The highlights include significant information wrt margins and how they expect the G71 to be able to fit in the marketplace, as well as their cost advantage. They also leak very interesting tidbits wrt G80, and precise the RSX royalty timelines (so much for a spring PS3! *grins*)
Uttar
There ya go - enjoy. This took me more than one hour to write etc. and it's VERY near a full transcript (31:21 call), so... there ya go people!- Good morning everyone and welcome to our session with NVIDIA. My name is [...]
- Gross margins: significantly improved, expecting further improvements, what are the factors there?
=> The first thing that improves gross margins is superior products from a cost-pov.
=> performance per squared-millimeter
=> GF6, followed by GF7, are very efficient architectures, good gross margins at all pricepoints.
=> Key point: revenue by wafer. That's the primary factor here because wafers they're fixed-cost, not related to yields
=> with reconfigurability, you can sell it at a lower pricepoint if it has defects => 500 points.
=> We still feel there is good progress to be made, we think Q1 margins will be above Q4's.
- What's the timeline for 45%? Reconfigurability, etc. - what about 90nm, and how fast?
=> 90nm: all of our new parts that we are going to announce... soon... are 90nm.
=> We are making a very rapid transition in the GPU business.
=> Two benefits for the new parts: smaller (90nm) and better reconfigurability
=> GPU business' margins are currently above corporate average, in the 43% range, but further improvements.
=> Later this year, we will get benefit from the PS3 graphics part, 1+% margin improvement.
=> Is the 45% achievable this year? We'll see. It's certainly not out of the realm, but it'll require a lot of work.
- [unhearable]
=> [unhearable]
- Notebook GPUs: Double marketshare?
=> Starting in Q1, more evident in Q2, stuff rolling out will materialize.
=> You'd expect it to take 6 months to really take steam after the design wins
- What might still prevent you from getting 50-50 with your competition within 1 to 2 years?
=> Design wins windows are more critical that in the desktop because there's no notion of addin cards.
=> Primarly OEM market, no aspect of channel, so "timely" is extremely critical in the notebook world.
=> The OEMs typically start to compare the new designs now for the chips that will come out next year.
=> So your ability to deliver roadmaps that have high quality performance-per-watt now is critical.
=> That matters for perceived value too.
=> I think the situation could easily become more than 50-50.
=> One of the things that made the current situation what it is is that we were NOT timely at all.
=> GF6/GF7 went into production at the right time, so that's helping for share and still will.
=> Whoever manages to execute & have the good roadmap is going to win the business.
=> Focusing on "mainstream all the way up to enthusiasts". Goal isn't 50/50. (???)
- Traditionally high-end in the notebook market, getting high-volume low-end design wins, so... margins?
=> Significantly improved, but that's because of the GFFX=>GF6/GF7 transition.
=> Margins are "very very different" as I pointed out on the desktop.
=> Since it's an OEM business, margins tend to be very slightly lower, but not significantly at all.
=> Dominating that transition to GF6/GF7.
- What's happening in the notebook segment, low-end? What about inventories?
=> Inventory situation with notebooks is difficult, even more so now that we're ramping design wins.
=> We don't know exactly what's happening wrt mainstream notebooks because right now, we're not heavy in that market.
=> I'd like to be able to answer that question differently, but I can't, maybe a year from now...
- Desktop: performance segment, that helped the margins. 80% share. New competition products doing quite well -> ?
=> Improvement when it comes to margins comes architecturally.
=> Entire bucket of parts in the GF6/GF7 family has above margins than the corporate average => can't drag them down.
=> "We believe we have an advantage in architecture when it comes to performance per squared-millimeter"
=> And we'll get the added benefit of still having to move to 90nm.
=> The current competition part is much better, and a very good part.
=> But we're going to be extremely competitive, market expecting it to be too. Have to combine that with SLI.
=> [misc ranting about how nice they compete with SLI configurations]
=> Performance is the "sweet spot of the market", high price AND high volume.
- Cost advantage etc.? Do you think you'll be competitive?
=> Oh yeah, "speculations are a wonderful thing".
=> "We know we'll be competitive". 7800GTX is still competitive, benefit from faster process speed at 90nm next part.
=> With regards to how we price it, product that is faster & cheaper is ideal, we can capture more share that way.
=> Technology leadership, Performance Leadership, Leadership in SLI, and we have a huge cost advantage.
=> That's a marketing guy's dream.
=> There's no question that the market will bear very high MSRPs. So... "why would we drop the price?"
=> Lower price doesn't reduce the perception though etc., performance is what matters.
- We've had NV say "Crush the competition" in the past, and "leapfrog"
=> haha
- or things like that. So 30-40% higher performance etc., should we expect that again?
=> The only thing that matters in the enthusiast high-end market is that "it's the fastest in the world".
=> You don't have to CRUSH, you just have to be the fastest part bar none, even if it's fun. => SLI etc.
=> "But, hum, yeah... we'll crush them with the next part, hehe"
- Second half, 8 Series Product, another architecture... Is that going to be 90nm, architecture-level innovation etc.?
=> Hum, we don't want to pre-annouance parts. We will come out with a new architecture later this year.
=> I don't think we've said whether it will be 90nm or not, but obviously it probably will have to be.
=> It will be a "significant new architecture", we'll be using it for a couple of years.
=> Revenue [related to this architecture] for this year will be small, it's "primarly an architecture for next year".
- Okay so then a refresh for the desktop [for the GF7], and then for the notebooks
=> And then goodbye.
- Okay. Chipset business. One of the main growth drivers for the company, seems like it'll still be the case in 06.
- Potential of doubling year-over-year again. How much is related to AMD growing, how much is related to share gain?
=> Well, both. If AMD does better, we do better. We have expanded our product portfolio to include all segments on AMD.
=> So the AMD share gain & our share gain in that market is a very large portion of it.
=> We don't have very large marketshare on the Intel side. That's both good and bad.
=> Nothing based on that doubling is truly based on our expectations of Intel.
- What What are your expectations when it comes to AMD?
=> We feel our assumptions are consistent with what you guys have been saying about them (60-80M units etc. talk)
- Spectacular increase in chipset shipping in Q4. So, inventories in Q1? Chipset vendors cutting expectations?
=> On the Intel side, things are very different obviously, we need to differentiate both.
=> On the AMD side, I don't think we've been seeing anything here
=> The balance between AMD CPU and Chipset sales has been roughly consistent.
=> So if there's an imbalance, it'll be on the Intel side (because of production increases etc.)
=> On the AMD side, you're sharing the market with 3-4 people; not the case in the Intel market.
- IP/PS3 business. NRA changes?
=> The original NRAs was for the chipset in the PS3. We always said they would ask for additional desgins.
=> But there is no commitment. You know, I think some of the analysts expected a drop in Q1, but that won't happen now.
=> [unreachable]
- Royalties, Q2?
=> If we get royalties in Q2, it'd be at the end of the quarter, and not significant.
=> What they produce in June, we get in July, royalty-wise (one-month royalties at a time).
- Vista?
=> GPU, GPU, GPU. Scale it. Capabilities.
=> Have to scale it to integrated. Technology like ours is required for the best experience.
=> Reaction will be different based on whether it's consumer or enterprise.
=> Consumers will jump on it quickly, don't think so for enterprises. Focus on GFX.
=> As of Q2, I think the focus on "Vista Ready" will be higher, higher ASPs to computers being sold too.
=> Good for the computer industry in general.
=> What you might see in H2 is OEMs keeping their designs, motherboard/cpu/etc-wise but upgrading the GPU.
=> I don't think they want to change the base chipset, but improving the GPU or switching away from integrated, yeah maybe.
=> It requires a "GF6 or higher" to run Vista.
- Handheld business: revenue doubled, profitability remains behind, ATI is significantly ahead of you there.
=> "What do you mean by profitable?"
=> When you see the data from Q4, substantial difference from Q3.
=> Profitability not at corporate average, but improving. Big possibility to reduce the die over time => margins!
=> Long times between design wins to avaibility, and production cycles are longer too.
- Manufacturing side: 90nm ramping up significantly, are yields good? Relation to margin goals etc.?
=> So far, 90nm progress is very good, at both UMC and TSMC. Yields are good. Good job done by the fabs.
=> Have to work hard to transition. Yields are good, Costs are good, so nothing in that that could hurt our outlook.
=> The margins of the both new parts are better than those getting replaced; although that's what we expected.
=> Charter 90nm.. We're continuing to explore that, advantage that it's IBM[-based], and we would like to continue to utilize that.
=> As of right now, obviously, that's just work in progress.
- Thank you. Okay well thank you so much for joining us!
=> Okay, thank you.
[... breakout talk ...]
Uttar
