Slashdot Mirror


NVIDIA's 55nm GeForce GTX 285 Launched

Visceralini writes "NVIDIA is launching yet another high-end 3D graphics offering, an optimized version of their top shelf GeForce GTX 280 single GPU card, dubbed the GeForce GTX 285. This new GeForce is a 55nm die-shrunk version of the legacy GTX 280 with lower power consumption characteristics that don't require an 8-pin PCI Express connector, rather just a pair of more standard 6-pin plugs. Performance metrics are shown here in a number of the latest game titles including Fallout 3, Left 4 Dead, Far Cry 2 and Mirror's Edge. The new GTX 285 is about on par or slightly faster than a GTX 280 but with less power draw and some room for overclocking over the reference design."

82 comments

  1. Power Savings!! by Arthur+Grumbine · · Score: 5, Funny

    The new GTX 285 is about on par or slightly faster than a GTX 280 but with less power draw and some room for overclocking over the reference design.

    40W less while idle (vs. 280), @ $0.12 kWh, means if I can pick one up for $400 (I can dream, can't I?), it will have paid for itself - through power savings - in less than 10 years!! I know what I'm spending my tax refund on!!

    --
    Now that I think about it, I'm pretty sure everything I just said is completely wrong.
    1. Re:Power Savings!! by CDMA_Demo · · Score: 2, Informative

      in less than 10 years!! I know what I'm spending my tax refund on!!

      You can also use it to crack passwords even Faster!

    2. Re:Power Savings!! by Lord+Ender · · Score: 1

      Not exactly... but if you were about to spend $250 on the mid-range GTX260 card, you can now get the high-end GTX285 which will pay for the difference in power saving after a few years.

      --
      A slashdotter who didn't build his own computer is like a Jedi who didn't build his own lightsaber.
    3. Re:Power Savings!! by Anonymous Coward · · Score: 0

      It obviously doesn't help much with the bills however less power draw in your computer is always a good thing and looking at the tests given there is only one where it performs worse.

      01000111 01101111 01101111 01100100 00100000 01101010 01101111 01100010 00100001

      Thanks :)

    4. Re:Power Savings!! by isBandGeek() · · Score: 2, Insightful

      What enthusiast that spends big dollars on the latest cards uses their card for more than a year?

      Most people that rush out to buy these type of cutting-edge hardware replace them every few months or so. The savings realized by power conservation will never cover the difference between the two with the crowd these new cards (indeed, all new cards) are targeted at.

    5. Re:Power Savings!! by Anonymous Coward · · Score: 0

      What's the Nvidia GeForce GTX 285 all about? Is it good or is it whack?

    6. Re:Power Savings!! by Anonymous Coward · · Score: 0

      I tend to buy the very top of the line every few years just so it will last me for the next few years.

      In 3 or 4 years the GTX285 won't be high end, but it will certainly be able to run games still.

    7. Re:Power Savings!! by feepness · · Score: 1

      What enthusiast that spends big dollars on the latest cards uses their card for more than a year?

      Paid $400 for an 8800 GTX well over a year ago, wish I would have caught the 8800 GTSs. But anyways, I plan to keep it for more at least another 12 months. We'll see what's out then.

    8. Re:Power Savings!! by afidel · · Score: 1, Interesting

      Or you know you could go with a 9600GSO which uses about 45W PEAK and plays most modern games at decent framerates (Crysis 1280*1024 medium, 20fps). I use Rivatuner to underclock the card by 50% at windows startup and run it at about 20% overclock with game profiles. Since my PC is running Mediaportal and is on 24x7 this saves me quite a bit over the course of the year.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
    9. Re:Power Savings!! by Anonymous Coward · · Score: 0

      How is 141W for doing nothing even acceptable? At that price, the card should include a bare bones framebuffer/2D-accelerator chip and simply cut the power to the GPU completely, if they can't get the idle power consumption under control.

    10. Re:Power Savings!! by Anonymous Coward · · Score: 0

      How is 141W for doing nothing even acceptable?

      141W is total system draw. Not for the GPU alone.

    11. Re:Power Savings!! by Anonymous Coward · · Score: 0

      141W for a whole system, which does nothing (idle), is still far too much. A good portion of that must be the graphics card, because my systems don't even use that much under full load.

    12. Re:Power Savings!! by F34nor · · Score: 1

      Why not throw in a an atom too so when your are browsing the web etc. you draw even less. That's the thing about Moore's law. You never need to buy a new computer ever again if you don't play games, edit video, steal video, or render shit. 99% of the crap I do could run on my Nokia e71 (including Quake) if I had an external usb keyboard and video card.

    13. Re:Power Savings!! by jandrese · · Score: 1

      Shoot, I dropped the big bucks on the 8800GTX when it came out and it's still an able performer. It looks like it's going to be another 6 months or maybe even a year before I have to dial back from "ultra extreme" settings to merely "high" settings on new games (although I didn't try Crysis).

      --

      I read the internet for the articles.
    14. Re:Power Savings!! by Creepy · · Score: 2, Interesting

      The PSU requirement is apparently 550 Watts, and you can usually save a lot of money when you drop from 700 Watts to 550-600, however, I remember seeing a 700 Watt PSU at NewEgg for $50 after rebate, which is about what I paid for my 500 Watt at Christmas (after rebate - I'd planned for 600+, but PSU and a few other things fell to budget axe).

      If you're building a system from scratch you may be able to save additional money with a lower power draw card. Also, waste heat from the PSU is lower with smaller PSUs.

    15. Re:Power Savings!! by db10 · · Score: 2, Funny

      .. and if it's whack, is it wiggidy whack or the regular kind?

    16. Re:Power Savings!! by skyride · · Score: 1

      THE PSU ONLY DRAWS WHATS ASKED OF IT!!! Seriously, its quite silly. A 700 watt PSU will only draw 700 watts if the components in the PC ask for 700 watts.

    17. Re:Power Savings!! by lsatenstein · · Score: 0

      My electricity comes it at 4Cents, except when the temperature is -20F (or -12C, which ever is reached first). Then it is 23Cents /kwh. So, we folks do not use the dryer on cold nights, and we turn the house temperature down to 65F and wear sweaters. By the way, only about 2 weeks of such cold in a winter season. I don't think I could wait 20 years to get a Nvidia payback based on electrical savings. (PS, average year round, electricity is around 6 cents per kwh.)

      --
      Leslie Satenstein Montreal Quebec Canada
  2. State of the Market by youknowjack · · Score: 4, Informative

    - The current best performing single card is the GeForce GTX 295
    - The best performance setup was (before this card) a tossup between dual GeForce GTX 295s (quad SLI) and three GeForce GTX 280s (three-way SLI).
    - The overclocking potential of the GeForce GTX 285 & reduced power consumption might make a three-way 285 setup preferrable to a dual 295 setup (for enthusiasts)

    1. Re:State of the Market by Anonymous Coward · · Score: 0

      All I care about is best performance under 50W and under $100. Is there anything out there better than the 8600GT?

    2. Re:State of the Market by youknowjack · · Score: 2, Informative

      All I care about is best performance under 50W and under $100. Is there anything out there better than the 8600GT?

      The Radeon HD 3870 is extremely similar in terms of price and performance. I guess it depends on what you can get the best deal on.

    3. Re:State of the Market by Anonymous Coward · · Score: 0

      However, ATI tends to have much shittier drivers on Windows and elsewhere.

    4. Re:State of the Market by jdb2 · · Score: 3, Informative

      The overclocking potential of the GeForce GTX 285 & reduced power consumption might make a three-way 285 setup preferrable to a dual 295 setup (for enthusiasts)

      You do know that the GeForce GTX 295 has the same overclocking potential and reduced power consumption as the 285 because both use the same chip(s)?.

      jdb2

    5. Re:State of the Market by Anonymous Coward · · Score: 0

      Correction: The GeForce 9800 GT and Radeon 3870 HD are comparable in performance, power, and price.

      Note: They use about 100W.

    6. Re:State of the Market by DigiShaman · · Score: 4, Insightful

      Unless you live way up north or play games only in the winter, dealing with 840 Watts of heat is going to be problematic for a dual GTX295 setup. Summer is worse in that you now have to pump out that heat through the AC system.

      People often will bitch about their cable/DSL bill, but have they ever tried to calculate the monthly cost of electricity their gaming rig racks up alone?

      --
      Life is not for the lazy.
    7. Re:State of the Market by therufus · · Score: 1

      Are you kidding me? Nvidia drivers are terrible. I installed the latest chipset (nvidia based) and VGA (8800GT) drivers a few weeks ago on my Windows XP partition. Now every time I go through the drivers to adjust anything from overclocking options to simple fan speed adjustments, next restart the picture on the screen gets corrupted and the PC locks up. You have to restart in safe mode, delete the video cards (SLI setup) out of device manager and re-install them.

      These are brand new drivers! My next card will be an ATI.

      --
      You moved your mouse. Please restart Windows for changes to take effect.
    8. Re:State of the Market by Flentil · · Score: 3, Informative

      By brand new drivers, do you mean the beta drivers? Just curious because a lot of people do that and don't get why they don't work 100%. They're beta, that's why. Still being tested to iron out the bugs.

    9. Re:State of the Market by afidel · · Score: 1

      Yes, 9600GSO. 96 8800 series stream processors(G92) uses about 45W peak. Be sure to get one with 384 or 768MB of ram, they reused the name for a new G94 part with much lower performance. Should run you under $75.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
    10. Re:State of the Market by Anonymous Coward · · Score: 0

      http://en.wikipedia.org/wiki/GeForce_9_Series#Geforce_9600_GSO says it (the G92-150) uses 105 watts. The 9500 GT is roughly identical to the 8600 GT in price, performance, and power.

    11. Re:State of the Market by afidel · · Score: 2, Informative

      http://archive.atomicmpc.com.au/forums.asp?s=2&c=7&t=9354
      This lists it at 46W which is much more in line with my experience. The TDP might be 105W for some crazy reason, but my system with Athlon 64x2 4200+EE, 2x7,200rpm HDD and a 9600GSO uses about 150W when gaming so there is no way in hell the GPU is pulling down 100W by itself.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
    12. Re:State of the Market by afidel · · Score: 3, Interesting

      Hate to reply to myself but I just figured out what they did, they pulled the specs from an OC 8800GT G92. The 9600GSO has a single 6pin PCIe power header which means absolute power draw is limited to 75W(cable)+17W(slot) or 92W, and it would likely be unstable well before hitting that since power lines don't like to be pulled to their limit.

      --
      There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
    13. Re:State of the Market by Zephiris · · Score: 0

      8600GT usually bests the 3870 for a similar price. Recently the 9600GT was priced similarly to the faster/better 8600GTs at ~$85.

      The 8600GT can slightly edge out in a variety of games (including GTA4), whereas the 3870's higher processing power can beat it in a few things that don't use complex shaders at all. The 9600GT utterly demolishes both in virtually everything, was within 10-15% of the 9800 GT, at $60 cheaper. Both are slightly more expensive with the big sales season over, though I'm sure there'll be more deals soon.

      The 3870 supposedly has better hardware than both, but basically all recent drivers (since at least 8.8) appear to botch the GPU load (with fixed clock speeds mind, no power throttling) so that most games only use 30-50% of the GPU's capacity. Obviously the 3870 has more a great many more shaders, but they're simply not getting utilized properly.

      It's a bad feeling when you get a low price on a decent 3870 GDDR4, only to have to have your older 8600 consistently performing better and providing superior gameplay experience.

      --

      "A Goddess rarely smiles for she is forced by others to be an island unto herself." - Zephiris
    14. Re:State of the Market by Zephiris · · Score: 1

      I've heard NForce motherboards tend to have a fair number of compatibility problems. Don't be so quick to blame the Geforce (VGA) drivers or the Geforce card.

      'Anything from overclocking options to simple fan speed adjustments' are handled by the NTune options, not the Nvidia geforce options. It integrates into the same control panel, but they're different things. There are also usually enough quirks with NTune itself to prefer relying on Rivatuner (driver overclock, hardware direct fan speed control) if you're going to tweak those. It has nothing to do with the drivers or their quality, as they're not a part of the drivers.

      Have you tried configuring things NOT in an SLI setup to narrow down the diagnosis of a root cause?

      --

      "A Goddess rarely smiles for she is forced by others to be an island unto herself." - Zephiris
    15. Re:State of the Market by mxs · · Score: 1

      If you care about GPU memory, the GTX295 only delivers 896mbyte per GPU (it's a dual GPU card), while the GTX285 delivers 1024mbyte. If you intend to do stuff with CUDA this may be the deciding factor for you -- and, to a limited extent, also for 2560x1600 gaming on 30" displays. The /only/ game where this would actually come into play, though, would by Crysis: Warhead, right now.

      Power draw of a 3-way 285 SLI will likely be more than a "Quad"-SLI 295. Cooling might be an additional problem (there is not exactly much room between the cards to take in air, etc.) Single-GPU-performance of the 295 will be less than single-GPU performance of the 285 (it is more comparable to the GTX260 on the 295).

      If you are going to go with a Tri-SLI-Setup, you will probably need a 1200W power supply, ESPECIALLY if you intend any sort of overclocking, and considering you'll likely want to be running this in tandem with a Core i7 920/940/965EE; Otherwise, you will very likely be CPU-bound instead of GPU-bound. The Phenom might make a platform for it too, but I have not done much proper research into that area.

      When doing Tri-SLI, the cards will be running in 16x / 8x / 8x configuration. While this is not optimal, the X58 chipset simply does not have more lanes than that to spare -- if your application is Memory GPU-Transfer-heavy, this might become a problem. All current games are not (it makes zilch difference in benchmarks). Spending money on a NF200 to get a "true" 16x / 16x / 16x configuration is wasted money -- it is not actually true, but provides 16x / 16x PCIe-communication between two cards, which are then connected to the X58 via one PCIe 16-lane connection (so communication with the third card is still bottlenecked, as is communication with the CPU/Memory). In current implementations, this simply does not have any benefit at all.

      And to stay on the ground in all this : if all you want to do is play games, you will be served well with a single GTX285 in almost any situation, for now. The only game that can actually dip is Crysis Warhead. Other games will still play fine even at 2560x1600.

    16. Re:State of the Market by NekoXP · · Score: 2, Interesting

      I wonder if he has tried NOT overclocking the card or changing the fan speed? :D

      Overclocking is the stupidest, stupidest thing people can do on modern hardware. By designing a graphics card or CPU that overclocks you're pandering to the statistics freaks who want to get that extra 1% performance increase and therefore "more bang for their buck".

      What a f**king ridiculous market. Processors and graphics chips go through sorts and testing for a good reason; they're not rated to go any higher because there is a very good chance they WON'T. Depending on the exact chip you get, at which time in production it was made, and the quality of the PCB it's soldered to (especially if you're overclocking a memory bus, which also relies on the quality of the memory) every card can and will be WILDLY different. Sometimes overclocking by 20MHz is going to completely screw things even though some guy said he got it past 200MHz on some review site.

      Simply stop doing it and guess what, the chip won't overheat, and the graphics card will get broadly the same performance give or take a frame per second in some game that your monitor is not fast enough to even display anyway :)

      Sometimes you have to put down your money, be happy with what you've got, and enjoy your 1-year statutory warranty which clicking ANY of those overclocking buttons automatically voids.

    17. Re:State of the Market by Anonymous Coward · · Score: 0

      Except each 285 has it's own cooler. The 295 shares it's cooler between two chips. Sure, the cooler is probably more efficient, but there's a limit to the amount of heat you can remove with just air in a two-slot cooler, and you'll hit that limit a lot sooner with a 295 than a 285.

    18. Re:State of the Market by jdb2 · · Score: 1

      Except each 285 has it's own cooler. The 295 shares it's cooler between two chips. Sure, the cooler is probably more efficient, but there's a limit to the amount of heat you can remove with just air in a two-slot cooler, and you'll hit that limit a lot sooner with a 295 than a 285.

      Yeah, I was going to mention that, but you beat me to the punch. :) Anyway, this is exactly why you need one of these. Custom designed for solving the above mentioned problem.

      jdb2

    19. Re:State of the Market by sw155kn1f3 · · Score: 1

      TDP is not how much heat chip generates, it's how much chip package is designed to handle.
      Most recent E8XXX Intel processor have 65W TDP, but it's nowhere near how much they really use, mine E8500 uses 21W under full load IIRC.
      Check it yourself: http://en.wikipedia.org/wiki/Thermal_Design_Power

      --
      - Arwen, I'm your father, Agent Smith.
      - Well, you're just Smith, but my father is Aerosmith!
    20. Re:State of the Market by JustNiz · · Score: 1

      >>If you are going to go with a Tri-SLI-Setup, you will probably need a 1200W power supply,

      Baloney. I have a 8800GTX SLI setup. It draws 400W at the wall, which goes up to 500W under full load. Furthermore its watercooled os includes all the pumps etc. Seems to me that 8800GTX draws at most 180W. That means if I put another 8800GTX in my box to get 3-way, it would still olny consume 650W under load.
      The GTX285 cards are 55nm so draw even less power.
      Even allowing for some overhead an 800W PSU would still be overkill.

    21. Re:State of the Market by mxs · · Score: 1

      Baloney.

      I call the same on your comment.

      I have a 8800GTX SLI setup. It draws 400W at the wall, which goes up to 500W under full load.

      Of course you do not cite what other components are in there. But Let's gop with your 8800GTX SLI non-OC setup. The 8800GTX has a TDP of 145W per card with reference clocks -- so with 2-way SLI that would be 290W there. Note that this is not the actual maximum power the card could consume, but it is fairly close. If you overclock, you are going to need more. If you had a 3-way SLI setup, you would have to account for 435W from the Geforce cards alone. A reasonably beefy CPU will have a TDP of around 130W. Without any other components, you will have to account for 565W. Note that we are not talking about "idle" here, we are talking full load. Very few applications/games come close to that on beefy enough hardware -- but to account for a stable system, you'll need it. This is all at spec-clocks, overclocking will yield a lot more draw.

      Furthermore its watercooled os includes all the pumps etc. Seems to me that 8800GTX draws at most 180W. That means if I put another 8800GTX in my box to get 3-way, it would still olny consume 650W under load.

      How 500+180 makes 650 is interesting, to be sure. You also seem to assume that the efficiency of your PSU is close to 100% -- when in reality, you really should not be counting on anything above 80% -- even for ones rated at 85 and above, especially when they are not at 100% load, you will not usually get more than 80% efficiency.

      The GTX285 cards are 55nm so draw even less power.

      Baloney. I mean sure, your nice 8800GTX is 90nm, but it also only has ~680 million transistors. The GTX285 has 1.4 billion. Oh, and it has a spec-TDP of 183W, too (higher if you get the factory-overclocked versions of cards). Let's be conservative and say it really is 183W. Here we have a lower bound 549W draw under load from the tri-SLI setup alone, plus the CPU -- another 130W at stock speeds. Plus all the rest that makes up your computer. Minus the efficiency hit you will take. 689W TDP for processor+GPU alone and you want to stick a 800W PSU in there ? Laughable.

      Even allowing for some overhead an 800W PSU would still be overkill.

      We could have this discussion with a 1000W PSU. Proposing 800W for this setup is ludicrous, and you should know better.

      An argument could be made that a 1kW PSU is enough -- and indeed, there are even (VERY FEW) manufacturers who will offer you a 1kW PSU with enough connectors for the tri-SLI setup (usually you only find that on the 1.2kW models). Don't forget hard disks (~10W while in use, up to 30W on startup), optical drives (idle ~10W, in use 20-25W). Add a single-digit amount of watts for every fan. Don't forget the motherboard and what it can supply on the USB ports. It adds up. And then there's water cooling.

      You could maybe go with a well-made 1kW supply and some extra connectors. I'd opt for a 1.2kW one in that scenario. Note that this does not mean that the machine will draw 1.2kW -- it will do so only when it has to. Efficiency curves of power supplies are not nearly as bad as they used to be, either.

    22. Re:State of the Market by therufus · · Score: 1

      Overclocking is the stupidest, stupidest thing people can do on modern hardware. By designing a graphics card or CPU that overclocks you're pandering to the statistics freaks who want to get that extra 1% performance increase and therefore "more bang for their buck".

      Although you are completely entitled to your opinion, I have to disagree. If you were a racing car driver, would you only put your throttle down to 80% of its capacity just in case you might push the engine too hard? These graphics cards/CPU's/Mainboards/etc are designed to have extra headroom for stability purposes just like the car you may own. People overclock to get the performance out of the unit that it's capable of. Same as people who get aftermarket parts for cars. Suspension, exhaust, engine management computer chips, the works. You can keep your stock computer (which is probably a mac anyway). I'll have my awesome gaming rig thanks. And I'll push it as hard as it can go. My gripe is not the hardware, it's the software drivers.

      I have ceased all overclocking while this problem is occurring. Besides that point, I've only overclocked my cards by 1-5%. With default settings I've tried raising the default fan speed only up to 100% for testing. On restart, same issue. Both in SLI and non-SLI forms.

      --
      You moved your mouse. Please restart Windows for changes to take effect.
    23. Re:State of the Market by NekoXP · · Score: 1

      Overclocking by 1-5% gives you how much of a performance benefit, really? You're far more likely to get a decent speed increase with simply better drivers and more efficiently written games.

      For the car analogy; just because your car's RPM meter goes up to 10000, doesn't mean you should hammer it at that constantly, or even ever. A racing driver would never keep his car running at full tilt, nor even try and get there, because the simple fact is that when you do the whole thing becomes unstable and the engine generally starts to burn. Formula 1 car engines are put through rigorous testing to make sure they can actually perform at the performance level they need to, but once they meet that level, they push it to see when it will die. Then they work out the safe level for it to be at and the driver knows this exactly.

      What you never see, though, is the results of this, neither for the car engine (trade secret) or the results from the tests at the fab. So YOU have no clue whatsoever what that safe limit really is.

      Being in the electronics industry I fully understand WHY these chips go through bin sorts and so on, and WHY you shouldn't actually do anything with them. A good example is the chips Apple used in their late-model PowerBooks. The MPC7447A chip they were using NEVER shipped at higher than a 1.4GHz clock speed. Freescale actually specially sorted out chips which would run at 20% greater voltage and 1.66GHz at the cost of only being specified for a 5 year (standard 10-year) running life at a junction temperature of 85C (standard 105C). They're laser marked on the die, still, as 1.4GHz chips.. this is a case where overclocking is okay, in fact even sanctioned. But they draw a ton more power, generate a ridiculous amount of extra heat, and require a cooling system to match. That's the price of 233MHz extra processor performance.. barely 10% in the real world.

      Get a CPU or GPU that says "1.0GHz" on it, and yes, it may run at 1.1GHz with a bit of tweaking, it may even run at 1.5GHz if you are clever with cooling. It may be that your graphics card is one of the lucky ones that can take a good 5% or even 10% clock speed increase, or your RAM may overclock FAR past the JEDEC specs it was designed to meet (however a vast majority of RAM is designed to meet it to within very strict limits, to the point that they use older processes and cheaper production techniques and get it within 1% tolerance, IF THAT - because this is exactly what the JEDEC standard says it has to be). But there is a chance your chip has been validated and tested at that clock speed and voltage, and badged at that clock speed and voltage, for the simple reason that it will not be reliable at anything higher, confirmed by the burn-in at the fab.

      Since you simply cannot tell what the actual tolerance is (and those "render something real fast until it starts corrupting the display" tests actually serve to damage the chip), it may well be that your chip could be only capable of running at 0.5% past it's rated value, it could be the lucky 10% chip. You're taking a big risk in even trying, and in the end, getting half a frame per second out of some game isn't worth losing a $300 graphics card over. So you want $3 extra bang? $30 extra bang? Come on..

      As for what computer I own; it's a stock Asus P4P-800 which is perfectly good for overclocking, a Pentium 4 HT 2.4GHz which is fine at 2.4GHz, and an ATI Radeon X800 with OverDrive turned *off*. I also have a VIA EPIA, PowerPC G4 (not a Mac, just a board with a G4 in it), PowerPC MPC8641D (which has all the switches to let me configure the entire gamut of bus ratios and core clock speed), a bunch of other embedded chips (PPC, ARM) and a Vaio laptop. I've never had a Mac in my life..

    24. Re:State of the Market by JustNiz · · Score: 1

      Why do you assume I'm not overclocking? I have a core 2 extreme so of course I overclock. Also ny memory is overclocked, both vid. cards are OC versions, I have 2 10k rpm drives, Yet still, I'm dragging only slightly over 500w at the wall under full load (3dmark running a benchmark).
      I've checked this with a power monitor and 2 different multimeters and they all agree.
      Explain that.
      BTW my PSU is a galaxy extreme 1000w.

    25. Re:State of the Market by mxs · · Score: 1

      I do not need to explain it. I am going by the manufacturer's claims of TDP under full load. This may be an upper bound, but if I can supply it, under full load, for all components, I can be sure that the machine won't brown out when I stress it to the extreme on all components. YMMV, and you are free to follow a different approach.

  3. And for real people... by caitsith01 · · Score: 5, Insightful

    ...who lack unlimited funds, the best buy at the moment are the ATi HD 48x0 series cards, which have ridiculously good price/performance and will run any current or near-future game easily at high detail.

    --
    Read Pynchon.
    1. Re:And for real people... by wild_quinine · · Score: 1

      Not to mention the fact that nVdia have so many faulty chipsets out there right now, and have lied time and time again about the scale and scope of the problem - to the degree that I simply won't touch them any more.

    2. Re:And for real people... by javilon · · Score: 1, Insightful

      Until ATI gets their act together with their linux drivers, I'm not buying ATI.

      Also, Nvidia has added MP4 video acceleration to it's linux drivers, so I can see full HD with my old P4@2.4GHz. When we have something similar from ATI I'll reconsider.

      --


      When his defense asked, "Which computer has Jon Johansen trespassed upon?" the answer was: "His own."
    3. Re:And for real people... by Anonymous Coward · · Score: 0

      ...not to mention that GT200 cards are HUGE. None of them would fit my Antec Solo case without case mods, which I'm not willing to do. A GTX260 equivalent ATI HD4870 would fit without a hitch.

    4. Re:And for real people... by MistrBlank · · Score: 1

      ......... ATI is pretty much there too. Might want to read up on the state of Linux drivers.

    5. Re:And for real people... by SlightOverdose · · Score: 1

      I just bought a GTX260 after comparing price/performance ration of it vs it's nearest ATI card. The nVidia card turned out to give more bang for my buck.

      Can't speak for low end cards though.

    6. Re:And for real people... by Anonymous Coward · · Score: 0

      Too bad case makers don't specify the cases that are out of ATX spec, such as the Antec Solo. (Doesn't take full length PCI cards.) At least they should specify "half length PCI" to follow the specification.

      Blame them, not these cards. The GTX295 is still smaller than a Voodoo2 or a Gravis Ultrasound of days past.

  4. Obligatory by Voyager529 · · Score: 0, Redundant

    But can it run Vista?

    1. Re:Obligatory by jo42 · · Score: 2, Funny

      Barely.

    2. Re:Obligatory by forkazoo · · Score: 1

      You just need to write an X86 emulator as a fragment shader. Part of me actually wants to see somebody attempt this, and get a PC OS running on a video card. (Larrabee doesn''t count, using x86 to start with spoils the fun.)

  5. Where are those GTA IV performance metrics by Anonymous Coward · · Score: 0

    Pff, Fallout 3 as benchmark.

    I want to see hardware benchmarks that fits the software description of "Software written for the future machines", or isn't this future enough?

  6. Power usage of a gaming rig by redstar427 · · Score: 1, Troll

    Unless you live way up north or play games only in the winter, dealing with 840 Watts of heat is going to be problematic for a dual GTX295 setup. Summer is worse in that you now have to pump out that heat through the AC system.

    People often will bitch about their cable/DSL bill, but have they ever tried to calculate the monthly cost of electricity their gaming rig racks up alone?

    Some of us don't care. :D

    --
    "Two things are infinite: the universe and human stupidity; and I'm not sure about the universe." Albert Einstein
  7. I just bought one by TheThiefMaster · · Score: 0, Offtopic

    My 8800 GTS died yesterday, wonderful vertical lines down the screen (lines of characters in text mode) and unexpectedly booting in VGA res.
    I looked online for a new card, saw a 285 being sold for cheaper than any 280, and looked it up. I saw that it was basically a 280 v2, so I ordered one. Even at 9:40pm I was offered next day delivery by ebuyer, so I took it. I got the order dispatched email at 10:20pm.

    I didn't realise until a little later that its release date was yesterday! That's some crazy timing.

  8. A false state by unity100 · · Score: 1

    the parent post misses the fact that for dual and more 280 setups, you need a huge case and a kickass power supply. ati 4870s are much easier to fit in a single standard case, and give similar or better performance.

  9. your claims are fishy by Anonymous Coward · · Score: 0

    Preface: I own none of the cards in question, nor any of their relatives, and am not a partisan observer or even a gamer. I keep up with graphics cards to assess their utility for co-processing.

    8600GT usually bests the 3870 for a similar price.

    Shenanigans. I've never seen a review or review site discover this in testing or make such a claim. Perhaps you meant to say "3650" instead of "3870" because that one fits your description. The 3870 is in a higher league entirely. The specs aren't even close; there's no way for [nvidia|ati]-friendliness in an engine to overcome so large a gap:

    Model in question-Gshader/s-Gtex/s-GB/s
    8600GT, (DDR2) -- 37.8 - -- 8.6 -- 12.8
    8600GT (GDDR3) -- 37.8 - -- 8.6 -- 22.4
    9600GT - - - - - 104 - - - 20.8 -- 57.6
    9800GT - - - - - 168 - - - 33.6 -- 57.6
    3850 - - - - - - 213 - - - 10.7 -- 53
    3870 (512MB) - - 248 - - - 12.4 -- 57.6
    3870 (1024MB) -- 248 - - - 12.4 -- 72

    Recently the 9600GT was priced similarly to the faster/better 8600GTs at ~$85.

    What? The slowest 9600GT is faster than the fastest 8600GT. By a lot. It was a huge disappointment upon introduction, falling far short of the performance level expected of a "mid-range" offering when the 8 series was released.

    The 9600GT utterly demolishes both in virtually everything...

    Again, shenanigans.

    The 9600GT 3870 are tightly matched in present games, but the 3870 has 150% more shader power despite the 9600GT's 70% texel advantage and thus will perform much better as shader programs continue to get more complex and play an even larger part in scene rendering.

    [the 9600GT] was within 10-15% of the 9800 GT...

    Not likely. Look at the specs. Give a source; if one supports your claim, it was obviously a memory-bound game (poor engine implementation) rather than texture- or shader-bound, because the 9800GT has at least 50% more of each on an otherwise similar design. Most games don't suffer from engineers botching such a basic element of engine design; memory bandwidth is by far the scarcest resource (i.e. the bottleneck) for your game, always.

    The 3870 supposedly has better hardware than both...

    A lot better. As long as your game/settings want an normal amount of shading power, you use AA/AF, your app can exploit instruction-level parallelism as easily as thread-level parallelism, etc. The relation between the 3870 and the 9600GT is borne out in tons of games and non-games. But, you say:

    ...basically all recent drivers (since at least 8.8) appear to botch the GPU load (with fixed clock speeds mind, no power throttling) so that most games only use 30-50% of the GPU's capacity.

    Appear to? To whom? This would represent an enormous crippling of performance on a very visible product (and after competing favorably with the 9600GT in the game hardware press and elsewhere) and subsequent failure to remedy. There was a ~20% performance hit with the 8.8 release compared to 8.7, but your "30-50% capacity" implies it was nowhere near "capacity" before that. In other words: source? reasoning? definition for "capacity"? [I know the 38x0 overspent silicon on the memory implementation that didn't benefit as much as they thought it would, but you're still being very ambiguous with that term.]

    It's a bad feeling when you get a low price on a decent 3870 GDDR4, only to have to have your older 8600 consistently performing better...

    Buying new stuff that doesn't work better than your old stuff when you thought it would is certainly a bummer. But I just don't buy that an 8600 anything is faster than a 3870, ever, because both the specs and the real-world testing in the press show the latter with such a colossal advantage over the former. The 8600GT was a lacklustre, first-gen 8-series that slipped from its target to a main

    1. Re:your claims are fishy by Zephiris · · Score: 1

      I did say quite a bit that it was more of a driver parity issue than anything else.

      The 9600 is faster than 8600. I meant that the price point of $85 was for the 'higher quality' 8600s. $85 was also a happy price point for decent 8600GTs over the holiday. Pre-overclocked, better fans, other things that would supposedly warrant a few dollars extra, compared to the $45-60 'standard' 8600GTs.

      The 3870 has better specs, obviously. Specs don't mean much if the driver doesn't take advantage. On Nvidia drivers, some driver versions have been better for some hardware than others, newer hardware can take a few versions for both Nvidia and ATI to exploit the hardware better.

      Rivatuner can present the GPU load statistics (as represented by the internal performance counters on the ATI GPU) along with framerate and temeprature via its statistics monitoring, with an overlay similar to FRAPS so you can see it while you play. OpenGL tended to squarely hit into 30% or less range unless blatantly simple. Various games, including GTA4, rarely moved above 50-55%, just so happened to get same or slightly better framerate on the 8600 (usually 2-4 times faster on the 9600). Things that only used basic shading or none at all (including emulators) tended to represent the highest percentage of load on the 3870 (and highest actual performance).

      The games where GPU load was low, also happened to be where atrocious unplayable performance was overwhelmingly likely to occur.

      The 8600GT didn't beat the 3870 in everything. Most obviously, in games that took better advantage of 512MB, the 256MB on the 8600GT limited it. There was a rather distinct trend across a wide variety of games made since 2004 or so (I play around with a lot of things) that the 8600GT would usually get the same framerate, in some cases better. Quality-wise, the 3870 would usually have minimum framerates that sagged lower and usually there'd be a significant lag in games, particularly with vsync+triple buffer enabled (a necessity with LCD).

      Half Life 2 (heck, the 8600GT got ~180 frames on Ep2 with max settings, including AA 4x, 3870 got ~40), GTA 4, Far Cry 2, Crysis, Crysis Warhead, Brothers in Arms: Hell's Highway, Fallout 3, San Andreas, Mercenaries 2, No One Lives Forever, Oblivion, UT2004, World of Warcraft, Tabula Rasa, among others.

      People say 'specs specs specs', but considering a single driver revision can change performance for a particular game for 15% or more, drivers matter a great deal. You could have a really fast car with high horsepower and a novice driver...who crashes it, or has to drive slower to keep it on the road, compared to a Ford Focus, which they drive more easily and confidently, and rack up a higher average speed.

      The press tends to use static demos and benchmarks. A benchmark in a 'real game' doesn't much matter if it uses a prescripted static environment which doesn't change and is specifically designed to provide higher framerates than is at all representative of the actual game proper. Very few sites who publish reviews do so in actual Real World game conditions, where they're actually playing in a meaningful circuit and record detailed framerate statistics per second to file. The 3870 scores pretty well on nearly every benchmark, including OpenGL ones. It's only a 'few hundred' points shy of the 9600GT replacement on 3DMark, despite there being no similarity in any game I've yet found. Lies, damned lies, and benchmarks.

      The performance counters only specify what percentage of GPU resource is being used, not how exactly they're defined, and don't relate how different drivers affect it. Hardware counts, how well the drivers can actually spread the load around would generally make the load percentage (and performance) go up. Presumably, the counters count based on how many shaders/render/texture pipelines are being utilized, or some subset thereof, though certainly shaders are being counted.

      If you're getting 15 laggy frames per second on a game that should be pushing as hard as the graphic

      --

      "A Goddess rarely smiles for she is forced by others to be an island unto herself." - Zephiris
  10. A review that's fun to read by Anonymous Coward · · Score: 0

    For a review that has all the game benchmarks AND is also fun to read, with insightful analysis and clever commentary, check out The Tech Report's comparo of the GTX 285 to SIXTEEN competitor cards and configs. Why settle for just benchmark results and technical jargon when there's humor in the pixels?

  11. lolZ by SM177Y · · Score: 1

    sry but those benchmarks are not even close to right. since when does the 280 score 2000 points more than the 4870 1 gig. sry not a chance. when you guys post benchmarks, at least post correct ones. and 1920x1200 at 4x aa and 16x af is not a valid benchmark because those are not 3d marks default (and online comparable for that matter) settings. this is not a nvidia fan boy cry, i run nvidia myself lol but i know what the 4870's can really do.

    1. Re:lolZ by db10 · · Score: 1

      cry smo nub

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

      Aiming for the Pulitzer, huh?