What's going on with NVIDIA?

[Rick67]

They haven't posted a new driver on their site since v23.11. Are they on vacation or what? I'm currently stuck with the v27.42's and there's all kinds of problems with these! They could at the least post a new driver for all the new GeForce 4 owners!! Come on NVIDIA get with the program! I paid $400 for this card and these drivers are NOT acceptable IMO. :disgust:

 

[Swanny]

There are plenty of drivers out for you to try. Why not go back to the 23.11s for a while? Which ones shipped with your card? Anyway, return it if you don't like the product.

 

[JellyBaby]

Early adopters always face issues. Clench up your butt cheeks and press on.

There should be 27.5x beta drivers to try. I hear these are better.

 

[BFG10K]

I'm currently stuck with the v27.42's and there's all kinds of problems with these!

So feel free to try the 2750s, 2752s, or 2770s.

 

[EdipisReks]

maybe instead of bad mouthing nVidia you should look a little harder. there is an inumberable amount of nVidia drivers, and i'm sure that one of them will work for you.

--jacob

 

[Rand]

Nothing is wrong with nVidia, they've always been rather slow to update their official drivers. If you don't like the 27.42's then try one of the later beta drivers after the 27.42's.

 

[johndoe52]

Where are you finding these drivers?

 

[AA0]

<< Early adopters always face issues. Clench up your butt cheeks and press on.

There should be 27.5x beta drivers to try. I hear these are better. >>



I love it, when new ATI cards have problems, it is ATI's awful drivers, when new nvidia cards have problems, its normal.

 

[vss1980]

Nice sentiment AA0!

 

[Insidious]

comprehensive list of NVidia Drivers can be found here

(I'm having great luck with 23.12 and the 60Hz fix of RivaTuner which is also available at the above site)

 

[tenoc]

Right on! NVidia drivers suck! ATI rules!

lol!

 

[Metalloid]

AA0, that is one of the truest statements I have ever heard. In fact, most of the time that Nvidia releases a new driver, it is worse than the last released one and everyone switches back. But noone ever thinks much of it. Nvidia is smart to release so many drivers so that the big swarm to choose from hides the many bad ones.

 

[JellyBaby]

I've never ripped ATi (much) for poor drivers, although I have ripped them on blatant benchmark and image quality cheating in their drivers.

IMO if you compare only official drivers nvidia has a better track record. If you compare beta drivers...well why are you comparing beta drivers? NV has more products than ATi and driver update frequency matches that fact.

That said ATi is only getting better (although they're STILL cheating to achieve high benches).

 

[LongCoolMother]

<< Right on! NVidia drivers suck! ATI rules!

lol! >>



uh oh...

 

[BD231]

Carefull tenoc, your words are flame bate. Try the 27.50's, they are the best out of all the current driver's available right now. My guess is that HWQL is taking their time with nVidia?s new driver?s, this type of driver support isn?t like nVidia.

 

[tenoc]

Sorry all! Didn't mean to raise a ruckus. I've a R8500 in my main unit using the latest supported
drivers and it is solid as the Rock of Gibraltar. (Cut scene - R of G after massive series of quakes
strikes Iberian peninsula.)

NVidia folk are constantly slagging ATI for their "pos" drivers and I just found it amusing that
someone admitted having problems with NV drivers! Always wondered why it was necessary
to upgrade them every week or two days - or whatever - if they were so wonderful to start with.

Btw, I use a GF2 gts pro in my backup unit, so I don't feel I'm an ATI fanboy!

That said, I do feel the R8500 Retail is THE bang-for-buck solution. Most current reviews show it
outpointing both the GF3 200 and 500. The GF4 4600 is 10 to 15% faster, but costs more than
twice as much.

imho.

 

[nemesismk2]

Nvidia drivers are usually pretty good IMO, however it's abit annoying that they take so long to update the drivers on their web site. When you compare them to PowerVR who usually have the latest Kyro2 driver release on their website almost monthly then it does seem abit lame that Nvidia can't do the same.

I have limited experience of ATI products so I can't comment on their driver releases!

 

[MemnochtheDevil]

I think different companies have different philosophies on drivers.

Nvidia has their official releases, which are usually more stable than the betas but don't have the newest performance enhancements/fixes. They also 'leak' many of their beta drivers giving all the tweakers out there tons of choices and bits to play with. I would say their driver release frequency is great if you like trying betas to good if you don't. (Plus I think their current cards are more of a mature tech, no revolutionary changes since the GF3 launch).

ATI has improved greatly from their past performance (I had a bad experience a while ago trying to help a friend). Their current releases have been much more reliable and timely. I still think the 8500 should have significant room for improvement as the hardware doesn't appear to be fully utilized by current drivers. I hope they manage to drive the chip to its full potential sometime in the near future. I'd say they rate a good or better.

I thought that the Kyro had to maintain constant driver updates because of the limited developer support for their unique architecture. If you didn't get new drivers every month or so most new games would have issues with their cards. (This is the company I have no personal experience with and have done minimal research on so I could be completely off on this one.)

A closely related side issue (at least to me) is ATI's current lack of credibility. They've been caught cheating on benchmarks and the wording on their response seems to indicate that they'll continue doing so. I don't view that as acceptable and I worry how it will affect future driver releases. Stability and real world performance are the only important factors and their taking control over the visual quality of a game from the gamer concerns me. We'll see what their future drivers show about their corporate attitude.

Just remember competition is very good for us users. The more ATI and Nvidia compete, the better the products and service we get. One of the reasons I'm hoping SIS's new chipset is a decent performer.

Anyways, just my thoughts on the matter.

Mem


edited to correct spelling and clarify in a few places..

 

[JellyBaby]

<< They've been caught cheating on benchmarks and the wording on their response seems to indicate that they'll continue doing so. >>

And yet nobody seems to care. They secretly reduce Quake 3 image quality to raise benchmark numbers but nobody cares. They don't apply anisotropic filtering properly to inflate benches but nobody cares. They revert to fixed AA patterns when fog is enabled to increase benches and nobody cares. They've been doing these things since the 1980s folks!

I can only guess they get a pass because they're the underdog and can do no wrong. Well, not in my book. I hate cheaters.

 

[JonDude]

Maybe if you ignore JellyBaby he'll go away ...

 

[JellyBaby]

The truth hurts doesn't it?

And really you don't want me to go away.

 

[Rick67]

Considering NVIDIA has released a whole new line of cards since posting the v23.11 drivers on their site, doesn't anyone else think a new set of drivers should have been posted by now?

 

[MemnochtheDevil]

Maybe, but not necessarily. The new card doesn't introduce any new features that need to be supported by new drivers. And the previously mentioned beta drivers that are all over the net hopefully provide options for people with stability problems.

I don't know when a new certified release is coming but even if it isn't soon, plenty of betas will be leaked. A new product doesn't always need new drivers.

I do admit it can be a pain trying out the betas, but I perfer it to being stuck with one 'official' driver and being out of luck if it doesn't work with your system.

And why is JellyBaby a pain? ATI should be getting their hands slapped for what they've been trying to pull. Wouldn't you be annoyed if you bought one of their new cards and brought it home, turned on your favorite game, cranked the detail levels up and found that it looked much worse than your neighbors computer, just so ATI could win a benchmark?

ATI has good hardware right now, they don't need to cheat but they've decided its a marketing strategy, I guess. :disgust:

Anyways, I don't see why nvidia needs to crank out an new official driver right now, what problems are you having that need to be fixed?

Mem

 

[JellyBaby]

I believe a big part of why initial NV reference drivers are always in a state of flux is due to the fact NV doesn't make the GF cards. Other manufacturers buy the GPU perhaps packaged with memory and either use NV's reference design or, more commonly, a custom design.

Just about everyone uses NV reference drivers, not the board maker's drivers. So NV has to support all the custom designed GF cards and surely that takes a lot of small driver tweaks hence all the beta drivers. Every time a VisionTek or LeadTek or MSI makes a slightly different board, NVs reference drivers may need small tweaks to best work with it.

ATi designs, builds and sells their own brand so they don't face these issues (but yes they have plenty others to deal with).

 

[vss1980]

Actually, that isn't strictly true. As all of the chips in a set range (say Geforce3) have the same pin layout (no matter which manufacturer makes the card), any design/layout changes on the cards must be directly compatible with each other and the reference drivers. Remember, its nvidia who make the chips not MSI or Asus, etc.
The main design changes that may be tweaked out in the drivers is the differences in the BIOS used on cards as few people rarely update their graphics card BIOS so I assume a lot of fixes will be worked in at the driver level.

Also, as for so called cheating in drivers, ATI are not the first and I doubt will be the last:

1) Sticking with Quake3, the S3 Savage 2000 had stunning performance in Quake3 in 32-bit when compared with its technical equal (the GF1 SDR card).
Why? Mainly because by default it never used a 32-bit Z-buffer and hence economised on memory bandwidth. Clever, but definitely a 'cheat'. Although you could always tweak a higher Z-buffer...

2) Nvidia's turn I think. Their success today mainly relies on the fact that the TNT and TNT2 were such good cards for their time. Without a doubt, these are the chips that lifted them above the competition and started them on their way to market domination. However, just dont plan on getting any good texture filtering from the card. Fair enough, they didn't declare it could do proper tri-linear filtering, but making the option appear as so is really a 'cheat'.
Also, I still think that their IQ for the TNT/2 and early Geforce cards was pretty bad compared to others (no matter what settings you chose - ie. Best quality or Best speed - basically LOD adjust) which is a bit of a 'cheat' especially as it was always noticable bad.