AMD Q415 results

[Sweepr]

Considering how gimped most Carrizo machines are, that's quite impressive, especially when you compare it to how well mobile Kaveri did in the same space.

I don't think so, the fastest SKU running dual-channel RAM should be able to easily outperform any Skylake-U GT2 notebook in real games, but hey, there's always 3DMark 11 for marketing slides.

The humiliation is that Carrizo is a 28nm planar product and its competitive at low power environments of 15W TDP against Intels 14nm FinFet products.

The humiliation is that despite your efforts to downplay Intel graphics at every chance, reality shows Carrizo was overhyped and underdelivered.

 

[AtenRa]

there's always 3DMark 11 for shinny slides.

Yeap,

 

[Sweepr]

Yeap,

http://legitreviews.com/images/reviews/2185/intel-haswell-deck-4.jpg

It used to favour Intel in these comparisons, now it's the opposite.

 

[mrmt]

The humiliation is that despite your efforts to downplay Intel graphics at every chance, reality shows Carrizo was overhyped and underdelivered.

Wow, that's an interesting sales pitch. I concede that "the chip that humiliated Intel" might sound awesome in a crowd of AMD fanboys. "put $50 in my pocket and humiliate Intel" sounds truer.

Now back to the real world, Lisa Su is basically writing off Seatle as a revenue provider, so regardless of how many SATA ports your AMD reseller count in front of you it won't be effective as a server product. In Lisa's words she sees a modest contribution in 2016, which for AMD standards means token revenue not worth the hassle to report.

Current console units will provide smaller revenue than this year (which slightly up from 2015 despite a double digit increase that year in units sold) and lower operating profits. They expect the business to beat 2015 with the help of a new design they plan to ramp up in the second half.

Lisa Su is saying that they will gain share next quarter but if we apply AMD guidance for Q116 and apply the seasonality factor for the consoles, that mean they will lose share on the PC market. Someone is screwing up here, either AMD marketing isn't providing the numbers or the execs can't read the reports. More important, the growth drivers Lisa Su is mentioning were supposed to appear on this quarter because of holiday sales, and not in Q1. I think her forecast is flawed.

This is especially huge because a lot of information she gave investors were relying on these assumptions, and if they can't model it it's not good, as investors will follow their models and management will lose even more credibility.

On Zen, she confirmed that it will arrive first on desktops and that servers only in 2017. She's reporting design wins for Zen but without gauging the interest on the market this might be just hot air. What I find interesting is that whatever interest she's reporting on server and workstation markets is not enough to absorb all the units of the initial production run, so my hunch is that this interest in Zen is not big at all. Zen will be a flop on servers.

 

[AMDisTheBEST]

BS. The "x86 market" is a mere subset of the worldwide market for computer processors.

But it's about the only processor used in desktop and laptop. Arm might dominate mobile but at least they license it out so others can create and sell their own.

 

[AMDisTheBEST]

Reason Intel sell more is less about technology and the quality. They are simply better at marketing, with more cash to throw at marketing. Nvidia is the same story. Intel has been bribing the OEMs back in the 2005 not to use amd chips. They were charged with monopolistic practice for doing that. Amd CPU Athlon was clearly superior than Pentium 4 back then

 

[ShintaiDK]

Reason Intel sell more is less about technology and the quality. They are simply better at marketing, with more cash to throw at marketing. Nvidia is the same story.

Keep telling yourself that. Both NVidia and Intel beat AMD due to better products.

Even AMD uses Intel CPUs to bench their graphics cards.

 

[mrmt]

But it's about the only processor used in desktop and laptop. Arm might dominate mobile but at least they license it out so others can create and sell their own.

x86 is not a totally closed market. IBM, can manufacture x86 processors at will and you have VIA. That these companies to not make substantial investments on x86 should tell you how profitable it is to invest on it. So if AMD goes under and the FTC forces Intel to open x86 to new licensees, I don't think there will be many companies queuing up for it.

One of the reasons is that building x86 chips is to compete against the guy making the instruction set, which means you'll be always behind in terms of implementing new features of the instruction set, and you'll also only be able to implement your features only if you have a solid backing from dominant software companies like Microsoft or Oracle, which is not easy to get.

Breaking up Intel is not a choice too, the company is too tightly integrated, break it up and you'll end up with two dead parts.

So in the end if AMD goes under, what changes? Nothing. The FTC won't do anything, the prices will stay more or less the same and life will go on.

 

[MrTeal]

At this point he still is on the dark as it s not an information he need to manage AMD s current finances, only the CEO and technical staff are on the know of what is after all a strategic info.



Process is the same so CPU dpt should be on the same level.




Thing is that OEMs must be sampled CPUs quite earlier than GPUs, if as announced they are on schedule for a 2016 release then ESs must be delivered no later than in a few months, wich point to revisions being already in the work, that s surely what Kumar was talking about, that is, that the definitive revision is still not taped out.

So, if I can recap your points here.
1. A CFO just handles finances, kind of like the treasurer of your local PTA, and is not on a need to know basis for a major project milestone for his company's single most important project.
2. Sampling in 2016 means partners will be releasing in volume in 2016, so Engineering Samples must be available in a few months at the latest.

 

[mrmt]

1. A CFO just handles finances, kind of like the treasurer of your local PTA, and is not on a need to know basis for a major project milestone for his company's single most important project.
2. Sampling in 2016 means partners will be releasing in volume in 2016, so Engineering Samples must be available in a few months at the latest.

Suppose you are a banker, and suppose you are negotiating a loan or structuring a debt placement on the market.

Now in the first meeting you ask a few things to better understand the company and model the cash flow of the operation. But then you have a nice surprise, the CFO just knows finance, he cannot tell anything about the financial milestones of the company or how future projects will impact cash flows of the company. What do you do then?

 

[MrTeal]

Suppose you are a banker, and suppose you are negotiating a loan or structuring a debt placement on the market.

Now in the first meeting you ask a few things to better understand the company and model the cash flow of the operation. But then you have a nice surprise, the CFO just knows finance, he cannot tell anything about the financial milestones of the company or how future projects will impact cash flows of the company. What do you do then?

Seems like the banker would just leave the cheque value blank, and when the CFO goes back to his technical leads and asks how much money they need he can just fill it in himself.

 

[mrmt]

Seems like the banker would just leave the cheque value blank, and when the CFO goes back to his technical leads and asks how much money they need he can just fill it in himself.

That wouldn't work.

Just check Intel Q&A for example, just check how Stacy Smith knows deeply Intel business and Intel technical tidbits. You can see a similar trend elsewhere, in companies like Qualcomm or ASML. That AMD puts someone in charge that doesn't know the business very well is more of a symptom of the state of disarray of that company than of a general business trend.

 

[pablo87]

I thought the quarter wasn't too bad for AMD. They bled $100MM by my #'s. Inventory finally came down as did AR which is why the cash went up but at the same time, it reduces their future borrowing base.

The main problem they have is gross margins, they somehow need to raise them to the high 30's in order to break-even (was in the mid 40's historically). If they can continue to whittle away at their old stocks, it should improve but by that much?

Outlook wise, its cloudy. The PC market is likely to nosedive in 2016 and even if Zen is a success, by the time it comes out AMD could be in dire straights.

Does anyone know if the deal for selling their back end has been consummated and if AMD has received the cash already? That would be a big help to AMD that's for sure.

 

[senseamp]

I'd buy some once it drops to $1.50
Purely on speculation it gets bought out. Look at how much Altera went for.
AMD has strategically valuable IP, and with the declines, management of companies like Qualcomm needs to do some "strategic" sh!t to justify their overpriced salaries. Once there is an offer on the table, there could be a bidding war from Samsung, Apple, Microsoft, etc.

 

[MrTeal]

That wouldn't work.

Just check Intel Q&A for example, just check how Stacy Smith knows deeply Intel business and Intel technical tidbits. You can see a similar trend elsewhere, in companies like Qualcomm or ASML. That AMD puts someone in charge that doesn't know the business very well is more of a symptom of the state of disarray of that company than of a general business trend.

I think you might have missed the sarcasm.

Regardless, that's only true if you accept Abwx's assertion that knowing things about icky nerdy technical stuff isn't part of a CFO's duties. I would take the other tack, and say that when AMD's CFO says that they're taping out Zen for servers in the next couple months and that investors should look to that as a key milestone, AMD is actually looking to tape out Zen in the next couple months.

 

[Phynaz]

At this point he still is on the dark as it s not an information he need to manage AMD s current finances, only the CEO and technical staff are on the know of what is after all a strategic info.

So you're saying the CFO is nothing more than an accountant.

Your lack of knowledge is showing again.

 

[Arachnotronic]

Does anyone know if the deal for selling their back end has been consummated and if AMD has received the cash already? That would be a big help to AMD that's for sure.

AFAIK the deal hasn't closed.

 

[dark zero]

AFAIK the deal hasn't closed.

And it was cancelled BTW...AMD has darker times now since they hace that debt... Even selling themselves won't pay all of them.

 

[Arachnotronic]

And it was cancelled BTW...AMD has darker times now since they hace that debt... Even selling themselves won't pay all of them.

Stop spreading FUD, that deal wasn't cancelled.

 

[dark zero]

Stop spreading FUD, that deal wasn't cancelled.

But it passed enough time to consider that chance.

 

[Arachnotronic]

But it passed enough time to consider that chance.

Just stop. Major deals usually take ~6mos to close. ALTR acquisition was announced in June, but did not close until December.

Don't mislead people with false statements like this.

 

[mrmt]

Does anyone know if the deal for selling their back end has been consummated and if AMD has received the cash already? That would be a big help to AMD that's for sure.

Nope, it was not. I think they are counting on it to withstand the pressures for the looming implosion of their CPU business.

 

[krumme]

I thought the quarter wasn't too bad for AMD. They bled $100MM by my #'s. Inventory finally came down as did AR which is why the cash went up but at the same time, it reduces their future borrowing base.

The main problem they have is gross margins, they somehow need to raise them to the high 30's in order to break-even (was in the mid 40's historically). If they can continue to whittle away at their old stocks, it should improve but by that much?

Outlook wise, its cloudy. The PC market is likely to nosedive in 2016 and even if Zen is a success, by the time it comes out AMD could be in dire straights.

Does anyone know if the deal for selling their back end has been consummated and if AMD has received the cash already? That would be a big help to AMD that's for sure.

At the time zen hits servers is well into 2017 and the market is full of a72 plus perf arm cores and established new players like qcom samsung and amazon. All have their own core. Ares is on the footstep.

Lowend servermarket is then red ocean.

Now if amd brings something that is 80% of the lower part they imo get stuck in a very bad place between a lot of money, strategic long term interests, and perhaps to much Intel capacity.
Thats not a nice place to be. Not for anything and not for higher margins.

Amd needs to partner up. What are they waiting for?
The solution is not on product side. Its a pipe dream even if zen is very very good with solid ipc and excellent perf/w.

Are they waiting for falling oil prices to do their job?

 

[Vesku]

That "Zen will compete in 80% of server market." could mean no quad socket platform.

 

[AMDisTheBEST]

Keep telling yourself that. Both NVidia and Intel beat AMD due to better products.

Even AMD uses Intel CPUs to bench their graphics cards.

I am pretty sure ppl choose intel HD over a discrete because of better products huh? Consumers have no choice over the matter. Many just buy whatever OEMs stuff inside their machines which happen to be mostly intel chips with intel graphics.