NVIDIA GTX 295 Brings the Pain and Performance

Vigile writes "Dual-GPU graphics cards are all the rage and it was a pair of RV770 cores that AMD had to use in order to best the likes of NVIDIA's GeForce GTX 280. This time NVIDIA has the goal of taking back the performance crown and the GeForce GTX 295 essentially takes two of the GT200 GPUs used on the GTX 280, shrinks them from 65nm to 55nm, and puts them on a single card. The results are pretty impressive and the GTX 295 dominates in the benchmarks with a price tag of $499."

Comments

Holiday Shopping

[Szentigrade]

Just in time for holiday shopping!

Re:Holiday Shopping

[Anonymous Coward]

No, just too late. It will not be available until January 8th. My guess is that NVidia wants all you gamers to hold your chrismas pennies until that date.

So, no hardware in your stockings this year!

Re:Holiday Shopping

[Anonymous Coward]

So, no hardware in your stockings this year!

You have no idea how wrong that sounds.

Re:Holiday Shopping

[Anonymous Coward]

get your mind out of the gutter ...or Santa won't be sliding down your chimney this year.

Really, though.

[Anonymous Coward]

I don't think I (or anyone who isn't ludicrously rich) will have any need of buying this, at least until Crysis 3: The Crysising or whatever comes out. ...not that I wouldn't buy it if I had the money or anything.

Re:Really, though.

[FredFredrickson]

Have you tried GTA4? I get a nice 3fps with a Crysis killing machine. I think I'll need two of these in SLI.

Re:Really, though.

[terranwannabe]

GTA4 kills proc and not GPU so much - I have a GTX260 Core 216 and a Quad-core Q9550, and I get an easy 40 FPS with medium textures enabled. I hear it's supposed to go higher when Nvidia releases their new drivers. In any case, Steam Forums was having a shootout in many threads when GTA4 was released and there's a pretty clear correlation between processor power and higher framerates. It's also supposed to be 64-bit optimized so I might be getting a boost from running 64-bit Vista.

Re:Really, though.

[redscare2k4]

GTA4 is processor dependent, not GPU dependent, cos it's the crappiest console port we've seen in years.

Re:Really, though.

[Frools]

You get 25fps in GTA4 and 40fps in Crysis
Now, which one (given those numbers) would you expect to look better?
Which one actually looks (a lot) better?
Theres your problem.

Also it has a spectacular memory leak that sees it using up all available physical memory after a while, grinding to a halt and refusing to load any new textures, which is actually pretty funny the first time. Driving around in a flying car avoiding invisible buildings :D

Re:Really, though.

[poached]

the DRM in GTA 4 may be slowing it down (If you are using legit version, that is).

Re:Really, though.

[yoyhed]

Definitely. Runs pretty shitty here, with a 3.0ghz Core 2 Duo, GTX 260, 4GB DDR2-800. I hear the quad-core people have it good.

Re:Really, though.

[floodo1]

GTA4 isn't the only CPU dependent game out there. All the Valve fps games are massively CPU limited. Things like Left 4 Dead, TF2, half-life 2 etc

Re:Really, though.

[LatencyKills]

I've been working on a game review for GTA4 (shameless plug: game-over.net). My machine meets literally the minimum system requirements and I'm running it at a steady 50fps in 1024x768. Many, many people are reporting problems with GTA4, and there seems to be no rhyme nor reason to it. Smoking fast machines can't run it, my machine can. Rockstar really pushed this game to market well before it was ready.

Re:Really, though.

[Anonymous Coward]

[...] and their own physics driving around in crysis.

to be fair, I don't think video cards do much with regards to the physics. amiright?

Re:Really, though.

[Anonymous Coward]

Not the drm, got it cracked and a E8400, still dog slow.

Re:Really, though.

[lysergic.acid]

not quite. physics calculations are very well suited to vector processing, which is why most Physics Processing Units are vector processors, just like GPUs, array processors, DSPs, stream processors, etc.

a GPU may be more specialized to handle 2D/3D graphics rendering, but they can also perform physics calculations quite well because of their SIMD architecture. GPGPU stream processors in particular are of great interest to the scientific community because of their superb performance in scientific modeling applications which require crunching large sets of numbers.

of course, there are some vector processors that are better suited to physics calculations while others may have architectures & instruction sets specialized towards graphics or digital signal processing/video encoding, but sometimes the distinction is purely marketing-based. so if you had to choose between using a general-purpose scalar CPU or a vector GPU on a graphics card to perform physics calculations, it would be far wiser to use the GPU.

Re:Really, though.

[thepotoo]

Please cite that.

I'm running L4D on my (very) old computer, a 1.6 Ghz AMD single core with a 7600 GS and 1.5GB ram. The game runs fine at medium settings (despite the fact that I am way, way under the minimum system requirements), and when I briefly swapped out the 7600 for a 7900, I was able to turn a few of the settings from medium to high (1024x768, textures low, medium effects -> 1280x1024, textures medium, high effects) and still get a stable 20-25 average frame rate.

Not quality benchmarks, I know, but the engine hasn't changed drastically since HL2 except for graphical improvements (=GPU limited), so claims about it being CPU limited haven't been true since the first public version of the Source engine, and that's assuming they were even true back in 2004.

Re:Really, though.

[IorDMUX]

Though TF2, HL2, and the like may be CPU limited, the bottleneck is so low (think: early P4) that any decent gaming rig from the past few years can produce much better graphics quality and framerates on these Source titles than they can with other games out at the same time, such as Oblivion.

Re:Really, though.

[karstux]

People need to understand that sometimes, those detail sliders aren't meant to be cranked all the way up on current hardware. They're there to "future-proof" the game, so that it can still look pretty 2 or 5 years down the road. Wing Commander 4 did it for example.

Of course, it's not a huge incentive for developers to future-proof a game when all they get for it is a forum-bashing like "omg the game sux i can't get 50 fps on my 1337 rig".

Re:Really, though.

[billcopc]

I don't know very many people who play the latest games in 1024x768 anymore. Heck, I was playing Q3 that high in late '99. Today it's all about 1680x1050, 1920x1200... maybe a bunch of peeps on 1280x1024 laptops or CRTs.

Maybe I should try running Crysis in 320x240 for shits and giggles.

Re:Really, though.

[billcopc]

They do, ever since NVidia implemented the PhysX API in their GPU drivers (for GeForce 8 and up).

On my PC, the PhysX control panel lets me choose between the GeForce or an AGEIA card. They rolled this out a few months ago, works on XP and Vista as far as I know.

Screenshot

Re:Really, though.

[Spatial]

The DRM isn't the problem, I think. It's just that they rushed it out for the Christmas shopping period knowing people would buy it. There isn't a demo so you can't even try it first to see if you're in the apparent majority that has severe problems with the game.

Actually, I don't think Rockstar 'get' the PC platform anymore. They actively prevented people from changing the game files so mods are impossible without hacks. They put mechanisms in place to prevent trainers from working so you can't have fun with silly cheats. The game requires you to get a Games For Windows Live account, and a Rockstar Social Club account. The atrociously poor performance, poor performance scaling and awful configuration options are just the icing on the cake.

Re:Really, though.

[Applekid]

According to the Steam Hardware Survey last updated October this year, about a quarter of it's user base have a primary resolution of 1024 x 768. Surely there are some that set certain games to higher resolutions, but I would expect the percentage of those people in the 1024 x 768 bracket to be similar to the percentages in the other brackets. Except those at 1920 x 1200, naturally.

Re:Really, though.

[FredFredrickson]

I call BS I am beyond minimum requirements, I don't get near that on 800x600.

What are your specs?

Re:Really, though.

[UncleTogie]

I am beyond minimum requirements.

...but does your system meet the "Recommended" requirements?

Re:Really, though.

[Molochi]

HL2 based games only use a single core on my system and frequently don't even max that out. Hardly massive demand. TF2 runs well even on my 4 year old notebook.

Re:Really, though.

[robthebloke]

not quite. physics calculations are very well suited to vector processing

Correction, *some* physics calculations are well suited to vector processing, but not all.

Simple particle physics is trivial to throw at a vector processor. This is pretty much limited to adding a bit of prettyness to a scene. HavokFX is a really good example, and a lot of the hardware accelerated parts of PhysX implements this stuff (eg fluids, cloth etc). Since these are mainly visual effects, it makes a hell of a lot more sense to put this stuff onto a GPU rather than a PPU.

Basic primitive collision gets a bit of a speed kick in PhysX. There are however two problems that i've seen with this though. Firstly, take a look at the ageia hanger of doom demo to see how adding lots more physical boxes to a game, doesn't translate to better gameplay.

The second problem is far more severe. Attempt any kind of feedback in your simulation (character control for example), and you'll find that disabling hardware acceleration gives you significantly better performance.

So, if particle physics is better suited to the GPU; simulating loads of boxes isn't needed for better gameplay; and the kinds of physics that people want to see in games, is better off without a PPU; One could fairly easily argue that the ageia physX card is the first and last PPU we'll ever see....

Re:Really, though.

[Anonymous Coward]

I've been meaning to try it, but no, I haven't. I meet/beat the recommended requirements, but if your Crysis killer can't run it very well there's no chance that mine can (what's your CPU speed? I hear GTA4 is more of a drag on the CPU than the GPU.)

Re:Really, though.

[KovaaK]

I prefer 800x600 because my CRT has much higher refresh rates at that resolution. Even if I get solidly high FPS at 1680x1050, running the game at 60-75hz seems so jerky compared to the 160hz that my eyes are used to. That being said, I'm fully aware that I'm in the minority.

Re:Really, though.

[afidel]

The reason they don't allow modding is that whole "hot coffee" thing you might have heard about, you know the lawsuit against the company based on a third party mod that almost shut them down.

Re:Really, though.

[lysergic.acid]

it has absolutely nothing to do with whether something is a "visual effect" or not. the GPU does not care whether its calculations are for generating 3D graphics or a table of numbers. and physics problems definitely aren't limited to gaming.

the fact of the matter is, physics calculations generally involve applying the same operation to a large set of numbers, thus greatly benefit from data parallelism. that is why SIMD processors are perfect for physics calculations.

and the reason why Ageia PhysX chips are the only dedicated PPUs on the market is precisely because there's very little practical difference between a PPU and a GPU, which is why most PPUs are GPUs (and vice-versa). there just isn't a market for a dedicated PPU when your GPU's stream processors will do the exact same thing. the Cell processor's SPU stream processors are used for both graphics and physics calculations. likewise, the NVidia GPUs in next-gen graphics cards will likely all have PhysX support (even current gen graphics cards without the PhysX chip will soon have PhysX support through firmware upgrades).

and i really don't see what gameplay has to do with anything. whether a game is well designed has nothing to do with whether or not consumers have dedicated PPUs or not.

Re:Really, though.

[BlackSnake112]

Your eyes can tell the difference between 160hz and 100hz refresh rates? I thought the human eye topped out around 60hz-75hz. Your super human.

Re:Really, though.

[floodo1]

http://www.google.com/search?q=tf2+cpu+limited&ie=utf-8&oe=utf-8&aq=t

Basically if you had increased your processor from 1.6 to say 2.0ghz you'd notice a big difference. Performance scales with video card, but also significantly with CPU.

Most other games are limited by GPU much more and don't show the same scaling with CPU.

Re:Really, though.

[floodo1]

Just because it is playable doesn't mean that it doesn't scale with CPU power. Try running their games on a P4, then swap out only the processor and watch as your framerates go up significantly. Most other games are so GPU limited that they do not show this same scaling framerates in response to CPU change.

Seriously google it. Or just try it yourself.

Re:Really, though.

[KovaaK]

My super human indeed...

I actually didn't state that I can tell the difference between 160hz and 100hz, although I generally can. I always hear people say that human eyes can only see 30-60 frames per second, but they forget that there is a lot more going on for a computer than just the screen refresh rate. If the game is in a hectic area that requires more computation, FPS will be lower in the game, and if your FPS drops below 60 while your refresh rate is at 60, it will cause some frames to be shown more than once per refresh. The higher the refresh rate, the less likely you are to notice such events.

Also, if your refresh rate is higher and you aren't using "wait for vertical sync", the your mouse inputs will make a difference on the screen quicker. I also happen to run my mouse at 1000hz...

Once again, I'm fully aware that I'm in the minority.

ugh

[larry bagina]

this is like the razor wars (double blade! triple blade! quad blade! pento blade!). With OpenCL and DirectGPU (or whatever MS is calling it this week), this should be good for anyone trying to build a cheap superGPU cluster.

Re:ugh

[Spazztastic]

this is like the razor wars (double blade! triple blade! quad blade! pento blade!).

Clearly you don't value a close shave.

Re:ugh

[Windows_NT]

Its pretty sweet what they can do with these GPUs these days. If i had $499 sitting around, id go buy one. But first id have to know ... how do the graphics look with xfce4? I dont want my AMOR guys looking pixelated! id make a pun about if it can run aero .. oops to late ;)

Re:ugh

[zehaeva]

maybe he's a software engineer who wants to be successful like a few of these gentlemen?

Re:ugh

[nbert]

At least those pento blades don't consume a whooping 289 W while you use them. Somehow most previews don't even mention power consumption. The author of the article linked above actually states about this card, that it "properly balances GPU performance and power efficiency". By that logic everything which does not start to burn is power efficient...

Re:ugh

[Anonymous Coward]

obviously cutting edge tech.

Re:ugh

[ArsonSmith]

Somewhere there is a real press release that is of this same tone from Gillette. Of course there was less cursing and such, but almost as funny. I have a feeling this was written as an exaggeration of the real one.

Re:ugh

[ThePhilips]

Somehow most previews don't even mention power consumption.

Had you RTFA properly, folks have mentioned that card is not yet officially out and nVidia asked to withheld further details as BIOS might still get tweaked.

By that logic everything which does not start to burn is power efficient...

This is not an absolute metric (or is it "yardic" in US?). I presume they compare it to 4870 on which the infamous DDR5 alone - even when idle - draws whooping 40W. 4870x2 has already tweaked factory BIOS and yet twice more DDR5 still consumes same 40W. Yes - RAM alone consumes 40W.

Re:ugh

[Mathness]

I like A Close Shave. But I still think The Wrong Trousers is the best of the films with Wallace & Gromit.

Re:ugh

[yoyhed]

I'm still using a Mach 3, after all these years. I refuse to upgrade to these newfangled quad- and penta-blade razors.

Re:ugh

[windsurfer619]

Just for comparison, my whole laptop, with brightness on full and high CPU usage, doesn't even consume 15 watts.

Re:ugh

[Cynonamous Anoward]

Ah, 4 blades is pointless, but the first 2 additional blades made perfect sense to you, did they? nice logic on that one...

Re:ugh

[Wintergr33n]

Yes in fact here's the list of the requirements nVidia insisted on for all the sites that got to take a first look at this early sample, which is why you'll notice they all review the same games and why none of them mention power, noise etc...

Top 5 Games benchmarks only

Plus one other title of their choosing

No other benchmarks period

No acoustic measurements

No power measurements

No PhysX testing or benchmarks

This is as stated by one of those sites that got a sneak peek

Re:ugh

[billcopc]

Well for one, this is a "preview", which usually means NVidia sent a bunch of propaganda for them to disseminate to the masses. These things come with NDAs and a variety of other restrictions, in exchange for getting the "scoop" on the new product, and raking in a few extra yen from the ad networks.

You will get a real review once the product hits the shelves, and a real person performs real tests. Anything else should be taken with a grain of salt, as most "review" sites exist to make money first and foremost, with the factual content being only second in importance.

Re:ugh

[BOFHelsinki]

I would laugh but I'm crying because I just saw a Braun electric razor ad built around the highly topical concept of "Magic Christmas". Clearly, a manual razor cannot provide Magic Christmas like a Braun can.

(I mean, really, is Ol' Santa the best figurehead they could find for razors?)

Re:ugh

[COMON$]

Most slashdotters cant grow pubic hair you insensitive clod!

Re:ugh

[Khyber]

Why are the engineering samples allowed out for advertising when you know that more than likely the release product will never perform as well as the engineering samples? Doesn't this constitute false advertising?

Re:ugh

[Anonymous Coward]

What's with all the whooping? Does one of you have a cough?

Re:ugh

[yoyhed]

Have you ever tried a one-blade razor?

Re:ugh

[Anonymous Coward]

Diminishing returns my friend, diminishing returns.

Re:ugh

[dave562]

I know I shouldn't feed the trolls, but... It comes down to diminishing returns. The difference between a single blade disposible razor and a Mach 3 with three blades and those nifty strips is obvious. The difference between three blades and four blades with all other factors being equal is pretty questionable.

Re:ugh

[abc_los]

At least check out the Mach 20 before you judge the razor wars.

Re:ugh

[Repossessed]

I very much doubt that.

My server is built from top to bottom with high efficiency parts, and it still consumes 40 watts. I could cut it down to 33 if I took out the backup hard drive.

Though it occurs to me that I'm measuring peak consumption, and you may be claiming the idle power?

Re:ugh

[Valdrax]

Clearly you don't value a close shave.

That's the ticket! My desktop isn't ancient and decrepit. I'm just rugged.

Re:ugh

[p!ngu]

I personally use a one blade razor (a safety razor), and find the performance well and truly above that of a Mach3. Plus, I bought 500 blades for 100AUD, so I'm set for...a while.

Re:ugh

[dave562]

To each their own. A buddy of mine goes to the barber shop three times a week to get his shave. Go figure.

Re:ugh

[windsurfer619]

Well, I just did a test, and ACPI reports a 15.2 watt draw when running two "yes" processes on my dual core. My (relatively crude) AC power meter measures around 19 watts on my power adapter when my laptop's battery is removed.

Idle can go as low as 9 watts, depending on how much I want to shut off. 12 watts if I leave peripherals on.

I am running an Intel Core Duo at 1.6 GHz, with a 945gm integrated vid card. Bluetooth is on. USB is not suspended.

Re:ugh

[Anonymous Coward]

don't consume a whooping 289 W

draws whooping 40W

I'd get those coughs looked at if I were you...

"whopping"... not "whooping"

Main Entry:
        whopÂping Listen to the pronunciation of whopping
Pronunciation:
        \ËhwÃ-piÅ, ËwÃ-\
Function:
        adjective
Date:
        circa 1625

: extremely large

Re:ugh

[MonsterTrimble]

Hate to tell you, but Mach 3 is at EOL. I couldn't find Mach 3 blades anymore and thinking 'well, why would they swap handle designs?', picked up a pack of fusion blades for my Mach 3 handle. Of course, they changed handle designs.

I hate planned obsolescence.

Re:ugh

[billcopc]

No, this constitutes Business As Usual (tm).

As for the performance, the biggest difference is usually overclocking capabilit. The review samples always shoot for the moon while the retail product falls a little short, but do you really care about hitting 850mhz vs 862mhz on that NVatia Gedeon N+1 ?

And three years from now...

[Anonymous Coward]

Its just another card for $50 at a yard sale. Seems like such a lot of effort too.

Impressive Card

[Stereoface]

I might consider upgrading from my 2MB VGA after seeing it in action... :)

Re:Impressive Card

[ServerIrv]

I remember when my friends at school called me a liar because I said our graphics card had 4MB of RAM. It was a sweet ATI VLB card for my dad's architecture workstation.

Re:Impressive Card

[meringuoid]

I might consider upgrading from my 2MB VGA after seeing it in action... :)

And what exactly is wrong with 1024x768x24bpp?

(No, seriously. I used that very setup from 1995 to 2001. Then I got a PCI Voodoo4 cheap because 3dfx had just gone bust, and then got drunk one night and bought a 19" CRT on an auction site, and discovered the joy of 1600x1200.)

Re:Impressive Card

[Stereoface]

Well these days streaming media is more 'girthy' then it used to be. Especially when you have multiple tabs open each containing a 'part' of a tv show or some erotic presentation. I mean these cards are geared for gamer world- intensive graphics and real time rendering, but the bare OS needs some power as well. Vista is a fatty if you leave the whizz bang graphical stuff enabled. I make XP look like 98, nice and gray and sexy. It really inspires work flow...

Re:Impressive Card

[smoker2]

Unfortunately, you have to spend a good deal more to get a TFT that can handle 1600x1200. I "upgraded" from a 19" CRT to a 19" tft and am stuck at 1024 (at 32bit admittedly). But the tft cost more than the CRT at the time. Good old Dell.

Re:Impressive Card

[Anonymous Coward]

You haven't looked at modern TFT pricing in the last two years then. Right now you can get bargain brand 19" TFT's with 1600-1920xsomething resolution for under 150 dollars, and name brand stuff for between 200 and 250 dollars.

Re:Impressive Card

[Anonymous Coward]

1600 x 1200 on a 19" CRT?
Hope it did more than 75hz and you weren't using a distro set at 92dpi!

People used to ask "How can you use 1280 x 960 x 85hz on a 19 CRT"? that'd make be blind!"

and now I have the decadant luxury of a 24" TFT :P prices are cheap, these days

Re:Impressive Card

[Valdrax]

I remember when my friends at school called me a liar because I said our graphics card had 4MB of RAM. It was a sweet ATI VLB card for my dad's architecture workstation.

I remember when my roommate was working for a server middleware company and still had a 486 desktop. He worked with network cards that had a better processor and more RAM than his desktop at home. Good old Arctic 960! Ah, nostalgia.

Re:Impressive Card

[Anonymous Coward]

I think you should get drunk more often.

Pain

[Anonymous Coward]

NVIDIA GTX 295 Brings the Pain and Performance

A fight to the death? No! A fight to the pain!

Anonymous Coward

[Anonymous Coward]

Tell me when they have the fastest graphics card with passive cooling with ~50W max in 3D, at most 10W in 2D mode for less than 150â/180US$.

Blarg.

[Anonymous Coward]

I hope these won't die in 4 months, like many other Nvidia products.

Re:Blarg.

[Khyber]

I don't see why this is modded flamebait. The GeForce 8 and some of the GeForce 9 series have this exact issue of failing because of faulty die packaging, it's been around for almost a year, and it still hasn't been fixed.

I hope that these cards don't die. Mine died within three weeks and forced me to send my laptop in for repair.

i hate fans

[marcuz]

how do you people buying this stuff get over the noise of the fans? and how often do you have to exchange the fans? i buy only hardware with speed-to-noise ratio near infinity.

Re:i hate fans

[Anonymous Coward]

These are targeted at the high end gamers, most of which will have headphones on blocking out their computers noise.

I used to hate fan noise too, but I find that the lower I get the noise from my computer the more I'm aware of every other noise in and remotely near my house. Until someone invents a usb powered Cone Of Silence, I'll take a little fan whir and put on my closed ear sennheisers to muffle it all out.

Re:i hate fans

[gEvil (beta)]

how do you people buying this stuff get over the noise of the fans? and how often do you have to exchange the fans? i buy only hardware with speed-to-noise ratio near infinity.

Easy--they're deaf. After years of working on building (near) silent PCs, I've learned that what many people/reviewers consider to be 'quiet' is nowhere near my definition of 'quiet'. I'm not quite sure how loud some gamers have their sound systems turned up, or if they play with the window open or what, but I simply can't trust a review on newegg or most websites when someone says a piece of equipment is 'silent'. There are a few websites like silentpcreview.com that do a good job, but if a piece of equipment isn't reviewed there or in the forums, you're SOL (or you get to be the guinea pig).

Re:i hate fans

[pohl]

I'm with you on this one. The only card I'd be interested in seeing the benchmarks on is the fastest one with a mere heatsink.

Re:i hate fans

[ServerIrv]

The same people that I watch movies with. I will have the tv volume at about 1-2 (out of 10), and inevitably, one of my friends will join me and have to turn up the volume because they cannot hear it.

Re:i hate fans

[ArsonSmith]

Equal amount spend on the audio system.

Reminds me of an old joke, "I spent $500 to get my muffler fixed on my car, but now I can hear all these other things that are wrong and it's going to cost me another $500 to get louder speakers."

Re:i hate fans

[Anonymous Coward]

No, it's because it's "white noise." Which always sounds quieter than it really is. until, of course, you sit near it for a few hours. Then you either tune it out or tel yourself, "damn, I gotta tune this out."

But the problem is that since they didn't notice the noise when you bought it (masked by ambient at the store, or the white noise phenomenon), they think they *have* a quiet PC and are just being whiny.

Re:i hate fans

[nuclearpenguins]

Some of us like having our systems sound like bad motel air conditioners you insensitive clod.

Re:i hate fans

[yoyhed]

Like this?

Re:i hate fans

[amoeba1911]

The fan is controlled by a thermal sensor so it's not running full speed all the time and that means it's fairly quiet most of the time. It's not much louder than a standard power supply fan.

Re:i hate fans

[Kjella]

Or maybe you have the "low-fat chips" phenomenon. Yes, it's somewhat less fat than regular chips but they're still chips, not celery roots and nowhere near healthy. A CrossFire/SLI paired with a high-end CPU and maybe overclocked too have only two options, loud and really loud. Wattage and DB ratings are a much better bet, and "fanless" is the magic word.

Re:i hate fans

[Anonymous Coward]

This is the card for me. And even if it is not as fast as the new nvidias, its just the matter of time. i am looking forward to see this new nvidia gpu series shrink in size so that they could be cooled passively.

but ok, i am not a hc gamer and i use my pc in a living room as a media center.

Re:i hate fans

[Anonymous Coward]

you can have your system almost absolutely silent if you buy the right things. buy only passively cooled components, ad 2 very silent fans (in/out) and make them turn even slower, put your new and silent harddrive in som soundproof case.
if you have those 2 fans quiet, the whole pc would be very quiet. and by that i mean you would not hear it from 5 feet.

Re:i hate fans

[Fallingcow]

Don't the heatsink-only ones rely on better-than-average airflow through the case, which means more and/or stronger case fans?

Re:i hate fans

[Anonymous Coward]

...I've learned that what many people/reviewers consider to be 'quiet' is nowhere near my definition of 'quiet'.

Absolutely. After years of dealing with noise pollution I just can't stand it anymore. I often use earplugs, and I wish there was some way to turn my hearing on and off.
Though, it seems to be worst with certain frequencies and irregularity of the sounds. Small diameter fans have a higher pitch than larger diameter ones, and large ones are less irritating. Hearing somebody's music through a wall is right up there with nails across a chalkboard.

Without fans in the computer, I'd probably notice the harddrives. This coming year, I'll make something to connect the interface of my computer with a 100ft optical cable, and stick the computer in a different room. Most of the interface doesn't make noise, and the machine doesn't need to be sitting 10ft away. At least then, I'll only hear the monitor whistling... I'm an embedded developer, and I have some of the components lying around already. I'll toss the schematic for it on the net when it's done (named project optizon).

Re:i hate fans

[networkBoy]

or you go for relatively high flow (I use a squirrel cage blower in the PC for exhaust) puss the exhaust air through a muffler, and dynamat the inside of the PC for noise abatement. Works a treat. All intake is passive, no fans, air comes in via the PSU and a vent in front of the HDDs. The air is then routed through the case and over components via ductwork made from plexi and terminates at the blower intake. overall the machine runs at a fairly substantial negative pressure, and very quiet.
-nB

Re:i hate fans

[PJ1216]

I have very few fans in my computer, but I do have watercooling which is much quieter (though obviously, still has fans). However, with people who work around these computers a lot, it almost becomes background noise. I need to turn on a fan at night because its too silent when I turn off my computer.

Re:i hate fans

[Spatial]

It's actually a pretty fast card. About the same as an 8800GTX, which will run any game you care for at quite reasonable settings and resolutions.

Re:i hate fans

[Anonymous Coward]

My 4830 was ridiculously loud, but I hooked it up to a fan controller and got it to be quiet. It is still the loudest component in my system, but it is a tolerable hum. I adjusted the fan just right to avoid any of the weird grinding or whiny noises you get sometimes. Of course, now it hits 75C on load instead of 55C, but I really couldn't care less. It is designed to run at over 100C anyway. Turning up the fan doesn't allow me to overclock it any further and I never get crashes or anything like that. I don't see why they ship these cards with these terrible fans.

By the way, for thirty bucks you can get a fanless heatsink that will cool a 4850. You can attach silent fans to it if you want to lower your temps or if you have bad airflow in your case. Arctic Colling S1 Rev2.

Re:i hate fans

[billcopc]

I've yet to find a squirrel cage blower that does not make an absurd racket at all speeds. Do tell where you find yours, as there are many applications where such a blower would be far more appropriate than conventional fans.

Re:i hate fans

[gEvil (beta)]

I got a pretty good deal on a 3850 a while ago. The reviews (real reviews and customer reviews) consistently said that it had a fairly quiet fan on it, but there were a fair number of people saying it wasn't really quiet at all. I got mine and I have no idea how anyone can consider that shrill whine to be 'quiet'. I tried cutting the fan down to 15-20% to see how that was, and surprisingly, that made the fan spin faster (I guess it was at ~10% by default). I ended up ordering an Arctic Cooling Accelero S1, which has been absolutely wonderful (my older passive VGA cooler wouldn't mount on the 3850). I've used a few of their case fans and proc fans and I can't recommend their stuff enough. Very QUIET high quality parts for fairly cheap. Oh yeah, just the cooler running passively has lower temps than the stock cooling fan, so you don't even need the fan add-on most of the time.

Re:i hate fans

[Tetsujin]

how do you people buying this stuff get over the noise of the fans? and how often do you have to exchange the fans?
i buy only hardware with speed-to-noise ratio near infinity.

Well, the Sox fans can be a real nuisance around here... But I hear they're nothing compared to Packers fans...

Re:i hate fans

[Khyber]

get one that doesn't use bearings, they're almost dead-silent minus the light whir that you can't hear after you're about three feet away.

I use them for providing airflow over my indoor garden of hot thai peppers and catnip and lemon cucumbers. Works like a charm.

Re:i hate fans

[Sentry21]

It's all relative. Our office is quiet until the HVAC turns off, at which point you suddenly realize how loud it was by how quiet it suddenly gets.

Re:i hate fans

[Anonymous Coward]

This is easily one of the top three reasons I only buy computers from Apple.
I've owned or worked on all of the machines from the Intel era except the Air, and even the Mac Pro is virtually silent, though that one does depend a bit on which harddrives you're using.
Also, if you update firmware on it, any room smaller than 30sqm will sound like a loaded datacenter -- but I believe that happened to me three times in total.

Mac mini? Well, when I boot it and it checks to see if there's anything in the DVD drive, it makes a bit of noise. Playing HD movies from disk or over the network (yes, my old 1.66GHz dual core, 1GB mini plays 1080p without a problem).

Macbook? I'm normally awake in the middle of the night, so there's not a lot of ambient noise competing with it.
So long as the vents aren't covered by anything, I can't hear it.

iMac? Sadly, I have one of the 24" ones where the display hums depending on brightness -- other than that? Noise free, though the attached external drives do click every once in a while.

Mac Pro? See above.
When we unboxed it at the office, everyone complained about the noise it was going to make.
Plugged it in, clicked the powerbutton and people though it was DOA. Until the classic Apple boot sound came on, and they were wondering how it could be more quiet than their dell 620s, dell optiplexes and whatnot.
Of course, they were going "holy fuck" when I installed the first firmware update ;)

MacBook Pro?
Pretty much the same as the Macbook.

I'm looking forward to see the updated Mini and iMac at Macworld, and hopefully an updated Mac Pro as well.
I'm thinking of getting a Mac Pro at home for photo editing.

I could easily build a 99% compatible Hackintosh, but that comes at a cost. The noise level would be unacceptable.

Re:i hate fans

[Anonymous Coward]

Easy--they're deaf. After years of working on building (near) silent PCs, I've learned that what many people/reviewers consider to be 'quiet' is nowhere near my definition of 'quiet'.

Sing it brother Evil! Sheesh.

I ordered this 9800GTX from the egg because of the numerous reviews stating how quiet it was. Bullshit. The thing generated a faint high pitched whine that about drove me insane. Back it went.

I ended up with this 9600GT since it's passively cooled. Yes, I dropped some performance... but it's serviceable and quiet. When I'm not attempting to be a one-man pwnzer division online I tend to program and work with network simulations. Excess noise is unacceptable.

Re:i hate fans

[default luser]

I hate noisy fans too, so for years I've simply been buying the card and replacing the cooler. Got a vf700 for my GeForce 6600GT, and a vf900 for my 7900GT (that fan was at %100 all the time!). But unfortunately, that voids the warranty of the card.

I recently purchased an HD 4850, and I've tried using the stock fan, and here is my impressions: with a properly-ventilated case, the idle fan speed of the 4850 is %10 or less, which is almost inaudible. It gets a bit noisier in the summer when the house temperature sometime tops 27C, and even though this accounts for only a few months out of the year, I'm still considering a third-party heatsink.

That said, if you want similar performance with much less idle power consumption, look to the HD 4830: it uses less than half the power of the 4850 at idle, so the fan never spins-up.

Re:i hate fans

[Rudeboy777]

Why settle for a ratio near infinity when it can BE infinity? Video cards with fans are for chumps only. Vote with your wallet.

http://www.newegg.com/Product/Product.aspx?Item=N82E16814125097

http://www.newegg.com/Product/Product.aspx?Item=N82E16814134050
(Look at the size of the heatsink on this motherfucker!)

http://www.newegg.com/Product/Product.aspx?Item=N82E16814161266
(A bit more elegant than the last)

I agree. Bring on the Pain

[BearGrylls]

A helathy does of competition in this market is going to drive the efficiency and performance up and at the same time keep the prices down to a reasonable degree. Now if only AMD would man up and make a core i7 killer then the pain will have been brought.

Prob. in order 4x RV770 from ati ?

[unity100]

2 years later, and we will be able to saddle a graphics card and fly with it in skies.

Re:Prob. in order 4x RV770 from ati ?

[lymond01]

All I know is that my graphics box (I call it a graphical) houses a nice little motherboard with a cute Intel chip, some hard drives, and I think I even have a sound card plugged into it.

I remember when the graphics "card" was simply part of the computer -- these days, all the other components are part of my graphical.

It's great that there's a market for this stuff...

[JoeMerchant]

I'm glad that people are out there buying graphics cards that can render the latest games in QuadHD resolution at 120 frames per second... it makes the integrated graphics in eee class PCs that much better when the tech trickles down 5 years later.

Drivers drivers drivers

[je ne sais quoi]

As a somewhat mystified recent purchaser of a GTK 260 from eVGA, I was amazed to discover that NVIDIA has such problems with their linux drivers. I owned one of their older cards before and built a new computer and thought it was a no-brainer to pick NVIDIA for linux (freedom issues are notwithstanding, but I decided to go with the pragmatic choice). Only after I ran afoul of the powermizer slow switching crap, or other weird issues such as the misreporting of the screen refresh frequency, did I start digging and realized how many problems there are. As it is, I've got the beta 180.16 driver installed and it's better but I still had to do some tricks to shut off the powermizer feature. Just this morning had some other weird problem with screen corruption that's never happened before with my old hardware but more or less the same software on top of it.

For me personally, I could care less if the card hardware is great if the drivers suck. NVIDIA, fix your linux drivers please. Next time I'll give a much harder look at amd.

Re:Drivers drivers drivers

[pak9rabid]

Next time I'll give a much harder look at amd.

I'll save you 5 minutes of research....stick with nVidia.

But in all seriousness, I agred with your point. It seems like their Linux drivers have taken a shit compared to previous releases. Personally for me, I have a lot of artifacting issues in KDE4 that are apparently related directly to nVidia's drivers from what I've read.

Re:Drivers drivers drivers

[DoofusOfDeath]

I've still been buying nVidia for my Linux boxes. Is ATI finally a better choice?

Re:Drivers drivers drivers

[Anonymous Coward]

unfortunately AMD's Linux drivers are actually worse... OpenGL + Pretty candy desktop = flicker flicker flicker... don't even ask about games.. bleh.

Re:Drivers drivers drivers

[DoofusOfDeath]

But in all seriousness, I agred with your point. It seems like their Linux drivers have taken a shit compared to previous releases. Personally for me, I have a lot of artifacting issues in KDE4 that are apparently related directly to nVidia's drivers from what I've read.

Performance issues, by any chance? I've been baffled why my KDE 4 performance is so terrible compared to my Gnome performance, and I have a reasonable nVidia notebook chip (Quadro NVS 140M).

A lot of the forums have similar complaints, but most people seem to indicate that their problems went away with the 177.80 drivers, which I have installed.

I was hoping the forthcoming nVidia driver will help, but from how people are talking, I've got to wonder if it's even wise to install it when it's released.

Re:Drivers drivers drivers

[Anonymous Coward]

Switch to Windows!

- Sincerely,

Bill.

Re:Drivers drivers drivers

[cecom]

I don't know about ATI, but NVidia is a _terrible_ choice! I also used to buy NVidia for all my Linux boxes and recommend it to everybody. I was wrong. The problem is that the free NVidia drivers are extremely slow. They are actually slower than using the generic VESA driver with Intel graphics.

I don't know why - I suspect it has something to do with reading from display memory - but it is a fact. I have a relatively fast quadcore machine, and yet when I am using the free NV drivers, it is unusable for Internet browsing. Scrolling of some websites is impossible (mostly ones who have a patterned background). Flash is about a magnitude slower for many things. Lots of the snazzy JavaScript demos are too slow to be impressive.

As soon as I installed the nonfree drivers everything improved drastically. However I am using 64-bit kernel with 32-bit userspace (this is the most reasonable choice if you have more than 1 GB of RAM), and the non-free drivers don't support that configuration...

So, short term I recommend only Intel. Given that ATI released some documentation recently, I recommend ATI long term. Sorry, NVidia, we had a nice run, but until you release detailed documentation, you will stay off all my machines.

Re:Drivers drivers drivers

[Kjella]

Next time I'll give a much harder look at amd.

A good advice - don't. Maybe when their open source driver rolls around but the fglrx driver has a lot of really bad issues, take a look at the Phoronix forums whenever a new AMD driver rolls around. Personally I run KDE3 on a 8800 GTS and all the 3D acceleration and whatnot is working just fine for me. No tearing in video (a problem that seems to plague AMD users endlessly), but then I use it in a desktop machine.

Re:Drivers drivers drivers

[Anonymous Coward]

(this is the most reasonable choice if you have more than 1 GB of RAM),

[citation needed]

and the non-free drivers don't support that configuration...

The *installer* doesn't support that configuration, but it's pretty easy to unpack the driver package and install it yourself. It works fine.

Re:Drivers drivers drivers

[DoofusOfDeath]

Generally what I've done is preferred Intel for my work laptops, because I wanted to good quality of the Intel open source drivers.

But on my home computers, which I sometimes use for gaming in Linux and/or Windows, it's really a choice between nVidia and ATI. I don't care much wether I use the open or closed source drivers on those cards, as long as they work well.

The problem for now seems to be that at least in KDE 4, nVidia's closed drivers aren't a good match for the implementation of KDE 4, but ATI is okay.

That's a bit of a reversal from past years when the ATI linux drivers (both open and closed) were far behind the functionality and speed of nVidia's closed drivers. So the question in my mind, when buying new hardware, is whether ATI's current lead in driver quality is a long-term trend as well.

Re:Drivers drivers drivers

[Splab]

Bad advice, the new AMD cards runs fine under Linux, since radeon 9600 I've never had any problems getting ATI/AMD to run under Linux.

Re:Drivers drivers drivers

[je ne sais quoi]

Hah ha, whoops -- that should be GTX 260. Too much time using gtk apps. :)

Re:Drivers drivers drivers

[Chandon Seldon]

So... you're recommending against ATI without having actually used it?

I'm currently using Nvidia (on price), but I was on an ATI card 6 months ago and it worked fine. The Nvidia drivers were slightly better when I switched, but only very slightly.

Re:Drivers drivers drivers

[Anonymous Coward]

My Radeon 9600xt caused me to buy two nvidia cards in a row because of how terrible the drivers were(the card itself was amazing though, quiet, efficient, fast and cheap. $150 including Half Life 2 was a pretty good deal at the time).

I was very happy with my 6600GT and 7800GTX though, which worked perfectly under linux. Last week I bought an ATI 4830 though and it also works just fine in Ubuntu. I didn't even have to touch anything, I just booted up and it ran off the drivers that I was using for my integrated 780G graphics. Windows Vista needed to uninstall the old drivers and then install new ones. Windows XP just put up a blank screen and I had to uninstall the old drivers in Safe Mode.

Re:Drivers drivers drivers

[Mad Merlin]

However I am using 64-bit kernel with 32-bit userspace (this is the most reasonable choice if you have more than 1 GB of RAM)

I'm sorry... what? I don't see how that's a reasonable choice at all.

Re:Drivers drivers drivers

[cecom]

I'm sorry... what? I don't see how that's a reasonable choice at all.

The 32-bit Linux kernel cannot efficiently handle more than 1 GB RAM. Linus Torvalds himself is outspoken on this subject and recommends using a 64-bit kernel with more than 1 GB (alas, I can't find the reference right now, but he has written many times on the subject in RealWorledTech.com's forums).

On the other hand, using 64-bit in userspace is pointless for most users. Most applications don't need more than 3GB of RAM. Also, there are binary plugins, etc, which only work in 32-bit.

64-bit kernel and 32-bit userspace is the most efficient and practical choice for the vast majority of users.

Re:Drivers drivers drivers

[pak9rabid]

God, don't even get me started on the performance issues. I still have a hard time understanding how the nVidia drivers can affect the performance in KDE4, but they apparently do. nVidia's acknowledged it and have even released updated drivers to tackle the issue (they even say 'KDE4' in their changelogs). I really hope they get their shit together soon, as I'm really liking KDE4.

Re:Drivers drivers drivers

[dave420]

Oh noes! If all linux desktop users don't buy NVIDIA, they've lost a few percent of the market! Sorry for the sarcasm, but seriously - maybe if more people used linux on the desktop, the drivers would be more of an issue to NVIDIA. We can't really expect a company to spend more money on developing and supporting drivers than they'll make back from selling hardware that will use them - it would be nice, but it would be a gift to the community, not something they *should* do. And they might not be able to simply open-source them and let folks have their way with the source because of licensing restrictions, which regardless of how anyone feels about them, still matter a lot to any company that has to use someone else's licensed technology.

Re:Drivers drivers drivers

[Anonymous Coward]

So... you're recommending against ATI without having actually used it?

For me, I recommend against ATI for high-end cards only. I have only ATI cards myself, so I can't compare with NVidia.

Here are my reasons:
- fglrx driver still hangs the system on some AGP setups. For my 9550, it can hang when I simply move the mouse from one display to the other.
- search for "checkerboard of death". Since catalyst 8.5, certain full-screen applications (Wine, MythTV) cause the fglrx driver to display only garbage. It is still not solved in 8.8

I have long since given up on the closed-source ATI drivers. I tried them once again half a year ago, to support my HD4850. That was the last time I will even try them.

On the other hand, the open-source drivers work perfectly. They support all non-3D operations and can run compiz and Xv without issues. If you're lucky enough, fglrx might work for you as well but I would not recommend any high-end ATI card for use under Linux.

Re:Drivers drivers drivers

[__aabvlw4075]

I think you mean that you couldn't care less, not that you could.

Sorry to be anal, but that's one error that's not going to catch on and become standard if I can help it.

It's great that there's money for this stuff...

[Ostracus]

It's not just about the graphics. GPUs are being called upon to do much more, from AI to Physics, to folding@home. Even encoding and decoding audio and video formats.

Re:It's great that there's money for this stuff...

[Anonymous Coward]

Does helping folding@home make that portion of your power bill tax deductible? Otherwise.. who cares? Donate to a non-profit science foundation.

Re:It's great that there's money for this stuff...

[LotsOfPhil]

Does helping folding@home make that portion of your power bill tax deductible? Otherwise.. who cares? Donate to a non-profit science foundation.

You are also kind of donating the hardware, which is a much bigger cost than the power. $10 worth of electricity will do more of these calcs than a $10 donation would enable.

Re:It's great that there's money for this stuff...

[JoeMerchant]

As a programmer who does a great deal of data crunching, I sincerely hope that Intel's 80+ core CPU comes along quickly to crush the silliness out of people who are trying to find applications for GPU "cores."

Re:It's great that there's money for this stuff...

[Ostracus]

In order to do that Intel would have to adopt some of the architecture that is a GPU. Hence become the very thing you dislike. Right now GPUs are here and being used. Your fictional core isn't and with present limitations most likely will not be showing up for some time.

Re:It's great that there's money for this stuff...

[vux984]

You are also kind of donating the hardware, which is a much bigger cost than the power. $10 worth of electricity will do more of these calcs than a $10 donation would enable.

Except donating 10$ cash being a tax deduction costs you less than $10 electricity which isn't. So you could donate say, $14 cash and have it cost the same $10 electricity.

Further, that's not $10/year. For a high end GPU running folding at home, you are easily pushing 300 Watts. In most places doing that 24x7 will run you upwards of $25/mo. ($300/year)

I suspect if everyone participating in F@H donated what it cost them in electricity, the MILLIONS that would result in could be put to better use. But that's me personally... if your into F@H, have fun with it, its your money. But you really should know what its costing you. And if you are running F@H on someone elses dime (e.g. your running it at work or in your mom's basement) you probably should stop.

It runs most people over $100 year. If you've got SLI/crossfire and a playstation 3... and live in an expensive state like New York or California you could be pushing $1000/year.

Re:It's great that there's money for this stuff...

[jgtg32a]

http://www.ddj.com/architect/207200659

Re:It's great that there's money for this stuff...

[caerwyn]

Take a look at the latest issue of Computer- there's an interesting (if technically fairly light) look at performance ratios of various multicore designs- the present "cpu surrounded by lots of simple processing cores design, as implemented by a CPU + GPU combination, turns out to have the best performance/space and performance/watt ratios.

Sure, there are some situations where you can do massive parallelization but each individual thread needs the flexibility of a full CPU, but there are at least as many, if not more, situations where most or all of those threads just need simple serial execution over the data. The current model does wonderfully for that.

I say this, btw, as a programmer who does a great deal of data crunching. I'm very much looking forward to OpenCL's widespread availability, especially under linux- I've got a system in place now that could greatly benefit from it.

Re:It's great that there's money for this stuff...

[JoeMerchant]

I signed up for CUDA when it launched (early '07?), due to conditional execution, only about 20% of what we were crunching at the time would have benefited. We were far better off optimizing and parallelizing to take advantage of 8 slightly memory bound cores that could actually do a conditional execution.

Re:It's great that there's money for this stuff...

[JoeMerchant]

All depends on your app... if it's Crysis, these things rock.

Re:It's great that there's money for this stuff...

[raftpeople]

For me it was the opposite, I started with the cell but because the SIMD requires the operands to be adjacent to really gain something (I could have loaded from multiple locations and then moved them all into the 128 bit reg but by then gains would be lost), I switched to cuda and it's working although I do have to jump through hoops on the conditionals to make sure the threads are staying in sync with their instructions.

Re:It's great that there's money for this stuff...

[raftpeople]

By "silliness" do you mean the fact that my first pass (simple and unoptimized) at running my calcs on GPU gave me a 10x speed up (including memory xfer)? I can get through thousands of generations of my simulation in hours instead of days.

If intel brings out their 80 core proc I will be one of the first to compare to gpu to see which I should continue to use, until then I will use my gpu for real performance gains.

Re:It's great that there's money for this stuff...

[lysergic.acid]

yea, and that general-purpose CPU will also consume about 80x the power needed by a vector processor that is twice as fast at performing: 2D/3D rendering, scientific modeling, financial calculations, video transcoding, data compression, and all the other tasks that GPGPU/stream processors are used for.

a general-purpose scalar CPU will never replace specialized vector co-processors because it's the wrong tool for the job. most applications that vector coprocessors are used for involve processing very large data sets, a task that data parallelism was specifically invented for, so why would you use a scalar CPU to perform these tasks when an SIMD processor would not only be able to process the data in fewer clock cycles, but it'd also use far less power (and produce less heat)?

there's absolutely nothing silly about using GPUs/stream processors for applications out of 3D graphics. it's sillier to have the CPU perform calculations it really wasn't designed for while most people have a powerful vector processor far more suited to the task sitting on their video card. that's why companies like Adobe are interested in GPGPU and are writing their applications to take advantage of the vector processing power of modern GPUs. the gaming industry has been exploiting SIMD for nearly two decades, so have CAD & DSP software developers and embedded systems manufacturers.

Re:It's great that there's money for this stuff...

[JoeMerchant]

You can claw your way up to 16 cores today if you work with 4 socket motherboards...

The data I work with isn't typically "fully massaged", there are a lot of sparse areas to consider quickly to identify areas that need more attention.

I still find cases where the algorithms can be sped up more than 10x by eliminating un-necessary work - this might not happen in more mature fields, but especially in cases where the new programmer has implemented something, there's usually more speed to be gained in a good code review than by throwing any amount of hardware at the problem.

Re:It's great that there's money for this stuff...

[JoeMerchant]

There are cases where CUDA is brilliant, just not my cases :-(

Re:It's great that there's money for this stuff...

[JoeMerchant]

I suppose the big question is what is the computing need out there.

I called "silly" on the massive parallelizing of very limited SIMD units because of the limited scope of applications. Financial? Yes, there is one options pricing algorithm that can be tackled very nicely with massive SIMD, but the rest of the financial industry needs less specific computing hardware. GPU cores are getting more capable, and they will eventually find a sweet spot where they are useful for video encoding and other things that lots of people actually do. Kudos to NVidia for trying to make GPUs more accessible, it will find some interesting niches, but what I see listed in the parent's post is pretty much a copy of the applications that were known two years ago, before CUDA launched.

Re:It's great that there's money for this stuff...

[Ostracus]

That's not why you got dinged. You got dinged for implying that pretty graphics were the only reason for GPU development. Your followup "silliness", simply reinforced that viewpoint. As me and others have pointed out GPUs ARE being called upon to do more than that. Directly related to "pretty graphics" is AI and Physics. Areas games traditionally have been weak in. Secondarily there is CUDA which promises speedups on tasks people are presently doing and I don't mean the rarefied world of academia but ordinary folks. And most important of all GPUs are HERE and available NOW. And economically to boot. Why shouldn't we take advantage of something we've already paid for?

Re:It's great that there's money for this stuff...

[raftpeople]

I looked at all of my options about 6 months ago and if I could have bought a 16 core machine without paying for server licenses/hardware and/or moving to linux (my app needs to run under windows right now), then I probably would have done that, but for the price, the gtx280 is working well.

Re:It's great that there's money for this stuff...

[Anonymous Coward]

Yep- I really Really like Qt for that reason - run on Windows when you need to, recompile for Linux when you don't.

Re:It's great that there's money for this stuff...

[Anonymous Coward]

I spent a few months writing software for DOS3, then almost a year in DOS4, and the upgrade to DOS5 was mostly painless, by DOS6 we got smart and employed a PharLap layer that insulated us from _most_ of the treadmill, which lasted for a whole 3 years before we upgraded to Win32... Looking back across those 6 years, I'd say that over 30% of our programming time was spent on "treadmill tasks" - just keeping up with the shifting OS requirements. CUDA and the hardware it runs on strikes me as DOS 0.2 territory. If your software is simple enough to rewrite in a couple of days every time CUDA shifts underneath you, then go for it, but if you're developing something that takes a year or two to get complete - you'll be lucky to be able to buy new compatible hardware to run it on after 2 years.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Comment removed

[account_deleted]

Comment removed based on user account deletion

fix

[doti]

NVIDIA, fix your linux drivers please.

NVIDIA, open your linux drivers please.

Re:fix

[Anonymous Coward]

That will do SHIT NOTHING as we see from the Intel and ATI area. *NO ONE* will do any real work to get the drivers working better. It will be just more 10 year old with "I hax0r3d NvIDIA drivers!". Seriously, who is the main contributor to the free software out there?? Random people?? Or people that benefit from doing what they are doing? (eg. Linux, OpenOffice, Mozilla, even PostgreSQL)

And who benefits that nVidia GPL drivers support some ASUS 375 GPXLS v2.1 card??? The few hundred people that bought that exact card? nVidia! So, basically, you want nVidia to GPL their drivers AND keep working on them because no one else will.

Sorry, that doesn't cut it.

Re:fix

[Symbolis]

Well, that's not going to happen.

Even if they did own all the bits and pieces that are in it.(Psst. They don't.)

ATI linux drivers have been improving a bunch.(Still not open, though.) My radeon 4850 works just fine for me, under Kubuntu 8.10.

Re:fix

[nschubach]

So by your argument, nVidia is still working on the drivers... still paying their devs the same... but they can accept tweaks from the community that improve them for free. I don't see the issue here.

Re:fix

[Anonymous Coward]

I'm currently on an internship, working in NVIDIA's driver department. I use linux at home, have for years, and I've bounced between ATI and NVIDIA for my cards. But I have to tell you, after being here and seeing what really goes into that driver? Demanding perfection from them (actual, real people) is absurd. Trust me, even with all the bugs you've heard about, there are so many more things that -could- go wrong, and don't, that it blows the mind.

NVIDIA's not going to open their drivers, for the reasonable reason that they have a lot of innovations they don't want competitors catching up on. I'm not saying that a little documentation wouldn't be nice for the opensource devs to get a leg up, but really. Making a graphics card driver that can support a wide range of cards is not quite as simple as a bash script. Now that I have a good understanding of how much work these guys actually are putting into making quality products? I'm absolutely going to keep buying NVIDIA stuff for myself.

Re:fix

[shutdown -p now]

They didn't do it in, what, 5 years? despite numerous pleas and petitions. Why do you think they are going to listen or care now?

The ironic thing is that there is really very little the Linux crowd can use as an incentive for NVIDIA to change its ways. Its bleeding edge cards are usually bought first by powergamers, who mostly don't care much about Linux (because it doesn't run their games). By the time the cards become mid-range, Linux drivers stabilize, and by the time they become low-end and appear in netbooks and the like (where Linux actually reaches the average consumer), the drivers are positively rock solid. Meanwhile, NVIDIA gets to cut some cost on beta testing by doing it live. Everyone's happy, except for Linux power users.

Microstutter

[Anonymous Coward]

I wonder if this card will suffer from microstutter. The 9800GX2 benchmarked very well but real world performance was lacking because the card essentially varied between very high fps and very low fps, so it still lagged even though it got decent average fps.

With these dual cards it's best to look at their low fps rating. An average fps is often misleading.

Re:Microstutter

[jgtg32a]

That's very true, I recently purchased a 4850 1gb over the 512 because while in the reviews they both showed identical frame rate, actual game play was not equal. "One" of the problems crysis is having is a zoom lag when you use a scope or binocs, because it has to load the textures.

The 512 had a bit of a lag, but the 1gb did not.

Re:Microstutter

[jerAzevedo]

The OP isn't talking about the amount of the amount of vram.

When you have two separate GPUs sometimes they work very efficiently together and you get a really high spike in framerate and other times resources aren't allocated efficiently and you get short periods where framerate is very low. When you take the average framerate over a second or so you end up with a fairly stable fps reading because these ups and downs usually average out.

However if you're actually playing the game and it's taxing your video card you'll notice "microstuttering" in the graphics or gameplay because you're not getting a stable framerate, your framerate is actually dropping low enough to cause short periods of lag several times a second.

Part of the problem with dual GPUs is that sense they don't share the same pool of vram the actual amount of usable vram is cut almost to half. Multi-core GPUs, if they scaled entirely, would probably perform better than dual-GPU cards because they would be able to share the same vram. The reason I think you don't see dual-core GPUs (there have been a couple in the past) is that the actual core on a GPU is very large and hard to shrink. They have all those shader and floating point units so it's probably easier, and definitely more efficient, to just fill up as many arithmetic units on a single GPU as possible then to make two cores each having half (or less) total number of arithmetic units.

Re:Microstutter

[Jthon]

You don't see multi-core GPU's 'cause there's no point. The GPU design is nothing like a CPU.

The GPU is already a collection of parallel cores, often called "shaders" (though NVIDIA and AMD are now branding them cores). These are what actually do the work. You can keep cramming more shaders into the chip and keep getting better performance, but at some point the chip gets too big to actually manufacture.

So basically there's no point of taking two GPU cores as they stand now and shrinking them together. You'd just end up with a less efficient design.

Of course the dual-GPU cards have some issues since you now have to distribute work between two different processors, yet they are hidden from the software stack so the dev only sees a single GPU. In this case you have a bunch of classic NUMA programming problems, but the software/game dev often doesn't take this into account. The drivers devs do their best to work around bottlenecks but they can only do so much.

This is often why some games can scale very well with SLI/Crossfire and others see no benefit. Those that see the biggest performance boosts avoid features/algorithms that don't scale well when adding multiple GPUs.

Re:Microstutter

[shutdown -p now]

RTFA. It's actually a pretty good review when it comes to numbers - they don't just give a single FPS figure, they give min, max and avg, and also the actual graph of FPS vs time, so you can see the spikes for yourself.

A spoiler to save you time: it doesn't seem to do any worse than ATI.

Re:Microstutter

[jerAzevedo]

That's basically what I meant. The next best thing is a multi-gpu design which is less efficient as far as data access and communication is concerned. But when you reach a limit as to how fast you can make your gpus run there's really no other way to go.

Another example of the law of diminishing marginal return.

This is all so 1998

[Junks Jerzey]

Ten years ago the video card wars were in full swing. Each generation brought amazing new advance and impressive technology.

But nVidia and ATI haven't realized that we passed the point of diminishing returns years ago. Mobility and battery life are what matter. And I know there are hardcore PC gamers out there, but there are only a handful of companies even bothering to push the high-end graphics, so you buy a $500 video card and there are exactly ZERO games that take full advantage of it. Wait a year or so, and you may find that one or two of the few high-end PC game makers decide to throw you a bone and add support. And as a bonus, you get SIGNIFICANTLY increased power consumption, and the video card addicts are just wasting resources so they can all whack-off to Shader 30.0 soft shadows on eyelashes.

It's a weird, captive, completely pointless market unless you're doing 3D rendering for a living (for movies, for commercials, for product design, etc.).

Re:This is all so 1998

[TheKidWho]

See this is where you are uninformed, the new GTX's have lower power consumption than the 9000 series at idle and for 2d applications.

Re:This is all so 1998

[Jeff DeMaagd]

I'd say let the early adopters go for it. It is usually the early adopters that help pay for the lion's share of the development anyway. A year from now, the same performance will be had for less than half, and there should be several games that can play it.

Re:This is all so 1998

[JasterBobaMereel]

No the render farms do not need this level of graphics either ... they render off-screen at seriously low frame rates (frames per minute not per second) they go for quality not speed

They do use the GPU's to render but they are using it as a general processor that is fast (in a cluster) not as a GPU at all ...

The only people using high end graphics cards are Gamers and a very few graphic artists ...

Re:This is all so 1998

[LandDolphin]

If there is enough of a market for them to develope these cards, and stay in business then what the problem? There is a need, and they are filling it.

Re:This is all so 1998

[nschubach]

I agree with you, but I also think that raster 3D is hitting a downward slope in "realistically programmable" feature sets and hopefully ray tracing or hybrid rendering will start to pick up in it's place. I actually think keeping the bleeding edge market going is a good thing toward both obtainable real time tracing and lower power consumption. Even today, I think (even today) nVidia/ATI have to reduce energy costs to go faster.

Re:This is all so 1998

[ceoyoyo]

ATI and nVidia know all about the mobile, embedded and low power markets. That's where they make most of their money. You only see the product announcements for the latest and greatest, super fast, gamer wet dream cards because those are the only ones that make the news.

ATI and nVidia have to do R&D, develop faster cards even if they are impractical for their target market. Then, once those features have been tried and tested, they get put into regular, production cards. Just like Formula 1, they've discovered that they can actually MAKE money off what is little more than a research prototype by selling it to crazy people who have to have the latest and greatest (or want to watch the latest and greatest, in the case of F1).

Re:This is all so 1998

[Anonymous Coward]

Mobility and battery life are what matter.

Then I think you're reading the wrong review. You probably meant this one: http://www.anandtech.com/video/showdoc.aspx?i=3478&p=1

Re:This is all so 1998

[Chandon Seldon]

And I know there are hardcore PC gamers out there, but there are only a handful of companies even bothering to push the high-end graphics, so you buy a $500 video card and there are exactly ZERO games that take full advantage of it. Wait a year or so, and you may find that one or two of the few high-end PC game makers decide to throw you a bone and add support.

This is blatantly false. What a decent ($200+) graphics card buys you today is being able to play games at decent resolution. To play todays games at 1600x1200 resolution or better, you need a decent card. And increased resolution really does mean better game play; don't imagine that it doesn't matter if you haven't spent time with it.

I say this based on personal experience. I got Far Cry 2 a couple days ago and it only runs at 1280x800 with the graphics setting cranked down on my 26" monitor. That's with an Nvidia 9600 GT, which sells for $100 today. I should really upgrade, but I hate to do it for just one generation of cards. Eh, that's what I get for buying mid-range.

Re:This is all so 1998

[TheKidWho]

Not true, a lot of engineering departments depend on computers running high end Quadro video cards. The difference between a computer with, and a computer without, is staggering in high end modelling software.

I also expect in the near future to see accelerated CAE/FEA built into new CAD packages that utilize the power of the GPU for processing.

Re:This is all so 1998

[Spatial]

But nVidia and ATI haven't realized that we passed the point of diminishing returns years ago. Mobility and battery life are what matter.

In the context of a card designed for desktop PCs, specifically for people who play games on gigantic monitors? You can't be serious.

And I know there are hardcore PC gamers out there, but there are only a handful of companies even bothering to push the high-end graphics, so you buy a $500 video card and there are exactly ZERO games that take full advantage of it.

Most modern games can use the full capabilities of this card. It's designed for people who want to play the latest games at extremely high resolutions with maximum quality settings. That takes an unholy amount of processing power. It's really the only use for a card like this. Most people won't buy it.

And as a bonus, you get SIGNIFICANTLY increased power consumption, and the video card addicts are just wasting resources so they can all whack-off to Shader 30.0 soft shadows on eyelashes.

Nobody who buys a dual-GPU card gives a singular shit about their power consumption, I assure you. They probably have hugely overclocked quad cores that require a Mr. Fusion to use on top of the absurd card.

I can see where you're coming from, the card is slightly silly. But the reasons you stated don't really apply; it's a niche part absolutely not representative of the overall state of video cards. They just market these things that way because it gives them a "Leading edge" image which goes over well with their typical market. I doubt it even helps 3D rendering, as far as I know that's a CPU-oriented field.

Re:This is all so 1998

[bigstrat2003]

This is blatantly false. What a decent ($200+) graphics card buys you today is being able to play games at decent resolution. To play todays games at 1600x1200 resolution or better, you need a decent card. And increased resolution really does mean better game play; don't imagine that it doesn't matter if you haven't spent time with it.

I have to disagree here. I have played exactly one game where more resolution led to better gameplay, and that was WoW, because of all its user interface mods. When I had a crappy computer, I used to run games at 640x480 resolution. Now I run the same games at 16900x1200, and the gameplay is improved not one bit. It doesn't even look that much prettier.

Re:This is all so 1998

[Pulzar]

When I had a crappy computer, I used to run games at 640x480 resolution. Now I run the same games at 16900x1200, and the gameplay is improved not one bit./blockquote?

16900, that has to be a 900" wide screen monitor? :)

But, seriously, try something like Civ4, Galciv2, etc.... games where a lot of information is displayed on the screen. The higher the resolution, the further you can zoom out and still be able to see the details. It improves the gameplay dramatically.

Re:This is all so 1998

[Chandon Seldon]

Now I run the same games at 16900x1200, and the gameplay is improved not one bit. It doesn't even look that much prettier.

I take it you play in a 12" monitor and have crappy eyesight, yes?

If not, do me a favor and try this experiment: Load up any reasonably modern outdoor FPS (where you have to aim at medium to long range - Far Cry, Battlefield, etc) and play it at 1600x1200 and then 800x600. If you don't notice a significant game play difference then either you're on crack or we have completely different sensory systems - maybe you evolved from one of those blind deep sea fish.

Re:This is all so 1998

[bigstrat2003]

I play both of those games, and I disagree with your assessment. Sorry.

Re:This is all so 1998

[bigstrat2003]

Or maybe you're on crack? Far Cry is, indeed, one of the games I've played on both systems. My eyesight is fine. My monitor is 22" widescreen. I'm sorry, but it just makes no difference to the gameplay.

Re:This is all so 1998

[Khyber]

Actually, I get better performance running at 800x600 for FPS games, at least as far as killing things goes. At lower resolutions harder shots tend to be easier because of the lack of resolution, so the game has to kinda 'guess' where you're shooting.

It's pretty much been like this since the days of Q2 with the railgun. High-res players got better graphics while I stayed on 640x480 and hit them halfway across the map while they couldn't aim for jack because of the 'pinpoint precision' of higher resolutions.

Re:This is all so 1998

[Chandon Seldon]

It's pretty much been like this since the days of Q2 with the railgun. High-res players got better graphics while I stayed on 640x480 and hit them halfway across the map while they couldn't aim for jack because of the 'pinpoint precision' of higher resolutions.

Games where everyone has a single hitbox and you get to aim with crosshairs are a bit different. I didn't make any claim about those. I'd recommend doing the test I suggested with a recent iron sights / hit locations game.

Re:This is all so 1998

[balthan]

so you buy a $500 video card and there are exactly ZERO games that take full advantage of it.

Clearly, you've never played Crysis.

Re:This is all so 1998

[donglekey]

Every time a video card or processor comes out, there is a comment cut from the same cloth, because someone thinks that because they aren't going to make use of this incremental increase in processing power then nobody is. Give me a year, and I will show you mind blowing new software. And you will say you don't want it, and you don't want to pay for it, but others will, and thats why millions of R+D went into this. Quit worrying and sit back and enjoy the ride. You've seen this right? http://www.youtube.com/watch?v=ZgoDypGMV50&feature=related

Re:This is all so 1998

[Khyber]

I did, with CoD2 and CoD4. At lower resolutions it's far easier to headshot than at higher, especially using the iron sights.

And just so you know, I run a fairly beefy system. half-gig 9800GTX+, 2.6GHz AMD Athlon64X2 5200+, 4GB DDR2-800.

Re:This is all so 1998

[Chandon Seldon]

And you're arguing that poor hit detection due to huge pixels improves gameplay?

Re:This is all so 1998

[Khyber]

In a multiplayer match, youbetcha. Good players know their games and exploit their weaknesses in order to acheive maximum pwnage potential. Same thing with like the divisible by 3 framerate glitch in Q3 that allowed things like 90, 99, 333, 666, 999 FPS to affect the player's actual movement speed.

Re:This is all so 1998

[Chandon Seldon]

Good players know their games and exploit their weaknesses in order to acheive maximum pwnage potential.

I agree. But the gameplay is even better if the exploits don't make the game look like shit.

Point: Missed

[jrronimo]

A while back, AMDTi said that they were not competing at the high-end anymore: "There were also very specific admissions that AMD/ ATI isnâ(TM)t competing at the high end with Nvidia, nor do they intend to match up to the GTX 280 with a release of their own uber-chip." source. So to say "ATI had to combine two cards to be on top!" kind of completely misses the point. (emphasis added.)

For the interested, there's a great article at anandtech talking about how the R770 came to be pretty awesome... Really, though, it's not a super-high-end part.

Re:Point: Missed

[Anonymous Coward]

Given you can now pick up a HD4870 for less than EUR200, it's a pretty awesome part in terms of price/performance.

Re:Point: Missed

[jgtg32a]

Hey you forgot to mention the best part, while the Nvidia card is about 15-20% faster the ATI costs about 1/2 as much.

Re:Point: Missed

[FugitiveMind]

That's odd, because ATI has the fastest (consumer) card on the market right now, the HD4870x2.

Even with its two GPU cores each with 1gb of ram, I only get about 20-25 fps in GTA4. (We're complaining about it up above. Rest of the system is: 2.6ghz Athlon X2, 6gb DDR2, 1tb hdd, Vista X64 Ultimate) I'm waiting for the patch that adds multi-core rendering support to GTA4. I might actually be able to play it at that point.

Re:Point: Missed

[DisKurzion]

[citation needed]

The new Nvidia is debuing at $499

Cheapest competing AMD (Radeon HD 4870 X2 2GB) card I could find was $480

Re:Point: Missed

[DisKurzion]

Doh! That's supposed to be debuting.

Re:Point: Missed

[Spatial]

Since GTA4 is an absolute whore for CPU performance, that's probably the limiting factor. If you're up for it, a C2D E8400 should improve performance a lot. Unfortunately you'd have to get a new motherboard.

Re:Point: Missed

[AcidPenguin9873]

The funny thing about graphics is that it's entirely about throughput, not latency. You could have a pretty dog-slow "stream processor", but as long as you build enough of them on a chip and/or stick enough chips on your board, and have enough memory bandwidth to feed all of them, the performance of a single shader core pretty much doesn't matter.

Like you said, DAAMIT built a pretty middle-of-the-road RV770 (compared to the raw performance of the NV chip) but was able to tie two of them together in a reasonable thermal/power/bandwidth/cost board, and took the graphics performance crown. I wouldn't be surprised to see a 4870 X3 or X4 sometime soon as long as those other factors (thermal/power/bandwidth/cost) remain in line.

Oh boy!

[DurendalMac]

Does it come with a small nuclear reactor and phase change cooling kit so you can actually run the damn thing without having it burst into flame after 10 seconds?

Taking back the performance crown?

[tangent3]

Nvidia never lost the performance crown. AMD did not even bothered to compete with Nvidia for performance at the high end.
Read this excellent article.

What AMD did with the RV770 series was to totally pwn everything below the super high end.
When the 4870 was released at $299, it was generally worse than GTX280, but it easily beat the GTX260 which was priced at $399.
When the 4850 was released at $199, it easily matched the 9800GTX which was priced at $249

Re:Taking back the performance crown?

[Chandon Seldon]

No, the funny thing here is that AMD *did* have the performance crown, even though they had planned to give it up. Before the GTX 295, the Radeon 4870x2 was the top of the pile for single-card graphics.

Re:Taking back the performance crown?

[Anonymous Coward]

You're wrong. The GTX 260 and 4870 are almost exactly matched for performance and the 216 core variant of the 260 is slightly faster.

Re:Taking back the performance crown?

[Anonymous Coward]

You're either stupid or ignorant. How the fuck did this get modded "Informative".

Nvidias top 5

[ijakings]

Unfortunately until its released all these "Special previews" are just mouthpieces for nvidia, notice all the special previews use those 4 or 5 games nvidia has been telling them to use.

Of course they come out on top when you control the tests. Im not a fanboi either way, it just annoys me when people make decisions based soley upon reviews that are clearly biased.

Try this...

[Civil_Disobedient]

It's been recommended on several OC forums to disabled clip recording in-game--apparently this is offers a substantial performance increase... YMMV, but it's free to try. Unlike this new NVIDIA beast.

The way they name new hardware pisses me off.

[Anonymous Coward]

Till these guys start back naming their cards by names that make since, I'm never going to buy another one of their cards again. I'll stick with my console for games. The way they name cards it's impossible to tell which ati or nvidia card is better even when comparing ati to ati or nvidia to nvidia. Intel/AMD has done the same thing with their cpu's the names/numbers are so screwed up, it's not even worth trying to keep up. I miss the 8086/80186/80286/80386-geforce/geforce2/geforce3 days.

Re:The way they name new hardware pisses me off.

[Spatial]

Just look at one of the hundreds of graphs available that directly tell you which one is better. You have to do this perhaps once every two or three years. Sure, the naming scheme is stupid but not buying them because of it is ridiculous.

And really, it's not that common. AMD CPU: higher number is faster. AMD GPU: Higher number is faster. Intel CPU: Higher number is faster. It's just Nvidia with their anal marketing.

Re:The way they name new hardware pisses me off.

[bigstrat2003]

The same rule holds true for nVidia cards. They shifted direction in their numbering scheme after they got to the 9000 series, it's true, but ATI did the same thing when they need a successor to their 9xxx cards.

The Real Question

[Nom du Keyboard]

The real question is: Did Nvidia simply make enough of these (~30 - 200) certainly very carefully hand-picked chips to homologate them as a valid offering in claiming the top spot on the performance charts...

...or are they actually producing them in quantity such that anyone who wants one can buy one at the stated price?

Personally I'm betting on the former being far more likely than the latter.

Re:It's great that there's a market for this stuff

[IorDMUX]

it makes the integrated graphics in eee class PCs that much better when the tech trickles down 5 years later.

This is also good news right now for the "sweet spot" gaming PC builders. Each time these new bleeding-edge-$500-200W-XXTREME cards come out, the previous two generations of graphics cards tend to suddenly drop drastically in price.

When I built my current middle-of-the-road gaming computer, I put in an ATI HD3850 for $150, with the expectation of adding a second on Crossfire once the price drop occurred. Looking at Newegg.com, the 3850's have hit ~$55.00 this week. My computer looks to be due for an upgrade after Christmas, it seems!

4870 X3 Anybody?

[Nom du Keyboard]

ATI 4870 X3 anybody?

Re:4870 X3 Anybody?

[Spatial]

Unlikely, if they're focused on results and not marketing. From what I've seen, adding more GPUs results in rapidly diminishing returns in the majority of cases - have a look at some benchmarks of Nvidia's tri-SLI for example.

-Nvidia-? All the -Rage-?

[Tetsujin]

You guys have got it all wrong... ATI cards are all the rage... Or, I guess they were until the Radeon...

No benchmark against GTX 280?

[Kagura]

I have a GTX 280, and I think it's suspicious that it's absent from their benchmarks, to say the least. Why would you include GTX 260 and an ATI card, but not GTX 280?

Re:No benchmark against GTX 280?

[gormanly]

Because this new card uses the 260's memory bus and memory effects dominate at higher resolutions with AA turned up? Because its core clock is the same as the 260? Yes, its cores have the same number of shaders as a 280, but if you run at 1920x1200 or above it's probably more like a pair of 260's in SLI than a pair of 280's - and there's no point in having a $500 card if you play on a crappy screen.

Re:No benchmark against GTX 280?

[Kagura]

Because this new card uses the 260's memory bus and memory effects dominate at higher resolutions with AA turned up? Because its core clock is the same as the 260? Yes, its cores have the same number of shaders as a 280, but if you run at 1920x1200 or above it's probably more like a pair of 260's in SLI than a pair of 280's - and there's no point in having a $500 card if you play on a crappy screen.

You must have completely missed what I was saying. I own a GTX 280 card, and I would like to see how my GTX 280 compares to the new GTX 295. Why would you benchmark a graphics card without comparing it to the card that was top-of-the-line for the last several months straight?

Re:No benchmark against GTX 280?

[arootbeer]

I believe the GP was saying the answer to your question is

the writer of the article is a shill for NVidia

Re:Palin

[Tetsujin]

NVIDIA GTX 295 Brings the Pain and Performance

A fight to the death? No! A fight to the pain!

Misread that for a second as "A fight to the Palin"...

But then the image of Michael Palin slapping Sarah Palin made it all worthwhile...

Re:It's great that there's a market for this stuff

[Jeng]

Low end graphics have not advanced much at all.

After all, isn't that one of the big things regarding the Vista Ready debacle?

When on board graphics cannot even display desktop eye candy I think it needs to be admitted that there is no trickle down for low end or on board graphic cards.

The low end and on board graphics cards are just as weak now as they were 5 years ago, I'm not positive, but I think they might even be worse than the old intel 810's.

What's a good middle-level card for Linux?

[Anonymous Coward]

I'd like to run things like Compiz but I can't with my current card (an earlier Nvidia GEForce).

But I don't have the money to spend on a really high end card.

So, what's in the middle? (I use Linux).

Re:What's a good middle-level card for Linux?

[Anonymous Coward]

NVidia 9600GT or ATI 4670/4830.

Get them with at least 1 gig of VRAM if you're going to bother, and some of the 9600's are showing with 1.5gigs for 150 or less.

32-bit and 64-bit kernel, userspace

[Tetsujin]

I'm sorry... what? I don't see how that's a reasonable choice at all.

The 32-bit Linux kernel cannot efficiently handle more than 1 GB RAM. Linus Torvalds himself is outspoken on this subject and recommends using a 64-bit kernel with more than 1 GB (alas, I can't find the reference right now, but he has written many times on the subject in RealWorledTech.com's forums).

Here, I was real curious about this issue, so I looked it up:

http://linux-mm.org/HighMemory

Basically, when a system has more than 1GB of memory, the kernel can't directly map all physical RAM into its own address space (because 3GB of the virtual address space is reserved for applications...) - which makes certain operations more complex than they need to be. The kernel needs to take some extra steps when dealing with more than 1GB of physical RAM when exchanging memory addresses between different modules, for instance... The problem gets worse as you go beyond 4GB RAM and have to deal with PAE, etc...

You can get around that by changing the amount of virtual address space reserved for the kernel vs. user space - for instance a 2GB kernel/2GB user split of virtual address space on a system with 2GB RAM, or 3GB kernel/1GB user split on a system with 3GB RAM - but of course that's not ideal as it limits how much virtual memory (RAM + swap, etc.) is available to user processes...

On the other hand, using 64-bit in userspace is pointless for most users. Most applications don't need more than 3GB of RAM. Also, there are binary plugins, etc, which only work in 32-bit.

I don't know that I agree with that. Among other things, in x86-64 there are twice as many general-purpose registers. This makes it a lot easier for GCC to optimize compiled code effectively. Additionally the other features present on modern processes can be safely taken as assumptions by the compiler and standard libraries. The binary-only problem isn't as bad as it used to be, either - video drivers are available in 64-bit, flash is available in 64-bit, Java is available in 64-bit... I just recently built a new system and I decided to go fully 64-bit - I think it's worth it, and so far there's been only minimal difficulty. I guess the main disadvantage of a 64-bit userspace is the increased size of pointers and std::size_t, etc. using more RAM. To me the tradeoff would be worth it.

64-bit kernel and 32-bit userspace is the most efficient and practical choice for the vast majority of users.

Re:32-bit and 64-bit kernel, userspace

[cecom]

Well, the choice of 32-bit vs 64-bit userspace boils down to convenience. If 64-bit works for you, great! On the other hand, I have carried over my desktop Linux installation through several different machines and only the last one is 64-bit capable; I'd rather not reinstall everything, given that none of the applications I use need 64-bit address space.

Also, don't forget that 64-bit userspace applications need more memory (pointers are twice as big). They have larger cache footprint and this alone could cause them to be slower than 32-bit despite of having more registers.

Re:It's great that there's a market for this stuff

[Anonymous Coward]

I'm not sure about 5 years ago, but 10 years ago I was having definite problems with desktop graphics being able to show simple graphs quickly... that is certainly not the case anymore.

The low end is driven by need instead of desire - Vista will be increasing the "need" level of the next generation of integrated graphics (kind of like all PCs had 8259 (1 byte buffer - i.e. sucky) serial ports until the internet came around - then the chipsets suddenly upgraded to 16 byte buffers.

Headphones (was Re:i hate fans)

[Anonymous Coward]

After years of working on building (near) silent PCs, I've learned that what many people/reviewers consider to be 'quiet' is nowhere near my definition of 'quiet'. I'm not quite sure how loud some gamers have their sound systems turned up, or if they play with the window open or what...

Nah, we use headphones. Better sound at lower volume, and blocks outside noise.

Re:It's great that there's a market for this stuff

[Anonymous Coward]

Well the technology trickles down to the integrated chips, it just gets castrated. ATi and Nvidia have offered dx9/OGL2 on their integrated chips for about 4 years but aside from running aero there wasn't much you could do with them.

The main sin of integrated video (which the i810 was perhaps the most egregious offender) is the shared system memory. Just a 1650x1020 framebuffer uses around 400MBPS of memory bandwidth. So 5 to 8% of your memory is tied up just showing you a desktop. Ask the GPU to actually do something and you may have a problem.

The old i810, by comparison, couldn't actually generate 400MBPS in total memory throughput. A 1024x768 desktop would cripple an EMachines Monster P3-800.

Re: CUDA Programming...

[Douglas Goodall]

There is really two things going on here. They are trying to build manycore GPUs and trying to produce commercially usable video cards at the same time. If they can do both successfully, it will be good for all of us, but solid hardware, and good drivers all across the board will be required. The new C1080 card has 240 cores and plugs into a PCI Express slot. Contemporary NVIDIA Video cards have less. This is the new game. A frame with four of these boards is now a supercomputer. I just wish I could get one of these into my Mac Pro (I must be dreaming).

Passive cooling! Sorta.

[Anonymous Coward]

The last few machines I've built have been "silent" and run at cooler temperatures than with stock heatsinks/fans. It is fairly easy to do:

1) Make sure there is good air flow. At the very least have a fan at the front of the case pushing air in and a fan at the other side pushing hot air out.

2) Use Passive cooling products on everything, like the Arctic Cooling Accelero S1 Rev. 2 on your video card.

3) Use low speed, quiet fans on the things you passively cool. This might be optional depending on what you use the computer for or where it is located.

My current gaming computer is budget priced but does decent with this generation of games. The Crysis demo runs at High settings. It played Warhammer Online and City of Heroes at max everything @ 1680 x 1050. Lately I play mostly older or indie games, and at ~$600 this machine crushes them.

My BFG 9800GT OC hits 100 degrees Fahrenheit at the very highest after many hours of taxing it. The processor cores (Core 2 Duo 8400 @ 3GHz) top out around the same temperature (100 degrees Fahrenheit).

The machine itself gives off about 22 dBA.

-AJK

Re:It's great that there's a market for this stuff

[Anonymous Coward]

Don't know where you were, but everything I had from the mid 90's up was 16450/16550 based Serial Ports. Think the last system that I had that might've been an 8259 was 486 based and I'd have to fire it up to check.

error in post

[SM177Y]

this statement....it was a pair of RV770 cores that AMD had to use in order to best the likes of NVIDIA's GeForce GTX 280....is false. im sorry to break it to yous but one 4870X2 mops the floor of even two GTX280's in sli. and dont try to say im just a whiny fan boy. this is not the case as i am running a 9600GT. i buy whats worth the money for the performance im getting. With that being said, ATI has been raping the market when it comes to price for performance for the last 2-3 years non stop. NVIDIA needs to get into the game and stop selling their inferior cards at such rediculous prices. If they make something that owns ATI's cards, well then they should at least make the price feasible. not dropping a 600+ price tag on it. thats not the way to sell a video card when u can get a 4870X2 for under 500 bucks. Come on NVIDIA, we're waiting for a real performance/bargain.