Geforce 280.28 OpenGL 4.2 Driver

Yes and no. GLView says the driver/hardware is 91% OpenGL 4.2 compliant, but you make a good point. NVIDIA says:

In any event, I don't know of anything that we need OpenGL 4.2 for at this time (unless you're a developer).

 

Well, I'd argue that this isn't really an OpenGL 4.2 driver unless it is 100% compliant but that at least explains why applications are detecting it as OpenGL 4.1 if they use the same criteria as 4.1 *is* 100% compliant! GPU Caps Viewers states that the OpenGL version string itself, presumably read from the driver, is 4.1 as well.

I'm not especially bothered about it, I was just surprised that it wasn't being detected as OpenGL 4.2. It is a developer driver though.

 

This driver includes a 4.2 string as well. I doubt NVIDIA is lying about this being a OpenGL 4.2 driver. Perhaps the key words are "on capable hardware".

 

What's your point, idc about maximum perf. mode in their observations, yea they said if gpu goes to lower 3d clock in games then of course i would force maximum perf. (which i do in all games anyway) but not in adaptive apps.

And I said the problem is with adaptive apps entering in P0 full3D power mode all the time, to many times and for far toooo long >> End result heating gpu and making EXTRA UNWANTED NOISE, Get it????:infinity:

Afterburner pulling doesn't matter i used default 1000ms too and i know how it behaved before and how it started with 280.xx, you won't persuade me otherwise with your preacher talk kapish:nerd:. And nv admitted its not ok with r280.xx like they thought, so yes they're gonna change it.. + now i see alot gpu's suffering from driver stopped responding crap, because of this adaptive apps unoptimized gpu freq. issue, im lucky that don't have such lame issues or it would drive me mad.
and why would gpu run worse if they change it back like it was? It doesn't matter because this "new technique" isn't faster in games anyway, actually by some games it was worse, lost 1-3 fps here and there compared to older 270/275 driver.. I don't play DragonAge2 or Witcher2>> although it preformed just fine with 270.90, so yea i saw no boost with 280.xx, except sun flicker fix in COD5 (i had to alt+tab with 270 and 275 drivers, to fix it).

And to worry about my reputation, for what? Do you think i bitched and did all those videos/screens just for fun? heh please.. You better worry about yours lol

anyway im done talking to you too.


@Darren

check with Aida64 it reports it correctly.

 

thats coz opengl extensions viewer is bugged (yes even the latest version, see here.. http://www.geeks3d.com/20110812/glview-4-02-released-with-opengl-4-2-support/ )

and GPU Caps Viewer needs an update before it can report v4.2..

OpenGL 4.2 was a surprise for everyone apart from kronos and nvidia..
no developers even knew it was in the works, hence no apps know how to report it..

@Cru_N_cher: dont waste time responding.. i dont even bother reading his posts anymore...usually full of illogical crap.. ya better off :bang: !!

 

Thanks IKnowJack.

Haha, I'm relieved that others feel the same way I do.

 

Um, it's not trying to crack the Enigma code - all the app does is read the info from the driver...pretty simple.

Kinda like saying an old version of CPU-Z wouldn't be able to report my CPU info unless it was updated - similar concept.

 

Selecting GeForce GTX 580/PCI/SSE2 from the pulldown at the top of OpenGL Extensions Viewer v4.02 redetects OpenGL 4.2 properly. Apparently, there's a bug in the app in the form of a missing Forward Context 4.2 hence why, by default, it doesn't detect OpenGL 4.2. The Report list then shows 100% compliance for OpenGL 4.2.

 

I kept getting windows explorer stopped responding when using these driver, I switched to the WHQL and the problem went away

 

I just checked mine with AIDA64 and it says I'm 57% 4.1 compliant lol

:3eyes::3eyes:

 

Mine says -57% OpenGL 4.1 compliancy, as in negative 57 percent :infinity:

It also says 100% OpenGL 4.2 compliant even though I'm currently using 275.89 drivers.

 

because you're missing dx11.

this is with 275.89Q still at 4.1.

 

It says I'm 100% 4.2 compliant..

 

AIDA64 Extreme Edition beta Version: 1.80.1498 (Aug 12, 2011)

support for OpenGL 4.2....

 

I don't say that the behavior is ok but it doesn't look like a bug like you say. Nvidias Engineers are surely not as dumb and have a reason for lowering the P-state slower now, after moving it up faster (why should they leave that bug in 3 driver revisions even being told it exists,that makes no sense).
And im not sure why you disappointed the Problem here with most Cooling solutions is that they extremely hard set im idling @ 30° C with my GTX 460 @ 40% Fan Speed (Twin Frozr2) and sure for me it's a heavy thing as they Spin up much to fast even for that small Temperature Increase which is far away from anything that could hurt the Chip but i realize the better Performance even in the Broswer that you seem to not see @ all when running any HTML5 tests.
The only Problem is that Scrolling also activates it because it's utilizing the GPU like any App that uses the GPU does there are no "Adaptive Apps" like you like to call them the Adaptivness are the P-states and the Driver the Apps are just Apps that run and Utilize the GPU and based on the Utilization the Driver decides when to switch into the next P-State that are defined in the Bios since the Fermi series and also when to leave them.

This system is what prevents your Fan to Speedup to fast like it would if you disable it in the Control Panel and go Full Performance, though it also costs Performance but Saves Power.

You can even define the threshold that Nvidia now lowered in their 280.xx Drivers yourself (with any Driver revision that supports P-State switching) via the NVAPI using for example the "Multi Display Power Saver" from Orbmu2k also known as RMclock for Fermi

The Change in the 280.xx driver highers Performance at the cost of lower turn back time (though i don't looked @ the time it takes for 275 to switch back to the old P-State yet, but it definitely was less Performant due to lower reaction time) the issue you reported correctly though doesn't seem to be a bug but wanted behavior, or it never became apparent in 275 as the lower reaction time also impacted it)
I could see a increase in several Applications Performance (GPU Encoding,3D Engines, Browser with the 280.xx Drivers so for Users that have Water cooling, and or Passively cooled cards the small increase for the short time shouldn't be a big problem @ all and for others they could adapt their Fan Speed to increase slower or even faster to compensate the Heatup.
The whole driver has just become more responsive, it is logic that it gets hotter now as Voltage changes faster, and the Fan is now trying to cool it down though @ the same linear Increase speed as before that Nvidia didn't changed and is now to slow to compensate the faster Voltage change Nvidia caused due to the faster Reaction change depending on the Card and Setup)

 

o yes they can be dumb like a watermelon on a stick.

And i said what i had to say in other thread, so you can continue to defend this flawed design or what ever you're trying to do here, but it won't change my pov or anybody else for that matter.

I thought they would have known better and learn from the mistakes in the past, 186.24 until 191.00 and then again 195.xx to 197.45 but no, its the same sheit all over again.


:infinity:

 

This Driver gives me a very big improvement in Crysis 2 before the 280 drivers i had 30.5 fps on Time Square now it's 47
every other benchmark shows also an improvement GPu Encoding also profits from the better reaction time but really the Crysis 2 result is the most amazing

Ultra

8800 GT = 17 FPS
460 GTX Pre 280 = 30.5 FPS
460 GTX 280.28 = 47 FPS

That's a really heavy improvement hardly explainable with in measuring error range and my Hawk is absolute stable no crash yet neither under Heavy Furmark load. Those are now my first days with the 460 GTX and im amazed like i was back with the 8800 GT.

 

For me 280.26 has "Driver Stopped Responding" issues. Can anyone else who experienced that confirm that 280.28 don't have it?

Else i will just install them myself to find out

 

Don't bother...same problem.

 

i didnt have any, but i had the same abnormal gpu "overheating" because of rapid P-state switching and longer times in P0 state (full3d freq) in adaptive applications like Aero (not so much but i happened frequently) and Firefox (longer P0 then usual, end result much much higher temps and higher fan noise because of it).

Also i didnt saw any boost in my std games, actually some were slower with this rls or any 280.xx atm.. Although the only positive thing about 280.xx for me was sun flicker fix in Cod5 and smoke flicker fix in Fear1 combat, both are broken with 270.xx and 275.xx. I can eliminate it in Cod5 by alt tabing and by Fear almost (98%) if i select 0 frames to render ahead.