Could we get a nV interview about their opinion of some tech

[Arun]

Hey,

Just an idea I had...
We had a lot of nVidia interviews about the GFFX. None of them revealed amazing details, probably because they don't like talking about something which might hurt their reputation even further.

So, what about an interview talking about nVidia's opinion on some coming technologies?
For example, AGP Pro. They aren't supporting it in their Quadro FX because they think it would reduce the customer base. But does that mean they are against the whole idea of that tech, and never will implement it? Or at their planning to implement it in the future for a niche market, if the support is there?
Also, is there a possibility for a workstation-nForce with support for AGP Pro in the future? ( not sure, can AMD mobos actually implement that anyway? )

And what about FP24? Do they think doing something *between* FP16 and FP32 ( such as FP24 ) could be interesting? Or maybe putting support for both FP32 and FP24 instead of FP32 and FP16?

Water-Cooling: While nVidia doesn't force cooling solutions, is it possible that in the future they suggest Water-Cooling for their high-end instead of traditional or Flow FX cooling?


Just an idea... I'd like to see those type of interviews. Of course, not with too much PR answers... Just serious opinions about some technologies ( is it too expensive? Too buggy? Useless? )
Ah well, but then again it's unlikely nVidia ever responds seriously to those type of questions. But it could be worth trying, no?


Uttar

 

[Grall]

I don't think it would be possible to get a straight, non-PR answer on any of these (or similar) questions out of Nvidia, or some other 3D company either for that matter.

As for AGP Pro, why should they support it? There's no point in it. A floppy/harddrive-style power connector is just as effective and reliable anyway for all but the most clumsy/incompetent of users (and these would probably be afraid of removing/inserting cards anyway).

One problem with the floppy-style connector ATI uses is that it's quite possible to insert the connector the wrong way, thus sending 12V into the 5V connection (I've actually done this with a floppydrive, and it died instantly with an audible pop). Now, it might be that the 5V lead isn't connected at all (since it would be less efficient anyway than using 12V only to feed the card), but what happens when one sends 5V into the 12V power transformer? It would result in a voltage that is like 60% lower than it should be I would think.


*G*

 

[Dave Baumann]

Really there's not much point in pursuing AGP solutions anymore - its going to be outdated by PCI Express soon so makinf AGP more compatible with current graphics requirements. The real question lies in whether NVIDIA, ATI and the other have had sufficient input, and a loud enough voice (they will, afterall, be contending with Intel), for the bus to handle future graphics needs sufficiently.

 

[lwells]

Really there's not much point in pursuing AGP solutions anymore - its going to be outdated by PCI Express soon so makinf AGP more compatible with current graphics requirements. The real question lies in whether NVIDIA, ATI and the other have had sufficient input, and a loud enough voice (they will, afterall, be contending with Intel), for the bus to handle future graphics needs sufficiently.

Perhaps an interview with engineers (please, no more PR bunnies ) from a number of leading IHVs on their thoughts on the future, including such things as PCI express, etc?

Lucien.

 

[Ante P]

Re: Could we get a nV interview about their opinion of some

Water-Cooling: While nVidia doesn't force cooling solutions, is it possible that in the future they suggest Water-Cooling for their high-end instead of traditional or Flow FX cooling?

I asked them about it some time ago.

[NH]: We've recently seen manufacturers like Hercules releasing products that are "water cooling ready" are you currently planning any similar extreme cooling solutions?
[nV]: When it comes to cooling that's entirely up to our third party manufacturers. We give them a reference design that can be improved upon in many ways.

http://www.nordichardware.se/artiklar/Intervjuer/English/nVidia/

Not that this answer is crystal clear hehe, well when I spoke to them they sounded pretty uninterested in the idea.

 

[Ante P]

BTW a little bit off topic: the 256 MB FireGL X1 supports AGP Pro while the 128 MB version doesn't.

 

[gkar1]

One problem with the floppy-style connector ATI uses is that it's quite possible to insert the connector the wrong way, thus sending 12V into the 5V connection (I've actually done this with a floppydrive, and it died instantly with an audible pop). Now, it might be that the 5V lead isn't connected at all (since it would be less efficient anyway than using 12V only to feed the card), but what happens when one sends 5V into the 12V power transformer? It would result in a voltage that is like 60% lower than it should be I would think.


*G*

Well the connector can only go in one way(with the indentation pointing toward the card's PCB, AND it already comes attached to the card anyways. You would have to be the stupidest person in this planet to take the connector out and put it backwards again.
So yeah, your point is valid if you have had a lobotomy or some kind of mental retardation.

 

[John Reynolds]

Perhaps an interview with engineers (please, no more PR bunnies ) from a number of leading IHVs on their thoughts on the future, including such things as PCI express, etc?

Lucien.

Nice idea but I guarantee you company policy wouldn't allow it. And this isn't just Nvidia. For example, one of 3dfx's PR guys heavily edited an interview Gary Tarolli, co-founder and CTO, did in fall of '00. There was definitely a stark contrast between Gary's original answers and those that were published.

 

[lwells]

Nice idea but I guarantee you company policy wouldn't allow it. And this isn't just Nvidia. For example, one of 3dfx's PR guys heavily edited an interview Gary Tarolli, co-founder and CTO, did in fall of '00. There was definitely a stark contrast between Gary's original answers and those that were published.

We can but live in hope, no? My reasoning is, if they are not talking about specific future products, rather just technologies in the hazy future, they might be freer to speak.

Lucien.

 

[pocketmoon_]

Nice idea but I guarantee you company policy wouldn't allow it. And this isn't just Nvidia. For example, one of 3dfx's PR guys heavily edited an interview Gary Tarolli, co-founder and CTO, did in fall of '00. There was definitely a stark contrast between Gary's original answers and those that were published.

A long way from the days when Mr Tarolli used to post on the 3dfx voodoo1 newsgroups. Once he described how he'd fixed a performance problem with line drawing...