EggXpert.com

I don't really get this question... graphics cards are installed in sockets. PCIe sockets. Most everyone is using them now so they are easily swapped and the amount of info the fastest cards puts out doesn't come close to pushing against the ceiling of PCIe x16, let alone 2.0.

I would really like to push my first question again, though:

How can one measure the amount of data passed over the PCIe lane and how much headroom do your flagship cards now have in a PCIe 1.1 x16 slot?  

Here is some good info on how the transfer rate is calculated.  then that would leave how much transfer rate are current video cards using on that calculated limit?

----(from wiki)

First-generation PCIe is often quoted to support a data rate of 250 MB/s in each direction, per (x1) lane. This figure is a calculation from the physical signaling rate (2.5 Gbaud) divided by the encoding overhead (10 bits per byte.) This means a sixteen lane (x16) PCIe card would then be theoretically capable of 250 MB/s * 16 = 4 GB/s in each direction. While this is correct in terms of data bytes, more meaningful calculations will be based on the usable data payload rate, which depends on the profile of the traffic, which is a function of the high-level (software) application and intermediate protocol levels. Like other high data rate serial interconnect systems, PCIe has a protocol and processing overhead due to the additional transfer robustness (CRC and Acknowledgments). Long continuous unidirectional transfers (such as those typical in high-performance storage controllers) can approach >95% of PCIe's raw (lane) data rate. These transfers also benefit the most from increased number of lanes (x2, x4, etc.) But in more typical applications (such as a USB or Ethernet controller), the traffic profile is characterized as short data packets with frequent enforced acknowledgments.^{[citation needed]} This type of traffic reduces the efficiency of the link, due to overhead from packet parsing and forced interrupts (either in the device's host interface or the PC's CPU.) This loss of efficiency is not particular to PCIe.

---

Your right MrMonroe

I worded the question poorly.

So now I will more on to explain the idea with a picture. (Source) 

This is a CPU expansion board out of ASROCK. I don't think there very popular but do think it's cool design route for graphics cards to take. Using this kind of design you could mix and match GPU chips and the GPU memory (within compatibility constraints) to make a custom graphics card on a universal platform far cheaper then any other way. So is this idea possible or do GPUs change in design so quickly that the idea becomes impractical?

Reactor89 

---

Those that say they know everything, may lose everything they know.

tallguy1618:

It seemed like they were advertising as well as answering the questions. So many things are unparalled that I've lost sight of what the true parallel is...

I agree most of the questions were looking for a small glimpse of whats coming in the future but nvidia only responded with things just coming out and we probably know more about than the person in the interview. The questions that they did not just push current technology we already know about on they said they could not comment on. I cant blame them as NDA are a bi......sucky thing for use curious people, but this definitely was not as interesting as I had hoped when reading the questions submitted.

Awesome post, thanks!

I must say, I'm a little perplexed (and disappointed) that the 9800 GX2 will initially only be supported by an AMD-capable mobo.  IMHO, Intel has jumped ahead in CPU performance, and I hope it isn't long before an Intel mobo is available to use this beast!

 That said, my rig is AMD (because they were still contenders at the time I built it).
 

RoboBro:

I must say, I'm a little perplexed (and disappointed) that the 9800 GX2 will initially only be supported by an AMD-capable mobo.  IMHO, Intel has jumped ahead in CPU performance, and I hope it isn't long before an Intel mobo is available to use this beast!

Where did you hear this? It's certainly not true. 

NVIDIA:

The first graphics card to support Hybrid Power is the GeForce 9800 GX2, which just launched March 18. The first motherboard to support Hybrid SLI is the nForce 780a for the AMD platform, which will launch in early April.

Kudos to the Egg Team, we all know you folks are doing this in our better interests. I too though am dissappointed they chose to clam up like a politician during an ivestigation hearing on Captial Hill.

If it's true there's more questions to be asked, I don't think they're going to expound on anything of great weight though. Sorry for my pessimism, but they couldn't have been anymore elusive on a majority of the key stuff that was asked.

I think in the next interview, they should be made aware that some of these issues we're really concerned with, and would like a bit more than canned alpo. Wonderwrench poses some really good concerns, kudos dude/ette (?). Also, are we all dupes, he said we can all get by with mid-range CPUs, just what is his cut off point for which grade CPU is over kill? I mean is anyone with Quad Core just tossing their $$$ away? I hate to think they are, or myself, someday.

In respect,

Penewah!

RoboBro:

NVIDIA:

The first graphics card to support Hybrid Power is the GeForce 9800 GX2, which just launched March 18. The first motherboard to support Hybrid SLI is the nForce 780a for the AMD platform, which will launch in early April.

Check it out, from the last two sentences of the first answer given.  I have no doubt that an Intel-based mobo will support it in the near future, but why should ATI be first?

AMD is the first to release a mobo with crummy integrated graphics that can use hybrid sli with the 9800gx2. The 9800gx2 still works fine on Intel boards. The only real cool thing about hybrid SLI is th energy savings because suposibly the computer will be smart enough to turn off the power hungry graphics card when running 2d apps  

Hey guys, just a quick update.  We should have part 2 of the interview by the end of this month. Nvidia said they're also going to provide a  bit of follow up to some of your Q's from part 1, for clarification, etc. 

Stay tuned...

-ning 

EGG: In the near future, will upgrading a video card be just like adding a new CPU and/or RAM to your motherboard? It seems that the more we demand out of these cards, the more real-estate they will need. So are swappable video processors and video ram even a thought?

NVIDIA: Users can and do upgrade their GeForce graphics cards with ease. SLI technology allows anyone with an nForce SLI-ready motherboard to simply add a second graphics card to the PC and install a new driver. The future is now!

I got a kick out of this response. Upgrade a video Card by purchasing another one! Now why didn't we think of that?

I give an A for effort to newegg, they really did a good job.

Nvidia however gets a D-. On one hand it was very cool of them to give the time and energy to newegg for the interview, but come on guys stop with the cookie cutter responses and answer the real questions. We know you can’t be so dense that you don’t understand what the questions are really asking.

I like how the part of my question concerning ageia and licensing the tech was answered by the simple "as well as future business". Meaning  what again?

Anyway, The fact that most of the questions went relatively unanswered, showed that we really did have some tough questions collectively, and most media outlets won't get into these types of questions. The press is supposed to make the interviewee sweat a bit and press the good stuff... don't see that done often enough anymore. I think this is a great way for both the community and the company to benefit and possibly see more product changes that consumers really want, not what their marketing department thinks we want.  That is, provided that these 'interviews' fall into the proper hands within the company. :-p

anyway, I enjoyed being a part of this thread, and may just have to go come up with something good to ask intel in the upcoming interview.

 

Well put, Nachowarrior. We just want to thank nVidia for stepping up to the plate, because we all know that most companies would skirt this type of opportunity to publicly answer questions from users.

Anyhow, we're looking forward to Round 2 of the Q&A session, which according to my watch should be in any time now!

-ning

are you going to ask them about the failure rate on the 790, 780, and 750 chipsets and why the last driver release for the 9800 cards didnt fix the fan speed reverting issue like they said it would?

dont get me wrong here, i use is nVidia products and love them for the most part(2-9800gtx sli, 1-9600gt512, 8800gs384, 680i a1 mobo) but these are the questions they need to be asked.

Their answers are so cheesy...

 "Using DirectX 10 vs. DirectX 9 doesn’t mean trading off performance."

Of course it does.