Slashdot Mirror
Nvidia Waiting In the Wings In FTC-Intel Dispute
The NY Times has a Bits Blog piece speculating on some of the fallout if the FTC prevails in its anti-competition lawsuit against Intel. The Times picks out two among the 26 remedies proposed by the regulator, and concludes that they add up to Nvidia being able to license x86 technology. This could open up 3-way competition in the market for combined CPU-graphics chips. There is a good deal of circumstantial evidence pointing to the possibility that Nvidia has been working on x86 technology since 2007, including the presence on its employment rolls of more than 70 former Transmeta workers.
Why does the remedy appear to be more harmful to AMD - an Intel competitor - than to Intel themselves?
Can you be Even More Awesome?!
What computer were you using at the turn of the century?
One with an x86 instruction set. Same as now.
I'm jaded enough to realize someone says so and so will be getting into the CPU market soon every few months. I've heard Creative and NVIDIA, probably some others I've forgotten. The thing that stands out to me is that VIA gave up. IBM gave up. Motorola gave up. Maybe the FTC can change things, but if they do it will probably break a few patent laws apart or force some fairly broad cross licensing agreements. Anything monetary is really just some fodder for the bankers to burn.
lol: You see no door there!
Intel x86. Serving all of us since 1978.
There's no reason to believe that this is going to change. Motorola's 68k never went anywhere, and PowerPC is dead. IBM's Cell went nowhere. AMD? Well they make a clone, and have 15% versus 83% marketshare, and one-fifth the revenue. Cyrix? Well they went belly up and got bought by NS, then Via. We're talking scraps. less than 2% of the market here.
Oh yeah, and AMD is teetering into bankruptcy. Primo competitive environment eh?
Since we are still stuck with Unix 40 years later and still will be 40 years from now, I can see that we could be still stuck with x86 for a long time. To the Computer Science graduate, they are flawed designs, but in the real world they work and work good enough not to merit a costly change.
Yes there are CPU architectures, but are they significantly better to warrant a change? Even Apple after touting the merits of PowerPC succumbed to the x86 train. Even Intel tried multiple times to bring an alternative to its x86 line (iAPX, i860, i960, Itanium), but without success. RHEL abandoning Itanium is one more example. Sun offers x86 hardware in addition to its SPARC line, so does IBM and HP, and every other server vendor. There were a time when x86 was laughed at and not considered server-class. Now most servers and super computers use x86 processors.
In the Unix-haters handbook, the refer to the original Macintosh OS as a better OS with better GUI than Unix and X, now Mac OS X is Unix, and if you jailbreak and ssh into your iPhone you'll find a familiar Unix under all the eye candy. Most servers either run Unix or Linux, so does most super computers. All assumed flaws of the Unix architecture accounted for nothing in the real world.
GENERATION 25: The first time you see this, copy it into your sig on any forum and add 1 to the generation.
If AMD and Nvidia can truly make competitive products, then having more of a non-Intel option makes that option seem much more mainstream.
I posted some of my thoughts on this topic here:
http://www.pcper.com/comments.php?nid=8143
Why would NVIDIA want to dive into such a complex product line when the GPU is becoming more and more important in general purpose computing anyway and that is obviously where their expertise is.
It's just an instruction set.
The modern CPUs you call x86s use a non-x86 core with an instruction decoder bolted on to make it run the x86 instruction set. It has been that way since the Pentium Pro, the NextGen chips and the AMD K5.
The AMD K5 in particular was pretty much identical to the Am29000 RISC processor. AMD just put a decoder on it and sold it as an x86.
CISC type instruction sets are considered to be the most optimal for code density (better cache and memory usage). So we pretty much have the best of both worlds. The instruction set is CISC so we get the memory benefits and the code is run as RISC via an instruction decoder which makes it easier to pipeline and for parallelism.
Actually, it's just the opposite. There WERE plenty of better architectures in the early days of x86. Today, x86 is just simply THE chip. The one that's left, competing for the high-end, pushing economies of scale, being all things to all people, and most importantly, with a healthy ecosystem of competitors continually trying to one-up each other.
Everything but the kitchen sink gets thrown into x86, to try to increase performance on various tasks. If there was a better chip out there, it would get integrated into x86 in no time. FPUs come to mind. x86-64 and SIMD instructions come to mind. GPUs seem to be the next big deal, with AMD looking to have an x86-64 CPU in one socket and a GPU in the other...
In short, if anything better comes along, it will quickly get integrated to Intel/AMD/VIA CPUs, and then there once again won't be anything "better"...
Slashdot gets worse every day... Pipedot: News for nerds, without the corporate slant
Intel didn't help themselves by making IA64 expensive to license and program for.