Let's start with the obvious. Everyone is spoiling for a fight to end Intel's hegemony in the server processor racket. But two decades ago we backed Intel to oust that arena's ruling silicon families kicking - and the chip giant successfully invaded servers around the world. The question we now face is this: can ARM do to Intel …
"I may be wrong, but I am not in doubt."
Easy: religious faith
He has no evidence to support his theory, but has complete faith.
Intel are very good at performance. They are getting tolerable at power consumption. I cannot see them competing on price. I am sure Intel will produce some excellent high speed low power chips with their next generation super expensive manufacturing process. ARM partners will do the same with the current manufacturing process which will be cheap as chips by then.
2001 Linux has AMD64 support working on a simulator
2003 real AMD64 CPU's for sale
2005 Windows XP Professional x64 edition (AMD64) released
200? AMD64 Windows drivers and applications
Anyone going to hold the breath until Windows RT64?
Not so sure about this bit either...
"Moreover, according to Gelsinger, when ARM server parts finally do come to market, the difference between a low-voltage x86 part and a high-end ARM part will not be significant."
Anand's review of the new Samsung Chromebook, which is the first device to market with an A15, states that versus an admittedly older Atom design the A15 has 40-60% better performance, costs $20-30 per chip vs. $86 for the Atom, with significantly lower power consumption (given the method of comparison it is difficult to say how much exactly... my guess is in the 30-40% range).
Marketing being what it is... ARM's claims on the A53 and A57 64 bit designs imply that they're at least serious about "real"/non-mobile/non-embedded applications for their chips. If their claims of 3x the performance at 1/4 the power (vs. the A9 IIRC) are accurate, of course, remains to be seen... but if they can get to a point of "good enough" performance per socket, paired with lower cost per chip (a certainty), and a competitive if not superior performance/watt equation (a possibility)... this could present a real challenge to Intel. Please not that I refrain, however, from even guessing at a winner.
Competition is a good thing.. for everyone. So cheers on that.
An interesting thought...
What about the home data centre?
Does this mean he envisages that the desktop/laptop in the home could see a rise in small centralised data servers in the home that serve the edge devices such as tablets, TV's, etc. ?
Re: An interesting thought...
32 bit is probably fine for a home server - its about availability, not cpu crunching.
WTF is the second graph? Why is this a line graph? What does it MEAN if a bit of revenue is halfway between "other mobile" and "x86 mobile?"
I question the ability of anyone who came up with such a horrid graph to understand the data they are trying to present.
I don't understand why he didn't use time along the x axis and coloured lines for the different catagories.
Agreed I was about to post the same thing I don't know what the continuum means, nor can I understand why time isn't the X axis? It is truly weird. It could only have come from an arts graduate consultant, not from an engineer, it is meaningless to anyone with any basic maths background.
I baulked at the first paragraph then bailed out half way throught the second. Editor, help, please?
Resting on laurels
That graph of shipments and revenues would seem to indicate that intel have been doing their own laurel resting (aka price gouging) because it shows shipments flattening out while revenues continue to rise which is a sure sign of a company without any serious competitors.
- Review Reg man looks through a Glass, darkly: Google's toy ploy or killer tech specs?
- MEN WANTED to satisfy town full of yearning BRAZILIAN HOTNESS
- +Comment 'Stop dissing Google or quit': OK, I quit, says Code Club co-founder
- Apple tried to get a ban on Galaxy, judge said: NO, NO, NO
- Banking apps: Handy, can grab all your money... and RIDDLED with coding flaws