Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Jul 16, 2018.
I have had my Vega 56 scent Nov 2017 and been rock solid so far in gaming
On top of this, the actual number of failures related to gaming GPUs were;
AMD - 4 fails.
Nvidia - 15 fails.
However, 10 of those failures were on a GTX 1060, which skews the results here. I quote "Almost two-thirds of NVIDIA’s GeForce failures came from one card.".
I also quote "We noticed that for both companies that most of the failures happened on the workstation side. The Radeon Pro cards had 27 of AMD’s 31 failures and then the Quadro series had 61 of NVIDIA’s 76 failures.".
As the "under-dogs", it's understandable why AMD did this, sure. However, I'm also sure if it was Nvidia doing this, that there would be more backlash towards Nvidia.
The out-come of this (pr stunt) is that we should hope that Nvidia works towards improving their drivers for both their consumer and professional products.
Currently, I'm on my 9th month of continually upgrading my W10 install, one after another (literally every insider build so far) since November 2017. Most of my problems have been W10, Edge browser, appstore/games and audio related. My graphics card has given me the least problems (the main "problem" is that I have to re-install the Nvidia driver after every upgrade for all games to detect it properly).
Sorry schmidtbag that not true with Crimson Edition or Adrenalin Edition vs Catalyst Software Suite which at lease give you a Control Panel under Linux
If I remember right my first gpu was NVIDIA Riva TNT2 M64 Vanta, followed by NVIDIA GeForce2 MX 200. I don't remember having problems with the drivers. If I had a problem was with the slot 1 from the mb that was pure trash.
Then Nvidia 7300 gt, again no problems with the drivers, ATI HD 4350, HD 4650 no problems with the drivers HD 5670 except the UVD bug (watching something that used UVD clocks and trying to play a game was resulting in the card using the UVD clocks that were considerable lower hurting the performance, ATI/AMD refused to fix it in the drivers (they could had added an exception on for 5xxx and make the card ignore UVD clocks if it was in 3D mode; only fix was to change the UVD clocks to max 3D clocks in the bios) no other problems, then hd 7750 and here the nightmare starts until 13.2 beta texture corruption in dx9 games, drivers past 13.5 up to 16. make the games not feel smooth (doesn't matter what game), UVD bug after 16.7.2 no clue if they fix it or not I gave up at some point and I bought a GT 1030 that is working fine as long as I don't use newer drivers, newer drivers = stutter in PoE (if there is something that I hate at the GT 1030 is the lack of nvenc).
At least in my case newer stuff from both of them => problems.
"it only affected a subset of people" WTF. It was a common issue for the NV useres who used that specific driver.
The reality is, since Crimson and Adrenalin, AMD developed the overall quality of their drivers meanwhile NV just got worse and worse. I don't say that AMD drivers are better now, but at least are on par with the green drivers.
Yet the NV fanboys would probably say even today that you shouldn't buy AMD because it has shxxty drivers. lol.
The reality is you will believe any BS from AMD to justify your purchase. Fact, end-of-story. No research required.
I would not attribute "justify your purchase" mentality to someone with calculator graphics.
Yup, pretty much sums up this thread, look at all the triggered posts.
the facts do not lie
amd drivers don't totally suck anymore they just suck as bad as NVidia now
They both have their issues. Simple as that. If they didn't, you wouldn't see posts on our forums, other tech communities forums, and even AMD and Nvidia forums on driver issues. If drivers from both didn't have issues, you wouldn't see a list of driver hotfixes and bug fixes in the update manifest.
Fact is, AMD has had their share of problems with their drivers. I have an RX 480, and I've had small issues I won't lie, but nothing game breaking. Some people have experienced game breaking issues with AMD drivers though. The same could be said for Nvidia though. When I had a Titan, had small driver issues but nothing ground breaking. There were people with ground breaking issues on Nvidia and some still do. Ask anyone running a Fermi card still today how their DX12 and Vulkan support is. Ask anyone with an original Fury(not different revisions) how it is overclocking on them. Ask someone how their GTX 980 TI and multidisplay works. Ask anyone with a Polaris card what a black screen looks like.
In this day in age, they both have issues. Heck even similar issues sometimes even. Architecture and design wise Polaris is completely different from Pascal, but considering these 2 are programmed for the same API, same functions and feature sets, because they do have to have a similarity in their design on this aspect with both you most likely are going to run into the same kind of issues if not the same exact issues as well depending on the circumstance.
Ain`t speaking about bugs in drivers is complete off topic (since article and OP is about stability)?
LOL... so only one driver was tested (397.64) and most of the issues concerned one card. And you're absolutely correct, if Nvidia had commissioned this test and provided the samples to be tested, the sharks would go into a feeding frenzy. And probably an [H] article exposing the scandalous methodology used.
Remember that Nvidia caused about 40% of all crashes in Vista due to its drivers. People tend to forget this though.
But you can still hear jokes about red bus drivers being bad from the Ati days. It's not as common as it used to be, but it occurs.
There have also been reports of burned motherboards on laptop due to drivers and more recently drivers which have caused overheating in separate GPUs.
But, Nvidia is people's favorite brand.
Certainly AMD deserved a lot of criticism for bad frame timings, I'm glad this became part of the hardware testing and not just FPS rates.
I didn't play a lot then (too much work) but I do remember seeing the difference in Skyrim as quite amazing. Much smoother FPS.
Now I have on occasion trouble with waking up the monitor once I have only watched through the TV, but I'm uncertain if this is because of AMD drivers or the OS or a combination. Games are problem free.
No, part of stability issues can be caused by bugs. Even still, it's relevant to the topic.
actually it was just less than 30%
though vista? really?
that was a horrible OS, i will say it was the worse OS until Win10 (in many aspects).
and just like then, in win10 1803, nvidia drivers don't play well at all esp. compared to 1709. not an excuse, just pointing that out.
but i am curious why 397.64 was used when a newer driver was available before the AMD drivers 18.5.1 (that were used) were released.
there is not much i see in this . ."testing" that seems . . unbiased.
Vista was not bad, it was just a resource hog that required good disk performance and plenty of memory. Windows 7 was basically Vista 2.0 and its often heralded as Microsoft's best OS, which I think is overrated. Mellenium edition and windows 2000 were much worse imho.
This is pretty funny. AMD paid for this study and directly supplied cards to QA consultants, while the nVidia cards were made by Gigabyte and and provided to the QA by AMD. The youtube video has the comment section closed, and doesn't allow likes or dislikes. The motherboard and PSU were crap, and the Quadro is failing more than the vega.
That was also between 3-4 months after the release date, so the drivers had time to stabilize by then. Also depending what you do, your anecdote might not mean much. There is real evidence that Vega's drivers at launch date were shaky, to put it lightly.
With the open drivers, it very much is true. I would gladly sacrifice that mostly useless bloatware of a control panel for better performance and stability any day. Even in Windows, I don't install the additional control panel stuff.
All that being said though, I do understand some people do have a use for it, and Linux's open source drivers could really use something other than the old and cumbersome driconf. Also to my recollection, I think the closed-source Vulkan drivers tend to perform slightly faster.
Re Vista, were their equal numbers of AMD and Nvidia cards in use?