Devinder Kumar - AMD Chief Financial Officer at 2018 Deutsche Bank Technology Conference Call
Transcript
7nm node:
Q: You mentioned a little bit about the process technology, so why don’t we check that box as well. Last week or the week before, we saw global foundries throwing the towel on the 7-nanometer node. Talk a little bit about, holistically, your view on how AMD uses different foundries and what that change means via your WSA?
A: Yes. So if you go back to the context, and I know we talked about in the 2016 timeframe. When we laid out the multigenerational roadmap in terms of server, data center, commercial, we talked about having access to leading edge process technology. In 2016, we modified the WSA with GlobalFoundries and that gave us the flexibility in terms of having access to leading edge process technology. If our products are on time, we want to make sure that process technology was not a constraint in terms of introducing the products to the customers. And that’s exactly what the 2016 modification was about. As it turns out, it played out. Today, as we sit here, TSMC has done a very good job with execution on the 7-nanometer technology node.
We said back in 2015 and said in 2018 timeframe, we think our competitor is going to have. We thought they already have the 10-nanometer node out there. And we were prepared to go ahead and have our 7-nanometer products in the 2018 timeframe. We've stayed the schedule, their schedule has slipped. Today, with the GlobalFoundries evolving their strategy from a process technology standpoint, we are targeting all the 7-nanometer products at TSMC. And like I said earlier, sampling the GPU 7-nanometer second half of this year later this year; and then going ahead and launching it this year; and then in the server CPU space, launching that in 2019. So that’s playing out exactly as we had targeted and we’re very pleased with being able to stay on track with process and product technology.
Semicustom (consoles):
Q: That provided a ton of great revenue; whether it was the Sony side, the Microsoft side; now you have a Chinese game counsel builder as well; but great revenues to allow you to have the operating and earnings to invest in these other areas. But how do you think about the semicustom, going forward. Is that something that should be declining over time as this generation of counsels has peaked out? Or are you optimistic that there is going to be refreshes and/or new versions of semicustom opportunities?
A: We like the semicustom model a lot. Semicustom model is one of those; as you observe the game consoles, you win the designs; some of the engine and the expenses gets depraved by the input from the customers; we go ahead and get the chip out; and after that, it’s a mutually exclusive deal where you can predict revenue. Going back to 2012, 2013 timeframes, we’ve had predictably somewhere between $1.5 billion to $2 billion of revenues coming from the game console business, both Sony and Microsoft and that has allowed us to invest in exactly the roadmap that is delivering right now. We like that business a lot. We are competing for the next generation product. But Sony and Microsoft have to make their decisions and then taken we'll take it from there. But we like it a lot from an overall standpoint.
GPU excess stock, competition & Turing:
Q: Last question on the graphic side of the C&G. How do you view the competitive environment? Now that Nvidia has Turing out, it seems like they would at the very least, introduce a new high price point but push the prices down for their last generation chips. And that might be more of a direct competitive comparison for you/ Are you seeing any changes in the competitive dynamic?
A: The view is, first of all, the introduction of the product, the timing is very interesting. I think both companies are seeing elevated levels of graphics inventory in the channel space. We need to work through that over the next one or two quarters. And then obviously the ASPs for the new product that comes is very high. And I think the volume -- only when you get to the volume skews is they're going to be a benefit from a new product standpoint. We continue to have a roadmap in terms of introducing the 7-nanometer GPU for the data center, because that’s where the largest opportunity is for us from revenue and from the profit standpoint, and we’ll come out with the product from the competitive standpoint. I feel pretty good from a competitive standpoint in the graphics space. We have gained market share, overall, over the last 12 months or so, going below 30% or 33% and we'll continue to be comparative as we look forward from here.
7nm consumer GPU:
Q: And the absolute last question on the graphics side, 7 nanometer Vega coming to the data center side of it, you've talk about that before at the end of this year. When should we expect 7 nanometer to occur on the more traditional gaming…
A: We haven’t missed that piece. I think, if you look at it from what we have stated, we have 7 nanometer data center GPU launching later this year; we are sampling the 700 CPU this second half of ’18 and then launching in 2019; after that, we'll have the client piece of it; we haven’t been specific about the timing; and graphics will be coming out later than these products.
GPU computing ecosystem:
Q: NVIDIA, we talk about CUDA and the ecosystem around the programming to do the GPU computing side of things. How do you compete with that ecosystem from a software perspective?
A: I think, first of all, we have to invest in that area, which we have continued to invest. You’ve seen OpEx go up for the company and the largest area of investment is R&D. And in R&D the largest area is machine learning and software, that’s an area of investment. We have the hardware obviously coming out. We are investing in a big way on the software side of it. And then the other thing that I think is going to play out is the Open Source as opposed to the way CUDA works. And if you go back and look at literature, not in the financial columns and all of that, in the technical literature working with mega data center customers in particular, because they like the open software solution too, and now there’s a lot of discussion even by a competitor about open software as opposed to continuing with CUDA forevermore.
LinkBack URL
About LinkBacks
Reply With Quote
Originally Posted by CAT-THE-FIFTH
