Using Intel's video drivers pros and cons

[hughlle]

I've never been given a reason to so much as consider updating or changing my video drivers.

 

[neurohax]

As a long time TabletPC user (long enough that I used the term 'TabletPC'), I avoid vendor drivers from past experience. In the past, the Nvidia and Intel graphics drivers I've installed have worked 95% of the time. Day to day use was fine and some apps performed better, but display rotation and external display setups would almost always be partially broken. For example, rotating the built-in display would rotate the image on an external monitor or projector, with no way to change the rotation independent of display.

I'm sure things have gotten significantly better, but I just can't risk running into unexpected issues like that when doing presentations or working out of town. So, 'installer beware': If you're going to be messing with GPU drivers, make sure to do a dry run on any external display presentations. (I have heard of the updated drivers causing issues with wireless displays, so maybe my fears are still founded. Not sure if it is the latest, latest, but a recent version.)

 

[Ordnas]

As a long time TabletPC user (long enough that I used the term 'TabletPC'), I avoid vendor drivers from past experience. In the past, the Nvidia and Intel graphics drivers I've installed have worked 95% of the time. Day to day use was fine and some apps performed better, but display rotation and external display setups would almost always be partially broken. For example, rotating the built-in display would rotate the image on an external monitor or projector, with no way to change the rotation independent of display.

I'm sure things have gotten significantly better, but I just can't risk running into unexpected issues like that when doing presentations or working out of town. So, 'installer beware': If you're going to be messing with GPU drivers, make sure to do a dry run on any external display presentations. (I have heard of the updated drivers causing issues with wireless displays, so maybe my fears are still founded. Not sure if it is the latest, latest, but a recent version.)

These drivers are the official ones from Intel. Not 3rd party drivers. They also make photos and videos look much better.

 

[vxm]

Interesting @neurohax , wasn't aware of that. I rarely use wireless display or present anything, so I haven't considered that. That said, in my line of work, when I'm using external display, tablet is in fixed position and I haven't ran into any problems so far.

 

[codrut]

As I tried recently to play games on SP3, for me, the Microsoft driver works better, with a Balanced power plan. I don't know why but it seems I get better frame rates. Don't forget that Intel drivers are generic, while the one from Microsoft might a custom one and maybe slightly modified. It would be quite interesting to hear more opinions about this.

 

[neurohax]

These drivers are the official ones from Intel. Not 3rd party drivers. They also make photos and videos look much better.

Yup, I know. So were the ones that I had trouble with. They may not do full testing on every release, and only the vendors and Intel know which have been fully regression tested. The versions I would install would be minor updates (like from v10.12.1 to 10.13.1), and bugs would be introduced that were wiped out in the later releases from the manufacturer (with similarly small increments), and even sometimes still not up to the most current. So the bug that is in 10.13.1 might not be in 10.12.5 and that was the vendor supplied "most current". It was rarely a problem with common stuff, or even resource intensive stuff. Just, almost always, tablet-y stuff.

Like I said, this was common 5+ years ago, but just a word of caution that Windows tablets are still relatively niche. 99% of users will not rotate their display, unplug a USB driven display at the same time as a DP driven display, and perform a bunch of other tablet specific actions, but I do those multiple times per day. Those bugs may not be caught quickly or may be tested for explicitly by MS (or other vendors) before rolling them into a "Firmware" (driver) update.

God, I sound like an old fuddy-dud. I guess I rely on my SP3 more than my past tablets.

 

[malberttoo]

Once a week or so I would try to watch something on Netflix or a movie from a thumb drive, and the blacks were just dark and so deep, if a scene was even remotely dark, then I couldn't make out anything at all that was going on. It was very frustrating. The Intel driver at least gave me the option to adjust the Gamma, which I know nothing about, but at least "turns down" the super inky blackness of the blacks.

Also if someone knows a more "proper" solution to my issue please let me know.

 

[vxm]

As I tried recently to play games on SP3, for me, the Microsoft driver works better, with a Balanced power plan. I don't know why but it seems I get better frame rates. Don't forget that Intel drivers are generic, while the one from Microsoft might a custom one and maybe slightly modified. It would be quite interesting to hear more opinions about this.

I too am using balanced power plan- "hacking" registry to add high performance for me only caused CPU to work on higher frequencies when idle and did not help with performance in games. For gaming I just use driver from Intel's website which adds control panel and allows me to set GPU for maximum performance rather than quality or balanced state. Also color correction came in handy because of my work in graphics.

As per performance, I did notice that default drivers were giving me lower framerates- I tested it on EvE Online and Guild Wars 2- can't be precise about frames per second difference, but on default driver and same settings games were choppy.

 

[neurohax]

@malberttoo I totally understand. There are definite benefits to trying new drivers. I mean, they wouldn't be releasing intentionally worse drivers. I guess my message is that there are things that go on behind closed doors that we, as consumers, are not privy to. So, before everyone has the knee-jerk reaction of "They're just slow to approve new software" or "They've given up on my product to promote the new version", we should keep in mind that there might be unintended consequences (and those operating on mission-critical machines should error on the side of caution).

For all we know, each x.x1 revision is not tested for interrupts between rotation and resolution change and each x.xx.5 is specific optimizations for IHDG 5xxx series. Usually, there are significant product-wide changes between each x.x release, but Intel is oooooold school and omnipresent, and I would not be the least bit surprised if they have some zany versioning system in place.

Last thought - I think many Surface users at least began with the Surface as an 'accessory computer' - something to use when away from their main workstation. So, it is either extra-tempting to install tweaks and interesting stuff to test, because worst case scenario is reverting to your main PC if things go haywire, or extra-repulsive to mess with a stable system, because you never know where you'll be when there's a problem. Both sides have their merits, and my perspective is that of someone that started out very much on the former and am now pretty much on the later.