Chip rumor-gasm: Intel to buy Altera! Samsung to buy AMD! ... or not Rumors flew fast and furious on Friday about possible acquisitions in the chip industry, with both Intel and Samsung said to be eyeing purchases that could strengthen their positions against rivals. First up, Intel is thought to be planning to gobble specialty chipmaker Altera, in what could become the largest acquisition in …
The problem with AMD is their never had the R&D capital or advertising money to compete with Intel. Back when AMD was better, Intel used illegal tactics to keep AMD down and all AMD got of it was a large settlement. But imagine if AMD had been allowed to compete on their own, AMD would have been better off and better competition is a good thing.
And today, AMD finds itself in another bad position. The current A-series line of CPU's is good enough for most people, but OEM's tend to put them in junk machines. Only HP puts the AMD chips in good machines, and those are only special order.
If Samsung bought AMD, they could get more capital, have more R&D, compete with Intel, forcing Intel to lower prices and work harder, which will force AMD to keep prices lower and work hard, and we all win.
This post has been deleted by its author
This post has been deleted by its author
Assuming Intel produces anything remotely worth being called "graphics chipset" must be a cynical joke. Intels chips transport pixels at the pace of ATI/NVIDIA products from the early 2000's.
So support for Intel graphics in Open Source system is moot, there is no such thing as Intel graphics. See this link for background info: http://www.notebookcheck.net/Mobile-Graphics-Cards-Benchmark-List.844.0.html
Perhaps Intel should buy AMD, so they get a credible graphics lineup.
A lot of companies use AMD processors. Lenovo is using them in their ThinkPad E555 series of laptops and their ThinkCentre M79 series of desktops. I also think that AMD shot themselves in the foot on more than one occasion. They had Intel running scared 10 years ago then they followed up K8 and K10 with the Bulldozer and that was the beginning of the end.
"The current A-series line of CPU's is good enough for most people"
This is true, however the equivalent performance Intel part is usually cheaper until you get down to the bottom end of the A-series line (which are the parts that end up in junk machines)
I really wish it wasn't so.
As a consumer who has switched between amd and intel over the years I can see some sense in this. I just bougth an amd because it was i5/i7 speed for i3 pricing (in a sale). AMD has a place in the market, they are doing some interesting things compared to intel, but they dont have the scale behind them. The question for us as consumers is would an acquisition by Samsung leave them better able to take on Intel? Access to current generation fabbing would work for AMD and bring even better economies of scale for Samsung. Access to all the GPU tech would make for interesting moves in the ARM products from Samsung. If Samsung can get them for a sensible price it could work out very well for everyone, especially us. I can dream anyway!
No one here is talking about the biggest problem.... x86 license which AMD bought from Intel .. is NON Transferable... so the only way to merge is AMD buys Samsung... i.e. Samsung gives charity to AMD so AMD could buy them.. hence license will remain valid. but the name SAMSUNG will have to die or become subsidiary of AMD.. a similar war broke out between Intel and AMD upon outsourcing fab to GlobalFoundaries ... I might be wrong but AMD manages to slip through by hair margin and even that is no more left.
>Finally, someone with the money to give Intel some real competition?
There's probably more to be made in integrating the two architectures. That might be true even if its only the ability to run on ARM and keep the x64 bit in deep sleep, waking it up when you really want it.
ARM needs some vision to make it to the desktop/laptop. I'm not sure if anyone has that vision matched with the finances to do it.
Errrrm, talk about cart before horse... have you actually seen the way PC sales figures are going in the last couple of years, and the way there's not enough money in the PC business any more (not for Western financial experts anyway)?
Intel needs some vision to make it outside the desktop/laptop/x86server world.
Intel has had plenty of time, plenty of opportunities, and blown plenty of cash.
Plenty of success? Well I guess there's Tesco's Hudl tablet. Any routers? TVs? Car/truck engine control units or similar? Consumer or professional equipment in general? Anything much outside the IT department's territory?
For some things in life that need a processor, there's x86 (especially if you need Windows).
For everything else, there's a better alternative (in volume terms, it's mostly ARM, occasionally MIPS, PowerPC, SPARC, etc).
Back in the 90-es and early 2000-es internal Intel politcs commanded that anything and everything that is not x86 (current) or Itanic (going forward) is to be burned at the stake. It had RISC CPUs which sold like hot bread for industrial controllers, but it deliberately starved them from investment. It involuntarily became an ARM licensee (via StrongArm), murdered that one as well.
When companies are performing well, they just ignore the Wall St talking heads, but as soon as there are problems, the companies must start doing whatever is deemed to be the right thing to be dooing, otherwise the shareholders give them hell and want boards fired etc.
That's much of what drove dot-bomb. Even companies with savvy boards were driven to buying up crazy start ups because that was what everyone did back then.
Now everyone is being forced to buy up other companies. Does Altera make sense for Intel? Maybe.
Even though Altera SoCFPGAs are ARM based and not x86 based and have little in common with Intel, an injection of Intel process savvy could really help the FPGA people make faster devices that use less power. (The current Altera board I'm working on eats about 8W and gets HOT.)
Intel has lost effective market share - their Atoms are crap and being killed off by ARM. Intel need to find some new business to get some growth and maybe they can do that in the FPGA/Embedded space.
With both the FPGA behemoths (Alteria & Xilinx) using ARM, that is certainly a threat to Intel. Not really the current cortex-A9 dual-cores which are just about good enough to run an OS, but the upcoming quad-core 64-bit ARM based parts that also have a full spectrum of "proper" peripherals that would enable them to be used as the heart of a computer system.
"With both the FPGA behemoths (Alteria & Xilinx) using ARM, that is certainly a threat to Intel."
It's only a threat to Intel's x86 business. There's plenty other impressive stuff for Intel to fall back on.
There's ....
Actually, what is there, exactly, in the last few decades, especially if you ignore stuff they bought in (AMD64, some StrongARM, VxWorks, McAfee???) and focus on stuff Intel developed in-house (IA64?).
[tumbleweed]
"As soon as there are problems, the companies must start doing whatever is deemed to be the right thing to be dooing, otherwise the shareholders give them hell and want boards fired etc."
Not so applicable to Chaebois though - and I can see Sammy regarding AMD as a risky buy.
A chinese company buying AMD would probably be blocked by the US gov on national security grounds.
AMD makes graphics cards which might be quite powerful for competition, but their software is terrible. Really, awful.
I have a few of the Intel Xeon-Phi's and all I can say is it is *beautiful* engineering, and with loads of information. Even compilers.
Nvidia is eating AMD's lunch and so is Intel.
Perhaps Samsung might see something we don't?
It's a shame, because we all know that uncontested markets lead to bilked customers.
And I *still* want 100 Tflops on a single chip...!!!!!
P.
AMD makes graphics cards which might be quite powerful for competition, but their software is terrible. Really, awful.
The worst thing for me is the installer. It's so full of ads that the thing is huge (as in screen space). try installing it on a fairly low resolution screen (such as a 720p TV) and you'll find that the installer is bigger than the screen, meaning it's not possible to click on the buttons (or even see them).
Leads to a non-lethal (but very annoying) form of Russian Roulette with the tab key.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
