Nvidia was cited as causing nearly 30 percent of early Vista crashes, while Microsoft itself was to blame for nearly 18 percent. Intel and ATI each accounted for about 9 percent of crashes, according to the documents in the Windows Vista Capable lawsuit.
I’m not really surprised on this. I cannot remember the number of times XP has crashed, only to be fixed by an nVidia driver upgrade.
Yeah, not that the upgrade always fixes it either.
Is it any wonder PC gaming is suffering? I don’t think I own a single game that won’t crash from time to time, and it’s always the graphics drivers to blame.
That’s not to say that Nvidia are any worse than anyone else – they just have the most market share. I think competition in the graphics sector has just pushed the technology too far too fast. Heck, even the consoles are having stability problems!
So let me get this straight, the company that develops DirectX blames the largest and one of their closest device driver partners for 30% of the crashes in their operating system?
Is it not encumbment upon Microsoft to make sure that through QA the device drivers not only adhere to their APIs, especially since they are used in the XBox/XBox360, but work without compromising the operating system and causing crashes?
Telling your customers that “they did it” isn’t solving any flaws in both the device driver code and the operating system code.
Mutual responsibility drives sales by fixing the problem(s).
It doesn’t look as if Microsoft was trying to spread blame around. This is evidence obtained during a lawsuit, not Microsoft “telling [their] customers” that someone else did it.
To wit, the very first line of the article is, There is a ton of interesting information in the documents that have been released as part of the Windows Vista Capable lawsuit. From the original article in the Seattle Post-Intelligencer, A federal judge today unsealed internal Microsoft e-mails that have been used to support the plaintiffs’ case in the lawsuit over the “Windows Vista Capable” program.
Curiously, the original article in the P-I is more concerned with an Intel chipset, and nVidia isn’t even mentioned. It even quotes one email: In the end, we lowered the requirement to help Intel make their quarterly earnings so they could continue to sell motherboards with the 915 graphics embedded. Now, there’s a boneheaded decision if ever I saw one. When Wal-Mart is complaining about lowering standards having the potential to cause customer confusion, you’ve got a problem.
P-I?: Would that be the Seattl Post Intelligencer? If so, I’m not surprised one bit. The Times and the P-I have been in the tank for Microsoft and Boeing for decades.
Instead of us having 3 or more Consumer Airline manufacturers in the United States we’ve got 1 giant behemoth whining about another US Company Northrup Grumman winning a contract by using Airbus and foreign parts [of course Boeing a huge whore with foreign parts and labor for engineering but we wouldn’t want to say that] and as a Mechanical Engineer from the Northwest I am rooting for Northrup.
More competition for contracts increases best of breed results.
I do not see why everyone seems to think this a big issue. Why blame a hardware company for a software computability issues. If Microsoft would not have locked down Vista ‘for security reasons’ to the point were I can’t even install iTunes without “being sure I want to do this” 5 times over, than maybe Macs would not be so popular right now. I speak with a biased as web designer against a company that just released IE8 beta, and yet again decided to set a ‘new and improved’ set of web standards.
I believe the original article had a pie chart of all the sources of reported windows crashes. With all the other issues Vista had I’m not surprised they let a lot of crap fall through.
Incidentally ati was 9.3% versus nvidia’s 28.8% (nvidia does have greater market share tho)
Even as far as aug. 07 I was getting crashes from the nvidia driver
Do you honestly expect Microsoft to audit every single driver and version of it for every piece of hardware?
I think it’s nVidia’s responsibility to ensure their drivers work correctly.
Every single one? No. But there are issues with WHQL tested drivers too.
That still doesn’t make Microsoft responsible for the poor quality of the drivers in the first place.
It might however demonstrate a problem in the qualification process to have drivers certified.
The whole point of a WHQL is it is meant to put these drivers under stress tests under many different conditions – Microsoft has also pushed the idea that if you ‘use an WHQL driver, you will have a reliable experience’.
It has been Microsoft who has been pushing the myth of WHQL and the ‘superior reliability’, not the hardware companies. When you hear WHQL, you assume that they’re driver which have been strenuously tested – the fact is, that isn’t the case.
With that being said, Microsoft’s fetish to make the new driver model a whole lot more complex than it needs to be, doesn’t exactly raise much hope in driver companies being able to write drivers which are of a reliable nature. When you make something complex, and ignore the human element of those who will be relying on that said piece of architecture, its doom to pain and misery.
Oh boo hoo. Nvidia doesn’t have that many drivers to be certified. Microsoft most certainly has the resources and they even charge for it.
So how much time and money should third parties spend to do Nvidia’s job?
”
So how much time and money should third parties spend to do Nvidia’s job?
”
They could release usable driver interface specs then see how much time third parties are willing to spend exploring what there hardware can actually do.
Fat-Lazy-Indignant corporate fingerpointing at it’s finest. “Your untested junk made our untested junk break.”
I just had a crash with nvidia drivers in my vista machine, it was after waking up my computer.
So, I must say screw you nvidia, next time I won’t be buying a nvidia card.
So did I on my Linux machine. Oh wait, my mistake. I didn’t.
Don’t misunderstand me, I don’t want to start any flamewar between Linux and Windows, my point just is that there must be something wrong with the underlying OS also if faulty display drivers can crash the whole system. Wasn’t that supposed to be fixed in Vista?
I don’t know what you’re smoking, but buggy nVidia and any other driver damn well will take down a Linux system, and I’ve had it happen to me and I am certain I’m not alone.
Claiming that the Linux driver architecture is somehow better than the current one in Windows also is hard to swallow especially where X.org etc are concerned. The fact that X mostly runs in userspace is completely irrelevant as it still plays with the hardware directly, bypassing any protections that would otherwise be afforded.
Google “Loic Duflot” for more information on this topic.
All current software sucks, and as always, drivers are especially irksome.
I’m afraid I’m going to have to call you out on that. Many times when nvidia drivers dont work you can still get the console, the change to vesa etc.
But many times I’ve had hard lockups where the screen freezes that have been fixed by driver updates, whilst those sort of problems did not occur with the frequency of the initial nvidia vista issues, they still occurred.
as for vista crashes, most of the time the driver would crash but vista would recover other times it would bsod. It depended where the problem was.
Edited 2008-04-01 12:24 UTC
Nvidia drivers in Vista have been bad for the past year, pretty much users have been beta testing them. You’ve been needing beta drivers to play the latest games proper and so putting your systems stability at risk.
Nvidia blame the fact that it’s a new OS and driver model, so it takes them at least a year to get to grips with it all.
Because I’ve seen two or three buddies with SIS chipset mainboards and ATI video cards that have never had a crash with Vista.
It doesn’t even last me 28 hours before I have to rip it out for being unstable. nForce 680i chipset mainboard with a 640 meg Ge8800GTS driving the center display and a Ge8400GS driving the two outers? If this is to be believed it could be why EVERY application I try to run under Vista randomly locks up and has to be killed. (even though the OS doesn’t recognize the program as not responding)
Not that it matters, I’m completely happy with XP x64.