Overclock.net banner

**186.18/191.07/195.62 Driver Review**

13931 Views 66 Replies 21 Participants Last post by  [CyGnus]
4
The following is my personal review of the last three major driver releases from NVIDIA. 186.18, 191.07, and 195.62. Crysis will be used as the main test for this review.

Test Setup

Motherboard: EVGA nForce 780i
BIOS Version: P09 (Latest)
CPU: Intel Core 2 Duo E8500
RAM: OCZ Vista Upgrade 4GB 800 MHz DDR2 (5-6-6-18-2T Timings)
GPU: XFX Geforce 9800GTX in SLI
HD: Samsung 80GB SATA
PSU: Delta Electronics 800 Watt
CPU Heatsink: Xigmatek Dark Knight HDT-1283V
OS: Microsoft Windows 7 Ultimate 64-Bit

186.18



191.07



195.62




As you can see the major difference between these is noted in GPU-Z each driver update adds one more feature to work with. However, for gaming performance it appears the 186.18 is the superior driver. I would like to note that the minimum FPS for 195.62 were greatly low (less than 6 FPS at best) and the minimum FPS for 191.07 and 186.18 was always above 16 FPS (most cases 18-19 FPS).

I hope this review has proved useful to some people and please feel free to comment at anytime.
See less See more
  • Rep+
Reactions: 5
1 - 20 of 67 Posts
191.07 is the 1st drivers I ever herd of that will damage hardware.
Quote:

Originally Posted by Pings View Post
191.07 is the 1st drivers I ever herd of that will damage hardware.
I'm not sure you got proper information because it is physically impossible for any driver (or piece of software) to damage hardware in any way (unless it's capable of modifying voltages that you have no control over).
See less See more
You should test out the XG drivers to see how they compare to the originals that would be cool.
2
Quote:

Originally Posted by Anth0789 View Post
You should test out the XG drivers to see how they compare to the originals that would be cool.
This is true. I never personally messed around with those drivers but it would be interesting to see how they compare. Other than performance I'm actually interested in what way they are "modified". Did they get rid of some content? Lessen quality? If this is true it might not be worth using for some people seeing as how the graphics being tested on wouldn't be as intensive as the stock ones. But I may consider checking them out sometime. Thanks for the suggestion.
See less See more
Quote:

Originally Posted by Open1Your1Eyes0 View Post
I'm not sure you got proper information because it is physically impossible for any driver (or piece of software) to damage hardware in any way (unless it's capable of modifying voltages that you have no control over).
Your wrong and right. Its very possible it does mess with voltages. Also it will break the automatic fan controller. It will get stuck at 40% causing you to overheat in no time. This driver causes more than one hardware problem. Here are some post from other forums.

http://forums.techpowerup.com/showthread.php?t=107715
http://www.yougamers.com/forum/showthread.php?p=1236840
http://forums.nvidia.com/index.php?showtopic=109913
http://www.hardforum.com/showthread.php?t=1462586
http://forums.techpowerup.com/showth...=105302&page=2
http://forums.nvidia.com/index.php?showtopic=109328
See less See more
Yup, the exact same results that I got in my testing. The 186's are the best for gaming.
O yea the 191.07 mostly effects NF200 chipsets motherboards. It has been know to damage motherboards to. Most likely a heating problem that ends up damaging your motherboard. The GTX295's have NF200 chipsets built into them so be careful.
Quote:

Originally Posted by Pings View Post
O yea the 191.07 mostly effects NF200 chipsets motherboards. It has been know to damage motherboards to. Most likely a heating problem that ends up damaging your motherboard. The GTX295's have NF200 chipsets built into them so be careful.
I personally use 191.07 on a Media Center PC I built with a GTX295. And I specifically ran a Furmark XtremeBurn Test and the fan speed does indeed go up as it should, so I am not sure what kind of issues the other people are having regarding the GTX295. I also monitored my in game temperatures and they never passed 75*C which is perfectly normal.

Also did several Furmark test on the system I used for the benches above and temperatures are perfectly normal. I see no issue here what-so-ever.

Another thing is that NVIDIA cards are designed to throttle at a certain temperatures and if they don't do so they automatically shut off. So as you can see, by design they make it almost impossible to kill because of temperatures (even if drivers would overheat the cards for whatever reason).
See less See more
I wish you would have also included 190.62 whql which in my testing is the best driver release to date. I recorded my highest vantage scores with it and haven't seen any other tests that reinforced my findings.
Quote:

Originally Posted by mtbiker033 View Post
I wish you would have also included 190.62 whql which in my testing is the best driver release to date. I recorded my highest vantage scores with it and haven't seen any other tests that reinforced my findings.
What has been said about 190 and 191 is pretty much similar in performance/issues so I didn't really bother with them but feel free to subscribe and I may update the thread with other drivers in the nearby future.
See less See more
2
Quote:

Originally Posted by Open1Your1Eyes0 View Post
What has been said about 190 and 191 is pretty much similar in performance/issues so I didn't really bother with them but feel free to subscribe and I may update the thread with other drivers in the nearby future.
sub'd, thanks!
See less See more
Did you run these with NVCP 3D settings set to quality or performance?
Quote:

Originally Posted by PizzaMan View Post
Did you run these with NVCP 3D settings set to quality or performance?
Actually I like to keep it on "Let the 3D application decide". That way I can manually set the settings for each game and I don't have to worry about NVIDIA overriding anything (just adds to confusion).
See less See more
I'm talking about the 'Texture Filtering', its choices are: high quality, quality, performance, high performance. It doesn't have 'Let the 3D application decide'. You'll notice a nice boost if set to high performance.
Quote:

Originally Posted by PizzaMan View Post
I'm talking about the 'Texture Filtering', its choices are: high quality, quality, performance, high performance. It doesn't have 'Let the 3D application decide'. You'll notice a nice boost if set to high performance.
True, but then it's sacrificing some quality so I'd rather keep everything running as it should and let it control itself and nothing else (just a personal preference).
See less See more
Visually, I never noticed a difference in the image quality. Though, the frame rate is noticeable. I never run NV driver with default settings. All of them run like crap at default. The adjustments I make are pretty much the same things XG does.
Quote:

Originally Posted by PizzaMan View Post
Visually, I never noticed a difference in the image quality. Though, the frame rate is noticeable. I never run NV driver with default settings. All of them run like crap at default. The adjustments I make are pretty much the same things XG does.
May not notice it but you know it's there. I can't say NVIDIA drivers run like crap at default, only the later ones do (specifically 195 series). 186.18 drivers are definitely the best running drivers that have been released thus far. There may be better ones but I like to keep things unmodified and running like it should (limits problems). Even if other ones don't have any, I like to be sure. Also, I'm not willing to sacrifice quality for performance (even if it's unnoticeable).
See less See more
Wouldn't the newer drivers be geared towards improvements in the newer cards. I would think a last gen card wouldn't see a performance increase, but a newer GTX2xx series card might.

I guess what I'd like to see, is a single gpu test (to rule out the effect the drivers may have on SLI scaling) on a newer card. If anyone has seen a test like this done, links please.

Thx
1 - 20 of 67 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top