interesting

      Since computer components run at extremely low power, the radiation shouldn't be an issue.

      At the moment, computers do cause some harmful radio interference if you leave the side of the case off. Since this is Slashdot, I assume there are several people reading this who have theirs off. However, even acrylic cases or case windows are enough to stop that radio interference.

      Even if the frequency picks up enough that we were getting microwave radiation, an aluminum case would still be able to block it. Even with the case open, the output would be so low that the worst it could do would be to ruin your cell phone reception if you were trying to talk with your head inside the case.

      --
      Computers need to explode more often.
    3. Re:EM emissions by Ford+Prefect · · Score: 2, Interesting

      Since computer components run at extremely low power, the radiation shouldn't be an issue.

      You joke, but I gather there were minor problems at Jodrell Bank when PCs' clock frequencies (and/or harmonics) happened to coincide with important radio frequencies used for radio astronomy.

      As you say, though it's hardly dangerous - but having done an undergraduate experiment there some years ago in which an FFT of pulsar data detected nasty big peaks at 50Hz, 100Hz, 150Hz etc. (mains power...) I'm wondering if all man-made alternating currents should be banned, for aesthetic and scientific reasons... ;-)

      --
      Tedious Bloggy Stuff - hooray?
  5. And what of... by Anonymous Coward · · Score: 5, Interesting

    All the motherboard manufacturers who dumped R&D into having to build alternative SLI solutions? One example being the Tyan S2895 which uses dual nForce4 chipsets to achieve true 16x pci-e in SLI mode. I'm hoping that nVidia didn't try to hold this information back from motherboard manufacturers otherwise we may see a lashback against nVidia. And considering I spent months hunting and waiting for a true 16x pci-e SLI solution I am a little disappointed in nVidia for waiting so long.

    1. Re:And what of... by pantherace · · Score: 4, Informative
      The Opteron/Nforce4 boards are still the only 16x/16x sli solutions available.

      I just double checked on Intel's website, and the best I could find was 8x/8x (3 x8 and 1 x4 PCI express slots (28 lanes total)) And with that it is not possible to have multiple x16 slots (Heck, it's impossible to have 1) (It's possible I missed a better one. I was looking in the server section.)

      The main reason that Tyan can do that is because of AMD's superior Hypertransport-based bus design in Opterons, over the shared bus favored by Intel. It's also the reason why Opteron scales a lot better than Xeon.

      The other reason Tyan can do that is that Nvidia realized how easy it would be to make very slightly different chipsets that facilitated that. Basically they are just Nforce 4 chipsets, that can operate in parallel, giving 40 Pci express lanes (2-way) or 80 PCI express lanes for a 4-way Opteron. (Note a maximum of 4 x16s, as the other 16 can only be a max of x4, due to the 20 lanes per nforce4)

      You can't do x16/x16 with any Intel Processor, as of now. (Though having seen how little x16/x4 or x16/x2 hurts benchmarks (vs standard x8/x8) I'm not convinced it's a big deal at all.)

    2. Re:And what of... by Slashcrap · · Score: 2, Insightful

      One example being the Tyan S2895 which uses dual nForce4 chipsets to achieve true 16x pci-e in SLI mode.

      The Tyan is an insanely specified server board with something like 40 PCI-E lanes is it's basic config. It's not like that because Nvidia wouldn't release specs. It's like that so you can run several high performance workstation level video cards. I don't even think it uses two NForce 4 chipsets, I think you've just misunderstood the specs. Do you have evidence to the contrary Mr Coward?

      And considering I spent months hunting and waiting for a true 16x pci-e SLI solution I am a little disappointed in nVidia for waiting so long.

      Now this is just funny.

      Are there any cards that can fully exploit a 16 lane PCI-E slot available? No.

      Are there going to be in the near future? No.

      So why are you so desperate to get hold of a motherboard that supports it? Because you don't really know what you're talking about. You just want to be fully buzzword compliant, even if it costs you thousands.

  6. Re:nVidia better than Intel by Skye16 · · Score: 5, Funny

    breathe?

  7. Re:nVidia better than Intel by scum-e-bag · · Score: 2, Funny

    With very little punctuation in your post I can definately see that from your definitions you are an nVidia fan!

    --
    Does it go on forever?
  8. nvidia by chrisnewbie · · Score: 2, Insightful

    Sure benchmark is good when it reflects what most gamers have at home. ---- Sure they score 20 gazillions points with 3dmark but it's almost a machine fit for nasa that would cost around 3000$ to buy. ------ why cant they use a normal machine like a pentium IV 2,4 ghz with a ultra-ata 166 and 1 gig of ddr 400.That's more common and more realistic.

  9. multi-everything by Cruithne · · Score: 2, Interesting

    Seems like we're trending towards multiple everything recently.. multicore CPUs, SLI.. how long before this propagates to everything?

    As a sys admin, I love the prospect of redundancy, but are there any benefits to bringing this multiplicity to anything else from a consumers perspective? Or does it stop here?

    1. Re:multi-everything by kannibal_klown · · Score: 4, Interesting
      Seems like we're trending towards multiple everything recently.. multicore CPUs, SLI.. how long before this propagates to everything?


      Wouldn't doubt it.

      You can only improve on things so long before you need a complete redesign. Adding more to the mix is a great stopgap that extends the usefulness of technology.

      At some point AMD and Intel are going to have to perform a MAJOR redesign (even bigger than the dual-core). Granted this might not be until we reach the 7GHz mark, but there is an invisible line somewhere.

      There is one big downside for the consumer though: increased prices. Dual-Core CPU's will be more expensive than regular ones. SLI graphics will require buying 2 cards. RAID storage requires multiple hard drives.

      Personally I think it would be cool if my next computer were dual-core with SLI video ports and a RAID setup. Whether or not I can afford it, that's another story.

      With the obvious effects of distributed and grid computing Sony's supposed cell tech might actually prove to be interesting (though I'd prefer it on a more local scale).
  10. Humbug! by Robotron23 · · Score: 4, Funny

    3D gaming, 3D schaming.
    Back in my day we had the Voodoo 2's and the ol' 6mb of ram, 12 if you were rich! Couldn't even get two separate sprites on the screen without extreme lag... but we liked it!

    1. Re:Humbug! by UWC · · Score: 2, Interesting
      Voodoo 2 came in 8MB and 12MB versions, the latter having an extra 4MB of texture memory. They were pass-through cards, requiring a separate, primary video card for non-3D stuff. They could be connected to do SLI (which at that time I think was Scan-Line Interleaving, the cards handling alternate lines of monitor resolution). I think with SLI, you could play your games at an astonishing 1024x768 in glorious 16-bit color. Single cards were limited to 800x600.

      I had a single 12MB one that I bought used on eBay. It wasn't much in 2000, but boy did Unreal Tournament and Counter-Strike play smooth as butter on that and a 450MHz K6-2. I didn't get back to that level of performance on a 32MB GeForce2 MX until I upgraded to a 1GHz Athlon Thunderbird. Wow. I have no idea where that K6-2 box is. I guess the Voodoo 2 is probably still in it, too.

      My best card at the moment is a 128MB GeForce 3 Ti200. I'm so behind the times.

      There was a 6MB 2D/3D combo variant of the original Voodoo called Voodoo Rush. I recall various claims that it wasn't very good. The original Voodoo cards had 4MB of memory.

      --
      Honor Among Slackers. A veri
  11. Nice motherboard, but... by supremebob · · Score: 4, Interesting

    Will it work with the new dual core P4 CPU's? It doesn't make much sense to buy a high-end motherboard if you can't get the high-end CPU to go with it.

    1. Re:Nice motherboard, but... by Billly+Gates · · Score: 2, Interesting

      Actually people who would buy such systems with dual cpu's and video cards would be video production profesionals (the parent poster), cad and 3d modelers, and engineers. Such systems are not cheap for the average user.

      At Liz Claiborne in the 90's when I worked there, the merhcandising team used dual vodoo's for studioMax when I worked there over the more pro video cards. They were very fast.

      All these apps fly on Intel cpu's if you look at any benchmark. This is because they contain hand written assembly optimized for intel chips. They are multithreaded and designed to be smp ready.

      But games it would be silly since they are not optimized for any one chip and are not smp ready.

      My guess is these units for AMD might appeal to the gamer but the Intel ones will be used by the professionals. An SMP core as well would make it a kick ass workstation.

      --
      http://saveie6.com/
  12. RAID 5 by gallard · · Score: 5, Informative

    The one thing the Intel version has over the AMD version of this chipset is RAID 5 support. A RAID 5 controller card by itself is over 100 bucks. Dammit this is going to make me want to turn over to the dark side.

    1. Re:RAID 5 by Anonymous Coward · · Score: 2, Informative

      It is "fakeRAID" so it is all done in software (however some fakeraid cards can do RAID1 in hardware) where as the 100 or so bucks one you can get normally has its own XOR processor (sometimes a few megs of ram).

    2. Re:RAID 5 by rpozz · · Score: 2, Insightful

      It's hooked up to the chipset, not the motherboard. That means a RAID array created on eg. a Silicon Image 3114 RAID controller can be transfered onto any motherboard with the same chipset.

      And besides, using mdadm under Linux to create a RAID array out of the same drives using exactly the same configuration will work fine. I've tried it.

  13. Another Intel-funded CPU comparison? by hirschma · · Score: 4, Interesting

    Hmm, let's see. Let's take an Intel processor with these characteristics:

    * Fastest consumer CPU they offer,
    * Priced at about $1100, street

    And compare it to the AMD offering, with these characteristics:

    * Second fastest CPU they offer,
    * Price of about half of the Intel offering.

    Yes, that is a most fair review. It makes perfect sense to conclude that, on mostly identical chipsets, that Intel is faster.

    How much are these sites paid under the table?

    1. Re:Another Intel-funded CPU comparison? by CDarklock · · Score: 2, Informative

      How about we take an Intel CPU at 3.73 GHz and compare it to an AMD CPU advertised to perform around that same clock speed? That gives us the 3800+, the 4000+, and the FX-55. Traditionally, AMD's estimation has been 5 to 10 percent high in terms of what Intel processors they can match (across the board, as opposed to gaming where the estimates are about right), so we can assume:

      - The 3800+ performs at 3.42-3.61 GHz which is too low.

      - The 4000+ should perform at about 3.6-3.8 GHz, which is about right.

      - The FX-55 is undoubtedly faster than the 4000+, owing to its 2.6 GHz actual clock speed compared to the 4000+ at 2.4 GHz. This is a 1/12 increase, so we can expect the FX-55 to perform at 3.9-4.2 GHz... which is significantly higher than the 3.73 we're targeting.

      So the FX-55 is too fast, the 3800+ is too slow, and the 4000+ is about right. An argument might be advanced for the FX-53, but I think those have been discontinued.

      Where exactly is the problem? I mean, if you RTFA, they say outright that if all you care about is gaming the AMD outperforms the Intel by around 10%, which gives us that 4000+ they advertise... but if you need to do other things, the Intel outperforms the AMD slightly. Which is roughly what you should expect, given the Intel and AMD core competencies. If the 4000+ specs out in the dead center of the range I estimated, it should come out at 3.7 -- which is just over 99% of the Intel clock speed, and should result in less than a 1% margin of loss. This is about what we see in the results.

      In other words, nobody paid me squat, and my own experience with processor comparisons predicts pretty much the exact results we saw. So any claim of "bias" seems ill-considered.

      --
      Microsoft cheerleader, blue flag waving, you got a problem with that?
  14. Re:Humans in my game by Ford+Prefect · · Score: 4, Insightful

    At this rate soon we will have processors that are capable of rendering real video instead of animation. Or say animation as real as videio footage.

    Hardly. Most game-style rendering today is mostly smoke and mirrors; while 3D graphics hardware has improved at a ridiculous rate over the last couple of years, there's still a long way to go before certain, everyday scenes can be rendered.

    Something I'd like would be a 'city-renderer', capable of rendering a decent-sized European city (i.e. not a grid) from aerial views down to individual rooms. While a clever level-of-detail system could go a long way towards this, there would still be an utterly horrendous amount of geometry for a typical skyline shot.

    Now add traffic, crowds of humans (typical FPS-style games give up after about ten or so, strategy games use crude mannequins for more), properly reflective surfaces and whatnot, motion blur and decent HDR and your quadruple-SLI Geforce 9000-Hyper-Pro-Matic setup will still grind to a halt.

    Things are slowly getting there, but I'm still waiting - but like a gas, FPS-style generic corridors will expand in processing requirements until they saturate even the greatest hardware. Look at Doom 3, for example... ;-)

    --
    Tedious Bloggy Stuff - hooray?
  15. Already has by elgatozorbas · · Score: 4, Funny
    Seems like we're trending towards multiple everything recently.. multicore CPUs, SLI.. how long before this propagates to everything?

    I have even heard about a guy with TWO complete individual PCs...

  16. Re:Another troll who didn't RTFA? by hirschma · · Score: 4, Insightful

    OK, I'm responding to a troll that called me a troll :)

    Normally, in a processor comparison, the processors are comparable for a reason - same positioning by the companies involved, same price point, whatever.

    In this case, it appears that the only reason why the AMD proc was chosen was to give Intel "wins" in close contests, like LAME MP3 encoding, and to not make Intel's best look too awful in the cases where AMD won.

    Point is, Intel was represented with its best game. Why should AMD be presented in a less favorable light?

    There is little journalistic integrity with these enthusiast sites.

  17. Nvidia RAID... not so good. by hirschma · · Score: 2, Interesting

    I wouldn't consider Nvida RAID a feature worth buying into at this point.

    I have one box under Windows using Nvraid, and it is just terrible. It drops drives from the RAIDs seemingly for fun, and configuring a bootable RAID is difficult under XP, and impossible under Win2k (even with an SP4 slipstream install, in case anyone was going to point that out).

    The management software is crude at best. It cannot, for example, email alerts when a drive drops off.

    My $.02.

    jh

  18. AMD board with RAID-5 by DamienMcKenna · · Score: 2, Informative

    Check out the Gigabyte GA-K8NXP-SLI board which has SATA-150 RAID-5 via an extra chip.

    Damien

  19. Physics >> FCC by Morgaine · · Score: 4, Funny

    Current microwave ovens operate at a nominal frequency of 2450 MHz, a band assigned by the FCC.

    I think you'll find that the physics of water molecule resonance had something to do with choice of this band.

    Funny how every other country in the world chose the same band, despite not being ruled by the FCC ... :)

    --
    "The question of whether machines can think is no more interesting than [] whether submarines can swim" - Dijkstra
  20. How is the nforce4 Linux support by Anonymous Coward · · Score: 2, Interesting

    I just bought an SLI board (AMD) and was wondering how well the raid and SLI video configuration is supported under Linux.

    I've booted my machine into it and to my suprise the ethernet devices worked out of the box with Xandros (based on debian sid). I still do not know about the raid or SLI video, however. I'm using a crappy old S3 PCI video card right now, but am about to receive two GeForce 6800 GTs in the mail. Can I use these bad boys in linux? Anybody know?

    1. Re:How is the nforce4 Linux support by hag3r · · Score: 2, Informative

      There has been quite a lengthy discussion on LKML on that topic. IIRC, there are some issues with NCQ and sound support, but it's an interesting thread well worth reading, if you're into nForce4-thingies that is.

  21. Open Real-Time Ray-Tracing by FiiKness · · Score: 2, Informative

    If you haven't taken a gander at it yet, you may want to take a look at OpenRT and projects using OpenRT such as Quake3 Raytraced. Also take a look at the hardware architecture as well.

    Ray-tracing presents a much more detailed rendering of a scene, but was always considerably slower than rasterization. If hardware-accelerated ray-tracing architecture grows in the market, you may see your skyline beautifully rendered in real-time .. with traffic, crowds, etc.

  22. Obstacles to Mainstream VR by podperson · · Score: 4, Insightful

    My wife operates a VR research lab (they still exist) and all the new hardware is just great for them. Unfortunately, it only addresses one of the concerns which must be addressed before we can all live in the world of "Snow Crash". The basic problems for mainstream VR are as follows:

    1) The headsets really haven't "tipped" price-wise. Kind of like LCD screens for a long time, they stay expensive (around $10k) while slowly improving in features (e.g. resolution, motion tracking). Until they get "good enough" the prices won't trend downwards. (There are cheap headsets, but they make you sick pretty fast. Even the pricey ones make you sick after 30 mins or so ... so you won't be playing WoW in them.)

    2) The big issues w.r.t. UI remain unsolved. E.g. a lot of VR setups involve complex motion tracking and setting aside a room for subjects to walk around in. Usually a second person watches the subject to prevent them from, say, running into a wall... There are rigs that allow you to suspend the subject to allow them to walk through theoretically infinite landscapes... we're talking six figures though.

    3) Behavior capture. The solutions to tracking movement remain pretty experimental and invasive. All the stuff we've talked about so far will, at best, get you walking around in a virtual landscape, capture your head movements (kind of), and maybe capture some of your arm and finger movements. Even assuming your $500,000 suspension rig captures all your body movements perfectly, we still to capture facial expression and lip synch. (So far, spacial 3d audio is pretty primitive too ... Teamspeak is a long way from a person's voice emanating from their position in a shared world with lots of people.)

    4) Force Feedback. All this VR is going to seem pretty lame when you can walk through solid objects or your hand passes through an item you're trying to manipulate. Arguably, this is a subset of item (3) above, but in fact just allowing people to walk around in an unlimited expanse is a big enough problem...

    There are plenty of finer grained issues to deal with, but the rendering of VR scenes (at least, so far) has turned out to be the easy part. At the moment, if you wanted to play WoW in VR you'd need to set aside a large room, buy an expensive HMD, and a really expensive suspension rig. (Luckily, WoW lets you run straight through people so the UI will match this perfectly.)