AMD Q3 Results

[CHADBOGA]

And 3/4 of this forum thinks all this is good news. Self inflicting pain. So idiotic its unbelievable.

How have people from this forum inflicted anything?

 

[senseamp]

If Intel doesn't attack this market, somebody else will.

By the way, I don't know how much I buy this "mix shift down" argument. People typically buy to their means, so it's not likely that anybody who normally buys $700 PCs with Core i5 processors will suddenly buy $199 Chromebooks as their primary systems.

Further, PC vendors like to make money too, so they will put the nicer components on the more expensive systems with the more expensive Core processors.

I think this view that Intel is "digging its own grave" if it sells appropriate, cheap-to-build chips into low-cost systems is too simplistic, and I think Intel itself feared this which is why Atom was so woefully under-powered until recently (and I think there is much more room for improvement).

People don't buy to their means in the PC space. They buy to their needs. I am typing this on a $199 Chromebook. In college, when I was much much poorer, I would be typing this on a PC that cost 10 times as much. Even people who spend a lot on PCs often don't feel a compelling need to upgrade until something physically breaks on their computer.
Intel didn't dig its grave with Chromebooks, it dug it with ultrabooks. By pushing good enough performance in thin form factor, it conditioned the market to think that performance is secondary to form factor. Low level of performance, especially IGP, in mainstream PC segment also set a very low baseline that PC software has been designed for, effectively allowing the mobile space to catch up, and also allowing console gaming to overtake PC gaming in many markets. So a lot of people who would previously game on their PCs are now gaming on mobile and consoles, and their PC is just their web reading device, which is where Chromebook comes in. PC's have been good enough for reading webpages for years, and it's completely ISA independent. So Intel's peak performance advantages and x86 lock in are rendered moot. It comes down to price and power consumption. Intel can compete on both, great, but Intel likes monopoly margins, and competing in a commodity space is not where they would like to be.

 

[DrMrLordX]

And 3/4 of this forum thinks all this is good news. Self inflicting pain. So idiotic its unbelievable.

The good news is that they aren't dead yet. By any indicator, the 16th was supposed to be the day when we saw the first of the wheels really come off of AMD. Now we're all waiting for the next round of layoffs and the horrendous (?) Q4 results. Surely then we'll see the final retreat of AMD from the x86 market, right? Layoffs, losses in market share, major reductions in revenue, delisting of their stock, it's all just ~3 months away . . .

Watching AMD die is like waiting for energy-positive nuclear fusion. It's always just over the horizon.

 

[krumme]

It does look like waiting for cold fusion.
Except we get charcoal if it happens.

It doesnt matter if its a camera, cpu, mobile phone whatever. Some consumers want their brand to totally dominate the competition - like they are in a computer game. They want to pay double for worse performance. Like in the good old days when we paid 400usd for 386sx while Intel was all along 486dx that was 2-3 years newer. The segmenting was crazy. Like we are on quad cores since core2 - just far worse back then.

People are celebrating they should pay Intel shareholders more. "Take my money" consumer identity. The stupidity have no ends.

 

[MiddleOfTheRoad]

The good news is that they aren't dead yet. By any indicator, the 16th was supposed to be the day when we saw the first of the wheels really come off of AMD. Now we're all waiting for the next round of layoffs and the horrendous (?) Q4 results. Surely then we'll see the final retreat of AMD from the x86 market, right? Layoffs, losses in market share, major reductions in revenue, delisting of their stock, it's all just ~3 months away . . .

Watching AMD die is like waiting for energy-positive nuclear fusion. It's always just over the horizon.

I mean seriously.... This Q3 Report really wasn't that bad. Was it a shock that Intel grabbed marketshare at a time when Intel is paying vendors to take Atom chips off their hands? Intel can flush a billion dollars like that, AMD doesn't have the scale to match them.

The financial dark days of AMD are behind them -- their balance sheet was actually pretty damn good for Q3. They do need new product, but APU sales were up overall for Q3 and Carrizo is only a few months away. I really hope they've got some new FX parts in the pipeline.

 

[mrmt]

I mean seriously.... This Q3 Report really wasn't that bad.

Done is done, there isn't much investors care for a done deal. But, from what AMD said in the Q&A, will its balance sheet be stronger or weaker for the next few quarters? AMD itself answered this question when they decided to go another round of cuts. It will be weaker, and that's what is bugging investors.

 

[GreenChile]

I mean seriously.... This Q3 Report really wasn't that bad. Was it a shock that Intel grabbed marketshare at a time when Intel is paying vendors to take Atom chips off their hands? Intel can flush a billion dollars like that, AMD doesn't have the scale to match them.

AMD does not have any meaningful presence in the space where Intel is paying contra-revenue program. Intel competes most directly with Qualcom in this space.

AMD is in somewhat the same boat as Intel with regard to low cost tablet space. Their BOM is too high to compete with Qualcom. Do they even have a chip in this space that is competitive? Not sure but they certainly don't have the cost low enough that that any company would choose it over a Qualcom chip.

AMD can't possibly compete in the tablet space until cost and performance is competitive with the market leaders. It really has nothing at all to do with Intel's contra-revenue program.

 

[MiddleOfTheRoad]

AMD can't possibly compete in the tablet space until cost and performance is competitive with the market leaders. It really has nothing at all to do with Intel's contra-revenue program.

AMD's marketshare dip actually does have everything to do with Intel contra-revenue program. Intel is paying vendors to take x86 chips for Tablets, 2 in 1 devices, etc -- So obviously, their x86 marketshare has grown. AMD's overall volume didn't change that much (heck, APU sales actually increased), but Intel simply grew x86 into low end tablets (and AMD isn't going to follow them down there where Intel is basically wrapping a 20 dollar bill around every Atom). A billion dollars isn't exactly a drop in the bucket.

 

[mrmt]

AMD's marketshare dip actually does have everything to do with Intel contra-revenue program. Intel is paying vendors to take x86 chips for Tablets, 2 in 1 devices

2 in 1 isn't covered by the contra-revenue program.

 

[positivedoppler]

AMD can't possibly compete in the tablet space until cost and performance is competitive with the market leaders. It really has nothing at all to do with Intel's contra-revenue program.

That is not true. Mullin is aimed squared at the tablet market, won best innovation award at ces and received very positive reviews all around. its sales is impacted by Intel's contra revenue. Lisa Su even admit to there being an impact during the earning conference. There might be a couple of reasons why AMD elect no action. 1. AMD is letting intel do the heavy lifting with the x86 adaptation and they stand to benefit later. 2. AMD and Intel are in the midst of their x86 and 64 bit x86 negotiation which might explain why Read is still around.

 

[GreenChile]

AMD's marketshare dip actually does have everything to do with Intel contra-revenue program. Intel is paying vendors to take x86 chips for Tablets, 2 in 1 devices, etc -- So obviously, their x86 marketshare has grown. AMD's overall volume didn't change that much (heck, APU sales actually increased), but Intel simply grew x86 into low end tablets (and AMD isn't going to follow them down there where Intel is basically wrapping a 20 dollar bill around every Atom). A billion dollars isn't exactly a drop in the bucket.

Source?

Cause I believe you are completely wrong about that. The 2 in 1 devices are not part of the contra-revenue program as mrmt stated. The program is only targeted at low cost tablet space where they are competing directly with lower cost ARM chips, not x86.

 

[GreenChile]

That is not true. Mullin is aimed squared at the tablet market, won best innovation award at ces and received very positive reviews all around. its sales is impacted by Intel's contra revenue. Lisa Su even admit to there being an impact during the earning conference. There might be a couple of reasons why AMD elect no action. 1. AMD is letting intel do the heavy lifting with the x86 adaptation and they stand to benefit later. 2. AMD and Intel are in the midst of their x86 and 64 bit x86 negotiation which might explain why Read is still around.

Sorry but you are wrong as well. AMD cannot gain any traction there because the BOM is too high. They are competing against low cost ARM chips in this area. If their costs are higher they will gain no ground but unlike Intel they don't have any money to help subsidize the higher BOM.

 

[mrmt]

That is not true. Mullin is aimed squared at the tablet market, won best innovation award at ces

Usually in the MPU industry sales results trump intentions and awards.

 

[III-V]

That is not true. Mullin is aimed squared at the tablet market, won best innovation award at ces and received very positive reviews all around. its sales is impacted by Intel's contra revenue. Lisa Su even admit to there being an impact during the earning conference. There might be a couple of reasons why AMD elect no action. 1. AMD is letting intel do the heavy lifting with the x86 adaptation and they stand to benefit later. 2. AMD and Intel are in the midst of their x86 and 64 bit x86 negotiation which might explain why Read is still around.

Mullins is aimed at the tablet market the same way Bay Trail is -- it's not. It's basically a netbook/low-power laptop processor downclocked for tablet use.

Usually in the MPU industry sales results trump intentions and awards.

No kidding. Who cares what award something wins? They're in the business of selling things, not racking up blue ribbons.

 

[NostaSeronx]

Mullins is aimed at the tablet market the same way Bay Trail is -- it's not. It's basically a netbook/low-power laptop processor downclocked for tablet use.

Mullins is the base die. So, Beema is a tablet processor overclocked for Laptop/AIO use.

This is the opposite of Temash which wasn't the base die. Your conjecture only works with Temash, and not Mullins.

 

[MiddleOfTheRoad]

Source?

Cause I believe you are completely wrong about that. The 2 in 1 devices are not part of the contra-revenue program as mrmt stated. The program is only targeted at low cost tablet space where they are competing directly with lower cost ARM chips, not x86.

My point was -- Intel is flooding the market with X86 chips in segments where they didn't previously reside. AMD's volume has been rock solid consistent, but the flood of cheap x86 tablets (which Intel is basically paying vendors right now to take them) has been the primary cause of the (somewhat artificial) Intel x86 marketshare increase.... It is ARM CPU's that are losing the most sales to Intel's latest push. If you look at AMD -- They made 1.46 Billion Q3 2013 / 1.43 Billion Q3 2014 = Nearly identical numbers. This doom and gloom is completely overblown.... Su is definitely going to run AMD more efficiently than Rory, though. And the layoffs will continue to make the company leaner and more efficient for the future.

 

[III-V]

Mullins is the base die. So, Beema is a tablet processor overclocked for Laptop/AIO use.

This is the opposite of Temash which wasn't the base die. Your conjecture only works with Temash, and not Mullins.

Is there an actual difference, or is it a semantic one with no substance behind it? Temash/Kabini is not far off from being the same thing.

 

[NostaSeronx]

Is there an actual difference, or is it a semantic one with no substance behind it?

Kabini's voltage range is ~0.8v to ~1.5v.
Mullin's voltage range is ~0.5v to ~1.4v.

Mullin's shaves off 300 mV at low power and 100 mV at high power. We also have to point out that, Kabini is on a High Performance node @ TSMC, and Mullins is on a Low Power node @ GlobalFoundries.

Temash/Kabini is not far off from being the same thing.

Kabini is the host die, Temash is the downclock.
Mullins is the host die, Beema is the overclock.

Off-tangent; Mullin's DVS wants me to think its FD-SOI, but it is probably just 28-nm SLP.

 

[III-V]

Kabini's voltage range is ~0.8v to ~1.5v.
Mullin's voltage range is ~0.5v to ~1.4v.

Mullin's shaves off 300 mV at low power and 100 mV at high power. We also have to point out that, Kabini is on a High Performance node @ TSMC, and Mullins is on a Low Power node @ GlobalFoundries.Kabini is the host die, Temash is the downclock.
Mullins is the host die, Beema is the overclock.

Off-tangent; Mullin's DVS wants me to think its FD-SOI, but it is probably just 28-nm SLP.

Interesting. But what about the problem of extra engineering dollars being needed to implement it, like Bay Trail has? And more board layers, routing challenges, etc.?

The ARM tablets avoid this, and I doubt Mullins does.

 

[NostaSeronx]

Interesting. But what about the problem of extra engineering dollars being needed to implement it, like Bay Trail has?

I do not know about that.

And more board layers, routing challenges, etc.?

FT3(Temash(Kabini)) -> FT3b(Mullins(Mullins))

Would probably reduce board layers.

FT3b(Mullins) -> FP4(Mullins/Amur), I do not know.

 

[mrmt]

My point was -- Intel is flooding the market with X86 chips in segments where they didn't previously reside.

(...)

If you look at AMD -- They made 1.46 Billion Q3 2013 / 1.43 Billion Q3 2014 = Nearly identical numbers. This doom and gloom is completely overblown....

And this is why the numbers must be broke down before any analysis. For AMD themselves they were selling 927MM in the consumer market with a corporate gross margin of 36% purged of non-recurring items, while today they are selling around 791 with a 33% corporate margins after purging non-recurring items. Also they were not forecasting a 13% drop in their revenues for Q413 as they are now in Q414. The situation has deteriorated a lot YoY.

 

[AtenRa]

And this is why the numbers must be broke down before any analysis. For AMD themselves they were selling 927MM in the consumer market with a corporate gross margin of 36% purged of non-recurring items, while today they are selling around 791 with a 33% corporate margins after purging non-recurring items. Also they were not forecasting a 13% drop in their revenues for Q413 as they are now in Q414. The situation has deteriorated a lot YoY.

Q1 2013
Revenue = 1.088B
Income (Loss)= (98M)

Q1 2014
Revenue = 1.397B
income = 49M

---------

Q2 2013
Revenue = 1.161B
Income (Loss) = (29M)

Q2 2014
Revenue = 1.441B
income = 63M

-----------

Q3 2013
Revenue = 1.461B
Income = 95M

Q3 2014
Revenue = 1.429B
income = 63M

-------------

Q1+Q2+Q3 2013
Revenue = 3.71B
Income (Loss)= (32M)

Q1+Q2+Q3 2014
Revenue = 4.267B
Income = 175M

Your doom and gloom in everything AMD is pathetic.

Edit : Lets estimate Q4 2014 and entire year

Q4 2013
Revenue = 1.589B
Income = 135M

Q4 2014 (estimation with -13% Revenue from Q3)
Revenue = 1.243B
Income (Loss) = (10M) ?? (estimation)


-----------

2013
Revenue = 5.299B
Income = 103M

2014
Revenue = 5.51B
Income = 165M

 

[Phynaz]

Your doom and gloom in everything AMD is pathetic.

Edit : Lets estimate Q4 2014 and entire year

Q4 2013
Revenue = 1.589B
Income = 135M

Q4 2014 (estimation with -13% Revenue from Q3)
Revenue = 1.243B
Income (Loss) = (10M) ?? (estimation)


-----------

2013
Revenue = 5.299B
Income = 103M

2014
Revenue = 5.51B
Income = 165M

Didn't listen to to the call? AMD predicted a $200M - $250M loss for 2014. Quite a ways away from your $165 profit.

 

[Ramses]

Time for a gubberment bailout!

all the cool kids are doing it

 

[hungtran]

Didn't listen to to the call? AMD predicted a $200M - $250M loss for 2014. Quite a ways away from your $165 profit.

For 2014, they are projecting cash flow positive. Are you saying that from the call, you heard they're projecting to have a $375M-$425M loss in Q4 on 1.243B in revenue? The company would be bankrupt today if that were the case.