Nvidia Editor's Day report

[hjs]

No topic yet ??

http://www.gamespot.com/pc/news/news_6077157.html

The panel then discussed Nvidia's comprehensive internal QA policy on optimizations, which states that the company refuses to optimize its drivers for specific benchmarks

Huang also made the interesting claim that although his company has recently experienced a loss of market share (Nvidia has traditionally sold the most graphics cards, from its entire product line, of any manufacturer), this loss was due not to competition from ATI, but rather, to competition from Intel's integrated graphics

There funny

 

[Trawler]

ideally, the company hopes to release as few as one driver update per year

 

[Tokelil]

ideally, the company hopes to release as few as one driver update per year

Well if their drivers where perfect there would be no need for driver opdates. (Not like thats gonna happen...)

I would have thought that Nvidia would have come up with some plan like Atis regarding driver releases. Almost all Ati users love that the drivers are updated so often, but Nvidia still sticks to "leaking" some beta drivers and hoping for the best. :?

 

[K.I.L.E.R]

Let me try and understand this.

They "HOPE" to release at least 1 driver update a year?

 

[jb]

Did you See HardOCP Stand on it?

Posted by Kyle 5:32 PM (CDT)

NVIDIA Event:
I have had more than a few of you email in and ask who is at NVIDIA this week seeing their dog and pony show about upcoming NVIDIA cards. I assume this was prompted by Anand's weblog post.


Right now Derek is out visiting NVIDIA and a bunch of game developers. I don't have a list of who made the final cut yet but there were some big name developers that were supposed to be at the meeting (I saw it as a response to ATI's Valve/Shader Day).

The fact of the matter is that no staff from HardOCP is at the NVIDIA offices for their event. I for one am growing a bit tired with some of the marketing techniques the video card companies employ and saw no reason for us to be there. After all the cheating scandals this year, drawing conclusions from what a timedemo tells us just is not good enough anymore, and it should not be for the consumer either. Besides we are busy with lots of GFFX 5950 and 5700Ulra work currently.

When it comes right down to it, there is nothing that NVIDIA can do or say this week that will change the gaming experience that their cards deliver. We have the card, we have the driver, and we have the games. That is all we need.. If there is information that NVIDIA needs to share with us that is going to change the end users experience with their video cards or is going to directly impact the way we evaluate that experience, then they certainly know where to find us in order to let us know.

We at HardOCP are all here working diligently, making sure we will have the best review we can write about the GFFX 5950 on Thursday morning. We are working diligently to remove the marketing from our reviews. We are working diligently to make sure our evaluations are reflective of real-world gameplay. This all takes a lot of time and we are hoping that our readers are truly benefited by the changes we are making and we hope that many others in our industry follow suit.

Also, as a side note, our 5700Ultra review will not be published till Monday morning. We need more time to evaluate the gaming experience that it delivers and are not going to push out a review too quickly so that we simply make an NVIDIA assigned embargo date and have it “posted firstâ€￾ along with a myriad of others. Simply put, we would rather get it right, than have it first. And we are banking on our readers appreciating that and showing their support.

Sounds good to me.

 

[John Reynolds]

Did you See HardOCP Stand on it?

Sounds good to me.

Sounds damn good.

 

[jb]

Guess I wanted to expand on it. I still not a fan of HardOCP. However I do like the fact that they would rather wait and try to get a review done "right" vrs rushing to mean the NDA experation deadline. That gives me some confidence. I also am glad to see that they are getting sick of the Dog and Pony show thats all PR. Yes everyone does it. Still they are right as all you need is the card, public drivers and the games to make your own conclusions

 

[Arun]

I'm VERY disgusted at NVIDIA right now. And I'm not the only one: Part 5 of my editorial which I sent for editing to someone came back even more bitter than the original - and saying I thought I'd complain I was being too extreme!

10 pages, 20K+ letters, and so on. You get the point.
I don't care how NVIDIA reacts to it. But I'm sure that if big investors read the text, they'd lost most of their confidence in NVIDIA.

Although if they did react positively to it, I'd be delighted. But then again, there's positively and "positively". Reacting in ways similar to the ones they have recently isn't a good idea IMO. It'd just put them in even more trouble than before.

Once again, I'll do my best to release P.S/ULE ASAP. And yes, I know I've been saying that "forever"


Uttar

 

[beyondhelp]

This is revealing...
"The panel then discussed Nvidia's comprehensive internal QA policy on optimizations, which states that the company refuses to optimize its drivers for specific benchmarks that emphasize features not found in real games"

"not found in real games" is the operative there as far as NV is concerned. That little turn of phrase lets THEM choose which are which and gives them an out with Games that DO use features found in real games. They don't bother Optimizing for Benchmarks that use functions Not used in games, (Yesterdays?, Todays?, Tomorrows?) but if the functions are used in games It's ok to Optimize for that Benchmark then? And what about Benchmarks that utilize Functions that may be used in FUTURE Games? , But are not presently used in Games, Or functions that are used in yesterdays and todays Games but will be dropped in future games for those new functions that will work faster, better in Future Games?

And this little tidbit was quite amusing...
"Huang also made the interesting claim that although his company has recently experienced a loss of market share (Nvidia has traditionally sold the most graphics cards, from its entire product line, of any manufacturer), this loss was due not to competition from ATI, but rather, to competition from Intel's integrated graphics."

Are they admitting to losing Market share to Intel? (Anything is better than admitting ATI is any good.) If that's the case, It must be petty bad when Customers prefer Intel Graphics over an Nv card... That's pretty sad. But then the Intel solution prob. does work better than FX5200

and Really? After seeing the quote about 1 driver update a year, I wonder what all the Gamers will think when they need a driver patch for some game that just came out and will have to wait up to a year for a fix.

They're still smoking something Halucinogenic, It's all Pipe Dreams! It's the same old Company Line for the most part. Yadda yadda yadda... Getting to be quite the Joke now!

 

[digitalwanderer]

I'm still laughing at the "one driver release a year" and waiting for them to issue a correction on that one later today.

I hope y'all like the 52.16 set, it's all you'll get for a while....

 

[digitalwanderer]

BetrayerX over at EB pointed out this great read at Tek Extreme, sort of a public rebuttal to the Gamespot article.

VERY amusing and cathartic. 8)

 

[Dio]

As the panel explained, the GeForce FX architecture favors long shaders and textures interleaved in pairs, while the Radeon 9800 architecture favors short shaders and textures in blocks

Long shaders like... those on DX9 applications? Seems that long shaders are better on Radeon 9800 to me

By the way, anyone should feel free on Radeon 9800 to do your texture loads any way you like. 'Blocks' aren't remotely important. Remember: "It Just Works"

 

[Xmas]

By the way, anyone should feel free on Radeon 9800 to do your texture loads any way you like. 'Blocks' aren't remotely important.

Except that blocks make it easier to spot when you hit the dependent read limit. And you should balance the number of texture reads and arithmetic instructions in every dependency level (ie. taking texture filtering settings into consideration, too), since internally, the R300 seems to be operating similar to the phase model of R200.

 

[jb]

Thats ok Uttar,

just PM me your article and all will be forgiven No? dang, worth a try.....

 

[Fred da Roza]

Did you See HardOCP Stand on it?

Sounds good to me.

Sounds damn good.

Well it seems Anand has picked up the ball. I am not suggesting Anand and previously Kyle are the only ones but why is it so difficult for some to see the blatant PR?

 

[Fodder]

I'm still laughing at the "one driver release a year" and waiting for them to issue a correction on that one later today.

I hope y'all like the 52.16 set, it's all you'll get for a while....

I'm sure every company is aiming to have compatibility/architecture/drivers of such high quality that they only need to update them once per year to support new products.

 

[Arun]

I'm still laughing at the "one driver release a year" and waiting for them to issue a correction on that one later today.

I hope y'all like the 52.16 set, it's all you'll get for a while....

Yeah, those guys are completely out of their mind, hehe. Their UDA is at fault here though.

jb: Well, I was seriously considering sending the final editorial 24 hours before the publishing to certain people ( not including you ) but I've finally decided that it's not worth it.
Sorry


Uttar

 

[GraphixViolence]

I'm sure every company is aiming to have compatibility/architecture/drivers of such high quality that they only need to update them once per year to support new products.

Drivers aren't just about making your hardware work, though. Even if a company managed to make perfect drivers that never had any issues even with newly released games (which is practically impossible), getting only one driver a year would mean waiting that long for any performance improvements or new features that are being worked on. This is why I think ATI's strategy of more driver updates instead of less is such a cool idea.

 

[Nazgul]

I'm sure every company is aiming to have compatibility/architecture/drivers of such high quality that they only need to update them once per year to support new products.

meanwhile, back in the real world....



seriously though, 3D gaming is far too complex an industry for any IHV to think they can get away with one driver release a year. Unless Nvidia can go a year or more without anyone finding a single bug in their drivers, they need to re-think that goal.

I'm still trying to figure out if the decision makers at Nvidia are a) complete idiots, or b) think we are. No intelligent, sane human being can possibly think another intelligent, sane human being would fall for this BS they're trying to pull.

 

[digitalwanderer]

I'm still trying to figure out if the decision makers at Nvidia are a) complete idiots, or b) think we are. No intelligent, sane human being can possibly think another intelligent, sane human being would fall for this BS they're trying to pull.

Both a & b are the correct answers, fueled by INCREDIBLE hubris. 8)