View Single Post
Old 03-28-10, 02:20 PM   #1
Registered User
Join Date: Jun 2009
Posts: 56,547
Post PAX East: Interview with Nvidia's Bryan Del Rizzo on the GTX400 series

After the Nvidia GeForce GTX 400 series launch event, I was lucky enough to grab an interview with Nvidia's Senior PR Manager of Consumer Products, Bryan Del Rizzo. Bryan was kind enough to give us sometime and answered each of my questions in a detailed manner, which we appreciated. Here is what he had to say‚?¶ What are your thoughts on the delays that Fermi was plagued with?

Bryan Del Rizzo: I mean, this is a chip that was designed from the ground up for DX11. It wasn't an after thought, we didn't take our old architecture and just tack on DX11 and then just, you know, bring it to market. This is a completely new design, a completely new architecture. It took us longer to get it to market than we expected, we would have liked to have it to market when Windows 7 shipped back in October. But we weren't able to do that, so we were late, and we admit that. But sometimes the wait is worth it. Were the causes of these delays things that you anticipated internally given the immensity of the architecture?

So, yes and no. When you're talking this complex a level of silicon, you never really know. You can design everything on paper first, and then you really have to wait for the silicon to come back. You have to test it, you have to make sure it validates and it's working the way you think it is. If it's not, you have to go back, you have to make some tweaks to the silicon, and that just takes time right? So I mean, it's just a matter of doing business. You generally never go to production with the first rev of the silicon anyway, if you do, you are awfully lucky. It is what it is right? We're not looking back, we're looking forward. Given the presentation today, it's pretty clear that you guys are putting a strong focus on gaming. But originally, before the product was officially launched, people were talking about how the GPU was really geared towards Tesla environments.

Bryan Del Rizzo: Yea, it's funny because, we had a show last year called the GTC (GPU Tech Conference), and so we launched the Fermi architecture, which we think of Fermi as the foundation for all of our chips, right? The issue is though, you can have Fermi-based chips in a bunch of different markets. You can have them in Tesla which is HPC (High Performance Computing), servers and things like that. You can have them in Quadro, which is content design, and GeForce, which is consumer based right? So we have this GTC, which is primarily an HPC conference, with engineers, designers, developers and guys doing cool stuff with GPU's, but it's not gaming centered. So people came to GTC or they looked at the presentations for GTC, and they said 'Well, what the crap? Nvidia's leaving the gaming market?' Haha, no! It wasn't a gaming conference, it was an HPC conference. So people didn't grasp that, they need to understand that we're doing all of this great stuff, and a chip can have many forms and functions, it can have many flavors. For some reason a lot of people didn't seem to get that. So they figured it just had one thing it was geared toward and that was it?

Bryan Del Rizzo: Yea, and the chip we're designing for GeForce is dramatically different than the chip we're designing for Tesla. In what ways?

Bryan Del Rizzo: Just in terms of the capability, functionality, things like that. And the usage model for what you're doing with GeForce is not the same as what you're doing with HPC. So there are hardware differences between GeForce and Tesla then?

Bryan Del Rizzo: Of course, the easiest one is double precision. There aren't any games, there aren't any user applications today that use double precision. But in the medical markets, in the HPC field, that's dramatically important to them. So we're not going waste the time and waste silicon space, and all that kind of stuff, providing a feature to gamers and incurring a cost for them, that they don't need or possibly want. Some news sources have suggested that there are more Fermi-based GPU's due sometime in June / early Summer. Is there any credibility to that?

Bryan Del Rizzo: Well, actually there probably is because I think one of our executives in the financial offices said that you can expect other variants of the GeForce family sometime in the second half of this year. So, based on what they said that probably is true. Can you give any insight into if these would be mainstream parts, or another high-end refresh?

Bryan Del Rizzo: It could be all, or any of the above. Certainly if you are looking at what Nvidia brings to market, they have a top to bottom product company right? Clearly the GTX 480, 470 fit into a higher end of the market, a higher end of the enthusiast class. So you can figure that we're going to have some more mainstream products with that, and you can figure that we can probably fit some cool stuff on top of that as well. So, expect the unexpected. Previously you've had dual GPU cards such as the GTX 295. Do you have any plans to do the same thing with the GTX 480 GPU?

Bryan Del Rizzo: So we can't comment on that today, we haven't announced any product. But I have to kind of laugh because when we launched the GeForce GTX 285, there were a lot of rumors that said we couldn't do a dual GPU product with that chip. They said it was architecturally impossible, it was too power hungry, it would just be impossible to do. And of course we brought the GTX 295 to market; it did really well! (laughs) So, if you look at the history of Nvidia, you can probably pretty much figure out what we're going to do, but I can't give you a confirmation on that today. The other day, SemiAccurate, another technology blog, reported that Nvidia was forcing resellers to purchase 80 pieces of your GT2XX series cards in order to get GTX 400 series cards. I have to ask, is there any truth to this that you are aware of?

Bryan Del Rizzo: If you look at the accuracy of that site, I think there's a reason why it's called SemiAccurate. Actually, I'd probably say it's really 'not accurate'. Given the immense fabrication of stories on that site, you know, it is what it is. Let me say this, if you go to Google and you search for Nvidia stories, stories from that site come up as either satire or parody. So, I'll leave it at that. Ok, fair enough, thanks.

Moving on‚?¶ No specs were given in regards to the TDP during the launch event, and the official reviews are just going live now. The rumors currently standing say that the GTX 480 is sitting at 250w TDP, is this accurate?

Bryan Del Rizzo: Yes, based on the press kits that we give out, for the GTX 480 it is 250, and I think for the 470‚?¶I need to double check, but I think it's 215 or something like that. So it's slightly higher than what the GTX 285's or 295's were. Did you guys have to do anything exotic with the cooling solutions in order to get them to run with stability?

Bryan Del Rizzo: Yea, I mean, it's a bigger chip‚?¶it's 3 billion transistors, so you gotta be crafty in terms of how you keep things running and cooling. It's a 4 heat pipe design on the GTX 480, the 470 is smaller so you've got to keep that in consideration. But, you also have to build yourself a great case around it, right? There's a lot of great chassis, from companies like Thermaltake, there's builders like Main Gear, I mean if you're building these great high performance PC's, it's standard cooling, standard air flow, standard stuff that you have to do. So it's not like it's radically different. I think some people are trying to make a big deal of it, but I think as an enthusiast you're probably less concerned about that than you are about performance. If you want balls to the wall performance, it's a slight trade-off right? It's slightly more heat, for awesome features, but basically just awesome performance, so take your pick. And if you don't want that, and you don't want the 480, then you can sit down with the 470, and that just comes down to user preference and stuff like that. What would you say is the GTX 4 series greatest performance feature in regards to gaming?

Bryan Del Rizzo: Tough question because I don't know if you're talking about gaming today or gaming in the future. Clearly we built the chip on DX11 tesselation, we have 16 tesselation engines that are scalable. So basically what that really means, the more tesselation in the game, the more performance you get out of the GPU. Versus say, the competitor products on the market today; they have a single tesselation unit. So what that means is if you have various degrees of tesselation, the tesselation performance stays flat. It doesn't matter what GPU you have from them, you're just gonna get the same level of tesselation performance. Which really doesn't do much for the developer, developing a game. Developers, well‚?¶they wanna crank **** up right? I mean, that's the whole point of having a scalable architecture. So I think for me, tesselation is killer, it's key to next generation titles, but they need the performance under the hood to drive that. So you think, for future titles especially, that tesselation is going to be a huge feature?

Bryan Del Rizzo: Yea, because I mean, the DX11 titles on the market today don't really take advantage of tesselation to a great degree. I mean, they claim DX11 support, but until the developers have an architecture where they can actually harness the power, and really play around with tesselation levels, I think this is what GTX 480 does. When you combine things like that with things that we're doing with ray-tracing, things like 3D Vision Surround, or just even 3D Vision. Our goal here is to change the level of the experience that you get. It's not too much about running a benchmark, it's about how you experience the game, how you immerse yourself in the game, how you feel in the game right? Everything we're building has that objective in mind, and I think that's what we're delivering. From a marketing perspective, would you say that 3D Vision is the most promising feature for you from a marketing perspective?

Bryan Del Rizzo: Well, 3D is hot. We've been shipping 3D Vision for about a year now. And the cool thing that we do with GeForce, for developers, they almost really don't have to do any work for 3D because we can do all of the graphic work ourselves in the GeForce graphics driver. So, all we're really doing is intercepting the signal to the game, and instead of rendering it once, we're actually rendering it twice, once to the left eye and once to the right eye, and then it syncs up and comes out as 3D. So from that perspective developers don't actually have a lot of work. What happened last year though is that they finally got a hold of our technology, they started playing with it and they said 'Holy ****, this stuff looks great!'. It's the first time that they'd seen a compelling 3D stereoscopic solution on the PC. People think of 3D, and they have this in their mindset, back in their head, they think of the red and blue paper glasses right? And everybody thinks that looks like ****. And it does, right? And that's the experience people are used to, so until they actually try it‚?¶ You know, the one thing I can say is that it's really hard to articulate and talk about it, you really have to see it. You do. We saw it today and it was extremely impressive.

Bryan Del Rizzo: It is. So we do all the real-time conversion ourselves in our driver, but developers like THQ and Blizzard and the folks that did Batman, and just a bunch of other developers, have gone back in and enhanced and optimized there games for it. That means they've fixed the depth, they rendered all of the backgrounds at the right depth level. They've made things look intrinsically more interesting when you get up close. You have the screen effect, in games like WoW, Batman and RE5, stuff will come off the screen at you as well. That's a completely different experience right? So, we're thankful that the developers like the technology, we're thankful that they are putting those customizations into the game, because today it's really about how does the consumer experience it. And if you play it on 2D and then you play it on 3D, you don't want to go back to 2D. If you play on a single 3D screen, and then you play it on three 3D screens, you don't want to go back to the single screen. It's just complete immersion, right? It's a whole new way to play the game. So would you say that a feature like 3D Vision gives you an edge over the competition?

Bryan Del Rizzo: I've always thought that we've had the edge on our competition anyway, because we're driving development of all of these cool technologies that developers take advantage of. 3D Vision specifically of course is one we've talked a lot about, ray-tracing, and tesselation. I mean, all of this combined just puts us head and shoulders above what our competition is doing. So at the end of the day it's not about running a single benchmark and saying..'well we're faster at x percentage', it's about‚?¶yea we want to be the fastest GPU on the planet obviously, right? And we are, but at the end of the day, it's more about how our consumers are experiencing the games that they're playing. And as long as we can drive the development and evolution of that, then we are doing our jobs. Sounds good Bryan. Thanks so much for your time!

News is offline   Reply With Quote