Slashdot Mirror


AMD Radeon HD 7970 Launched, Fastest GPU Tested

MojoKid writes "Rumors of AMD's Southern Island family of graphics processors have circulated for some time, though today AMD is officially announcing their latest flagship single-GPU graphics card, the Radeon HD 7970. AMD's new Tahiti GPU is outfitted with 2,048 stream processors with a 925MHz engine clock, featuring AMD's Graphics Core Next architecture, paired to 3GB of GDDR5 memory connected over a 384-bit wide memory bus. And yes, it's crazy fast as you'd expect and supports DX11.1 rendering. In the benchmarks, the new Radeon HD 7970 bests NVIDIA's fastest single GPU GeForce GTX 580 card by a comfortable margin of 15 — 20 percent and can even approach some dual GPU configurations in certain tests." PC Perspective has a similarly positive writeup. There are people who will pay $549 for a video card, and others who are just glad that the technology drags along the low-end offerings, too.

281 comments

  1. This would be really cool... by Anonymous Coward · · Score: 2, Insightful

    ...if most PC games weren't just shitty console ports these days. If you spend over $150 on a graphics card you're an idiot.

    1. Re:This would be really cool... by Hatta · · Score: 5, Funny

      Hush. Those idiots finance the advance of technology.

      --
      Give me Classic Slashdot or give me death!
    2. Re:This would be really cool... by AngryDeuce · · Score: 1

      Hopefully the prices on the 5000 and 6000 series start dropping after Christmas. My 4670's are starting to show their age...

    3. Re:This would be really cool... by Anonymous Coward · · Score: 1

      You'll need a bit more than your average 150 dollar card to max out teh pretty on most "consolized"

      I'll admit, though, that your average GPU nowadays has a much longer life than they used to.

    4. Re:This would be really cool... by Lumpy · · Score: 5, Insightful

      Says the idiot that only uses a PC for gaming.

      Adobe After Effects will use the GPU for rendering and image processing.

      --
      Do not look at laser with remaining good eye.
    5. Re:This would be really cool... by jellomizer · · Score: 2

      I think it is an issue that most game graphics have reached a peak with the current rendering technology. Where you need exponential work (In man power) to get a linear improvement.

      Black and White text... All find and good until we need a graph.
      Back and White graphics... Now only if it could do color.
      CGA... What bad colors.
      EGA... Looking a lot better if only we could get some shading and skin tones.
      VGA... Enough visible colors to make realistic pictures. But a higher resolution will make it better.
      SVGA... (The first good peak) OK static images are looking good, we can watch movies, but game animation is limited and 3d is getting popular.
      3d cards. if we could get more polygons per second... More textures... Alpha Channels... Smoothing...
      That is where we are now. Right now we have reached a limitation where we can display what we want to display. However the next steps will require new rendering methods that make graphics easier to create.

      --
      If something is so important that you feel the need to post it on the internet... It probably isn't that important.
    6. Re:This would be really cool... by LordLimecat · · Score: 2

      Arent you way better off with a workstation card for most workstation loads? From what Ive read, a GTX or ATI HD makes for a poor CAD or Adobe machine.

    7. Re:This would be really cool... by Anonymous Coward · · Score: 0

      You'll need a bit more than your average 150 dollar card to max out teh pretty on most "consolized"

      Uh... nope. I seem to be doing just fine with a video card under $100 (or, at least, it can max out Skyrim and various other games).

    8. Re:This would be really cool... by Anonymous Coward · · Score: 1

      Some of us use graphics cards for non-gaming purposes. For example, I develop CAD-CAM modelling and my programs rely on these math cop-processors which we call graphics cards to do thermal/structural analysis. Being able to rely on these 250€ worth of kit to do the job that it would take about 500€ of AMD processors and 2500€ of Intel processors, and all this wasting only a fraction of the energy and occupying a fraction of the volume, is an excellent good thing to have.

    9. Re:This would be really cool... by Endo13 · · Score: 2

      Not true. You just have to find the sweet spot of performance/$. My current card (I think it's a 6870 but I'd have to double-check to be sure) cost less than $150 a couple months ago and runs Witcher 2 quite smoothly with high settings. Haven't tried BF3.

      --
      There is no -1 Disagree mod. Slashdot.org/faq defines mod options. USE IT.
    10. Re:This would be really cool... by Lumpy · · Score: 4, Interesting

      Nope. Bang for buck this new card kicks the butt hard of the Workstation cards.

      --
      Do not look at laser with remaining good eye.
    11. Re:This would be really cool... by Baloroth · · Score: 1

      I was actually hoping they come out with a 7850 or similar soon. My 5770 is still pretty strong, but I wouldn't mind an upgrade soon, and I like to have all the newest features (I can do without the top speed).

      --
      "None can love freedom heartily, but good men; the rest love not freedom, but license." --John Milton
    12. Re:This would be really cool... by Anonymous Coward · · Score: 1

      What I hate are the console fanbois who ruin the gaming experience for everyone by refusing to play on anything except consoles, thus leading to most big games being developed for severely inferior systems.

      (P.S.: This post is a troll. The real PC gamers should know what I mean.)

    13. Re:This would be really cool... by Anonymous Coward · · Score: 1

      You're an idiot! I could/would/should use this GPU for "GPGPU" rendering using MATLAB. And then I would *still* benefit from using two or three of them in the same setup. Just because you're a gamer doesn't mean that everyone else is!

    14. Re:This would be really cool... by P-niiice · · Score: 2

      I'm a console gamer, but I would prefer more balance; it's too console-leaning right now and we haven't gotten any real advancement in gaming for a long time. Consoles keep the developers and publishers afloat, and great PC games would temper devs and force them to add more depth to games....although that didn't help Oblivion and Skyrim (I love those games, but they lost some of the nerd-appeal of Morrowind).

    15. Re:This would be really cool... by durrr · · Score: 2

      It's the fastest GPU in the known universe! surely it have to be worth something!

    16. Re:This would be really cool... by theantipop · · Score: 1

      You've been able to get a 6850 for darn near $100 for some time now. It's a pretty good deal unless you only buy high-end GPUs, which honestly seems like a waste anymore. My 4 year old 8800GT is just now starting to feel inadequate. I'm looking for the second coming of the 8800GT to emerge from this generation so I can hold on to it for another 4 years.

    17. Re:This would be really cool... by billcopc · · Score: 0

      Hey now! The last guy who called me an idiot got shit on by a transgendered midget.

      Depends on your resolution, but yes. Those of us with beefy GPU setups tend to be doing 3D, multi-display, or bitchy resolutions like 2560x1440. Or in my case, CUDA processing and 3D raytracing. A lot of people forget that GPUs can do a lot more than just games. For some of us, those non-gaming uses are our day jobs.

      --
      -Billco, Fnarg.com
    18. Re:This would be really cool... by Anonymous Coward · · Score: 0

      You should play Skyrim at 2560 x 1600 and ultra settings. Definately not a "console port".

    19. Re:This would be really cool... by billcopc · · Score: 4, Informative

      Depends on the type of processing. GTX and Radeon cards artificially limit their double-precision performance to 1/4 of their capabilities, to protect the high-margin workstation SKUs. If all you're doing is single-precision math, you're fine with a gaming card.

      --
      -Billco, Fnarg.com
    20. Re:This would be really cool... by Anonymous Coward · · Score: 0

      Make it so I can lend/swap my PC games again and I don't need a separate copy for my daughter (even if there will never be any co-op play) and we'll talk.

      Yeah, Steam sales are cheap, swapping XBox 360 discs with my friends is cheaper. I spend a shit-ton on games, but graphics aren't the end all and be all. I'm typing this on a machine with Crossfired XFX 6870s, so it's not like I don't like PC gaming, but PC gaming sucks purely due to PC gaming sucking, not "consoleized" titles (seriously, if you think Deus Ex HR is any more fun on PC than console this argument is not only probably lost on you, you're probably a moron).

      If you want to keep selling me PC games for 3 bucks in Steam and GOG sales, well then go ahead, but that's going to be most of my purchasing.

      Incidentally, many of the games that have made the PC great don't actually use a lot of horsepower (and didn't even when they launched). Go take a look at the GOG library if you don't believe me. Also think about Minecraft, Terraria, etc. in recent years.

    21. Re:This would be really cool... by dasherjan · · Score: 1

      Hey. I need all the power I can get for spider solitaire!

    22. Re:This would be really cool... by Anonymous Coward · · Score: 0

      Stuck with the 8800GT for a long time, but it falls short on newer games that are really shader heavy, unfortunately.

      Went to a radeon 6870. Amazing value.

    23. Re:This would be really cool... by Luckyo · · Score: 1

      Witcher 2 is a bad example anyway, because with ubersampling it will rape pretty much any modern graphics card when everything else is on ultra. You'll need SLI or the card above for that workload.

      That said, most modern 200ish cards will handle B3 high and W2 high just fine at sub-1080p. It's the best possible settings and very high resolutions where you need the high end offerings.

    24. Re:This would be really cool... by Luckyo · · Score: 1

      Still is. Granted I can't push the resolution beyond 1080p, but with everything maxed it, the fact that it's a console port is very visible. Because at that point the screen size is big enough that you start seeing the textures for what they really are - dirty smudges. You also start seeing them cheating on geometry in comparison to more PC-optimized offerings like BF3.

    25. Re:This would be really cool... by Bengie · · Score: 1

      For now anyway. MS is looking at Double Precision becoming the standard for some future DirectX. That's probably still a few years off.

    26. Re:This would be really cool... by Anonymous Coward · · Score: 0

      Might be you've got a bottleneck elsewhere. I have a Radeon HD 5870 in my tower, but it can't run L4D at 60FPS consistently because my CPU is a low-end Core2Duo OCed to 2.7GHz (from 1.9GHz base) running Windows XP Pro (so only 4GB of RAM). I game on my laptop now even though it's got a GTX 460M because the Core i7-2630QM, Windows 7, and 8GB of RAM make a huge difference.

      Still waiting for the price to be right to build a new tower. Mostly, I'm just waiting for Ivy Bridge before I sink a few hundred dollars on a CPU and mobo, and looking for sales on a good mid to full size case.

    27. Re:This would be really cool... by Bengie · · Score: 4, Informative

      Ray Tracing!!!

      We're also capped right now because of too many single-threaded game engines. A given thread can only push so many objects to the GPU at a time. Civ5 and BF3, being the first games to make use of deferred shading and other DX11 multi-threading abilities, can have lots of objects on the screen with decent FPS.

      The biggest issue I have with nearly all of my games is my ATI6950 is at 20%-60% load and only getting sub 60fps, while my CPU has one core pegged. My FPS isn't CPU limited, it's thread limited.

    28. Re:This would be really cool... by Anonymous Coward · · Score: 0

      Skyrim uses the Gamebryo engine, it's nearing 10 years of age. The default textures are 1024x1024 resolution or less. Replace those with hi-res textures from skyrimnexus.com and see how your computer runs it.

      Also, what resolution are you running the game in? That makes a huge difference. The specs to run 720p are drastically lower than 1080p (having more than twice the pixels of 720p), which in turn is entirely incomparable to 2160p (having 4 times the pixels of 1080p).

    29. Re:This would be really cool... by Mitchell314 · · Score: 1

      I'm still working with integer* based rendering engines, you insensitive clod!


      *16 integers. None of this new fangled 32 bit garbage kids play with these days.

      --
      I read TFA and all I got was this lousy cookie
    30. Re:This would be really cool... by bhcompy · · Score: 1

      This is a pretty good deal for a real good card(5870). Faster than the 6870, but more of an energy hog.

    31. Re:This would be really cool... by Lumpy · · Score: 1

      And even that can be fixed. the limit is in the firmware. I have a PC ATI card in my PPC mac that is running a workstation firmware that unlocked some serious processing power. This was back when ATI video cards for the quad core G5 were anal rape robbery pricing and the exact same hardware for the PC was going for $199.00 That system utterly screamed running shake and after effects back in 2007-2008

      --
      Do not look at laser with remaining good eye.
    32. Re:This would be really cool... by anonymov · · Score: 2

      This small step up for texture dimensions won't give much impact on modern hardware, and from a bit of testing Skyrim seems to be mostly CPU/VS limited, not pixel shading limited, anyways.

    33. Re:This would be really cool... by Anonymous Coward · · Score: 0

      Really? I run Skyrim at 2560x1600 with a GTX 590 and I find that when I turn up all the anti-aliasing and anisotropic filtering it runs slower than when they're turned off. So how do you manage to max out all of the settings?

    34. Re:This would be really cool... by Anonymous Coward · · Score: 0

      I have a few rules. I don't build a new system until 3 years have passed. Then I don't spend more than $100 on any single component. Reuse what you can. (monitor, case, PSU, speakers).

      In this way, I am able to play any current game at decent settings for the same price as a console gamer. Especially once you take into account all the red rings of death etc. that require you to toss the whole system.

      My 2 year old PC may not be able to play Crysis 2 at maximum, but it's still well above the minimum required. Cost me about 300 dollars.

    35. Re:This would be really cool... by Anonymous Coward · · Score: 0

      65535 should be enough for anyone.

    36. Re:This would be really cool... by Anonymous Coward · · Score: 0

      Why would you use "anymore" to mean "these days"?

    37. Re:This would be really cool... by Anonymous Coward · · Score: 0

      I have an ATI 6990, and while BF3 on highest settings won't slow anything down, large Civ5 games will slow down to a slideshow if there's a lot going on with maxed out settings.

    38. Re:This would be really cool... by hairyfeet · · Score: 1

      Actually if you look around you can get an HD4850 for around $60 these days and frankly its overkill for a good 75%+ of the games out there.

      Personally i'm kinda glad the consoles have long tails these days as some of us don't care for spending a $150+ every year just to be able to play the latest games. That's the way it was in the late 90s/ early 00s and frankly it seemed like just as I got my machine the way I liked it out would come games requiring a new card to run and there we go again. I've had my HD4850 for a couple of years now and frankly the graphics are plenty jaw dropping as it is. If I want to impress someone all I have to do is fire up Just Cause II and set some remote charges on the stacks and let them go. Seeing my character do the "Cool guys don't watch explosions" bit while the world crumbles behind him with all the smoke and particle effects is frankly pretty impressive.

      But don't worry there will ALWAYS be those 'must win teh benches!" types buying crazy priced cards and there will always be new features they can push us like 3D and Eyefinity. In the next year or so the new consoles will come out and I'll grab a 6850 when the price drops below $100 and I'll be set for another 3 or 4 years, hell of a lot nicer than sinking a couple of hundred a year on my GPU. With 2 boys that also game buying three cards at $150+ a piece gets pretty expensive so I'm happy with the way things are. Go AMD and keep cranking out good sub $100 cards please!

      --
      ACs don't waste your time replying, your posts are never seen by me.
    39. Re:This would be really cool... by Anonymous Coward · · Score: 0

      What are you talking about more like 250. You need at least a 6950 to max games out these days.

    40. Re:This would be really cool... by hairyfeet · · Score: 1

      Not really, I've set several Radeon cards up for guys using CAD and programs like Solidworks and frankly they work just fine. it all comes down to making sure the GPU supports the same OpenGL that the application supports since all those CAD and engineering programs use OpenGL, but as long as that matches up you're good to go.

      BTW if anyone is wondering SolidWorks seems to play nice with any Radeon HD46xx or better and flies on the HD4850 which can be had dirt cheap. Really takes a load of the CPU when you have a lot of little pieces in an assembly to have the rendering done by the GPU.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    41. Re:This would be really cool... by Anonymous Coward · · Score: 0

      "Workstation" card translates to: We're going to sell you the same silicon, usually clocked a little lower for better reliability and reduced power, with different drivers.

      There are two basic categories of "pro" apps which tend to benefit from this (that I'm aware of):

      1. GPGPU apps which use double precision FP (I know at least NVidia artificially restricts DP FP performance on the "consumer" versions of their cards, so you have to get the "workstation" cards for full performance)

      2. CAD apps which use legacy OpenGL APIs that are on the way out (stippled lines, etc.). These APIs are never the fast path in games, so mainstream game-oriented drivers do not provide good performance. I believe most CAD apps are slowly moving to modernized, mainstream APIs like DX10+ or OpenGL ES 2.0 +, so once that migration is complete, this will no longer be a real advantage for a workstation card.

      It comes down to the software you run. Some things will benefit greatly from a workstation card, others not at all. It's all a game to get deeper pocket customers to pay more. (Not without legitimate reason, mind you. Pro customers cost a lot to support and generate few sales compared to the wider market.)

    42. Re:This would be really cool... by Anonymous Coward · · Score: 0

      unfortunately Adobe pulled CUDA card (pun intended) only NVIDIA can FULLY accelerate Photoshop, unfortunately, since that was only reason i did not buy faster and cheaper AMD card

    43. Re:This would be really cool... by Anonymous Coward · · Score: 0

      huh i have PC AND a console, i pirate on both, i payed only one game in life (World of Warcraft), so if they start making games for console pirates will just move there m8

    44. Re:This would be really cool... by hairyfeet · · Score: 1

      Is that USD? Because I just looked in both Google and yahoo shopping and all the HD6850s I'm seeing are nearly $200. Personally I'm hoping for another steal like the HD4850s which i got at $60 a piece over a year ago, but I'd happily pay $100 for a 6850 if i could find one but that doesn't look to be the case.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    45. Re:This would be really cool... by hairyfeet · · Score: 1

      Is that so? Frankly I haven't pirated in years and neither has anyone I know, since the triple score of Amazon+GOG+Steam makes games cheaper than the bandwidth and hassle of pirating anymore. Hell I've even been buying up old games i pirated back in the day simply because its less hassle to have them all loaded in Steam.

      I mean when you can buy games like Just Cause II with ALL the DLC for $7 on Steam, Max Payne I & II for $3, and I got all the HL:2 series for $5? Why would you bother pirating, dealing with cracks, worrying about possible malware, losing MP, what's the point?

      Frankly the only piracy I see (other than the numbers publishers pull out their ass to ask for more draconian policies) is the kids that are too broke to buy squat and those that are making a statement like what happened with Spore and Assassins Creed over their frankly insane DRM. Everybody else just seems to be hitting the big 3 I just listed and have more games than they know what to do with for less than a meal for 4 at the Mickey D's.

      Oh just FYI the biggest pirates i know have hacked X360s they got off CL. they have the hacked one loaded with pirated games and keep their real X360 for the few they want to play online. Considering how badly the consoles get screwed on price compared to PC gamers I'm really not surprised, you aren't buying 3 or 4 good games for $5 for a console anywhere but a yard sale.

      --
      ACs don't waste your time replying, your posts are never seen by me.
    46. Re:This would be really cool... by tyrione · · Score: 1

      ...if most PC games weren't just shitty console ports these days. If you spend over $150 on a graphics card you're an idiot.

      This idiot needs those parallel units for OpenCL/OpenGL and it's many applications with Mechanical Engineering.

    47. Re:This would be really cool... by AmiMoJo · · Score: 1

      There are two differences between consumer and workstation cards. Firstly the driver artificially limits the double float computational performance since all games use single point and only CAD type stuff uses double. You can fix that with a driver hack.

      The second difference is that workstation cards are guaranteed to give correct output. They actually test that they render the correct colours, that the analogue output is right and so on. With games it doesn't matter if there are slight inaccuracies here and there, but for cad and image editing on a calibrated monitor it might. I say might because the reality is that for most DTP and CAD stuff it probably won't, so a hacked driver and consumer grade card is more than adequate.

      --
      const int one = 65536; (Silvermoon, Texture.cs)
      SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
    48. Re:This would be really cool... by Bengie · · Score: 1

      Unfortunately ATI/AMD has not implemented the required DX11 methods in their drivers for multi-threading. The Civ5 devs had a screenshot of a nVidia card with proper drivers, and a 12 core AMD cpu running about 80% load on each core. My ATI/AMD drivers only use about 4 of my cores, yet alone 12 cores.

      From what I've read, there really aren't many games that support DX11 threading(I think just those 2), so neither ATI nor nVidia had much of anything to test with until very recently. Even nVidia had a hard time working the Civ5 team to get threading to work.

  2. Only once have I splurged like that by halivar · · Score: 1

    I rebuild my machines every two years. My previous rig couldn't do Crysis as max settings so my latest system has dual 5870's that I got for $400 a piece. I'll never splurge like that on video cards again. Then again, 2 years later, I still max out the sliders on every game I get. It's great to have that kind of computing power... but maybe I should have waited 6 months? Those cards are going for $150 today.

    1. Re:Only once have I splurged like that by Moheeheeko · · Score: 1

      Pretty sure most systems cant run Crysis perfect at max settings, simply for the fact that Crysis is one of the worst optimized games ever developed.

    2. Re:Only once have I splurged like that by atrain728 · · Score: 1

      Any new machine with a graphics card can. It's an old game at this point.

    3. Re:Only once have I splurged like that by billcopc · · Score: 3, Interesting

      Pretty sure today's mid-range PCs trounce 2007's high-end with ease.

      Just for shits, when I got my current rig just a couple years ago, I played through Crysis again. On a single GTX260, it was butter smooth at 1680x1050. When I switched to quad-SLI 295's, it was butter-smooth in triple-wide surround.

      People who continue to claim Crysis is an unoptimized mess are:

      - not programmers
      - not owners of high-end hardware

      Could it be improved ? Sure. Is it the worst optimized game of the 21st century ? FUCK NO, not even close, and subsequent patches greatly improved the situation.

      --
      -Billco, Fnarg.com
    4. Re:Only once have I splurged like that by Luckyo · · Score: 1

      In perfect honesty, it's better to buy a single powerful card (to avoid early problems in games) for 200-250 range, and upgrade every couple of years. Cheaper and you should be able to max or near max all games that come during lifetime of the card.

      Obvious exceptions are the extreme resolutions, 3D vision and multi-monitor gameplay.

    5. Re:Only once have I splurged like that by Anonymous Coward · · Score: 0

      "Is it the worst optimized game of the 21st century ? FUCK NO, not even close..."

      I'm pretty sure Minecraft takes that honor.

    6. Re:Only once have I splurged like that by Talderas · · Score: 1

      Dwarf Fortress is pretty ugly about memory usage.

      --
      "Lack of speed can be overcome. In the worst case by patience." --Znork
    7. Re:Only once have I splurged like that by Anonymous Coward · · Score: 0

      Pretty sure today's mid-range PCs trounce 2007's high-end with ease.

      Maybe. I built my current system at the end of 2007. Q9550 running at 3.4 Ghz, 8 GB RAM, 8800 GTS 512, bunch of WD Black drives in RAID, etc. I have played around with more modern systems and can't tell a difference. My video card is aging but almost keeps up with current mid-range systems.

      My RAID array is as fast or faster than any mid-range system you can build today. 2+TB pushing about 600MB/sec read and 500MB/sec write and that's without SSD.

      Really the main reason I would upgrade now is because of the newer virtualization features that allow sharing PCI devices (VT-d or whatever it's called). Plus RAM is really cheap, I'd go for at least 16GB if not 32GB. I would keep my RAID array though, especially with drives prices like they are, plus you still can't get realiable drives any bigger than the ones I am using (640GB each).

    8. Re:Only once have I splurged like that by arun84h · · Score: 1

      The PC version of GTA IV.

      *shudders*

    9. Re:Only once have I splurged like that by Anonymous Coward · · Score: 0

      [citation needed]

      Please name some of the ones that are worse. Don't just make claims.

    10. Re:Only once have I splurged like that by Anonymous Coward · · Score: 0

      While I will agree that Dwarf Fortress uses a lot of memory, if you actually looked at how much it's doing, and consider the fact that the game is still SINGLE THREADED, you might find yourself more impressed with what it's capable of. Flow patterns for water, dynamic animal and dwarf movement, open space checking to determine if there's the possibility of a cave-in, etc.

      DF seems like a lot of unnecessary usage if you consider it based on graphics, but not if you consider it based on processing/datastructures. In that respect it's up there probably above most games out there other than maybe minecraft, and the few other world simulation games that exist.

    11. Re:Only once have I splurged like that by rmdyer · · Score: 1

      The original Crysis game (CryEngine2), combined with mods like Real Lifesys (or other extreme tweak mods aka Photoreal), combined with HD textures, and all viewed on 1920x1200 or above res display devices (especially combined with multi-monitor setups), will make your jaws drop.

      You really need pure GPU and CPU power to push this stuff over 50 FPS. Whoever tells you otherwise simply hasn't done it.
      See...

      http://www.youtube.com/watch?v=PivoSi2VvqA

      http://justsitback.deviantart.com/?title=Videogame%20Environments%20Realtime%203D&rssQuery=gallery:MadMaximus83/25304347

      http://www.facepunch.com/threads/714112-Photo-Realistic-Crysis-mod-AKA-Real-Life-sis-Orgasmic-Crysis

      http://www.youtube.com/watch?v=RPCJh8tYri0

      Granted, Crysis has it's problems, but it is far and away better graphics-wise than anything console, or console ported to PC, and it's almost 5 years old now. Console ported games by comparison look a bit "cartoon'ish", even the strange changes they made in the CryEngine 3 to port Crysis to consoles (reduced res and color textures, etc). I'm one of the guys who purposely didn't buy Crysis 2 because I didn't think the graphics were any better than the original Crysis. I've also skipped Skyrim, for the graphics (poor), but also because slaying dragons (and every other creature in the woods that wants to do you in) all day does get old. Why can't Skyrim graphics look as good as this? ...

      http://www.youtube.com/watch?v=M7I6EBc4mRc
      or,
      http://www.youtube.com/watch?v=kTuAw_i7ngI

      I'll tell you why. It's because Skyrim was made for the CONSOLES that came out 5 years ago!

      Personally I am a high-end enthusiast. I do run SLI and 1920x1200. I do have an overclocked 4.1 GHz machine. I do upgrade my graphics about every 2 to 3 years. I use my money to push the industry along. I honestly believe that if people quit buying high-end, then the state of the art in GPU/CPU will slow down dramatically and we will never reach pure photorealism as soon. Console designs will also suffer because of it. For example, the nex-gen consoles had better be coming with DX11 capable GPUs and 4+ Gig mem, or for this day in age they would fair poorly. I am the reason that this new AMD graphics card even exists.

      I'm the type of person who would buy this...

      http://proavmagazine.com/projectors/high-resolution-projector-digital-projection-dvis.aspx

      http://www.digitalprojection.com/BrowseProjectors/SeriesList/ProjectorList/ProjectorDetail/tabid/87/ProjectorId/170/MarketTypeId/10/Default.aspx

      ... just to game on cold nights in the winter, so I can hang out on a Crysis style beach projected on my living room wall.

      Sure, arguments can be made that the "game" is more important than the graphics. That is not entirely true, but it is why I still love a game of the now 11 year old Quake 3 Arena multi-player on occasion. But what I have been craving all my life is pure immersion. I'm an adult, and as such I'm looking not so much for a

    12. Re:Only once have I splurged like that by Anonymous Coward · · Score: 0

      Optimized.. that word doesn't mean what you think it means, if you're throwing 4 years worth of new technology to get a machine capable of playing max settings Crysis.

      Or, well, it does.. But you're optimizing your machine to play Crysis. Crysis is not optimized by adding better hardware.

    13. Re:Only once have I splurged like that by loosescrews · · Score: 1

      Have you tried the PC version of L.A. Noire? It makes GTA IV good like pretty good port. Some highlights are that it is capped at 30fps, requires admin rights to even launch, and crashes instead of exiting. And that is just the beginning...

    14. Re:Only once have I splurged like that by billcopc · · Score: 1

      Dear AC: go fuck yourself.

      Optimized, in this usage, means the game engine scales with upgraded hardware. There are numerous games that simply cannot use hardware efficiently, where throwing it 4 times more GPU power results in perhaps a 50% increase in rendering speed. There are even more games that chug as soon as you exceed 1280x1024 or some other token resolution. Crysis is not one of those offenders, because it can take advantage of those extra GPUs and deliver jaw-dropping graphics on displays that exceed its original design specs fourfold. The more power you give it, the more it impresses you with its capabilities. This, to me, as a programmer, says the engine is very well designed and efficiently coded. This ain't no goddamned Gamebryo shit show.

      Take any other game from 2007, run it on dream machine, and compare it with Crysis. Even the fanfared Bioshock and Stalker show very cruel signs of aging while Crysis could very easily pass for a current-gen title, while delivering higher framerates and greater photorealism than just about any 2011 release. Yes, even that mediocre CoD knockoff they call Battlefield 3.

      --
      -Billco, Fnarg.com
  3. Overpowerful. by unity100 · · Score: 4, Interesting

    Due to console gaming retarding pcs.

    im on single radeon 6950 (unlocked to 6970 by bios flash), and i am doing 5040x1050 res (3 monitor eyefinity) on swtor (the old republic), all settings full, and with 30-40 fps on average, and 25 fps+ on coruscant (coruscant is waaaaaaay too big).

    same for skyrim. i even have extra graphics mods on skyrim, fxaa injector etc (injected bloom into game) this that.

    so, top gpu of the existing generation (before any idiot jumps in to talk about 6990 being the top offering from ati ill let you know that 6990 is 2 6970s in crossfire, and 6950 gpu is just 6970 gpu with 38 or so shaders locked down via bios and underclocked - ALL are the same chip), is not only able to play the newest graphics-heavy games in max settings BUT also do it on 3 monitor eyefinity resolution.

    one word. consoles. optional word : retarding.

    1. Re:Overpowerful. by parlancex · · Score: 4, Insightful

      ... 30-40 fps on average, and 25 fps+ on coruscant (coruscant is waaaaaaay too big). same for skyrim...

      Looks like PCs isn't the only thing gaming consoles have been retarding. Most PC gamers would have considered 25 fps nearly unplayable, and 30-40 FPS highly undesirable before the proliferation of poor frame rates in modern console games. There are still many of us that are unsatisfied with that level of performance, but are unwilling to compromise graphics quality.

    2. Re:Overpowerful. by LordLimecat · · Score: 1

      Consoles support 5040x1050? Color me suprised.

    3. Re:Overpowerful. by Warma · · Score: 2

      This is highly a matter of preference. I feel that 25fps is just flat out unplayable and anything under 60 is distracting and annoying. I believe that most gamers would agree. I always cut detail and effects to get 60fps, even if this means the game will look like shit. Framerate is life.

      So no, based on your description, the top of the previous generation is NOT able to play those games in the environment you defined. You would need around twice the GPU power for that. The benchmarks suggest that 7970 won't cut it either.

    4. Re:Overpowerful. by Anonymous Coward · · Score: 0

      ...and with 30-40 fps on average, and 25 fps+...

      Haha... HAHAHAHA.

    5. Re:Overpowerful. by rodrigoandrade · · Score: 2

      Yes, it's really a shame that 30 fps became an acceptable framerate for games nowadays, thanks to crappy underpowered consoles.

      Funny, however, is that back in 1999 (Dreamcast days) any console game that didn't run at a solid 60 fps was considered a potential flop.

      This framerate crap is one of the many reasons I'll never go back to console gaming.

      Times change, no?

    6. Re:Overpowerful. by Jeremi · · Score: 1

      This is highly a matter of preference. I feel that 25fps is just flat out unplayable and anything under 60 is distracting and annoying.

      It's partially a matter of what you are accustomed to. I remember playing Arcticfox on a 386/EGA at about 4FPS and thinking it was awesome, because compared to the other games available at the time it was.

      --


      I don't care if it's 90,000 hectares. That lake was not my doing.
    7. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Agreed. It's ridiculous to state that current PC hardware is overpowered for games if you can't reach 60fps. Nintendo got that right with Metroid Prime and Mario Galaxy. Plus iPhone is definitely 60 fps. That's why the rest suck so bad.

    8. Re:Overpowerful. by unity100 · · Score: 0

      Wrong. I'll agree that over 60 doesn't really matter but as a former game developer I can tell you there is a HUGE difference between 30 and 60.

      yes. that should be why hdmi standard requires 24 fps as absolute minimum. because a game developer somewhere, and some percentage of gamers who equate the fps counter or the hardware in their machine with their egos, think they perceive differences in between 30 and 60 fps.

      no, you dont. unless you genetically modified yourself. any argument otherwise is bullshit, until proper research is referenced.

    9. Re:Overpowerful. by Warma · · Score: 2, Insightful

      No need to use strong language and unbacked opinions. You are simply incorrect.

      Put two FPS players of similar skill in front of a computer. Configure the computers so that the other shows 30fps and the other shows 60fps (easy to do in Quake and most other shooters). No matter what you may hope for the truth to be, the player with the 60fps display will have an enormous advantage.

      It is extremely easy to test this yourself, just go play a game and record your performance by some metric while alternating the frame rates. I am willing to test it with you, if you wish.

    10. Re:Overpowerful. by anonymov · · Score: 1

      Not really.

      First, different people have different perception, so 20 may be enough for some and 60 just right for others.

      Second, fast motion on low framerate requires some amount of motion blur to be perceived as "smooth". When shooting a film, this blur is already there thanks to the nature of filming. When rendering, developers have to care about it, and as it can be costly it's often dropped.

    11. Re:Overpowerful. by Anonymous Coward · · Score: 0

      As I always said: If you can win the game at below 30 FPS, the game is shit!

      I remember the the two most important changes that improved my skill:
      1. Play 30 minutes Quake 3 pro mode (CPMA) before the actual game. = Level 10,000 on the burst-top twitch scale
      2. Make sure you play at 60 fps. Not 30. Period.
      (Having a good cable mouse of the right type [laser vs optical] deserves third place, and having a screen with no lag is a fourth, although I still play on CRTs because they offer free resolution scaling and have neither lag nor smearing, so I wouldnâ(TM)t know.)

    12. Re:Overpowerful. by unity100 · · Score: 0

      No need to use strong language and unbacked opinions. You are simply incorrect.

      UNbacked opinions ?

      link me ONE research that shows humans are able to perceive differences in between 30 fps and 60 fps. ONE research. except from a percentage of 'fps gamers' thinking that they can perceive 30 and 60 fps.

      and then move on to explain why hdmi standard mandates 24 fps, and this has been the minimum requirement for 'smooth' since aeons in display related 'anything'.

      no need to use strong language ? why yes there is. there is a percentage of people who think that their subjective 'feeling' precedes SCIENCE.

      im waiting for the proper research to show that people are able to perceive 30 vs 60 fps, AND rebuke the hdmi specification.

    13. Re:Overpowerful. by unity100 · · Score: 1

      then bring here ONE research that 'people with different perception' can perceive difference in between 30 vs 60 fps. one is enough.

    14. Re:Overpowerful. by Charliemopps · · Score: 1

      There is a difference, a very obvious difference. But your television most likely runs at 30fps, and your monitor at 60, if you're lucky. Get an old CRT capable of 200hrz, run some old 3D game like decent that a modern card would make easy work of so you can hit 200fps and marvel at the difference. It really does make a huge difference.

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

      I think you should check your facts on this one. You are wrong.

    16. Re:Overpowerful. by Anonymous Coward · · Score: 0

      thats why hdmi standard used to mandate 24 fps.

      Movies are about 24 fps. But the frames in movies are SMOOTHED OUT. Each frame isn't a perfectly clear and sharp image of a single instant like with computer games.

    17. Re:Overpowerful. by Junta · · Score: 1

      His point being that game developers are conservative about pushing graphical complexity such that they don't even produce a workload that remotely challenges modern top-end cards. He attributes this to developers targeting weaker consoles. I think it's just because they want to perhaps have a slightly larger market than those willing to shell out $600 a year in graphics cards *alone*, regardless of game consoles. Now to push framerates down to a point where it looks like things matter, they have to turn up the complexity settings to max *and* break out three monitors at 1920x1080 a head *and* turn up things like AA to ludicrous values to show a meaningful difference for game playing.

      I personally look forward to a 'midrange' Northern Islands card that can pull off the 1920x1080 with basically all the settings cranked up without AA or multihead. I know I could get one already with either southern islands or fermi, but holding out for one more generation for just that much less power draw. My 8800GT just can't keep up with some of the new releases without turning a lot of settings down, despite the GP post opinion.

      --
      XML is like violence. If it doesn't solve the problem, use more.
    18. Re:Overpowerful. by Anonymous Coward · · Score: 0

      You do realize that perceive and feel mean the same thing right?

      Just because your conscious mind can't go as fast as 24 fps, doesn't mean that gamers can't perceive that difference especially having trained their bodies to make split second decisions.

      This is the difference between studying the mind, and actually using it practically and why so many of us disagree with you.

      Plus I find it highly dubious for any scientific person to say without a shadow of a doubt that no one can "see" anything higher than 24 fps. You are making such a blanket statement there it's not funny. Especially to those of us who can definitely feel monitor's refresh rates and the mechanical flicking of fluorescent tubes.

    19. Re:Overpowerful. by anonymov · · Score: 4, Interesting

      You keep talking about "research", may be _you_ care to provide a research that shows "24 fps should be enough for everyone"? (hint: it's not, and it's the reason for current studies for 50p/60p/72p film and television).

      Why, you can just go here http://frames-per-second.appspot.com/ and tell us "I don't see any difference". And then we'll just tell you to visit your eye doctor.

    20. Re:Overpowerful. by unity100 · · Score: 1

      yes, and hence, therefore people have the biology to be able to perceive the difference in between 30, 40, 60 fps.

      just because there materialized a percentage of gamers that think they do.

    21. Re:Overpowerful. by Kjella · · Score: 1

      Well, I'm sorry you're an idiot. You're somewhat right, actually the eye just takes 18-20 fps to feel smooth if the scene is motion blurred. Reality doesn't have frames, during that 1/20th of a second everything moves. A rendered screen is not motion blurred and will seem extremely stuttering. Yes, perhaps if you rendered at 60 fps and averaged down to 24 fps you wouldn't notice the difference, but having a graphics card that can only render at 24 fps is clearly insufficient. You should go see an optician if you don't notice it.

      --
      Live today, because you never know what tomorrow brings
    22. Re:Overpowerful. by Jerome+H · · Score: 1

      I can see when a CRT is set at 60Hz and even a faint flickering at 70Hz.
      Same as these LED lights blicking everywhere.

      --
      int main() { while(1) fork(); }
    23. Re:Overpowerful. by Warma · · Score: 1

      HDMI is for passive entertainment, not twitch action games. The dynamics and requirements are entirely different, yet you're using them just the same. The standards also exists because of legacy reasons concerning existing video material, processing speeds and data storage limitations of existing media. Compromises to provide best image quality with certain limitations for consumer use. Perhaps you would want to investigate what kind of framerates and response times the US army uses in their remote feeds and huds etc.

      You can do a double-blind test with your friends. Normalization can be done between having them use 30 vs 30, 60 vs 30 both ways and 60 vs 60. If you can wait until January (when I have time), I can help. If you really think the results will be news, you may want to publish, but I fear they are not.

    24. Re:Overpowerful. by Anonymous Coward · · Score: 1

      You are using research as doctrine, which is foolish and dangerous. You are unwilling to perform the simple experiment that other people ask, why? Because you know you are wrong when faced by the evidence. We do not need research to know we are right, because we have done it ourselves. If you did it, you would find that your statements are bullshit as well. Stop spouting research as if its the be-all-end-all source for knowledge. That is dangerous thinking and not what scientific thought is meant for.

    25. Re:Overpowerful. by Anonymous Coward · · Score: 0

      How about you just wiki that crap at first...
      http://en.wikipedia.org/wiki/Frame_rate

      Also read up on what motion blur does and also check why 24 fps film actually exists.

      Studies by Thomas Edison determined that any rate below 46 FPS "will strain the eye".

      I'm no Edison but I think it's safe to claim nowadays that it's probably even more than 46. :)

      Now go google your "research" yourself.

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

      I can perceive the difference between 60 fps and 200 fps (ibm crt/quake 3 arena ftw).

      (I miss real framerates, but at least the "3d-ready" flatpanels are finally getting above 60/75.)

      But don't trust me, or your own senses, have a science paper:
      http://web.cs.wpi.edu/~claypool/papers/fr/fulltext.pdf

      "CS Grad students score 1 more frag per minute when framerate goes from 30 to 60 fps."

      Captcha: "record"

    27. Re:Overpowerful. by billcopc · · Score: 1

      Depends on the game.

      Back in the 90's, many top-selling PC games ran at 30 or even 15 fps, and were perfectly playable. I played the fuck out of Doom and Quake at then-acceptable framerates, which today would be considered slideshows. I sometimes play WoW on my laptop, where it can drop to 20-25 fps during intense fights, and it's just fine.

      The only place where absurdly high framerates are mandatory are fast-paced shooters like Quake 3/4, Call of Duty, Team Fortress etc. Racing titles also benefit from a steady 60fps to enhance realism, but are otherwise perfectly manageable at 30.

      When I'm looking at graphics upgrades, I don't want higher framerates, I want better post-processing quality or increased resolution.

      --
      -Billco, Fnarg.com
    28. Re:Overpowerful. by unity100 · · Score: 0

      yes. that many justification, rationalization as to why and how, and yet there isnt any research to back anything over 30 fps.

      oooookay. twitch gamers, are a superbreed indeed. would believe it if i wasnt one of them.

    29. Re:Overpowerful. by unity100 · · Score: 1

      and you are not able to see a lcd/led set at 60 hz. why.

      the fact that your eye can send 60 hz flickering light on/off situation to your brain does not mean that your brain is able to interpret the picture in front of you at the same rate.

      and still, where is the research backing that 60 fps proposition ?

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

      then here I bring ONE research that 'people with different perception' can perceive in difference (one is enough):
      http://www.boallen.com/fps-compare.html

    31. Re:Overpowerful. by Luckyo · · Score: 1

      False. The human eye's focus is indeed typically incapable of sensing more then 24 frames every second. On the other hand, peripheral vision can in some cases distinguish over 100 images every second. This was very visible back in CRT days when monitors caused headaches as peripheral vision saw the blinking on the crappy 60hz and sometimes 75hz monitors stressing the hell out of your eyes.

      The issue dates from the way out eye evolved, the way it processes the image, compresses it and sends it to the brain via optic nerve. Our peripheral vision is almost purely light intensity based and designed to track movement, while focus is mostly color-based, far more populated with receptors then peripheral and designed to accurately access objects deemed important. As a result, when passively watching TV from long distance, you can usually fit entire image into focus = 24 fps is enough. When playing on a big screen near you and actively tracking movement over entire screen, 60 fps can be insufficient in some cases.

    32. Re:Overpowerful. by unity100 · · Score: 0

      yes. and now link this idiot ANY research that shows humans are capable of perceiving 30, 40 and 60 fps motion picture differences.

      3 people called me 'stupid, idiot etc' in this thread so far, ALL of them have recounted their PERSONAL perceptions/propositions, AFTER i asked them where was the research that supported that proposition just 1 post before.

      - 'humans can perceive 30, 40, 60 fps !!!'

      - okay where is the research ?

      - you are an idiot ! me and my gamer buddies this me and my gamer buddies that, bleh, blah and so on !!!!

    33. Re:Overpowerful. by Anonymous Coward · · Score: 0

      link me ONE research that shows humans are able to perceive differences in between 30 fps and 60 fps. ONE research. except from a percentage of 'fps gamers' thinking that they can perceive 30 and 60 fps.

      Why don't you link research showing the contrary? Or at least research validating your opinion of "24 fps, the magic number"?

      and then move on to explain why hdmi standard mandates 24 fps, and this has been the minimum requirement for 'smooth' since aeons in display related 'anything'.

      It mandates 24 fps because film is on 24 fps because it has been on 24 fps since practically forever. Actual film also provides a certain "look" that people want in movies. Frame rate conversion when not dealing with even multiples will cause you to lose quality. This is especially true when the conversion is being done in real time on some cheap $2 chip in a TV, rather than over a long period of time with high quality on a film-editing workstation.

    34. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Look at this video: http://www.mediafire.com/?ifmmynzl50m
      If you can't see the difference then you should visit your eye doctor.

    35. Re:Overpowerful. by Anonymous Coward · · Score: 0

      How about you so some simple testing yourself? There are no "research papers" for the same reason there are no research papers to see if the majority of people can taste the difference between water and chocolate milk.....it's extremely obvious.

    36. Re:Overpowerful. by anonymov · · Score: 1

      So it seems you don't have a research on hand to show us that 24 fps is totally enough for synthetic images without temporal smoothing.

      And you was talking so confidently, I almost believed you talk from knowledge :(

    37. Re:Overpowerful. by unity100 · · Score: 1

      that is not a research. it is some website saying the same thing. do you know what a 'research' is ?

    38. Re:Overpowerful. by Bengie · · Score: 2

      I use to play software rendered Quake @ 320x200 8bit color and sub 20fps.

      This is what it feels like to play on consoles, when coming from PC. Flat lighting, crappy models, poor special effects, and a FPS that makes it feel like I'm roller skating with a strobe light.

    39. Re:Overpowerful. by unity100 · · Score: 1

      http://web.cs.wpi.edu/~claypool/papers/fr/fulltext.pdf

      some dudes' master thesis counts as research now ?

      a study which polled the other gamer dudes in their department, and ASKED them whether they were able to perceive stuff by the way. yeah.

      if i do the same in an overclock forum, i can assure you that i can come up with higher than 100% percentage testimonies to that regard.

      that doesnt make it anything scientific. does not modify their biology/physiology either.

      im still waiting for the research to prove the bullshit. apart from some gamer dudes polling some other gamer dudes and submitting it as master thesis.

    40. Re:Overpowerful. by unity100 · · Score: 1
      are you aware that the wikipedia article you linked gives NO reference to the SAME claims that are made by gamers in the below form ?

      However, this question also does not have a single straight-forward answer. If the image switches between black and white each frame, the image appears to flicker at frame rates slower than 30 FPS (interlaced). In other words, the flicker fusion point, where the eyes see gray instead of flickering tends to be around 60 FPS (inconsistent).

      where is the reference ?

    41. Re:Overpowerful. by unity100 · · Score: 0

      You are using research as doctrine, which is foolish and dangerous.

      oh - is it ? why ?

      i could as well say that i was a ghostwhisperer, and there were other people too, and someone would ask me to link a research in that direction proving my claim ?

      and in response, could i tell them 'you can simply do this experiment' ?

    42. Re:Overpowerful. by anonymov · · Score: 1

      So you can't trust your own eyes? You only want to read smart words, experimental proof is too lowly for you?

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

      Hmm. I think I might not understand what you mean by "research."

      If you could elaborate a bit on what sort of research you're looking for, i can take a look around for you.

      Please define for me what you consider "scientific research." An example in a different topic would be fine.

      I'll check back at this comment thread later on today when I'm having lunch and see if I can find something that'd demonstrate it one way or the other.

    44. Re:Overpowerful. by unity100 · · Score: 1

      the fact that you are looking at a 60 hz led/lcd monitor and being comfortable with staring at it, whereas you were uncomfortable using a 60 hz crt monitor back 10 years ago, should be explicative enough for you.

      'synthetic images with temporal smoothing' -> oh yeah.

    45. Re:Overpowerful. by Jamu · · Score: 1

      Fairly easy to do if you wave your hand in front of the screen. It's noticable too, if you rotate your view in a 3D computer game. This makes it noticable in first-person shooters. I've not seen much research into it, although John Carmack did some for Rage I believe. He found 60 fps crisper than 30 fps, but little difference between 60 fps and 120 fps. This agrees with what I've found while playing games. 30 fps is playable, 60 fps is very smooth. It's easy enough to research this yourself.

      --
      Who ordered that?
    46. Re:Overpowerful. by unity100 · · Score: 1

      Why don't you link research showing the contrary? Or at least research validating your opinion of "24 fps, the magic number"

      you gave your answer in your retort to hdmi point below your post. rationalizing it like 'because film has been like that forever' is one step away from rationalizing it as 'film has been like that forever because it was the minimum frame rate that humans were able to interpret moving images forever'. not 5 fps, not 10 fps, but 24 fps.

    47. Re:Overpowerful. by Luckyo · · Score: 1

      It doesn't rebuke HDMI specification. The reason for 24 fps being the typical video speed is hidden in the fact that our brain stop perceiving individual images and starts perceiving concurrent frames as motion around that number.

      It doesn't mean that eye and brain are incapable of distinguishing or things in motion across the screen being "jerky" (i.e. tell that it's separate frames) without either significant smoothing effects typically used in movie industry or other similar methods.

    48. Re:Overpowerful. by unity100 · · Score: 1

      yeees. and where is the research that shows humans are biologically capable of interpreting 60 fps again ? nowhere ? i thought so.

    49. Re:Overpowerful. by YojimboJango · · Score: 1

      Here's a good one http://www.100fps.com/how_many_frames_can_humans_see.htm
      "So the conclusion is: To make movies/Virtual Reality perfect, you'd have to know what you want. To have a perfect illusion of everything that can flash, blink and move you shouldn't go below 500 fps."

      Also from Wikipedia to debunk the 24fps recording thing:
        Judder is a real problem in this day[when?] where 46 and 52-inch (1,300 mm) television sets have become the norm. The amount an object moves between frames physically on screen is now of such a magnitude that objects and backgrounds can no longer be classed as "clear". Letters cannot be read and looking at vertical objects like trees and lamp posts while the camera is panning sideways have even been known to cause headaches. The actual amount of motion blur needed to make 24 frames per second smooth eliminates every remnant of detail from the frames. Where adding the right amount of motion blur eliminates the uncomfortable side effects, it is more than often simply not done. It requires extra processing to turn the extra frames of a 120 FPS source (which is the current recording "standard") into adequate motion blur for a 24 FPS target. It would also potentially remove the detail and clarity of background advertising. Today, devices are up to the task of displaying 60 frames per second, using them all on the source media is very much possible. For example, the amount of data that can be stored on Blu-ray and the processing power to decode it is more than adequate. Though the extra frames when not filtered correctly, can produce a somewhat video-esque quality to the whole, the improvement to motion heavy sequences is undeniable. Many televisions now have an option to do some kind of frame interpolation (what would be a frame between 2 real frames gets calculated to some degree) using technologies like Trimension DNM. Sophisticated algorithms can utilize motion compensation information to achieve a very high degree of accuracy with few artifacts.

      Balls in your court now.

    50. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Except this has been proven wrong in 'blind' (ie no FPS showing) tests:

      http://whisper.ausgamers.com/wiki/index.php/How_many_FPS_human_eye_can_see

      You are the one who is full of bullshit. Repeatedly. Stop repeating this tired crap, and actually go and LEARN SOMETHING.

    51. Re:Overpowerful. by unity100 · · Score: 1

      yes ok and now link the biology/neuroscience research that shows human eye-brain is capable of interpreting 30, 40 and 60 fps.

    52. Re:Overpowerful. by Warma · · Score: 4, Informative

      Human eye does not see in frames per second. It has a certain data transfer speed, and the way brains process the information is also not as discrete as you might want to wish.

      For example, the flicker fusion point (inability to distinguish alternating black and white images) is somewhere around 60fps and the army has done experiments on showing images for a very short time to see whether they could be identified by pilots. The shortest intervals were way less than those postulated even by the flicker fusion point. It also matters greatly how large amount the object moves in your absolute field of vision between frames for your brain to understand motion and simulate smooth movement. Your brain has interpolation algorithms that piece together information streams to form smooth motion.

      This has some information, but not many references: http://www.100fps.com/
      Wikipedia has more stuff and it's a starting point to look for research. Here's some by BBC on fast-moving objects in sports: http://downloads.bbc.co.uk/rd/pubs/whp/whp-pdf-files/WHP169.pdf

      Also, you can test the difference of 30 and 60 fps here: http://frames-per-second.appspot.com/
      At least to me, it is blatantly evident.

      You were wrong, and acted like an ass over it towards me and other posters. Will you please apologize and shut up?

    53. Re:Overpowerful. by anonymov · · Score: 1

      Yes, and I used at least 100Hz capable CRT monitors for that exact reason.

      But wait, so you agree that 60Hz flicker is still noticeable and that's the reason non-flickering LCDs are better?

      And I thought that it had to flicker at 24fps for it to be noticed by the eye, right, right?

    54. Re:Overpowerful. by anonymov · · Score: 1

      Oh, and what did you mean by:

      'synthetic images with temporal smoothing' -> oh yeah.

      Some unfamiliar words? Don't be shy, ask if you don't understand something.

    55. Re:Overpowerful. by Anonymous Coward · · Score: 0

      I would love to find out what you think 'a research' is.

    56. Re:Overpowerful. by Bengie · · Score: 1

      I can tell when my games drop below 60fps all the time. Fast movements get "choppy".

      Try making fast movements. 1080p monitor has 1920 horizontal pixels. If an object moves across your screen in 3 seconds, that means it transverses 640 pixels per second. At even 60fps, that means the object is skipping ~11 pixels every frame. That is NOT smooth.

      When you're playing an FPS games, objects can move across your screen MUCH faster than 3 seconds. When something takes 1 second to cross your screen, it's hard to aim when it's skipping 32 pixels at a time.

      Back when I played Counter-Strike, I got to play on a 100hz monitor for a few hours. I actually found it easier to aim during *fast action* with the 100hz monitor than my 80hz monitor, because objects moved smoother across the screen. And yes, the video card was able to handle more than 100fps on average.

      I've been stuck on a 60hz lcd for a long while and can't wait for 120hz to come down because I miss the smoothness of 80+hz

      I would also like to point out "30fps is enough" is like the "24bit colors is enough" argument. Yes, it supports 16.8mil colors, but I see color banding any time there's a gradient, which is quite often. "but the human eye can't perceive that many colors". What ever. Obviously not measuring the same thing. The real world doesn't update in frames, it's fluid. The human eye may update an average of 30fps, but it's probably an analog update. Screen refreshes are a digital update (on or off)

    57. Re:Overpowerful. by Luckyo · · Score: 1

      Your assumption that everyone has a handful of conveniently placed online research papers to back up their knowledge is rather intriguing. I studied from dead tree books, and that's what they said.

      Considering that you're alone with your disagreement I would suggest that you instead present research that human eye cannot perceive more then 24 frames per second. Or you can go what I did - get an university education that touches on these things.

    58. Re:Overpowerful. by BlackSnake112 · · Score: 1

      There are many people who can tell the difference between 30 fps and 60 fps when playing a game. If the game changes fps while playing it is more noticeable. If the game is constant 30 (everywhere) you would not notice. That is the issue. There are fps changes. If you can get a GPU that keeps the fps around 60 you will have a better playing experience. The fps will drop in the really intense areas, but they will still be smooth to your eye. If you get a GPU that drops to the mid to low 20s in intense areas, you will notice this. That can affect your game play. The thinking is keep it around 60 fps everywhere. So if it drops a little, we will not notice.

    59. Re:Overpowerful. by Anonymous Coward · · Score: 0

      You don't understand how subliminal images work, do you?

    60. Re:Overpowerful. by Anonymous Coward · · Score: 0

      24 fps on TVs is equal to 48 fps on computers.

    61. Re:Overpowerful. by andydread · · Score: 1

      I would take that performance at 5040x1050 res What performance are you getting at that resolution?

    62. Re:Overpowerful. by Jerome+H · · Score: 1

      Well I see the rainbow effect when looking at such projectors http://en.wikipedia.org/wiki/Digital_Light_Processing.
      And I also see the new cars backlight LED flickering which can be quite annoying while driving.

      --
      int main() { while(1) fork(); }
    63. Re:Overpowerful. by squizzar · · Score: 1

      I would have put it as a convenience feature - there's plenty of 24fps source material out there and cross-conversion is tricky. One interesting point to note is that cinema screens tend to double- or triple-expose each frame (e.g. the light is flashing at 48 or 72Hz, the film frames are changing at 24Hz) to reduce the perception of flicker. So quite probably the point at which smooth motion is perceived and the point at which flicker is perceived (or irritating) is very different. Makes me think of strobe lights: there's a point at which the movement is no longer jerky, but the flashing is still very apparent.

    64. Re:Overpowerful. by willy_me · · Score: 1

      The human brain can perceive objects at 30fps. Great, but one has to generate the signal at twice the frequency in order to create an accurate reproduction for the human brain. Look up "Nyquist frequency" if you want some more technical details. Even higher frequencies are better but have little effect.

      While intuitively 30fps should be sufficient, the generated images are not "synced" to the brain. This is not a digital to digital transfer, information is lost. To accommodate for this images should be displayed at 60fps. The brain might only perceive 30fps but displaying images at 60fps ensures that the brain processes the correct information.

      There is a reason why Apple runs their iOS interfaces at 60fps. It does make a difference.

    65. Re:Overpowerful. by Anonymous Coward · · Score: 0

      yes. and now link this idiot ANY research that shows humans are not capable of perceiving 30, 40 and 60 fps motion picture differences.

      you have called everyone 'stupid, idiot etc' in this thread so far, ALL you have recounted are your PERSONAL perceptions/propositions (and some crap about HDMI which is completely irelivant), AFTER you have been asked where is the research that supported that proposition.

      - 'humans can't perceive 30, 40, 60 fps !!!'

      - okay where is the research ?

      - you are an idiot ! me and my lonely self this me and herp a derp, bleh, blah and so on !!!!

    66. Re:Overpowerful. by jeffreyrollins · · Score: 1

      I'll play. In On Frame Rate and Player Performance in First Person Shooter Games by Kajal Claypool and Mark Claypool the authors perform a study which includes measuring the performance and perception of players while they play a first person shooter game at varying fps. The perception difference is is slight and not statistically significant (figure 19) but I would like to point you to figure 17 where they evaluate the score of the players vs fps. Players playing at 60 fps significantly outscored players playing at 30 fps.

      It does not invalidate your argument about perception, but instead provides a strong rationale to exceed 30 fps.

    67. Re:Overpowerful. by parlancex · · Score: 1

      Actually, what you're thinking of 24 continuous frames per second, ie. what you see in most films. It looks smooth to the eye because each frame actually consists of the accumulated light during the 1/24th of a second duration between that frame and the next frame. The resulting "motion blur" can give the illusion of a higher frame rate than what there actually is. I can still EASILY tell the difference between 24 motion blurred frames per second and a higher frame rate, but I'll grant you that it is more subtle.

      Now, in video games we don't have this luxury. Although some modern games implement some form of motion blur no game that I know of would actually attempt to render the "sub-frames" required to achieve a kind of continuous motion blur, and even if you could, why wouldn't you just display those frames you rendered then? In video games each frame you are shown is an an exact point in time which is why you need a much higher frame rate to achieve perceptibly smooth animation.

    68. Re:Overpowerful. by Anonymous Coward · · Score: 0

      are you aware that your posts give NO reference to the SAME claims that are made by you?

    69. Re:Overpowerful. by Anonymous Coward · · Score: 0

      that's a published paper not a masters thesis

    70. Re:Overpowerful. by nschubach · · Score: 3, Insightful

      The argument for 60fps isn't about genetically engineered people. It's about spikes. If you are running your game at 30FPS, you'll turn a corner or some monsters spawn in the next room that drop your FPS below that. The reason people want high frame rates is because of these spikes.

      --
      Every time I start to have faith in humanity, I ruin it by driving to work between 7 and 8 am.
    71. Re:Overpowerful. by ponos · · Score: 3, Interesting

      You silly newb. HDMI uses 24fps for compatibility reasons and the initial decision was probably based on an quality-cost tradeoff back in the days when actual film was used and the NTSC/PAL specifications were defined. Using 60fps would mean that the tape would last half the time, for example. There is the famous "notion" that eyes cannot see over 24fps, but in fact eyes are very sensitive to some kinds of motion, colors and contrast and less sensitive to others, so you cannot generalise that 24fps is "enough" for all kinds of motion, image and people (ye, people are different too). Furthermore, even if the above were not true, in fact you need an average of at least 50-60 fps in most games to ensure that the MINIMUM will not go below 30fps, which is not only visible but also implies a between-frame reaction time of 30ms (plus ping, plus input lag, plus keyoard lag etc). In hardcore-land this mean PWNAGE for you and your silly rig.

    72. Re:Overpowerful. by Anonymous Coward · · Score: 0

      > why wouldn't you just display those frames you rendered then?

      Exactly. Why waste consoles' weak CPUs for caculation of intermediate frames, when you can barely spare them from physics and stuff?

      Human eye is easily fooled, and the fastest movement is usually due to sudden camera rotations, which can be emulated nicely enough with screen-space motion blur. Just throw in some stylishly drawn motion lines at few vertexes of fast moving objects and you're good.

    73. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Glastner in Principles_of_Digital_Image_Synthesis cites
          Robert Sekuler and Randolph Blake. Perception. Alfred A. Knopf, New York,
      1985.

      "The effect of temporal smoothing leads to the way we perceive light that blinks,
      or flickers. When the blinking is slow, we perceive the individual flashes of light.
      Above a certain rate, called the critical flicker frequency (or CFF), the flashes fuse
      together into a single continuous image. Far below that rate we see simply a series
      of still images, without an objectionable sense of near-continuity.
      Under the best conditions, the CFF for a human is around 60 Hz [389]. In
      contrast, a bee has a CFF of about 300 Hz. We note that as with most other visual
      phenomena, the flicker rate (that frequency at which flicker becomes noticeable) is
      dependent on many factors, such as ambient light, size of the visual target, and duty
      cycle between the length of time the image is displayed and the blank time (if any)
      between images. For one set of conditions, Figure 1.12 shows the sensitivity of the
      eye to different frequencies of flicker. Very early movies flickered because there were
      not enough frames displayed per second to cause the eye to integrate the images;
      they were perceived as a flickering series of still photos."

    74. Re:Overpowerful. by sydneyfong · · Score: 1

      and in response, could i tell them 'you can simply do this experiment' ?

      Yes.

      Why not?

      If you had a solid and repeatable experiment that shows the existence of ghostwhisperers, and easily doable using common tools and apparatuses, why wouldn't (some) people be interested in trying it out?

      Alternatively, you could point me to "research" published by cultists and snake oil merchants. Guess which argument is more convincing?

      --
      Don't quote me on this.
    75. Re:Overpowerful. by Anonymous Coward · · Score: 0

      That last link = case closed.

    76. Re:Overpowerful. by amorsen · · Score: 1

      If you think 24fps is smooth you haven't watched the beginning of Lord of the Rings in a cinema. I'll never forget that slideshow.

      --
      Finally! A year of moderation! Ready for 2019?
    77. Re:Overpowerful. by froggymana · · Score: 1

      there is no difference in between 25 fps, or 30 fps or 40 fps to the human eye playing a game. other than the counter fraps shows. there wasnt any difference in between these with respect to human physiology and eye-brain connections either. we were still perceiving 24 fps as smooth, and anything over 30 fps as very smooth. thats why hdmi standard used to mandate 24 fps.

      You gereralize way too much there... It depends greatly on what is being displayed. Please read : http://www.100fps.com/how_many_frames_can_humans_see.htm

      --
      "To prevent this day from getting any worse, I'll just read ERROR as GOOD THING" 1GJU8xLuDKDxEs4KLf8fAGyptoDsqvEsBT
    78. Re:Overpowerful. by Anonymous Coward · · Score: 0

      http://frames-per-second.appspot.com/

      The difference between 30 and 60fps when rendering graphics is right in front of you now.

      Case closed. Now stop being an asshole.

    79. Re:Overpowerful. by amorsen · · Score: 1

      Stunt Car racer on the Amiga would go below 3fps on some tracks. It was perfectly manageable. That doesn't mean I'd ever go back to that kind of frame rate, and I would gladly have sacrificed visual quality to get better frame rate. Unfortunately there was not much visual quality to cut. I suppose they could have switched to wireframes to save on the Z-ordering, but that's about it.

      --
      Finally! A year of moderation! Ready for 2019?
    80. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Doh!
      I misspelled Glassner.
      BTW Andrew has kindly put all 1400 pages of Principles of Digital Image Synthesis out on the internet for download.
      My above quote is from page 18

      Captcha is bumbler
        I swear this thing is sentient

    81. Re:Overpowerful. by Anonymous Coward · · Score: 0

      You probably want to start here:
      The neural determination of critical flicker frequency.
      Bartley, S. H.
      Journal of Experimental Psychology, Vol 21(6), Dec 1937, 678-686. doi:

      Alterations in critical flicker frequency as a function of age and light: dark ratio.
      RA McFarland, AB Warren - Journal of Experimental , 1958 - psycnet.apa.org

      Age and sex differences in critical flicker frequency.
      Misiak, Henryk
      Journal of Experimental Psychology, Vol 37(4), Aug 1947, 318-332. doi:

    82. Re:Overpowerful. by tompaulco · · Score: 1

      It must suck to have to watch TV or movies then. Especially cartoons. Most movie theater films are shown at 24 FPS, and cartoons are often shown at 24 FPS, but they show the same frame twice in a row, and sometimes three or even four, so you are only really getting as little as 12 or even down to 6 changes per second. I am not sure if any of the high resolution digital cable channels are yet using 1080p. They are mostly, if not all 1080i, which effectively means the whole picture only changes 30 times per second, but it looks more fluid because they change have the screen lines 60 times per second.

      --
      If you are not allowed to question your government then the government has answered your question.
    83. Re:Overpowerful. by UnknownSoldier · · Score: 1

      > there is no difference in between 25 fps, or 30 fps or 40 fps to the human eye playing a game.

      Just because YOU can't see a difference, doesn't imply that everyone else is just as blind. I can CLEARLY tell a difference between 30 Hz and 60 Hz gaming, and so can many gamers.

      1. You NEED 60+ Hz framerate so that you can GUARANTEE the _worst_ framerate stays ABOVE 60 Hzm such as when explosions/smoke, etc. is shown.

      2. Average FPS is a useless number compared to MININUM framerate as it completely masks problems.
      http://techreport.com/articles.x/21516/1

      Maybe you should try reading up on Micro Stuttering before looking like an idiot.
      http://www.tomshardware.com/reviews/radeon-geforce-stutter-crossfire,2995.html

    84. Re:Overpowerful. by Anonymous Coward · · Score: 0

      The Ordering of Individuals in Critical Flicker Frequency Under Different Measurement Conditions
      Olga W. McNemar
      The Journal of Psychology
      Volume 32, Issue 1, 1951

    85. Re:Overpowerful. by Anonymous Coward · · Score: 0

      http://downloads.bbc.co.uk/rd/pubs/whp/whp-pdf-files/WHP169.pdf

      Clearly states that there was a visable difference in +100 fps films and below

    86. Re:Overpowerful. by Anonymous Coward · · Score: 0

      if i do the same in an overclock forum, i can assure you that i can come up with higher than 100% percentage testimonies to that regard.

      Well aren't you the ethical scientist that isn't willing to make up statistics to agree with the conclusion you started with. Troll.

    87. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Who needs research when it's so obvious? Do you need a citation than men have a penis and women have a vagina? Seriously if you've seen any computer animation at 30fps and then at 60fps it's hard not to notice the difference.

    88. Re:Overpowerful. by hairyfeet · · Score: 1

      Sorry but I CAN tell the differences between say 32FPS and 40. Now on 40 up things are smooth but at 25-35 for me at least there tends to be a noticeable pop and jerk as i call it where things don't scroll smoothly. Watch a game like Just Cause II at 29-32FPS and you'll see little jerks and pops where things just don't "flow" for lack of a better word. But with my HD4850 I'm getting 60FPS with a min of 43FPS when there is a lot of fire and I can really tell the difference.

      its the whole immersion thing, the smoother the world moves the easier it is to suspend disbelief....well at least until I hookshot into the ground from a burning chopper and walk away without so much as a limp but hey, its a game right? Now if you'll excuse me I got the DLC on the Steam sale and I think I wanna see how quickly villagers get out of the way of my new monster truck. "Up on the sidewalk... bonk bonk bonk" sometimes you just gotta stop and smell the senseless death ya know?

      --
      ACs don't waste your time replying, your posts are never seen by me.
    89. Re:Overpowerful. by Anonymous Coward · · Score: 0

      SWTOR has trash for graphics. Forced max all settings, 16xAnistropic 16xAA, 80fps.

      It looks like garbage from 5 and 10 years ago depending on the aspect. Many textures look like crap from 15 years ago, a face texture should NOT BE UNDETAILED AND SMUDGED. Geometry for many obects have very low polycounts and are very noticeably jagged.

      The only things it has going for it is terrain geometry and face geometry/detail.

    90. Re:Overpowerful. by aaron552 · · Score: 1

      there is no difference in between 25 fps, or 30 fps or 40 fps to the human eye playing a game.

      That's really strange, because I'm sure that BluRay movies @ 1080p/24Hz look horrible on long, fast pans. And that's movies. Games are noticeably worse at the same frame rate dues to the high amount of similar motion (eg. fast mouselook in FPS games). 30fps is a bare minimum, barely playable. 60+fps is ideal.

      --
      I had a sig once. It was lost in the great storm of '09.
    91. Re:Overpowerful. by Anonymous Coward · · Score: 0

      Excellent site. Nice to know that the viewmeter on my monitor agrees with the output: 85fps is visibly different than 60fps, even after leaving the room to forget if 60fps is the bottom or top ball.

    92. Re:Overpowerful. by cheekyjohnson · · Score: 1

      Will you please apologize

      What would the point of that be?

      --
      Filthy, filthy copyrapists!
    93. Re:Overpowerful. by inasity_rules · · Score: 2

      Actually the flicker on those monitors was often traceable back to the flicker from the lights; when the 50/60hz mains '0' coincided with a scan rate, the flicker becomes obvious. It wasn't so much the refresh rate, but the fact that there was an interference pattern. At higher frequencies, the phosphor had less time to fade, so the dip in brightness became less obvious(except at low ambient light levels...). Remember a LCD has no fading effect,so much lower frequencies are now acceptable.

      However, he is wrong. The human eye can perceive all sorts of things. Often you won't be able to say what it is, but it is annoying and can cause headaches. 24fps can be compensated for in various ways - immersion, and using synced sound to distract the brain from the flicker. If the person is talking and the sound comes at the right time, your brain will smooth it out for you...

      --
      I have determined that my sig is indeterminate.
    94. Re:Overpowerful. by Billly+Gates · · Score: 1

      The graphics on a console blow the PC away. My son laughs at World of Warcraft as his Little Big Planet on his PS 3 is nearly photo realistic. Many games on the PS 3 are thanks to crappy intel integrated video chips (not real video cards) sold in 80% of all PCs. Even the xbox 1 from 2002 has better GPU performance than most PCs. Surprise, the games then are optimized for the crappy cards as the developers prefer the power of a console that has real 3d dedicated hardware where it does not have to interrupt the CPU and wait 1,000 cycles to access the system ram.

      Finally AMD is fixing it with the AMD llano but it still is not caught up with the 2007 era PS 3 by a long shot.

    95. Re:Overpowerful. by bemymonkey · · Score: 1

      For an RPG or RTS game, 30FPS minimum (not average) is absolutely no problem. Looks completely smooth.

      For shooters, of course, anything under 60 is pretty much unplayable... but that's not because of stutter or lack of smoothness, but because it induces mouselag.

    96. Re:Overpowerful. by sonicmerlin · · Score: 1

      Also what's wrong with 3 monitor setups that run at full speed? I think that's pretty awesome. If consoles are allowing that to be done without spending a fortune then awesome.

    97. Re:Overpowerful. by Shadow99_1 · · Score: 1

      Well the 24 fps playback of movies is dictated by a classic term called 'Critical Flicker Fusion' and is one reason movies used that rate for a very long time (the modern media uses different rates, such as blue rays output rate of 29.XX fps and many movies are now recorded in high digital fps). A lot of early research showed that an series of images only needs to reach this rate before the human perception is of motion and no longer catches the flicker (like one would see in a flip book).

      That isn't to say people cannot see a difference between fps rates, it's just a note on the rate between 'frames' were the effect occurs.

      --
      we are all invisible unless we choose otherwise
    98. Re:Overpowerful. by Anonymous Coward · · Score: 0

      The reason you can see a CRT at 60hz flickering is that an electrom beam goes through the whole process of creating the picture line by line 60 times a second (From top to bottom progressive, so it has to draw the whole screen everypass)
      LCD:s don't do it this way. They just change the individual pixel states, so the non affected pixels just stay the same.

      For reference the NTSC television standard is 60Hz interlaced, which means 30 fps, drawn every other line per cycle.
      The PAL television standard is 50Hz interlaced, which is 25 fps every other line drawn per refresh cycle.

      Many movies are/were shot at 24 frames per second.

      So you can't watch normal television or movies?

  4. I'll wait till it's $50 by na1led · · Score: 1

    This is good news for many who are looking for bargain Video Cards. I recently purchased a GT240 on Newegg for $40 which performs as well as a Geforce 8800 sold for $600 a few years ago.

    --
    -- By all means let's be open-minded, but not so open-minded that our brains drop out.
    1. Re:I'll wait till it's $50 by drinkypoo · · Score: 1

      I paid $100 for one of those (with 1GB) when I built my system almost two years ago. It's 75% of the performance of a GT250 at 50% of the power and (at the time) just over 50% of the money. It's a great card, or family thereof.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  5. Does GMA still stand for Graphics My ___? by tepples · · Score: 1

    and others who are just glad that the technology drags along the low-end offerings, too

    Has the advance of high-end NV and AMD GPUs dragged along the Intel IGP in any way, shape, or form?

    1. Re:Does GMA still stand for Graphics My ___? by Anonymous Coward · · Score: 0

      Thankfully, yes. As the very least, the current Intel on chip graphics do a nice job of video de/encoding on chip, and do offer modest dx11 acceleration.

    2. Re:Does GMA still stand for Graphics My ___? by yincrash · · Score: 1

      not by much, but it does mean that $50 for an amd and nv card will get you more power.

    3. Re:Does GMA still stand for Graphics My ___? by na1led · · Score: 1

      Intel IGP has the fastest encoding of all other GPU's if you have software that support Quick Sync. Intel has come a long ways but it's mostly geared for HD video not gaming.

      --
      -- By all means let's be open-minded, but not so open-minded that our brains drop out.
    4. Re:Does GMA still stand for Graphics My ___? by LordLimecat · · Score: 1

      Intel Sandy bridge graphics are enough for most things, and Ivy Bridge, is supposed to increase its performance by another 20%.

    5. Re:Does GMA still stand for Graphics My ___? by jandrese · · Score: 1

      Assuming you're willing to run at lower resolutions and framerates, yes.

      --

      I read the internet for the articles.
  6. Bitcoin by Anonymous Coward · · Score: 5, Funny

    Yes yes.. Rendering yada yada. How many Mhash/s does it average when bitcoin mining? And what is the Mhash/Joule ratio?

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

      If you got the card for free and you tried to pay for the increase of your power bill with mined bitcoins, it wouldn't cover half. But for some people, power is free when you mine from work. (In the late 80's, They Might Be Giants had an answering machine in Brooklyn that would play you something new every day. They just had a regular NY phone number, considered "long distance" at the time for most Americans, but they helpfully added the tip that the call is "free when you call from work". I just realized that I'm old and my nerdy references now need explanations. Sorry.)

    2. Re:Bitcoin by Guppy · · Score: 1

      I've never been interested in Bitcoin mining, but as it becomes less worthwhile, I'm hoping it will depress prices on the used graphics card market, as former miners liquidate their rigs.

    3. Re:Bitcoin by Statecraftsman · · Score: 1

      Not to speculate too much but we've probably passed the peak of mining hardware being listed for sale. Due to difficulty decreases and recently increasing prices, you may see increasing prices on GPUs in the next month or two.

    4. Re:Bitcoin by Kjella · · Score: 1

      Sadly I know the answer to this as it appeared someone asked in all seriousness. The new GCN architecture is better for compute in general, but worse for BitCoin as they switch from VLIW to a SIMD architecture. But please buy one and eBay it for cheap afterwards all the same ;)

      --
      Live today, because you never know what tomorrow brings
    5. Re:Bitcoin by Anonymous Coward · · Score: 0

      Overall compute power is still increased, and that's before overclocking, which has almost no thermal penalty at stock voltage on this card, so it should be good for miners, too. Also, I'm rather confused as to how VLIW4 can be superior to SIMD for just about any workload - the entire point of the switch was that VLIW4 had dependency limitations that prevented its compute power from being fully utilized at all times, whereas SIMD is closer to full utilization.

    6. Re:Bitcoin by Kjella · · Score: 1

      Also, I'm rather confused as to how VLIW4 can be superior to SIMD for just about any workload - the entire point of the switch was that VLIW4 had dependency limitations that prevented its compute power from being fully utilized at all times, whereas SIMD is closer to full utilization.

      BitCoin is one of the few algorithms that didn't run into those limitations. So SIMD is much better for computing in general, but actually a drawback for BitCoin.

      --
      Live today, because you never know what tomorrow brings
    7. Re:Bitcoin by Anonymous Coward · · Score: 0

      BitCoin is one of the few algorithms that didn't run into those limitations. So SIMD is much better for computing in general, but actually a drawback for BitCoin.

      An algorithm which "didn't run into those [VLIW] limitations" does not imply that SIMD is a drawback for said algorithm. Why do you leap to that conclusion?

  7. Linux Driver State? by chill · · Score: 5, Insightful

    What is the state of Linux drivers for AMD graphics cards? I haven't checked in a few years, since the closed-source nVidia ones provide for excellent 3D performance and I'm happy with that.

    But, I'm in the market for a new graphics card and wonder if I can look at AMD/ATI again.

    No, I'm not willing to install Windows for the one or two games I play. For something like Enemy Territory: Quake Wars, (modified Quake 3 engine), how does AMD stack up on Linux?

    --
    Learning HOW to think is more important than learning WHAT to think.
    1. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      How about the open source AMD/ATI drivers specifically?

    2. Re:Linux Driver State? by Anonymous Coward · · Score: 1

      I find the proprietary AMD graphics driver more than satisfactory. I run a compositing desktop, one 3D game and sometimes a bit of CAD or modelling software, on a Radeon HD 5700. Very smooth, utterly reliable, no problems at all, except that because the driver apparently conflicts with the GPL my distro vendor no longer provides it, so I have to rely on another repository. But that's okay.

      Frankly, I'm willing to use non-GPL-friendly drivers for a video card. Yes, it would be nice if they'd show more respect for the GPL and FSF, but one has to be pragmatic.

    3. Re:Linux Driver State? by Anonymous Coward · · Score: 3, Insightful

      They suck just like they always have. But don't feel left out, they suck on Windows as well.

      ATI/AMD may at times make the fastest hardware but their Acillies Heel has and apparently always will be their sucky drivers. The hardware is no good if you can't use it.

      They need to stop letting hardware engineers write their drivers and get some people that know what they are doing in there. They need solid drivers for Windows, Linux, and a good OpenGL implementation. Until then they can never be taken seriously with their broke-ass software.

    4. Re:Linux Driver State? by LordLimecat · · Score: 1

      Quake3 probably doesnt need a top of the line graphics card. Go with an nVidia, for years that has been the best move if you think you may use Linux at some point.

      $30 should get you a card that maxes out anything quake3.

    5. Re:Linux Driver State? by simcop2387 · · Score: 1

      I've personally had stability issues with the AMD drivers for my laptop. But this seems to be because they only try to support ubuntu and all the versions that they use, anything else it's hard to get a fix for until ubuntu either updates to your version or you find a work around in the community. The open source drivers aren't as performant but they are much more stable for me.

    6. Re:Linux Driver State? by chill · · Score: 1

      Bah! My mistake. It is a heavily modified id Tech 4 engine, which is Quake 4/Doom 3 -- not Quake 3. No $30 card will max that out.

      My fault.

      --
      Learning HOW to think is more important than learning WHAT to think.
    7. Re:Linux Driver State? by drinkypoo · · Score: 2

      The almost-but-not-quite-latest card is generally fairly well-supported by fglrx. If your card is old enough to be supported by ati then it may work but it probably won't support all its features. You're far better off with nvidia if you want to do gaming.

      Every third card or so I try another AMD card, and wish I hadn't immediately. Save yourself.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    8. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      I know that peoples experiences with ATI drivers and Linux are quite varied, but mine have been terrible.
      Using both their closed source, and the open source Xorg drivers, I have had nothing but stability issues with ATIs graphics cards. Xorg crashing, kernel dumps, system freezes, you name it and I've experienced it. These same graphics cards work flawlessly under Windows.

      So, these days, when I've got a Linux box with ATI graphics, I just use the VESA framebuffer drivers. If I need more than that, I order an NVidia card.

      Please bare in mind that I'm not saying ATI and Linux is always bad, I'm just relating my experiences.

    9. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      Good question. Which modern graphics card will increase your fps from 200 to 500 for Quake 3?

    10. Re:Linux Driver State? by div_2n · · Score: 2

      The closed drivers have serious quality issues with major regressions seemingly every other release.

      The open drivers are making great strides, but the performance isn't there yet for newer cards. If you are using a pre-HD series card, you'll find pretty decent performance that often beats the closed driver.

      Based on the progress I've seen over the last year, I would expect the performance for this new series of cards to be acceptable in a year or so for the simple fact that as they finish the code for older cards, much of the code base will help improve performance for newer ones.

      http://www.x.org/wiki/RadeonFeature

    11. Re:Linux Driver State? by saint0192 · · Score: 1

      I run Debian and Scientific Linux, and have used the Driver installer from AMD's page to configure the proprietary driver on both with no problems. My home desktop has an ATI 5770 running triple headed on Debian Squeeze and has been flawless and stable for over a year. The drivers are fine, IMHO.

    12. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      Better yet, what's the state of OpenCL support in linux? The only thing that is keeping OpenCL down is the absence of drivers to be able to use it in computers that matter. If OpenCL stays down then no one will bother to develop software for it and people like me are forced to go the multi-threaded way instead of taking advantage of a piece of hardware which already exists and is largely wasted on compiz and the like.

    13. Re:Linux Driver State? by karolbe · · Score: 3, Informative

      State of ATI/AMD drivers on Linux is rather poor, much worse than nVidia. My recommendation is to stay away from AMD GPUs if you plan to use Linux. If you are looking for more details about AMD & Linux read this article on Phoronix: http://www.phoronix.com/scan.php?page=article&item=amd_ayir_2011&num=1

    14. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      Every third card or so I try another AMD card, and wish I hadn't immediately. Save yourself.

      fglrx is great for newer cards. But the older stuff is better supported by the "plain jane" x.org radeon/radeonhd drivers. Bother yourself with fglrx as a last resort.

    15. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      The driver doesn't work with an AMD A6 fusion chip in my laptop so i doubt it works with a new generation of cards yet.

    16. Re:Linux Driver State? by makomk · · Score: 1

      They suck about the same amount as NVidia's Linux drivers in my experience, though it's compensated for slightly by software developers being more willing to work around NVidia driver bugs than AMD ones.

    17. Re:Linux Driver State? by makomk · · Score: 1

      It's probably more accurate to say that when NVidia break something it doesn't make Phoronix or get negative attention in the same way as when AMD do. For example, part of the reason I switched to ATI cards in the first place was because on several occasions NVidia managed to break their driver so that it caused system lockups on dual-core systems with the card I was using and then dragged their feet on fixing it. The final straw was their buggy rendering acceleration that made KDE 4 unusable - it took so long for them to fix it that I basically gave up and switched to the open source Nouveau driver before eventually buying an ATI card.

    18. Re:Linux Driver State? by LS · · Score: 1

      Fuck AMD. I specifically bought an HP D4-3011TX notebook due to it's relatively powerful Radeon HD 6750, and as it turns out the douchebags at AMD aren't going to put out a Linux driver that supports it. Their standard Windows driver doesn't even support it - you've got to use the old one from HP. And I DID do my research on this before purchasing. I couldn't find anything anywhere that stated that they wouldn't support it, and in fact everything on their site indicates that their drivers support the 6750. The laptop happens to also have an intel graphics processor, and AMD won't support chipsets that switch between two graphics processors. Assholes. I have to buy a new laptop because of this.

      --
      There is a fine line between being a cultivated citizen and being someone else's crop. - A. J. Patrick Liszkie
    19. Re:Linux Driver State? by Anonymous Coward · · Score: 0

      It's finicky if you don't know what the drivers are looking for, but Catalyst 11.11/11.12 added OpenCL support into the end-user driver package, and combined with the amd-stream sdk for clinfo, you can relatively easily check if it's working properly. If it gives a segfault, it means /etc/OpenCL/vendors/*.icd is botched. The gentoo experimental ebuild for example isn't currently set up to place the 32 bit icd file (which is just a one line plaintext file pointing to the shared library for a piece of currently available hardware). Once that is set up, it should be smooth sailing. I've tested a few of the opencl demos while I had an OGL app running without problems, and it does provide some pretty impressive benchmarks for your hardware if supported (4xxx minimum, OCL 1.0 with limitations, 5xxx for OCL 1.0 support, and I believe 6xxx is the minimum for full OCL 1.1 support, 5xxx had some extensions implemented, 4 had nothing, 6 has almost everything, plus the ability to properly segment OCL kernels so you can run more processes at once.)

    20. Re:Linux Driver State? by tyrione · · Score: 1

      The closed drivers have serious quality issues with major regressions seemingly every other release.

      The open drivers are making great strides, but the performance isn't there yet for newer cards. If you are using a pre-HD series card, you'll find pretty decent performance that often beats the closed driver.

      Based on the progress I've seen over the last year, I would expect the performance for this new series of cards to be acceptable in a year or so for the simple fact that as they finish the code for older cards, much of the code base will help improve performance for newer ones.

      http://www.x.org/wiki/RadeonFeature

      Ah yes, the great myth of myths. The Open Source version with zero 3D acceleration, OpenCL support and much more are great! The proprietary version with all support SUCK! This is getting very old.

    21. Re:Linux Driver State? by tyrione · · Score: 1

      The big Pile of Crap in the equation is X-Windows/XOrg Server. It's junk and it's mainly due to it being ductaped with functionality well beyond it's design thus the need for Wayland.

    22. Re:Linux Driver State? by drinkypoo · · Score: 1

      fglrx is great for newer cards. But the older stuff is better supported by the "plain jane" x.org radeon/radeonhd drivers. Bother yourself with fglrx as a last resort.

      The machine I'm using right now has R690M and I run Vista on it because the graphics drivers for every other OS are garbage. 7 will run with a Vista driver but only resume from suspend once, on the second try you get a free reboot. The free ati driver trashes the display even with renderaccel disabled and fglrx doesn't support the hardware at all, nor did it when it was released. I'm done with ATI forever, thanks.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  8. Yeah but compatibility... by Anonymous Coward · · Score: 0

    That would be fine, but my general experience is that the NVIDIA cards are more compatible with OpenGL, whatever the specs say, which is frequently used for indy games. Also the NVIDIA cards have been more reliable and compatible on the whole when used in workstations in my experience.

    Granted, I rarely have the budget to try the newest or best models... maybe it's different at the top end of the spectrum.

    1. Re:Yeah but compatibility... by Tr3vin · · Score: 1

      The only difficulty I have had with my 5870 was during the first week that Rage was out. It was a terrible mess. Once the drivers and game were patched, everything worked fine. I have never had issues with any Indie games though. "Compatibility" is typically measured in the number of patches made to the drivers to work around the various developers' interpretations of OpenGL or DirectX.

    2. Re:Yeah but compatibility... by Anonymous Coward · · Score: 0

      I think it's the reverse. If the drivers have to be re-written for every game with optimization/tweaks it makes me think game developers are adhering to the spec and the driver companies aren't. One driver should work fine with any game.

  9. It's too bad... by Reverand+Dave · · Score: 0, Troll

    ...that AMD can't put out a video driver that isn't total garbage.

    --
    I got here through a series of tubes
    1. Re:It's too bad... by jiriw · · Score: 1

      Well .. PC Perspective had to benchmark this card with some sort of drivers... Guess what; those probably were written by AMD personell. It's already faster than the competitors offering. If it had any major defects they surely would have mentioned it in the article. So if that's total garbage, it can only improve, no?
      I'm not that afraid the cards will be usable only as badly designed space heaters. Because apparently that's something they do badly... having a similar thermal envelope as the previous gen cards. The developers of high power PSUs will be the least pleased with this new product :P

    2. Re:It's too bad... by Dr.+Spork · · Score: 4, Insightful

      Which version are you having trouble with? Are you sure that you're not just mindlessly repeating a 7 year old meme? Are you also one of the people who switched to Chrome because "Firefox uses too much memory" when simple tests show that Chrome uses more? I know it feels like you're a part of the club when you repeat what you hear from the other club members. But don't confuse groupthink with truth - especially when it comes to the quickly-changing world of tech.

    3. Re:It's too bad... by drinkypoo · · Score: 1

      When I brought home my R690M-chipset-based netbook ATI had already abandoned the X1250 graphics in it and dropped them from fglrx (assuming they were ever in there) and they apparently haven't given the folks making the ati driver enough information to support it properly despite their claimed commitment to open source (IME intel has made good on this in more cases than AMD) so it craps all over my system if I try to run Linux, even with RenderAccel disabled.

      ATI Rage Pro stuff is the only ATI stuff that seems to work flawlessly under Linux for me... but how old is that? And how bad were the windows drivers for those cards when they were new?

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
    4. Re:It's too bad... by Reverand+Dave · · Score: 1

      I'm speaking as a technician that works in a shop that uses almost exclusively ATI cards and having experienced many many problems caused DIRECTLY by shitty ATI drivers.

      --
      I got here through a series of tubes
    5. Re:It's too bad... by Dr.+Spork · · Score: 1

      Yeah, but you know very well that the policies back then (many years ago) don't apply to ATI practices with new hardware, which is what this article is about. For starters, it's a different company now. Also, their relationship with the OSS community has changed a lot. Maybe they're not going back and fixing 7 year old problems, but that doesn't mean that their new stuff has the same problems. I'm not saying that everything is peachy now with the drivers, but I'm saying that I see too many people base their conclusions on what they know to be outdated reasons.

    6. Re:It's too bad... by Reverand+Dave · · Score: 1

      Read below before you start going all fan boy. I routinely have problems with that shitty catalyst software on a wide numbers of computers in the shop I work. It's experience talking. So I guess you could count your trolling as unsuccessful.

      --
      I got here through a series of tubes
    7. Re:It's too bad... by keith_nt4 · · Score: 1

      Which version are you having trouble with? Are you sure that you're not just mindlessly repeating a 7 year old meme? Are you also one of the people who switched to Chrome because "Firefox uses too much memory" when simple tests show that Chrome uses more? I know it feels like you're a part of the club when you repeat what you hear from the other club members. But don't confuse groupthink with truth - especially when it comes to the quickly-changing world of tech.

      Well there's been a few version changes in firefox since this time but at my last job 6 or so months ago I had to run chrome because the employer-issued laptop simply couldn't run firefox or couldn't anything *besides* firefox... Chrome on the other hand ran perfectly on this pathetic laptop. So chrome is taking up less of something. If not memory than...something else? All I know is on a laptop with very limited resources Chrome ran great and FF ran crap.

      --
      "UNIX is very simple, it just needs a genius to understand its simplicity." -Dennis Ritchie
    8. Re:It's too bad... by Anonymous Coward · · Score: 0

      I assure you ATI video drivers are still total garbage, at least that has been my experience building several thousand PCs every year for the last 20 or so years.

      I still give my customers what they want, but I try like hell never to sell an ATI card. For the 10% or so of the PCs I sell that have ATI cards, I have two full-time techs devoted solely to dealing with them. That's all they do - deal with ATI card and driver problems.

      So, yes, ATI drivers suck. It's not a meme. It is fact - one that costs me $80K/year in supporting them.

    9. Re:It's too bad... by Anonymous Coward · · Score: 0

      I assure you ATI video drivers are still total garbage

      Agreed. I used nVidia cards until they jumped the shark with Fermi (GF100) and switched to ATi after reading plenty of assurances that ATi's new drivers are good enough.

      Woo boy was that a lie. The drivers lack features of nVidia's, are glitchy and inconsistent (11.10 played Starcraft 2 fine, 11.11/12 have frame stutter) and some of my older games from 2005 and earlier barely work at all, let alone well. [One game I played with forced AA in nVidia, ATi can barely render it correctly without AA, turn AA on and half the animations become black boxes].

      Linux drivers? Suck. The open source ones are crap, the binary catalyst ones barely work and are a real pain to set up. The nVidia ones pretty just dropped in, modify Xorg.conf and you're done. The ATi drivers have /etc/ati and various crap that you need to fiddle with (did I mention that it crashed and locked up for hard reboot every 10 minutes until I worked out the magic voodoo that toggles the default-on "let's use TLS in a completely broken way" flag, just setting that in Xorg.conf isn't good enough, you also need to edit the ATi pseudo-registry using a command line tool with 50 modes).

      After 12 months, I can confirm my disappointment. If nVidia wasn't retarded and gave up their stupid broken CPU pretending to be a GPU then I'd buy one just to have everything work without issues again.

    10. Re:It's too bad... by Anonymous Coward · · Score: 0

      My 4850 will BSOD my W7 x64 when using new catalyst drivers. I can only use the driver provided by microsoft for the OS to boot properly.

      Firefox beta channel (ever since FF4) will crash every two/three days when it exceeds 2GB memory usage. Chrome releases memory when I told it to close a tab when firefox will keep them for the rest of the session until it crash or I restarts it.

      Your premise is quite flawed.

    11. Re:It's too bad... by drinkypoo · · Score: 1

      Yeah, but you know very well that the policies back then (many years ago)

      No, less than two years ago. AMD is the same company it was then, since that's post-merger. You're full of shit, and you don't know what you're talking about, and if you had looked the chipset I was talking about up with WIkipedia you would know that this was depressingly recent. You may comment again when you have done the requisite homework. Or, you know, when you stop working for AMD, you astroturfing tool.

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  10. That FERMI fire... by Anonymous Coward · · Score: 1

    ..keeping you nice and warm in the winter?

  11. What bugs me most by Psicopatico · · Score: 3, Interesting

    Why card manufacturers utilize (rightfully) new manufacture processes (28nm transistors) only to push higher performances?

    Why the hell don't they re-issue a, say, 8800GT with the newer technology, getting a fraction of the original power consumption and heat dissipation?
    *That* would be a card I'd buy in a snap.
    Until then, I'm happy with my faithful 2006's card.

    --
    Mastering the English language is fucking easy: all you have to do is to put an f* word in every fucking sentence.
    1. Re:What bugs me most by jandrese · · Score: 4, Insightful

      The problem is that speed is only one part of the equation. that 8800GT only supports DX10.0. DX10.1 games may run, but you'll find them crashing after awhile unless the developer was very careful (they were not). DX11 games won't work at all.

      You're much better off with a modern card that just has fewer execution units if you want to save money. They won't be out right away (the first release is always near the top end), but they will eventually show up. Since you're worried about saving money/power, you don't want to be an early adopter anyway. Oftentimes the very first releases will have worse power/performance ratios than the respins of the same board a few months down the road.

      --

      I read the internet for the articles.
    2. Re:What bugs me most by Anonymous Coward · · Score: 0

      You seem to be under the misapprehension that the 28nm process is being used by NVidia. I see no current NVidia products using 28nm technology. Therefore, there is no power efficient equivalent to the 8800 GT using the 28nm process. However, the DDR 3 GT 545 is a low TDP power part and seems to have decent specs (see the comprehensive graphics card charts at techarp.com). If low power consumption means that much to you the GT 545 could be a stop-gap solution for your needs until the first NVidia 28nm products begin to appear sometime in 2012.

    3. Re:What bugs me most by Anonymous Coward · · Score: 0

      What are you talking about? They do, in their value offerings. Some are actually passively cooled! In fact, Intel released its value 32nm chips before its performance chips (though this is CPUs, not GPUs). They can't release everything at once, though!

    4. Re:What bugs me most by Anonymous Coward · · Score: 0

      They do, routinely. Not usually for the high-end cards. The GT1XX series is a reissue of the 9XXX series at 55nm. The 3XX series is a reissue of the mid range of the 2XX series at 40nm. The current part of the 6XX series is a reissue of the mid range of the 5XX series at 28nm.

    5. Re:What bugs me most by bemymonkey · · Score: 1

      There are a few cards out there that have relatively low power consumption. The 8800GT idles at about 50W and uses about 110W under load... mid-range ATi cards like the 6670 or 6750 use 12-16W idle and less than 90W (70W in the case of the 6670!) under full load.

      My next card will probably be either a 6750 or one of the midrange 7 series cards (provided there's one that uses significantly less than 100W under full load)... should cut my power consumption nicely (currently running an 8800GT).

  12. But can it run Unity without lag? by captrb · · Score: 2

    But can it run Unity on two screens without lag? I suspect that whatever video card I buy, the modern Linux dualhead display will feel slower than it did in 2005 :-/

  13. switched to radeon, not thrilled. by doug141 · · Score: 0

    I used to use nvidia cards, but always assumed video card brand made no difference, since I'd never heard anything to that affect. I recently got my first Radeon card, a minor upgrade that was twice as fast as my old card but used the same wattage (cool, right?). About once an hour, the new card freezes for a full 2 seconds, and maybe 30 seconds later freezes again for half a second. Sounds like no big deal, unless you are playing modern warfare 2, because then you are dead, and your kill-streak is reset. Sometimes I come out of the 2 second freeze to see my avatar ran off a ledge and is falling to his death. Kinda funny, but not when it borks a kill-streak. Getting killed by your video card is just wrong. I googled around for a solutions, and just found forums full of people lamenting the bad quality of drivers for those cards. It's fine for most games, and it's fine 99% of the time, but that other 1% can be a cruel bitch. Now you know.

    1. Re:switched to radeon, not thrilled. by jandrese · · Score: 2

      Those freezes are probably the driver crashing and resetting itself. It used to be that a driver crash brought down your whole system, but now they can do it in the background silently and all you'll notice is some stuttering (or a short freeze).

      I suspect that ATI and nVidia have been able to use the silent-restart feature to sell more defective cards. If your system totally locks up every half hour when playing a game, you're going to return the card. If it freezes but then resumes silently you may be annoyed but not pin the blame on the card, and not be annoyed enough to actually take action. It could be the disk, motherboard, or something else too, you don't get an indication that it was a video card problem. The very first revisions of this feature used to pop up a box telling you what happened, but that doesn't seem to be the case anymore.

      --

      I read the internet for the articles.
    2. Re:switched to radeon, not thrilled. by Shippu · · Score: 1

      I had a similar problem with my nvidia 7600gt, where I would get lock ups every 10 minutes or whatever in most games, which was apparently caused by a hardware bug. There were also green dots and lines that would appear on the screen. Like they would start as dots and turn into lines if i scrolled down a web page for example.. So this type of shit isn't limited to ati/amd.

    3. Re:switched to radeon, not thrilled. by Shatrat · · Score: 2

      I'd recommend using OCCT to run a stability test on the card, and also note the temperatures and system voltages it hits when fully loaded.

      What you're describing could also be an overheating issue or a power supply shortage. If the temperatures are approaching boiling, that's a problem. If the voltage drops significantly when your CPU and GPU scale up, that's also a bad sign.

      Alternatively, yeah it could be crummy drivers.

      --
      09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0
    4. Re:switched to radeon, not thrilled. by toddestan · · Score: 1

      From what I have heard, the crashes come from the card clocking the memory speed up and down. This also can cause corruption and flicking on the screen when it happens too. The 'solution' is to go into AMD's overdrive utility and fix the memory clock to some value. I don't know what this does to the power consumption, and in some cases it doesn't seem to help either.

      I must say I don't really care for either ATI's or nVidia's products, but when your only other option is Intel's integrated graphics you've got to pick on.

  14. yes. 60 fps. by unity100 · · Score: 1

    because, there is at least ONE research that shows human physiology and senses are able to perceive differences in frame rates above 24 fps as smooth ?

    no. the question is rhetorical. there isnt one single research that shows humans are able to perceive a difference in between 40 fps and 60 fps. its total bullshit.

    hdmi specification requires 24 fps. not 60 fps. because, 24 fps is scientifically backed, whereas the only thing backing 'i can perceive 60 fps' is the self-propagated bullshit from gamers. nothing else.

    1. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      um, with high speed motion in video, you WILL be able to tell difference between 24 and 60fps. perhaps you should read some research papers and spend some time in the display industry.

    2. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      Are you stupid or something? The 24 fps spec is because of film, dumbass. And films have motion blur builtin because of how they are done. I gaurantee that you should be able to distringuish between 24 fps and 60 fps of cripsly rendered frames. If not, see an eye doctor right away because your vision needs help.

    3. Re:yes. 60 fps. by unity100 · · Score: 1

      yeah please link me one of the papers that say people can distinguish in between 30 fps, 40 fps and 60 fps and therefore 60 being the 'necessary norm' for pc gaming. im waiting.

    4. Re:yes. 60 fps. by unity100 · · Score: 1
      im going to just slap the same response :

      yeah please link me one of the papers that say people can distinguish in between 30 fps, 40 fps and 60 fps and therefore 60 being the 'necessary norm' for pc gaming. im waiting.

    5. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      You are an idiot. 24 fps was chosen because it matches the 24 fps of films.

      60 fps is the classic vsync refresh rate, and tends to work pretty good. It's possible for at least some people to tell when a CRT is set at a refresh rate of 60hz or lower, or something higher.

      24 fps in film is okay because thanks to the wonders of the physics of film, it has built-in temporal anti-aliasing, allowing the human brain to interpret it as a smooth video.

      If your video card has temporal AA turned on, then sure, a steady 24 fps would be fine. But otherwise you need more frames per second for your brain to smooth it out.

      Any gamer can tell the difference between a slideshow bouncing between 24 and 30 fps and something smoother.

      To re-iterate: you are a moron. "Scientifically backed" my ass.

    6. Re:yes. 60 fps. by Your.Master · · Score: 1

      If humans can perceive at most 24fps then you need to display at a minimum rate of 48fps (Nyquist rate) in order to reliably convey that information to a human; else you'll jitter.

      But humans don't actually take frame-like snapshots. The flicker fusion threshold for black & white is about 60fps with noticeable variation between individuals. Human beings can reliably identify an object flashed in from of them for only 5 milliseconds: effectively 200fps for one frame. Even though they cannot distinguish the same video played at 200fps and 100fps (the latter one skipping alternate frames to keep the same pace).

      HDMI was 24 fps because film is traditionally 24fps. Film gets away with 24fps because of motion blur. Somewhere around 60fps is pretty decent for images that aren't blurry.

    7. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      Sorry to burst your bubble, but I used to develop titles for Nintendo DS, and the choice for frame rate was essentially 30fps and 60fps. Now, if EVERYONE in the office was able to say when the game wasn't running 60fps after one change or the other, I really don't need any research papers to verify this. It's OBVIOUS that anyone with normal eyesight can tell the difference, and the requirement for our title was 60 FRAMES PER SECOND, because otherwise it wouldn't have been as smooth as we wanted.

      For the trained eye, it's possible to see the DIFFERENCE between 50fps and 60fps. And 60 fps is definitely better. I'm able to tell PAL Mario Galaxy from NTSC galaxy pretty easily.

      24fps is just the lowest acceptable limit, and it's there because of tradition and bandwidth limits. Not because it's some kind of limits for human perception.

    8. Re:yes. 60 fps. by unity100 · · Score: 1

      yes. im an idiot. and where is the research that supports the proposition that humans are able to perceive 60 fps again ?

      cant hear you ?! where ?

      nowhere i guess. other than subjective propositions and perceptions. 'i can see this i can feel this bleh blah'.

    9. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      Your views are extremely outdated, and if you can't tell the difference between 24fps and 60fps then I recommend getting your eyes checked.

      Tell me these examples are indistinguishable.

      http://www.tomshardware.co.uk/athlon-ii-x3-440-gaming-performance,review-31906-2.html
      http://www.boallen.com/fps-compare.html

      Here's some of that research you keep begging for:

      www.cs.wpi.edu/~claypool/papers/fr-rez/paper.pdf

      and perhaps 60fps itself isn't the cap people thought it was:

      http://downloads.bbc.co.uk/rd/pubs/whp/whp-pdf-files/WHP169.pdf

    10. Re:yes. 60 fps. by Luckyo · · Score: 1

      While I can't "link you papers" as I've studied from actual dead tree books back in university, the structure of human eye can be accessed pretty well just by looking at wikipedia. You have focus in the center, highly populated with slow and data-heavy color-sensitive receptors, and periphery highly populated with intensity-sensing fast receptors. Peripheral vision is designed by evolutionary process to track movement, and as a result is capable to distinguish a lot more images every second then focus.
      Another part of the reason why focus is largely incapable of tracking faster movement is because optical nerve compresses raw data extremely efficiently before it's sent to the brain. Much of the compressions is related to the "smoothing of movement" (and many of the so called "optical illusions" you can see on paper are usually a result of this compression). Nature is essentially saving on resources where they're not judged as necessary for survival.

      I'm not sure where you're getting the info that human eye can't track over 24 images per second when pretty much any doctor, or person who studied usability in terms of human perception will be able to tell you you're wrong. There are a couple of ways in which you can test it yourself on yourself if you're doubtful, such as looking at a 50hz CRT television with your focus, and then peripheral vision and noting how you begin to be able to tell that image is in fact blinking as it shifts to peripheral vision.

    11. Re:yes. 60 fps. by unity100 · · Score: 1

      The flicker fusion threshold for black & white is about 60fps with noticeable variation between individuals.

      FLICKER fusion threshold. not image interpretation threshold.

      which is why we are comfortable with 60 hz lcd/led monitors, as opposed to being uncomfortable with 60 hertz crt monitors.

    12. Re:yes. 60 fps. by Luckyo · · Score: 1

      The intentional blurring is a part of it, yes, but it's just a part. There is also the issue of how we as humans process the visual data (i.e. over 24 pictures per second gets interpreted as continuous movement).

    13. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      Quantitative recorded difference between 30fp and 60fps: http://web.cs.wpi.edu/~claypool/papers/fr-rez/paper.pdf

      Can you now provide current research (not the outdated 1950's research that stated 24 fps was enough) that says otherwise? No? Care to apologise to everyone you claimed was speaking bullshit for in fact spouting it yourself, while acting like an arrogant douche?

    14. Re:yes. 60 fps. by chrisG23 · · Score: 1

      Show me a study. No I am serious. My understanding is that around 24 fps is necessary to trick the human eye into seeing continuous motion, hence movies are shown at 24 frames per second, but also using blurring on individual frames. In other words if you look at a single frame of a scene in a movie where the action is happening quickly, the single frame will not be a static image, it will have blurring artifacts. If the single frame was a sharp static picture and was in between other sharp static pictures, the movie would look really weird playing at 24 fps whenever there was a lot of motion happening on the screen.

      I can see the difference in a game running at 30 fps and 60 fps. Both look like they are moving pictures and not a series of still images being constantly redrawn, but there is a distinction i can make between the two. You probably can too. Setup a blind seeing test if you want to find out.

      I also prefer games at a constant 60fps over games at a constant 30fps. I notice the difference. I like the 60fps ones more.

    15. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      How about you just play something at those frame rates you've mentioned yourself? You don't exactly need a laboratory study to see it yourself. If you really cannot distinguish between those frame rates, there is something seriously wrong with your eyes. Even the slowest eyes among people I know can discern 40 from 60.

    16. Re:yes. 60 fps. by Anonymous Coward · · Score: 0

      Here, a concise, explicit, detailed explanation of the actual issue. And its not that people can necessarily distinguish between them, rather that depending on the content of the images, intensity of the whites vs. blacks, etc. you can get different quality of viewing which includes flickering, ghosting, or other results based on that frame rate and content. The reality being that vision ranges / responses can vary widely in individuals does in fact have impact as well of course. If you don't have the few minutes to read it... or understand it then no amount of "help" will likely explain it to you. Of course the fact that none of you seem to be able to find a simple article/site like this one based on fact instead of opinion already pretty much sinks the likelihood of you being able to get to the included link much less understanding its content. *shrug*

      http://www.100fps.com/how_many_frames_can_humans_see.htm

  15. My NVIDIA driver died when I hit the comments page by kaychoro · · Score: 1

    Ironically, when I started reading people's comments, my driver failed and had to be reloaded in Windows 7 for my year old NVidia card.

    --
    //TODO: create a signature
  16. I don't get it. by pablo_max · · Score: 1

    How is this modded as insightful?
    What have Linux driver to do with this card? How are Linux users in any way the target market for a high end enthusiast GAMING graphics card?

    Perhaps once you can purchase BF3 or the like for Linux, then ATI and NV will spend more time writing drivers for Linux.

    I cannot imagine that anything more than an older HD48xx series will help you in any way.

    1. Re:I don't get it. by chill · · Score: 5, Insightful

      Because it was a question that people other than just me were curious about?

      Did you read the entire post? Or did your head just explode when seeing "Linux" in a gaming thread?

      nVidia already spends time on quality Linux graphics drivers. They run fine on both 32-bit and native 64-bit Linux systems. I was wondering if the AMD/ATI stuff had matured as well is all.

      Take a valium and go back to getting your ass n00bed by 10-year-olds on BF or MW.

      --
      Learning HOW to think is more important than learning WHAT to think.
    2. Re:I don't get it. by Korin43 · · Score: 1

      Linux people play games? I thought this was still the 90's!

    3. Re:I don't get it. by Anonymous Coward · · Score: 0

      nVidia already spends time on quality Linux graphics drivers. They run fine on both 32-bit and native 64-bit Linux systems. I was wondering if the AMD/ATI stuff had matured as well is all.

      a) nVidia's "quality" drivers will bend your Ubuntu installation over and ream it not-so-nicely without lube. Try un-installing the "stock" nVidia drivers sometime, and then installing the x.org drivers. Then try enabling desktop effects (aka Compiz)...watch as hilarity ensues.

      b) the OSS X.org drivers for ATI/AMD are just fine, thank you. I get decent 3D framerates while spinning my desktop cube in Compiz (about 30-50fps, average is about 40).

      tl;dr: nVidia installer on Linux has problems and won't uninstall cleanly, x.org Radeon/RadeonHD drivers are just find for all but the most recent 5xxx/6xxx cards.

    4. Re:I don't get it. by Anonymous Coward · · Score: 0

      Try un-installing the "stock" nVidia drivers sometime, and then installing the x.org drivers

      Why would you want to do this? Never had a problem with NVs proprietary drivers neither on Windows nor on Linux.

      With ATI my experience was exactly opposite - proprietary drivers just plain blew on Lin and had numerous annoying bugs on Win - last one I got to experience was HW cursor bug on my friends's PC. It was a year ago. It was fucking 2010 and their fucking devs and QA still couldn't avoid something as glaring as that.

      As far as I can see, they are indeed better off with just giving specs to open source community, as their internal dev team seems to be much less competent.

  17. Re:My NVIDIA driver died when I hit the comments p by Anonymous Coward · · Score: 0

    Now THAT'S a contingency plan!

  18. yes yes. by unity100 · · Score: 1

    its THAT obvious, THAT well known, THAT common, THAT normal, and yet among the approximately 6 people who called me 'idiot' in this thread after i asked them where was any research that showed people can perceive 30, 40 and 60 fps, have yet to still bring ONE research.

    oh, one of them linked to a random website/blog that says the same thing, from some dude's mouth tho. thats what they understand from 'research'.

    1. Re:yes yes. by anonymov · · Score: 1

      And you're yet to deliver a research showing 24 fps is enough, don't forget that.

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

      We did this research at our game studio with the DS. Are you _honestly_ telling me I'd lie to you about such a trivial matter? And all these people who can SEE the difference are wrong, and you're right? Grow up, or go fuck yourself in the meantime.

    3. Re:yes yes. by epine · · Score: 1

      Let me be number seven: you're an idiot.

      Since when is the HDMI standard is a good reference point for arguing human perceptual limits? Furthermore, many things which are quite obvious have no supporting research. Research is like open source, except that the itch has to be even bigger to successfully fund the project. Where's the incentive for pouring research dollars into quantifying the trench between 24 fps and 60 fps? Celluloid is not getting cranked above 24 fps for mainstream productions due to cost reasons (it was tried and failed more than once), and digital playback devices are standardized at 60 Hz already, which poses hardly any cost barrier just a few years down the road. For every paper on the fundamentals of visual motion perception, Google returns a hundred patents on motion estimation and interpolation devices screaming to be monetized.

      I've read several excellent papers over the years on the trade-offs in human acoustic perception (which the margins of my generosity are too small to fish up) where the golden ears demonstrate different trade-offs between fine pitch/duration perception, but always along the same contour.

      From Judder-Induced Edge Flicker in Moving Objects

      In natural scenes, smooth pursuit stabilizes or reduces the temporal rate of motion of a moving image on the retina. Smooth pursuit of a sequence of momentarily stationary images, however, produces a geometric shearing of the image on the retina

      This paper is talking about degradation. For my purposes, if I can discriminate degradation at 30 fps vs 45 fps, then I can perceive something valuable in one that the other lacks.

      In many theatres, I find the judder so severe in a rapid pan I can almost count the frames. Back in the day, 45 fps was the break point in many games where my motion estimation starts to feel creamy enough to nail a snap spin frag. I could image a small additional improvement all the way up to 60 fps, if you're after that last one percent advantage.

      Strangely, there in an ethos is science (which I generally laud) where many researchers tune out the ridicule and devote themselves to proving the obvious nevertheless. Perhaps the definitive paper you are seeking will make the cut on next year's Duh! list.

      From Duh! The Most Obvious Scientific Findings of 2010

      Guys also indicated that even with hook-ups (which are meant to be string-free), they feared their casual-sex partners would seek a relationship. Women indicated the opposite, wanting a relationship and worrying about becoming too attached to a noncommittal other. Who knew?

      Sometimes it is THAT obvious. And now, in 2011, we have a paper to cite.

    4. Re:yes yes. by Anonymous Coward · · Score: 0

      You're pulling a Glenn Beck here. The burden of proof is on you, as the established norm for video games is 60 fps.

      Half the reason the movie industry wants to stick to 24 fps is because of cost, the other reason is "look". When movies were still filmed on actual film, it was expensive. Going from 24 fps to 50 or 60 would more than double your film cost, pretty significant with those big old cameras. Now with the advent of CGI, there is now the worry of render time, which is also really expensive. Avatar probably would have cost millions more if they went for 60 fps, it was already double the cost because 3d rendering requires the whole movie to be rendered twice (once for each eye). I suppose now with streaming and downloaded movies there is also the doubled file size and accompanying bandwidth cost, already a problem because of netflix.

      As far as "look" goes, cheap crappy cameras can all shoot stuff in 60 fps now. High frame rates have become part of what makes up the shitty crappy look of a cheap DV camera, so now most people think a 60 fps movie looks "bad" because it reminds them of cheap cameras but they are not aware of it.

    5. Re:yes yes. by Anonymous Coward · · Score: 0

      Thank you for putting that "idiot" in his place, he is just trolling, I have been trollmodding him here (that's why the AC).

  19. it was a long wait, but... by Anonymous Coward · · Score: 0

    finally my radeon 7500 is state-of-the-art again! WOOHOO!

  20. Re:My NVIDIA driver died when I hit the comments p by Mashiki · · Score: 1

    Well nvidia just released a new beta driver the other day it seems to be stable and I haven't had a TDR since yesterday with it.

    Win7/vista 32
    Win7/vista 64
    XP
    XP 64 server and 2003 64 server
      The TDR problem has been on going with the 280 release and all that the 275.33's were the last stable release, it looks like the 290's are finally stable. Only took them 6mo.

    --
    Om, nomnomnom...
  21. I'll bite. by pepty · · Score: 1

    there is no difference in between 25 fps, or 30 fps or 40 fps to the human eye playing a game. other than the counter fraps shows. there wasnt any difference in between these with respect to human physiology and eye-brain connections either.

    Cite?

  22. What is science? by Warma · · Score: 1

    Did you actually read that thesis? The author defined his methods and subjects well, and created test situations with clear, measurable metrics in addition to just user perception. He then shows the data he got and you can do your own conclusions. I cannot understand, how that paper would somehow not count as research.

    You see, the point of science is not to about WHO did something and what were their credentials. That's the whole point. Only truth matters, and how you approach it. Not only did the author of said thesis postulate something, he actually tested it by meticulous experiment and proved his case. This is what SCIENCE is.

    After reading that paper, if you disagree, you really have to have something to show for it. Either do your own study or present us with a critique of the paper, because that undergrads just completely destroyed your position.

  23. Bitcoin by Rinisari · · Score: 2

    What's the Bitcoin Mhash/sec?

  24. Ummmmmmmm by neozed · · Score: 0

    Due to console gaming retarding pcs. im on single radeon 6950 (unlocked to 6970 by bios flash), and i am doing 5040x1050 res (3 monitor eyefinity) on swtor (the old republic), all settings full, and with 30-40 fps on average, and 25 fps+ on coruscant (coruscant is waaaaaaay too big). same for skyrim. i even have extra graphics mods on skyrim, fxaa injector etc (injected bloom into game) this that. so, top gpu of the existing generation (before any idiot jumps in to talk about 6990 being the top offering from ati ill let you know that 6990 is 2 6970s in crossfire, and 6950 gpu is just 6970 gpu with 38 or so shaders locked down via bios and underclocked - ALL are the same chip), is not only able to play the newest graphics-heavy games in max settings BUT also do it on 3 monitor eyefinity resolution. one word. consoles. optional word : retarding.

    --
    [url="http://www.rinconandroid.com"]aplicaciones android[/url]
  25. Mod Parent Up!! by Anonymous Coward · · Score: 0

    +1 Troll

  26. Here are over 53,000 by Anonymous Coward · · Score: 0

    Here are over 53,000, if you see no difference greater than 24 Hz there's something wrong with your brain.
    http://scholar.google.com/scholar?start=40&q=critical+flicker+frequency

  27. Here's something recent by Anonymous Coward · · Score: 0

    Is there an omitted stimulus response in the human cone flicker electroretinogram?
    J. Jason McAnany and Kenneth R. Alexander
    Visual Neuroscience (2009), 26 : pp 189-194
      "Omitting a stimulus from a t rain of repetitive stimuli, either by interrupting or by terminating the train, can elicit an electrophysiological response that occurs at the time appropriate for the omitted stimulus. This study investigated whether such an omitted stimulus response (OSR) is present in the flicker electroretinogram (ERG) of the human cone system. ERGs were recorded from 11 visually normal subjects in response to full-field sinusoidal flicker trains presented against a rod-desensitizing adapting field at frequencies ranging from 12.5 to 100 Hz. Recordings were synchronized with the onset of the stimulus trains, and the amplitude and relative delay of any additional ERG responses following the offset of the flicker train were analyzed. At stimulus frequencies below 35 Hz, the number of ERG responses always equaled the number of stimulus cycles. However, over the frequency range of 38.5 to 100 Hz, the ERG contained an extra response following flicker train offset. At stimulus frequencies from 38.5 to 62.5 Hz, there was a constant delay between the peak of the extra ERG response and the time at which the next stimulus would have occurred had the flicker train continued. This constant delay is characteristic of an OSR. In addition, an extra ERG response was apparent at these same stimulus frequencies if the flicker train was interrupted by omitting stimulus cycles from the middle of the train. The pattern of ERG findings is consistent with a recently proposed model of the OSR that attributes the phenomenon to a resonant oscillation in retinal bipolar cells."

  28. Crysis *is* an unoptimized mess. by Anonymous Coward · · Score: 1

    My high-end hardware doesn't run it right, and my hardware is better than yours and more expensive than yours.

    Likewise, ioDoom3 will be coming out soon with integrations from Doom3_Voodoo2 Patch and FastVoodoo so that you can see how shitty an engine truly is when all it's eye-candy are shutoff so that the Owner/player can experience the engine with the logical controls and quicker primitives. Whatever it takes to win is what proves how non-efficient an engine truly is, and Crysis is worse than Doom3.

  29. Nice to be able to turn it off by BlueCoder · · Score: 1

    I like the prospect of turning off the graphics card. I could use a KVM to switch to built in graphics. It's a shame they didn't actually put a low end chip with built in memory that could run on a couple watts. Seems short sighted to me.

  30. Re:This would be really cool...but no linux no fun by Anonymous Coward · · Score: 0

    its no good in linux though, they don't have OSS drivers for it, or even any hardware assisted high profile (L5.1) H.264 video encode/decode ,but then none of the AMD gfx chips from the whole last generation do that in Linux ether.

  31. Where have all the by uninformedLuddite · · Score: 1

    bitcoin comments gone?

    --
    The new right fascists are bilingual. They speak English and Bullshit.