"AMD Moves Away From PCs Amid Steep Losses"

[sontin]

AMD is down 11% 12% right now. It looks that the second strategy change is not accepted by the stock market. Market value only $1,63 Billion.

 

[AtenRa]

Jaguar is the successor to bobcat because they cancelled the 28nm Krishna and Wichita (28nm bobcat shrink) that were supposed to come out before Jaguar. Tick tock.

Yes but they have replaced them with Jaguar. What problem/s do you see with that ??

 

[NTMBK]

Yes but they have replaced them with Jaguar. What problem/s do you see with that ??

Because Jaguar won't be out for another year, most likely. If they have truly sped up their Jaguar roadmap a lot by ditching Wichita, then it was worth it, but if it doesn't come out until Christmas next year then they're stuffed.

 

[Idontcare]

Yes but they have replaced them with Jaguar.

They haven't replaced anything with Jaguar, Jaguar isn't out yet.

They replaced it on the roadmap, true, but roadmaps are not sellable products.

Yes but they have replaced them with Jaguar. What problem/s do you see with that ??

Make a list of everything you might foresee to be a problem if AMD cancelled Jaguar right now at this exact point in time.

Then write across the top of that list "Things that became a problem when AMD cancelled the 28nm bobcat shrink".

It really is one of those "all of the above" snafu type situations of snatching defeat from the jaws of victory.

But AMD recognizes this, hence the inclusion of the bullet point "significant execution problems" on their slide. No one can claim AMD has no process gap and no execution problems and no design complexity problems when AMD themselves acknowledge as much.

But this still doesn't answer the question "what does AMD think 'agile and flexible SoC methodology' means?"

 

[pelov]

AMD is down 11% 12% right now. It looks that the second strategy change is not accepted by the stock market. Market value only $1,63 Billion.

At that price, what's to stop someone like Samsung from coming in and taking over? That's around 1 quarter of their profits from Q3 alone.

I feel pity for Read. He's on a sinking ship that was run into a rocky harbor by the captains before him and now his job is to seemingly get it going again, but with a broken mast and a smaller and less competent crew.

Dirk's strategy to go gunning for the low/mid range of the market hasn't worked. If people go out to buy, or make, a PC within a limited budget, they're now presented with tablets at a cheaper cost that do pretty much everything their PCs did. It's no wonder AMD's hurting as bad as they are. If Joe Average walks into Best Buy, he's now looking at tablets rather than laptops/desktops, and those sales are cannibalizing the low end PC sales first. You can't replace a workstation and a high end enthusiast gaming rig with an ARM core or tablet, but those have been Intel's bread and butter for years now. With AMD stuck in the 'budget' arena, they're getting walloped.

When I saw the Trinity reviews for mobile, I thought the chip looked great. It sipped power less than Ivy and Sandy i5's, offered graphics performance equivalent to a 540m and had good enough CPU performance. Perfect laptop chip for me, as I was looking for very good battery life, and the most CPU intensive task would be 1-2 hours a week of gaming. Trinity is perfect for that. Then I went out looking for them in attractive laptop designs... and I'm still looking. The OEMs have lost faith in AMD and so has the consumer. An OEM won't put an AMD APU into a high-grade laptop design/chassis, and this stems directly from AMD's monumental screwups of yesteryear: delays, underwhelming performance, yield issues, and self-drawn high expectations that they never lived up to.

Going forward, they suffer the same issue Intel faces in the mobile end, insofar that they both lack attractive products. Intel can still make loads of cash in server/PC, but AMD can't. They can talk all they want about mobile-focused, SoC designs, but unless these chips are wrapped up in Samsung, Asus, Apple and Lenovo chassis that will sell, it doesn't matter. Reading those statements just reaffirms their dependence on Win8 succeeding, much like Intel. And that just isn't going to happen...

They did say one thing right, though

- Adopt industry partnership approach

Maybe these partners can help them get a decent product out, otherwise it's too little too late.

 

[AtenRa]

Because Jaguar won't be out for another year, most likely. If they have truly sped up their Jaguar roadmap a lot by ditching Wichita, then it was worth it, but if it doesn't come out until Christmas next year then they're stuffed.

It has been mentioned by Rory himself in the financial conference that Kabini (Jaguar cores) is coming H1 2013.

 

[blastingcap]

Even Intel is trying to move away from PCs... stagnant market, not as high margin as servers or as hot as mobile.

What do you guys make of http://www.zdnet.com/calxeda-lays-out-64-bit-arm-server-strategy-7000005979/ btw?

 

[Idontcare]

They did say one thing right, though

- Adopt industry partnership approach

Maybe these partners can help them get a decent product out, otherwise it's too little too late.

What is in it for the partners though? I mean I get why businesses jump on the HSA train, they don't want to miss the boat when it sets sail from the harbor.

But AMD only benefits from this if they are selling hardware/IP to those partners to generate money for themselves. But what does AMD have to offer any of the partners in return for that money?

^ If you look at the revenue streams of all these fabless players I don't see any of them hurting because they don't have a partnership with AMD. So what is AMD going to offer them that in turn will compel them to financially assist AMD?

This is the part I don't get by AMD's stated goal of "industry partners".

 

[Idontcare]

Even Intel is trying to move away from PCs... stagnant market, not as high margin as servers or as hot as mobile.

What do you guys make of http://www.zdnet.com/calxeda-lays-out-64-bit-arm-server-strategy-7000005979/ btw?

I think we are only seeing the tip of the iceberg of what Calxeda has in the works. They are in startup mode, hiring the best and the brightest most enthusiastic engineers and meshing them with a can-do management that isn't purely cost-reduction driven right now.

Calxeda is looking to grow out of the niche that AMD is merely hoping to shrink in size to fit into, like ships passing in the night. Only one of them is going in the right direction.

 

[AtenRa]

Dirk's strategy to go gunning for the low/mid range of the market hasn't worked. If people go out to buy, or make, a PC within a limited budget, they're now presented with tablets at a cheaper cost that do pretty much everything their PCs did. It's no wonder AMD's hurting as bad as they are. If Joe Average walks into Best Buy, he's now looking at tablets rather than laptops/desktops, and those sales are cannibalizing the low end PC sales first. You can't replace a workstation and a high end enthusiast gaming rig with an ARM core or tablet, but those have been Intel's bread and butter for years now. With AMD stuck in the 'budget' arena, they're getting walloped.

When I saw the Trinity reviews for mobile, I thought the chip looked great. It sipped power less than Ivy and Sandy i5's, offered graphics performance equivalent to a 540m and had good enough CPU performance. Perfect laptop chip for me, as I was looking for very good battery life, and the most CPU intensive task would be 1-2 hours a week of gaming. Trinity is perfect for that. Then I went out looking for them in attractive laptop designs... and I'm still looking. The OEMs have lost faith in AMD and so has the consumer. An OEM won't put an AMD APU into a high-grade laptop design/chassis, and this stems directly from AMD's monumental screwups of yesteryear: delays, underwhelming performance, yield issues, and self-drawn high expectations that they never lived up to.

Sorry but you have to read the Q3 2012 Results - Earnings Call Transcript

Rory P. Read - Chief Executive Officer, President and Director

Yet, against this backdrop, we saw a continued consumer adoption of our Trinity APU in the quarter. Trinity notebook unit shipments increased more than 70% sequentially and accounted for nearly 1/3 of our total notebook shipments in the third quarter. Although Trinity is targeted at mainstream price points, Ultrathin notebooks featuring the low-power APU are also competing effectively at higher system price points. As a result, we believe we gained share in the $600 to $799 retail notebook price band globally in the third quarter. More than 125 AMD-based systems are expected to launch with Windows 8, including tablets and several new Ultrathins. While we look forward to the introduction of Win 8, the fourth quarter will continue to be challenging, and we do not expect PC market conditions to improve for several quarters.

It is the APUs that hold them as of now and they will continue to be the major income for them in the future.

 

[pelov]

^ If you look at the revenue streams of all these fabless players I don't see any of them hurting because they don't have a partnership with AMD. So what is AMD going to offer them that in turn will compel them to financially assist AMD?

This is the part I don't get by AMD's stated goal of "industry partners".

100% agree. I think the only thing that they can offer at this point would be a hybrid x86/ARM device for somebody like Apple, who might desire one if they look to transition to a single overarching architecture. Or rather, that's what would make the most sense to me.

Selling off their Adreno line was perhaps the most idiotic move they've ever done. I'm not sure how small they can get their Radeon cores, but I highly doubt they can offer graphics blocks for licensing in mobile. Hondo has 80 Radeon cores at a low clock speed, but the entire APU is 4.5W. Unfortunately, they're stupider than Intel:

As for software support, the story with Hondo is much the same as it is for Clover Trail: the chip is made primarily for Windows 8, and AMD is not planning any Android or wider Linux support for the chip at this time.

So one of these partners is apparently Microsoft. Additionally, they have to work closely with the OEMs who put these products out in order to achieve such a low TDP. Just who these OEMs are who are willing to stick an outdated 40nm neutered APU into their tablets/hybrid devices? Who knows, but I guess they would qualify as "Industry partners." You're right in that the industry partners they need are the ones who are willing to bite on Jaguar and offer it in an attractive package. Just who that is and what will it look like? I have no idea.

It is the APUs that hold them as of now and they will continue to be the major income for them in the future.

And it's still not enough. The low end laptops and ultrathins are competing for the money coming out of the same pocket. Intel's workstation and enthusiast cash isn't at risk from ARM, but AMD's APUs selling for $400-$500 certainly are.

 

[Phynaz]

Jaguar is the replacement of Bobcat ... it will be available for low-end Desktop, Mobile, Tablets and more.

Which is exactly what IDC said, competing with Via.

 

[railven]

Its clear you don't understand how being a predator works.



Step 1 - give predator speech
Step 2 - lose lots of money and fire employees, but only for the first year or three
Step 3 - profit

Had to LOL at this. Kudos!

 

[AtenRa]

Which is exactly what IDC said, competing with Via.

By that logic Intel's ATOM and ULVs are competing with VIA as well, is that what you and IDC are saying ???

 

[pelov]

The biggest problem facing AMD will be getting this chip into tablets from major manufacturers; of the x86 Windows 8 tablets we've seen at IFA and other unveilings throughout the year, most manufacturers seem to be offering high-end models using Intel's Ivy Bridge processors and low-end models running either Clover Trail Atoms or ARM-based chips. AMD wouldn't provide information about which OEMs will be shipping tablets based on Hondo. Hondo's predecessor, the AMD Z-01, appeared in only one shipping tablet I can find, MSI's WindPad 110W.

One hurdle to Hondo's adoption may be that, unlike Clover Trail, the processor is not a system on a chip (SoC). The CPU and GPU are indeed one piece of silicon, but USB, SATA, and other functions are still handled by a separate chip called the Fusion Controller Hub (FCH). This particular FCH does include some features Clover Trail lacks (most notably native USB 3.0 support), but in tablets, space is still at a premium—having to use two chips instead of one takes up room inside the system that could be given over to a larger battery or shaved off entirely. The FCH also consumes extra power (according to AMD's own slides, between 0.55 and 0.68W during normal use) that has to be considered alongside the 4.5W TDP of the APU itself.

Hondo looks like a more or less viable tablet processor, but if the company can't get it into desirable tablets, it may not make much of a dent. To truly take on ARM and Atom and encourage adoption, AMD would do well to develop an SoC version of the chip manufactured on a 32nm or 28nm process. This would take up less space, consume less power, and could bring more GPU cores or ramp up clock speeds because of the extra thermal headroom. At least some of these improvements are currently rumored to be coming in "Tamesh," Hondo's replacement, which will supposedly launch at some point in 2013. Until that happens, AMD's tiny presence in the tablet market may further cement the chipmaker's slow slide into irrelevance.

Gunning for mobile is fine, but you need to have the right chip for it. Right now, they have no SoC and they're only option is a 40nm Hondo, which will chew up more power than 2 ARM tablets. You can neuter the I/O all you want, you're still fighting an uphill battle. And that's just for AMD! Now consider the added costs for the OEMs who have to put this into a tablet:

- larger PCB
- added components (WiFi, Fusion controller hub, etc)
- higher power consumption means larger battery or lower runtime
- it's going to be cramped, thus you can't have a sleek design or you can, but you'll have to compromise somewhere
- higher cost. The chip is going to be cheap, but the PCB, chipset and battery are going to cost even more.

Where is Jaguar?
What does it look like? Layout/integration wise
What's the performance like?
How far down in TDP can they get it?
When's it going to be here?

 

[mrmt]

Sorry but you have to read the Q3 2012 Results - Earnings Call Transcript

Rory P. Read - Chief Executive Officer, President and Director

Yet, against this backdrop, we saw a continued consumer adoption of our Trinity APU in the quarter. Trinity notebook unit shipments increased more than 70% sequentially and accounted for nearly 1/3 of our total notebook shipments in the third quarter.

It is the APUs that hold them as of now and they will continue to be the major income for them in the future.

(...)

You are pretty much naive trying to find rays of hope in the Q3 call.

In this case, Trinity shipments grew but Llano shipments dwindled. Among other things Trinity simply could not make up for Llano dwindling so revenues fell.

I never ever saw an execute team so incompetent as that of AMD. The analysts got to the point of discussing basic math the CFO, and Read, gosh, he can't give a straigh answer for none of the analysts questions. All that with lots of marketing buzzwords and giving fake excuses for their bad results.

How do they expect to have any credibility when they can't admit a simple thing that it's open for everyone to see such as their cash reserves burning?

AMD is getting blooded in the market today and much of this panic run should lie in the shoulders of that incompetent executive team that can't outline a strategy or at least appear in control of the situation.

IDC, you are being pretty optimistic thinking that they can downsize to compete with VIA. With this executive team, they are going to chapter 11 or chapter 7 soon.

 

[ShintaiDK]

AMD is down 11% 12% right now. It looks that the second strategy change is not accepted by the stock market. Market value only $1,63 Billion.

I remember someone saying before the layoffs and Q3 that stocks could only go up. Boy did that go wrong...

Their stock is at -14.5% right now. Actually -15.31% now before I could finish this post.

 

[frozentundra123456]

Sorry but you have to read the Q3 2012 Results - Earnings Call Transcript



It is the APUs that hold them as of now and they will continue to be the major income for them in the future.

That may be correct, but what counts is what total percent of the market AMD has. If they increased from a small percent to 2x that percent, that is a 100% gain, but not really taking control of the market. It is a trend in the right direction, but can they continue it?
Personally, I like trinity and Llano in laptops if the price is right, but you have to look at the market as a whole and see what kind of share you have. I guess what I am saying, mathematically, it is easy to increase a big percentage from a small starting point and still have a low total share of the market.

 

[AtenRa]

You are pretty much naive trying to find rays of hope in the Q3 call.

Trinity shipments shows that the product is doing fine, they just need to sell more not only in Mobile.

In this case, Trinity shipments grew but Llano shipments dwindled. Among other things Trinity simply could not make up for Llano dwindling so revenues fell.

Yes but the Total units shipped difference is small.

Q2 2012 Mobile

Llano (2 and 4 cores) = 1,262 units shipped (millions)
Trinity (2 and 4 cores) = 1,600 units shipped (millions)
Total = 2862

Q3 2012 Mobile

Trinity 1600+70% = ~2720 units shipped.

I haven't seen the Q3 numbers but Im betting they lost more Sales/Revenue from the desktop and Server segments than Mobile.

 

[ShintaiDK]

It is the APUs that hold them as of now and they will continue to be the major income for them in the future.

That quote says nothing besides that they sell more Trinity APUs than before compared to Llano. Expected since Trinity replaces Llano. And looking on the writedown. Thats pretty darn significant. basicly using your numbers they wcouldnt sell an entire quarters production of APUs.

I haven't seen the Q3 numbers but Im betting they lost more Sales/Revenue from the desktop and Server segments than Mobile.

AMD have been utterly destroyed in the server space for years now.

 

[ShintaiDK]

At that price, what's to stop someone like Samsung from coming in and taking over? That's around 1 quarter of their profits from Q3 alone.

It would cost billions to do that. You wouldnt be able to make x86 CPUs. So we are talking about a monthly expense of around 500mio$.

Then you got 10000 employees you more or less want to get rid of. That would also quickly cost a billion or 2. Then there are contracts and obligations, unknown amount there. But anything than pennies.
Lastly you got about 3.6billion in liabilities that also needs to be payed.

Compare that to maybe buy the IPs you want and hire the engineers and other people for maybe 100-250mio later on.

I´m kind enough to say AMDs cash=stock price in the overhead.

 

[v8envy]

I remember someone saying before the layoffs and Q3 that stocks could only go up. Boy did that go wrong...

Their stock is at -14.5% right now. Actually -15.31% now before I could finish this post.

Pretty violent reaction, that. Unfortunately, at $2.25 my options (pun intended) for building any kind of straddle to gamble on volatility are too limited to even try. Looks like people aren't buying the nearly zero content fluff "strategy" after all.

 

[pelov]

Compare that to maybe buy the IPs you want and hire the engineers and other people for maybe 100-250mio later on.

That's what I was getting at, actually.

I don't think any potential buyer would want the x86 license nor the headaches that come with it. But I can definitely see somebody offering a good lump sum for their graphics IP or graphics/cpu integration. Would AMD even be in a position to decline? Could they afford to say no?

 

[exdeath]

So basically we are stuck with i7-3960x as the top CPU for the next 20 years.

 

[ShintaiDK]

That's what I was getting at, actually.

I don't think any potential buyer would want the x86 license nor the headaches that come with it. But I can definitely see somebody offering a good lump sum for their graphics IP or graphics/cpu integration. Would AMD even be in a position to decline? Could they afford to say no?

Graphics IP cant be worth much really. And you most likely dont want to compete against nVidia either in a marketspace that is slowly going out. (Discrete cards.)

There is no GPU/CPU integration so to say. Its hard bolted on in an even more cruel way than Intel, that atleast shares ringbus and L3. But none of them are essentially integrated in any way besides being on the same die. SoCs are already the same.