Official AMD Q1 Earnings Thread -

Page 6 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Toggle sidebar Toggle sidebar
iraq_denial.jpg


"45nm in 1H 2008"
 
He also said "hats off to our competitors, they've done really well in closing the gap" ...one for Sharikou.
Click to expand...

I think AMD is still deluded in thinking they have a lead. Or at least, that's what they're telling investors.

"We can neither confirm, nor deny Intel's dominance".
 
Nice.

-r0ck

don't quote pictures, it uses more bandwidth and breaks links much quicker. 😀
Click to expand...

It does? I thought it downloads the image once, and when another instance is repeated on the page, the browser recognizes and reads from the cache.
 
Nice.

-r0ck

don't quote pictures, it uses more bandwidth and breaks links much quicker. 😀
Click to expand...

It does? I thought it downloads the image once, and when another instance is repeated on the page, the browser recognizes and reads from the cache.

You'd think, but I've seen it happen many times. It likely creates a copy. I don't know for sure, just warning you that is all.

Many people have complained before, so I just avoid quoting pictures just in case it is true. :wink:
 
He also said "hats off to our competitors, they've done really well in closing the gap" ...one for Sharikou.
Click to expand...

I think AMD is still deluded in thinking they have a lead. Or at least, that's what they're telling investors.

"We can neither confirm, nor deny Intel's dominance".

They did say say their competitors have a monopoly during the conference.
 
Conjecture.

It may be that the only thing which will reduce AMD/ATI red ink in Q2 will be a very strong showing from the R600 line in the $150 to $250 price point. That is of course assuming they can make good on the "We don't do paper launches" comment.

I say this because there appears to be a sizable hole(performance wise) in nVidia's DX10 line and their midrange products aren't overly impressive at the moment.

This is based on the "DirectX 10 For The Masses" article here at Toms and an interesting price segment breakdown/analysis done by X-bit labs titled "Integrated Chipsets Kill Low-End Graphics Cards Sales".

X-bit goes over the numbers of discrete graphics cards sold for the entirety of last year. The short version is about 2 out of 3 graphics cards sold last year where sold in the $150 to $250 range with the ASP being ~$245. The same price range accounted for on the order of about 3 out of 4 dollars in revenue for discrete graphics last year.

If AMD/ATI can deliver a killer product or a couple of superior products in the performance/midrange segment and keep up with demand then I think they will go a long way towards reducing the bloodletting and red ink.

Thoughts?
 
You won't likely see a huge effect in Q2. Your comments are valid and have merit, but the introduction of the R600 based cards will be too late to make a substantial difference in the Q2 numbers, hence why they estimate revenue to be about "flat" during the quarter. Q2 is usually a seasonally down quarter, so they expect the 1 month of R600 sales to help prop up the overall quarter, but as for saving a good deal of red ink, I don't think there is enough time.

If the launch holds true, on time, and with appropriate quantities, Q3 should be a much better quarter, maybe not profitable (very unlikely), but if they can limit EPS to a 25 cents then they will be doing fairly well. Mind you ML's EPS for Q3 is in the 40 cents territory (loss obviously).
 
Conjecture.

It may be that the only thing which will reduce AMD/ATI red ink in Q2 will be a very strong showing from the R600 line in the $150 to $250 price point. That is of course assuming they can make good on the "We don't do paper launches" comment.

I say this because there appears to be a sizable hole(performance wise) in nVidia's DX10 line and their midrange products aren't overly impressive at the moment.

This is based on the "DirectX 10 For The Masses" article here at Toms and an interesting price segment breakdown/analysis done by X-bit labs titled "Integrated Chipsets Kill Low-End Graphics Cards Sales".

X-bit goes over the numbers of discrete graphics cards sold for the entirety of last year. The short version is about 2 out of 3 graphics cards sold last year where sold in the $150 to $250 range with the ASP being ~$245. The same price range accounted for on the order of about 3 out of 4 dollars in revenue for discrete graphics last year.

If AMD/ATI can deliver a killer product or a couple of superior products in the performance/midrange segment and keep up with demand then I think they will go a long way towards reducing the bloodletting and red ink.

Thoughts?
Click to expand...

There are two problems here.

First, ATi revenue has historically been ~$1.2Billion/year. Profit on $400-$500Million, that includes a lot of low-end stuff, isn't going be the deciding factor if they are trying to pare down a $600Million+ shortfall.

Second, I wouldn't call the 8600 a hole exactly. It pretty much matches up with the top end of the previous generation on older games and beats them on Oblivion. The 7600 and 1600 can't claim that kind of performance. I'm not sure what Polkowski was expecting. He must have been looking at too many Core 2 benchmarks.
 
Apologies for typos, bias, etc., but I was typing as fast as I could. I'm clearly no stenographer. These are just the notes I took from the conference call. There was a little more, but I got bored. 😀

For 100% accuracy, listen to the AMD feed off their investor link.

Thank you and welcome...

<Introductions>

<Numbers to reach AMD telephone replay>

<Future events...potlucks...etc>

<Disclaimer of futures>

Dirk:

After 14 Qs of gaining share, Q1 is a setback
Need to understand problems and fix them
Business is more complex (more customers, more products)
Complexity led to challenges

Sufferred growing pains.
This due partially to growing global OEMs
Pricing pressure continued in CPU
Competitor did everything to protect their monopoly
Weakening demand in consumer electronics
Sum of all problems caused a perfect storm

Need to grow top line. Simplified S&M group.
Change cost structure - slowing the rate of F38 conversion. Limiting hiring to critical hires. Will leave year with lower HC, than started.
Deliver on roadmaps. Need to deliver on Barcelona. Customers are excited. Need to covert F30 to 300mm
Starting 45nm Q1 2008
Need to get and maintain GPU leadership

Unmatched tech portfolio for customers

Bob:
1st quarter unacceptable, but we can fix issues.
Blame ATI for some of the loss
Net loss of $1.11 per share. ATI expenses, tax expenses, and stock expenses.
Operating loss was $363 million
Gross margin was 28%
Lower microprocessor sales, lower microprocessor ASPs and ATI costs
Lower sales and revenues across all platforms
Chipset sales increased.
197 million revenue (up 19%) for graphics. 35 million dollar loss for net profit
Handheld and gaming decreased
Cash - 1.2 billion (600 million above min acceptable lows).
Selling 200mm fab equipment
Selling admin land and offices
Looking for financing opportunities
Inventories up due to sales misses. 65nm products were a huge chunk of inventory issue.
2nd quarter outlook - revenue to be flat, to slightly up. Cost savings should show up in 2nd half more than in 2nd quarter.
Q1 2007 was a terrible start, but confident in plan

Hector:
Thanks, Bob
Complete package is strong
This quarter was more than a miss.
Major inflection point in timing of innovation.
Only served to accelerate what we're doing
Beginning of restructuring is occurring.
Need to balance the different market segments.
Customer-centric innovation
Transformation should be larger than in 2002.

Mike for Q&A:
Tim from Lehman Brothers
Q: Could you clarify how we should perceive your pace of investment in capacity (vis a vis prio plans).

Also ...regarding...clearing inventory levels. Any updates with Barcelona?
A (Dirk): Capacity plan was to facility F36 and start F30 conversion in middle of the year. 38 to start in 2008. Only change is that we'll reduce rate 30 conversion. Still fully ramp F36. Barcelona - started shipping pre-prod samples to Customers. Should ship prod samples in Q2. Should ship actual for sale in Q3.
Q: Guidance for Revenue is stronger than competitor
A: Obviously we're disappointed in our performance. Channel should be increasing in Q2, where as they didn't in Q1 as much as we'd like. Launched new products in graphics business. So we have high confidence.

Q: Joanne Feeny - Like to hear more about restructuring. Like TI, perhaps use a more asset lite strategy
A (Hector): Broad and challenging question. Level of restructuring is significant. 4 years ago, had one product in the company - it could serve many segments. Now we have a broader mix of products. Products
are more matched up to their respective segments. We are changing from heavy channel to heavier OEM weighting. How we manage accounts is different. Will be announcing changes. Had good experience artnering with other companies, like IBM. IBM let us go without an R&D fab. Learned a lot from foundry partners. With ATI, we have a good perspective into asset lite. It's unique for us. What worked for TI
is tailored for them. We need something unique for us. We have to be in bleeding edge for server markets, but not necessary for emerging markets. Obviously can't do this in a quarter or two, but should
give more insight this summer.

Q (JP Morgan): Go through pricing expectations. Also talk about GM
A: Stabilization in pricing since the beginning of the year. Still incredibly competative, though. GM from high level, very disappointing. Plans in place. New products in GPU world next month. Still in early stages of 65nm. Takes 10 weeks to push product through a fab. Every day, more and more is 65nm. Die is 30-40% smaller (65 compared to 90nm). Watching ourselves closely in the R&D department. Hopefully
can be less in 2nd half

Q (UBS): Pricing question for Henri. ASPs for server didn't do well.
A: Server ASP price went up in the quarter. Barcy will help. We also want to be focused on volume. Q1 has been successful on design wins, despite revenue. Barcy can't come quick enough.
Q: Referred to channel share loss. When will you win that back
A: Predictability and availability important in channel. Shift is occurring from desktop to mobile. Advantage of having more inventory, we can supply channel better for their particular needs.
Q: On computing side, can you give MPU/chipset split?
A: No

Q: Pricing strategy was agressive. Will you back off?
A: Our strategy hasn't changed. We are pricing them for the value they provide in the market. But now we are a platform, so we can do more interesting things. Increased focus on graphics. Slightly different
pricing strategy in the future since we have platforms, rather than single point.
Q: Any future special products/vision?
A: Excited about future product peformance/watt/square of silicon. Idea of specific components is exciting, but we won't release information on our future products in that space
Q: Raising money?
A: 600 million is the deadly zone. We're still above that. Working capital needs to be as robust as possible. Focused on appropriate liquidity

Q (ML): Back to Barcy. Sample in Q2? Will we see some desktop parts in Q3? Curious about mix vs Brisbane
A: You will see desktop in 2C/4C in Q3. Won't get into detail about mix
Q: Looking at mobile. Turion at 65nm is up for refresh. When new arch?
A: Product lined up in 2008 - 65nm new CPU and platform. 1h of 2008

Q (Prudential): Cash flow side. How many grants and subsidies for Dresden? How much for 200mm tools?
A: In broad terms, they are north of 200 million dollars each. We've started a little in Q1, but more to follow through the year.
Q: Limitations on what you can outsource?
A: We have lots of flexibility in our agreements (near infinite). Narrowing down choices and opportunities with our partners. Already have joint development with IBM (mentioned before). Want to
structure business units so they focus on what's important. Perhaps whole business can be run outside of the US.

Q (AG Edwards): Mobile ASPs - what were they like?
A: Marginally down. ASPs were the small piece of the equation. Really, the unit drop QoQ across all segments was down.
Q: 45nm schedule?
A: Starting in dresden in 1h of 2008. Shipping product 2h, looking to pull that in.

Q (J&P Securities): Pricing successes in corporate desktop market?
A: No significant difference in the pricing market. Consumer market, huge competition. Corporation most competition in the server. Sales are growing, quarter after quarter. ATI design wins should help us with corporate desktop
Q: Private equity in AMD...getting out of public spotlight?
A: No prejudice to capital, as long as it makes sense. We are alert for any opportunities. Diff b/w public and private...if it makes sense, we will consider it.

Q (Citigroup): Q1 was difficult...given thought to changing target model for AMD (years out)
A: No. Business as usually and driving to past Thanksgiving model
Q: Slowing down F30 to F38 transition. What's the change to your output.
A: Won't change 200mm output - still taking down starts. Just lowering 300mm tool insertion.
Q: Any milestones need to see before Barcy shipping to market?
A: We'll be demo'ing Barcy in late this summer. Need to qual customer systems and get them out the door

Q: Restrictions based upon x86 license?
A: All models take that into account and we have flexibility. Won't talk about details of IP licensing
Q: Breakout b/w units and pricing?
A: Disappointed in pricing stability, but it's firmed up, main issue was volume. 80% of the challenge was volume
Q: Graphics - how to compare Q4 to Q1 to see the growth?
A: Revenue and ASPs down, units up.
Q: R600 coming out in May...is it for back to school or christmas?
A: R600 is discrete graphics at start, back to school is for integrated. Enthusiasts don't work on cycles and they are who will be buying the R600.
 
Apologies for typos, bias, etc., but I was typing as fast as I could. I'm clearly no stenographer. These are just the notes I took from the conference call. There was a little more, but I got bored. 😀

For 100% accuracy, listen to the AMD feed off their investor link.

Thank you and welcome...

...........
Click to expand...

If they can deliver what they say, I think Intel is now in trouble.
But...... in the past months, AMD did not deliver at all.
 
I don't think Intel is in trouble, but they may get a nudge from behind just to let them know AMD is still there.

I think AMD will be able to get close to matching Intel's top offering (as of now they aren't even close), but they won't take the crown. Honestly, I don't have much evidence of this, just my gut feeling (Jack knows alot more about this than I ever will).

If they can actually get the R600 out the door in May... that will go a long way to improving their perceived image. 😀
 
GPM will most definitely be below 35%
Click to expand...

In a ML report filed yesterday (of which I have immediate access to, as well as real time market data), the Q1 estimate for GP is 33.9%.

Given the movement of the stock and Intel's solid numbers, I would expect GP to hit 35% on the nose, with a slight understatement in operating expenses. I need to see if I can call in to the earnings release.

Earnings release is at 4pm CST, right after the closing bell. I am all reg'd up for the call. 😀

Well, I had no idea what to expect.... based on Q4 and the sequential decline I guessed at <35%, had no idea it would be less than 30%.... I was off by a factor of 2 on the loss as well.....

Where is Baron --- I want to see him explain how AMD is not bleeding :)

Hehe, no way Baron can spin this.

I was off as well, I said 35% and I wasn't even close. I forgot to think about ATI's historically lower GM as compared to AMD's numbers. When I was running ratios 35% seemed reasonable (after doing 3 years of ratios lol). Oh well, I actually see this as an optimistic outlook. This is about as bad as it could be, I only see improvement from here on out.

As for Barcelona, is what I said at least kind of accurate? It won't beat Penryn, but it will give Intel a tap on the shoulder and utter "Don't fall asleep at the wheel, I'm still here"?
 
There are two problems here.

First, ATi revenue has historically been ~$1.2Billion/year. Profit on $400-$500Million, that includes a lot of low-end stuff, isn't going be the deciding factor if they are trying to pare down a $600Million+ shortfall.
Click to expand...

Be a deciding factor, no.
Still be red but maybe not as bad fiscally(optimism), might turn a few opinions in the process.

Second, I wouldn't call the 8600 a hole exactly. It pretty much matches up with the top end of the previous generation on older games and beats them on Oblivion. The 7600 and 1600 can't claim that kind of performance. I'm not sure what Polkowski was expecting. He must have been looking at too many Core 2 benchmarks.
Click to expand...

If you compare the performance of all the video cards in the non-synthetic benchmarks in the Toms article I linked to you will find that there are two or three 10% to 15% jumps in performance across all testes.

I consider these to be small holes in the wide view of current performance. Admittedly there are other video cards on the market, though.

However, my original point was there is a sizable hole (30% to 50%* performance difference) between the 8600 GTS and the 8800 GTS in nVidia's DX10 line.

*Lower 40% typical, ignoring fliers on the high-end, more pronounced at high graphics settings and heavy load.

I personally think nVidia's 8600 GT and GTS line should be ~$120 and ~$160 respectively to be really competitive. Now that is a very harsh view on my part but this is coming from someone who refused to buy a 6600 GT because it's price stagnated for a year and a half at ~$150. I am a tough customer.
EDIT: As I am typing this you can get a X1950 Pro, which traded blows with the 8600 GTS and typically beating it at high detail, for ~$165 after rebate at newegg.
 
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom