AMD expected to announce 30 percent workforce cut next week

[Slimprofits]

http://news.cnet.com/8301-10805_3-57...f-its-workers/

and

http://allthingsd.com/20121012/exclu...-of-workforce/

A third source said that the scale of cuts may be lower, closer to between 10 percent and 20 percent of AMD’s workforce, which would imply a range of between 1,170 and 2,300.

[shiny!]

Re: AMD expected to announce 30 percent workforce cut next week

Oh noes! If AMD dies, Intel prices will go through the roof.

I need to build a new computer soon. Was waiting for AMD's next iteration APU after Trinity, but I think I'll pop for the Trinity now.

The Future of AMD's Fusion APUs:

AMD has been moving in the direction of a unified CPU+GPU chip for a long time — starting with the Llano APU — and Kaveri is the next step in achieving that goal of true convergence. AMD announced at the keynote that “we are betting the company on APUs,” and spent considerable time talking up the benefits of the heterogeneous processor. Trinity, the company’s latest APU available to consumers, beefs up the GPU and CPU interconnects with the Radeon Memory Bus and FCL connections. These allow the GPU access to system memory and the CPU to access the GPU frame buffer through a 256-bit and 128-bit wide bus (per channel, each direction) respectively. This allows the graphics core and x86 processor modules to access the same memory areas and communicate with each other.

Kaveri will take that idea even further with shared memory and a unified address space.

[Milton Kuo]

Re: AMD expected to announce 30 percent workforce cut next week

Oh noes! If AMD dies, Intel prices will go through the roof.

I need to build a new computer soon. Was waiting for AMD's next iteration APU after Trinity, but I think I'll pop for the Trinity now.

While things in the microprocessor field have certainly changed a lot in the past 15 years or so, it's still highly likely that AMD is not going to die because it would be in Intel's best interests to keep AMD alive. Intel still has a large enough market share for microprocessors used in desktop PCs, notebook computers, and servers that, should AMD disappear, the drums of anti-trust investigations could start beating again.

[aaron]

Re: AMD expected to announce 30 percent workforce cut next week

they just gave AMD 1.5 billion or so to settle all that anti-trust stuff. Intel has to pay again?
Simply put, AMD may not be able to compete in the new world order.

[Milton Kuo]

Re: AMD expected to announce 30 percent workforce cut next week

they just gave AMD 1.5 billion or so to settle all that anti-trust stuff. Intel has to pay again?
Simply put, AMD may not be able to compete in the new world order.

It's not about paying money to AMD for reparations; it's about making sure AMD is around to give the semblance of competition in the markets so the Federal Trade Commission doesn't start causing trouble for Intel.

Intel benefits if AMD is around as a weak competitor, is certainly harmed if AMD is a strong competitor, and is likely to be harmed if AMD disappears in a bankruptcy.

[FrankL]

Re: AMD expected to announce 30 percent workforce cut next week

Intel will be competing more and more with the likes of Qualcomm, Marvell and Apple in future, who have the expertise to develop/tweak their own ARM-based architectures. That's why you see Intel investing heavily in trying to cut down power consumption to ARM-like levels.

Desktop computing is yesterday's technology, and high-powered computing will mostly be used for servers in datacenters in future. The average household already probably has at least 10 times as many ARM or MIPS based processors in use as they to traditional high-powered CPUs (by Intel or AMD). This ratio is only going to increase.

I think that within 5 years ARM will also make up a substantial part of notebook-formfactor based computer sales. Their upcoming 64-bit implementation means that they'll want to address server use too in future. However, Intel is still very good with performance/watt at high-powered computing. Basically, ARM wants to become more Intel-like and Intel wants to become more ARM-like (high performance at ultra-low consumption levels). AMD will not be very relevant in this picture (unfortunately).

[Slimprofits]

Re: AMD expected to announce 30 percent workforce cut next week

http://articles.chicagotribune.com/2...-amd-and-intel

PC chipmaker Advanced Micro Devices said on Thursday it will cut its work force of nearly 12,000 by 15 percent, its second round of layoffs in less than a year...

[lakedaemonian]

Re: AMD expected to announce 30 percent workforce cut next week

It's not about paying money to AMD for reparations; it's about making sure AMD is around to give the semblance of competition in the markets so the Federal Trade Commission doesn't start causing trouble for Intel.

Intel benefits if AMD is around as a weak competitor, is certainly harmed if AMD is a strong competitor, and is likely to be harmed if AMD disappears in a bankruptcy.

Sounds like history might repeat itself in a way.

Anyone remember this one?

http://www.youtube.com/watch?v=WxOp5mBY9IY

BUT can AMD out innovate and out market Intel the way Apple eventually did with Microsoft?

I'm thinking no........

I'm thinking AMD will remain a convenient semblance of competition for Intel.

Does that make Intel a potential long-term investment due to the fact that AMD is no Apple and Intel is about 25% off it's 6 month high?

[Milton Kuo]

Re: AMD expected to announce 30 percent workforce cut next week

Sounds like history might repeat itself in a way.

Anyone remember this one?

http://www.youtube.com/watch?v=WxOp5mBY9IY

That announcement was made when it appeared as if Microsoft was going to take over the world. Back when I first saw that presentation, I distinctly remember thinking when Bill Gates' face appeared on the large screen, "Gee, this is eerily similar to Apple's 1984 commercial. 1984 wasn't like 1984, but maybe 1997 will be and Bill Gates is Big Brother?" Of course, it wasn't.

BUT can AMD out innovate and out market Intel the way Apple eventually did with Microsoft?

I'm thinking no........

I'm thinking the same thing. Apple, at least when Jobs headed the company, has always been seen as an innovator and typically were among the first, or the first, to introduce certain interesting or key new ideas to the mass market. AMD, as far as I am aware, has introduced nothing new to the markets of consequence. (x64 doesn't count since it's a logical evolution of the x86 platform that Intel ignored.)

The ATI graphics business that AMD bought might be more interesting for AMD than AMD's core CPU business. With the ATI business, AMD doesn't have to worry about a much larger, much richer competitor in a highly-capital intensive industry where the product must conform to very strict standards (the x86 instruction set and the baggage of backwards compatibility going back to the technology from the 1970s!) I feel that AMD would have a few more degrees of freedom to do something interesting with their ATI business than with their x86 business.

I'm thinking AMD will remain a convenient semblance of competition for Intel.

Does that make Intel a potential long-term investment due to the fact that AMD is no Apple and Intel is about 25% off it's 6 month high?

Personally, I wouldn't invest in Intel because the bulk of their profits still come from sales of x86 microprocessors. There is substantial risk, in my opinion, that this business could see a noticeable decline in unit sales over the coming years due to the so-called Post-PC era.

For many users in the Post-PC era, it is no longer necessary to own a general-purpose computer, which typically uses an Intel microprocessor. It is entirely possible for people to have all of their computing needs satisfied with devices using microprocessors based on the ARM or other architectures. The casual user who sends occasional e-mail, browses the Internet, or watches videos can get by just fine without a PC.

The Post-PC era, where a lot of the devices are mobile and thus rely on battery power, has also channeled hardware design to ARM or other architectures that do not have the baggage of the x86 architecture, which must dedicate silicon to providing backward compatibility. That extra silicon increases die size and power consumption: all things which matter greatly on mobile devices. The extra silicon also increases the cost of the processor.

Where it really gets interesting is that Microsoft themselves have revisited the microkernel architecture of Windows to come out with Windows RT, which runs on an ARM-based processor, for use in their Surface tablet. In light of how lousy Windows notebook computers are, I can easily see Microsoft releasing their own line of notebook computers. It is entirely possible Microsoft will release notebook computers that use ARM processors to give the machines better battery life than can be gotten from similar notebook computers using Intel processors.

Notebook computers based on x86 processors would only be needed if binary compatibility with old applications that are no longer updated is required. [Even that may not be necessary if some company comes out with a virtualization product that virtualizes across different processor architectures.] For applications that are being updated, a recompile should be all that is necessary to make a version available for a version of Windows that runs on a non-x86 architecture. A recompile won't even be necessary for applications that are developed for a managed environment such as Java or Microsoft's .NET Framework.

All in all, I see a lot of already-existing technologies out there that, if someone puts it all together, could really put the squeeze on the Intel cash cow. There is already a business case for doing so. As Intel still hasn't developed a significant line of business outside of its x86 processors, I see them in a difficult position and wouldn't want to risk my money on them, nevermind the additional risk already present in equities due to Bernanke's meddlings.

[FrankL]

Re: AMD expected to announce 30 percent workforce cut next week

I agree with the general contents of your post, but with the following observations:
- AMD did use to innovate in the Athlon (K7) era, but in the processor industry where you rely on your own semiconductor fabs, you can only grow as fast as you can build extra capacity (which takes many years). Even in their good years AMD never turned a huge profit. Their situation for the past 15 years has been described as 'profitless prosperity' by observers. Their investors keep pouring money in, afraid to have to write off their huge previous investments if they'd go bankrupt.

- x86 overhead is minimal on modern process technology. It's more likely that the low power design experience of ARM, which predates Intel in this area in the order of decades, is the cause if the discrepancy.

- Intel is indeed facing declining revenues in the post-PC world. Even when (not if) they catch up with the power envelope of ARM processors, they will not be able to compete. Why? Because ARM licenses their cpu designs (and graphics too these days), and their licensees integrate them with all kind of other technology onto a package called System on a Chip (SoC). Intel doesn't want to do that, as the license costs are much lower than the revenue they get on their current products, which are almost completely in-house designed and manufactured. The less of their technology ends up in the end-product, the less revenue they'll get. At one point, they will have to however, not to lose out completely on this business.

- ATI is not going to save AMD. They sold off the part with the most potential for revenue growth in 2008: their embedded graphics design department. A monumental mistake on AMD's part (one in a long series of big mistakes).

- I'm not sure if Microsoft will play an important part in the post-PC world. They have a convergent roadmap for all their platforms (phone, tablet/smartbook, PC/notebook) which is good. But they might have already missed that boat... Let's see how much impact their Windows Phone and Windows RT products will have on Apple's marketshare. I'm doubtful it's going to be a gamechanger. In that sense, Samsung, Apple and Qualcomm are Intel's biggest competitors, and they're doing just fine squeezing Intel without needing Microsoft's help.

[Milton Kuo]

Re: AMD expected to announce 30 percent workforce cut next week

I agree with the general contents of your post, but with the following observations:

Thanks for the corrections. Quite obviously, I have no training as a semiconductor engineer.

- ATI is not going to save AMD. They sold off the part with the most potential for revenue growth in 2008: their embedded graphics design department. A monumental mistake on AMD's part (one in a long series of big mistakes).

Whoops, that wasn't smart. I don't follow AMD too closely so I had no idea they did such a thing. As you said, one in a long line of strategic mistakes by AMD; it seems like it's part of the AMD DNA.

- I'm not sure if Microsoft will play an important part in the post-PC world. They have a convergent roadmap for all their platforms (phone, tablet/smartbook, PC/notebook) which is good. But they might have already missed that boat... Let's see how much impact their Windows Phone and Windows RT products will have on Apple's marketshare. I'm doubtful it's going to be a gamechanger. In that sense, Samsung, Apple and Qualcomm are Intel's biggest competitors, and they're doing just fine squeezing Intel without needing Microsoft's help.

I brought up Microsoft because desktop Windows and its ecosystem could be seen as something of a last bastion for selling large quantities of Intel's x86 processors. However, with the release of Windows RT, Microsoft appears to have dusted off and decoupled various Windows subsystems to realize the benefits of the NT microkernel design, which makes it very possible that they will release desktop versions of Windows that run on processors other than x86 as was the case when Windows NT first came out. [In its earlier days, I seem to recall Windows NT running on x86, DEC Alpha, MIPS, and Power PC.]

For people who don't have legacy applications that require x86-based hardware, the benefits of using computers built on different processor architectures can be appealing; e.g., lower cost, lower power consumption, SOC integration resulting in smaller computers.

Speaking of Apple, there are rumors that Apple might change the processors used in their desktop Macintosh computers. In the past, they've switched from Power PC to x86. Perhaps they're considering changing from x86 to ARM, thus having a uniform CPU platform for OS X and iOS? This is just another potential pain point for Intel.

[Milton Kuo]

Re: AMD expected to announce 30 percent workforce cut next week

- x86 overhead is minimal on modern process technology. It's more likely that the low power design experience of ARM, which predates Intel in this area in the order of decades, is the cause if the discrepancy.

I feel like kicking myself for what I originally wrote as it's quite obvious that, by itself, x86 compatibility isn't going to take a lot of extra silicon.

[FrankL]

Re: AMD expected to announce 30 percent workforce cut next week

Intel will be competing more and more with the likes of Qualcomm, Marvell and Apple in future, who have the expertise to develop/tweak their own ARM-based architectures. That's why you see Intel investing heavily in trying to cut down power consumption to ARM-like levels.

Desktop computing is yesterday's technology, and high-powered computing will mostly be used for servers in datacenters in future. The average household already probably has at least 10 times as many ARM or MIPS based processors in use as they to traditional high-powered CPUs (by Intel or AMD). This ratio is only going to increase.

I think that within 5 years ARM will also make up a substantial part of notebook-formfactor based computer sales. Their upcoming 64-bit implementation means that they'll want to address server use too in future. However, Intel is still very good with performance/watt at high-powered computing. Basically, ARM wants to become more Intel-like and Intel wants to become more ARM-like (high performance at ultra-low consumption levels). AMD will not be very relevant in this picture (unfortunately).

Seems like my prediction was off by at least 3-4 years, as Apple is now shifting all their CPUs to ARM-based in-house designed technology (+ licensed stuff for what they cannot design in-house).

I think Intel will have a very rough decade ahead of them.
They now:
- are lagging behind in process technology (which is extremely capital intensive)
- are experiencing resurgent competition from a better-than-ever AMD
- have completely lost the mobile and embedded form factors to ARM-based designs from competitors
- are about to slowly start losing market share in their traditional strongholds, notebook, workstations and datacenter.