Hehe. Okay, nice news to me.
But then, yes, I believe even though it is a lot of money intel still came off quite cheap. And so they admit their faulty deeds. Well.
Next thing is that the argument of intel having interest in keeping AMD alive is truly valid.
First, withouth AMD there would be far less innovation, intel would be lazy and we'd still have these power hungry P4 sold for prices up to the moon (I still remember the 486 times, they cut off 50% of price within a week).
And yes, I also think that anti monopoly agencies would quite hack them to pieces, which is probably intel's concern.
I mean, there'd only be VIA left and Transmeta is already out of business in the x86 field.
Even the governments heavily rely on x86 tech (I mean most people today really do, knowing it or not) and nobody there would want a price dictatorship.
Its probably good to have now at least these 3 players in the x86-market. Compatible in hardware but each one doing innovations and all of them having to keep prices on reasonable levels.
So I hope AMD will use the money well. If I wouldn't gift myself a little (but still relatively expensive) ARM machine for christmas it'd be another AMD. Well then, there is still new year that should hopefully bring me a little financial aid that I might invest in GlobalFoundries/AMD and some free software services/books.
True, this is something that they must fear considerably. =/ Maybe this is why MS doesn't buy out linux / mac os etc...
Originally Posted by birdie
I'm not sure how Intel could really be broken up in a way that diminishes its monopoly power. It's not like Microsoft where the proposed breakup would have separated two product lines that were independently viable and already being sold separately, or AT&T, where a national monopoly could be broken up into regional monopolies.
Originally Posted by birdie
You can't buy out linux.
Originally Posted by b15hop
CONGRATZ AMD/ATI & JERRY, THIS DEAL HAS BEEN A *VERY* LONG TIME COMING. :-D
Party happened, to mark this. :-)
Now, PLEASE hire some more software guys, so the FGLRX work can make progress faster.
Originally Posted by GreekGeek
fglrx should stick with embedded stuff since it can never offer any kind of benefits out-of-the-box (due to licensing issues).
R600G is where the future is. And that is where the bulk of development must be pushed.
Make their patents public domain. Who doesn't want hyperthreading in their Athlon II or Nano? :-)
Originally Posted by Ex-Cyber
Care to elaborate? AFAIK, Catalyst's license allows distributions to bundle the driver with them; it's just that most "mainstream" distros still keep their "only open source software enabled by default" policy.
Originally Posted by droidhacker
I don't. Hyperthreading is one of those concepts that sorta makes sense in theory, but in this era of multi-core systems, adds a lot of complexity for little gain in the real world use cases.
Originally Posted by Zapitron
On a single core system, it makes a lot of sense as it DOES help with perceived latency. On a dual core, it massively complicates scheduling, and the perceived latency benefits are reduced. There are also many benchmarks that show it slows down well tuned applications, even on single core. (This is completely ignoring the benchmarks where it brings a dual core down to 50% performance.)
Hyperthreading is, in my opinion, best summarized on modern multi-core chips with it's first 4 letters... Hype.
I was always under the impression that hyperthreading existed only to help manage MS's really really crappy process scheduler by taking some of that process scheduling away from MS.
Originally Posted by RobbieAB