Revenue & operating expenses in line with expectations, but gross margins miss and thus EPS was 36 cents instead of 38 cents. Stock was down 8%+ in after hours before the CC started, but then recovered when Marv said margins would increase by 100 basis points quarter over quarter, so it was clearly a one-time phenomenom.
Press Release: http://phx.corporate-ir.net/phoenix.zhtml?c=116466&p=irol-newsArticle&ID=1142512&highlight=
My Transcript/Summary:
Press Release: http://phx.corporate-ir.net/phoenix.zhtml?c=116466&p=irol-newsArticle&ID=1142512&highlight=
My Transcript/Summary:
Mike Hara
Good afternoon.
On the call today, Jen-Hsun Huang and Marv Burkett.
[standard disclosures and information]
Jen-Hsun Huang
Today, revenue of $1.15B. 43% growth YoY. GAAP net income grew 34%.
GPU grew [missed].
MCP grew 31%
PSB grew 44%.
GPU ASPs grew 10% YoY.
We gained share in virtually every product category YoY.
For the performance segment, we estimate our segment share is 90%.
Overall, our market share grew from 30 to 33%. With double-attach, our share is approximately 42%.
Marv Burkett
Today we'll be discussing both GAAP and non-GAAP results. The difference is stock-based compensation and its tax effect.
Revenue was down 4% quarter-over-quarter, in line with expectations.
GPU revenue declined 8% QoQ. 44% desktop growth YoY, 99% notebook growth YoY, 36% memory
MCP revenue declined 4%.
PSB grew 21% QoQ, including
ASPs were held relatively flat QoQ.
Gross margins was disappointing. We didn't manage our product transition as well as we liked.
GPU margins declined nearly 2% QoQ. We had hoped to make yield improvements in the new products. This did not happen.
Heacount grew by 373, mostly R&D.
$151M property purchase in Santa Clara.
$143M cashflow from operations.
Foreign exchange losses of ~$4M over the quarter.
R&D tax credit not passed again yet, which obviously affects taxes.
We completed three acquisitions for cash in the quarter, including $133M stock repurchases in the quarter.
Inventories were up due to increases in inventory of new products, both announced and unannounced.
These inventory levels are in line with our target.
On the outlook, Q2 revenue was historically the most difficult to forecast.
We have been favoured both favorably and unfavorably. Other than that, we have no reason to believe it will be anything but seasonal.
That's 5 percent decline plus or minus. Gross Margins: up by 100 basis points.
We intend to moderate any increases in Operating Expenses going forward. The rate for hiring will moderate. The result should be flat to slightly up.
Jen-Hsun Huang
G80 to G92 was challenging. [... missed ...]
Still some G80 inventory to work through, but good positioning.
[Optimized PC marketing & impact on ASPs]
[Growth potential for GPUs in Lifestyle PCs & Transcoding marketing]
[CUDA as a growth opportunity]
[Mentioned Chevron for Oil & Gas CUDA usage]
[Mentioned CEA supercomputer design win]
Q&A
1)
Q: Gross margins. It sounds like it's just a transitional issue. If you look forward into July and future quarters, what to expect with the coming products?
A: I'm not ready to announce the future products yet, but the gross margins on the new products are better than existing products.
Q: Linearity of orders in the past month or two. Seasonal?
A: Hahah. I guess I'm trying to figure out. We haven't seen Q2 yet. You know, Q2 is a seasonal quarter on the industry, on the other hand this is also the quarter that we sometimes, many times, we announced new products as we go to the Back-to-School season. So that tends to balance each other. It's hard to see exactly how it'll turn out. We'll report on Q2 when we're done on it.
Q: New product question.
A: The products announced in Q2 will layer into the existing product line.
2)
Q: When do you expect significant revenue from Tesla and Application Processors?
A: Both significant I would say later this year. The lead times/design cycles tend to be much longer because of the kind of markets.
We're seeing a lot of enthusiasm out there. For application processors, design cycles are also very long, but we think both are going to pay off handsomely.
Q: CUDA platform; how big that market can be?
A: I happen to think that within the next 5 to 10 years, no longer than 10 years, and probably around 5 years, the heterogenous multicore computing will be the way of processing for all computers. For graphics obviously, also more importantly for physics and artificial intelligence simulation. [talk about game production cost]
We shipped 60 million CUDA GPUs already. In a couple years, that's several hundred million PCs with heterogenous computing in it.
Just one more comment: in the last 20 years I've been in the computer industry, the CPU has improved in performance a staggering 1000 times. And yet people are seeing 20-200x speedups with CUDA. That kind of a discontinuity in performance has never happened. That's the reason why just about every supercomputing center on the planet are all looking into heterogenous multi-core computing.
Q: Their best friends in Santa Clare, have any of their talk or actions affected [...]
A: I think any dialogue about GPUs is good for me. I'd love to have Good Morning America cover GPUs. [...]
3)
Q: Troublesome production lead times still out there? Your inventory is still above historical norms [...]
A: 60 is where I want to be. 50 or below is too low. The real question at 60 is whether it's old stuff that's moving slowly or new stuff. And I'm very confident there.
Production lead times continue to be a challenge for us since we're still growing as a company, but our fab partners are doing the best for us. [talk about how complex GPUs are]
When I grow 42% YoY, I know none of the foundries grew 42% YoY. So they're working very hard for us, and they'll need to help us continue to grow.
Q: Consumer business?
A: PS3s are doing great. The Bluray vs HD war is over. They're benefiting from that, first-party titles, second-generation titles, etc.
PS3 is the best Bluray player, flat-out. And so I think PS3 is going to do great. They're in the process of transitioning to a new process node and so we look forward
to see how they ramp in the holiday season.
Q: Desktop GPU segment expectations; pricing in Q2 and incremental pricing actions by some of your competitors?
A: It's hard to exactly predict what happens in the short-term, but I know what the general trends are. The general trends are that the GPU is benefiting from an industry-wide recognition of its increasing value.
You'll see GPUs where you didn't expect them. Units, ASPs and overall TAM will grow because of that. We're seeing all of those trends.
We'll see new products, but I can't control what the competition does.
Q: OpEx in Q3/Q4?
A: Will grow as we continue to hire. We're hoping to keep it flat in Q2, we'll see how successful we are. Moderate tick-up in Q3/Q4.
4)
Q: Margin guidance being 100 basis points next quarter.
A: Well you know, 100 basis points should be... you know, we can do much better in terms of manufacturing. What we do need is superb execution internally. It's not dependent on mix. We'll see.
Q: General & Administrative depending on revenue in Q3/Q4?
A: OpEx is not closely tied to revenue, it's closely tied to the number of people we hire and the number of projects we have. So independent movements.
5)
Q: Revenue in line with seasonality. GF9 ramping in Q2. Do you expect to gain some share in the back-to-school season? And if so shouldn't you be guiding higher than seasonality.
A: [Huge laugh from Marv]. You're right that we have a very good competitive position. We have a great products.
[Talk about 3DMark Vantage and their extraordinary position there].
Second thing is we're going to announce the highest-volume CUDA application this quarter, physics processing. And later in the quarter hopefully, transcoding.
I think people are going to be nuts about the work we're doing in that area [...]
Having said all that, I think we need to show Q2 its due respect.
Q: Notebook & July quarter.
A: It'll certainly play out in Q2. There'll be a little bit of drag in Q3. The major portion of it will be in Q2.
The notebook GPU business is much smaller than the desktop GPU business. And the business we lost is mostly entry-level.
So it's a small percentage of a small percentage. We need to keep it in perspective.
6)
Q: Margin benefit from manufacturing. Favorable mix, yet margins were below expectations. Is there going to be another element that's really helping through July?
A: Jen-Hsun outlined this: the G80 to G92 transition has been difficult for us. The overall GPU margins were down 2% QoQ.
You're going to get the benefit of one of those because you don't have to repeat it. And the other part is manufacturing.
So even without mix helping in Q2, we think 100 basis points is reasonable.
When you introduce a new product, it's either higher performance or lower cost. With G92 it's both faster and cheaper.
Combination of both process improvements and architecture improvements. So that makes the transition really tough.
Q: So why the miss?
A: We were in allocation on G80 until the day we had to sell it at lower margins. G92 reviews were just so good. It's just hard to manage those transitions.
Q: Desktop-Laptop crossover. Cannibalization etc.? Hybrid Technology consequences?
A: It's important to understand what the crossover actually says. What the data shows is desktops are growing slowly and notebooks are growing fast.
Notebooks are mostly low-end growth. Basic trend seems to be that for every desktop in a family, you have multiple notebooks.
But we're not in the vast majority of the world's PCs yet, we still have a lot of TAM growth ahead of us.
So I think the growth opportunity is still pretty healthy out there.
7)
Q: Inventory. Did you write-down inventory in Q1?
A: We always write-down inventory, but nothing abnormal. It was normal.
Q: Inventory at channel partners and OEMs?
A: It's higher than normal, but I think at this time of the year it tends to be higher than normal. It correlates with seasonality.
At about this time last year, inventory grew in the channel then it fell with Back-to-School etc.
So higher than last quarter but I think what I'm expecting for this time of the year.
Q: [... missing ...]
A: [...] There's no harm in being a little bit cautious about Q2. [...]
8)
Q: Motherboard GPU business in terms of pricing in the second half. And platform design wins?
A: Let's see. Marv you already mentioned the MCP data, 30% growth YoY and 4% decline QoQ.
No reason dynamics to change. In order for you to be Vista Premium certified, you need to have DX10 by July or August.
[...]
I don't expect anything except that graphics will be harder and harder to do.
Our lead is increasing. The rate of our lead increasing is increasing.
[...]
The VIA processor is very good. It's capable of running Vista Premium.
For $30 it's good enough. In less than a year or so, $10 will be good enough.
Q: Mobile computing/APX 2500?
A: Everyone is excited, just think about the kind of product you could build, etc. [...]
9)
Q: PSB. How many of that revenue was already Tesla?
A: The PSB revenue does not include Tesla.
Q: If I look at PSB, what are the drivers for this business going forward? AMD's FireGL threat, not that significant but?
A: At the highest level, the most important thing to recognize is that in every single industry, they're going digital!
Whether you're making comic books or designing buildings in China, everything is going digital.
We need to make the technology good enough to be able to replace the old alternatives too.
Car rendering needs to be incredibly realistic for example to replace clay models.
By changing people's pipeline, we can increase our TAM. Still early phase.
We're starting to see Quadro in showrooms. Every car company wants to do this now.
[...] There's so many ways this can still transform the industrial workflow.
And don't forget about off-sourcing, India is an opportunity there.
Regarding FireGL, as you probably realized, our business isn't called GPUs. It's called a solutions/services business.
It took us 10 years to get here. It's not about an add-in card or how big your framebuffer is.
All that just shows these people don't understand the business.
Q: Capacity constraints for PSB?
A: In terms of total units, it's not insignificant, it's very large. We ship more Quadros solutions in one year than every workstation company in the history of the industry ever shipped.
But in the totality of our business, it's too small, it'd never get in the way.
This is an intellectual capacity business. I would love to have a 100% more people who understand that business tommorow.
10)
Q: Robust in workstations, what drove the up-tick. And Q2, declines in the different units. And normal seasonality in the back-half of the year? In the macro context?
A: I don't think there's any specific segment that did well. They all did well.
It has exceeded everybody's expectations and it's still growing. We're not following any market research in terms of the size of the market. We are beyond market research.
It's not that easy to predict this quarter or next year, but you just have to go back to first principles to figure out how it'll do going forward.
Q: July quarter?
A: Notebook, as it makes the transition to Montevina, probably decline. It's a transition quarter. Not just share.
[...]
Q: Back-half seasonality?
A: 10%+ growth in Q3. We've had many years of 20%. And then in Q4, it's slightly up from Q3. Single digits.
Q: Getting that this time?
A: You tell me what's going to happen to consumers worldwide.
Q: Gross margins for the continuation of the year after the weakness in the first quarter?
A: We'll talk you in mid-August about Q3 margins.
Q: Inventory question.
A: [...]
Q: ASPs in general, solid this quarter, how do you generally see the framework for ASPs through the calendar year?
A: Has been stable to slightly up for years now. 10% growth in ASPs YoY, no reason to see this change.
