The purpose of RAS is so that people like Anthony don’t have to waste their weekends fixing broken servers. ;^)

The features of Power and Itanium (which are currently about equal in revenues) that make them attractive as servers are RAS (redundancy and stability.) Many of those features are being moved into x86, which is why x86 server sales continue to grow while the RISC server sales have peaked.

As far as PowerPC goes, there were many people at IBM, Apple and Motorola expecting it to be 2-3X faster than x86. In fact, the performance difference was a wash – which killed it on the desktop. Modern x86 chips offers a RISC like core with the cache efficiency of CISC. The orthogonal RISC instruction sets are inherently memory inefficient by design.

You’re right, game consoles are loss-leaders, sold by game companies in anticipation of licensing revenues from game sales. However, the various components (chips, controllers, etc…) are not loss-leaders, but revenue centers for the various companies that make them. If it weren’t so, they wouldn’t bother as they are not entitled to licensing revenues. In some cases these profits are subsidized by the game console vendors, but they are profitable nonetheless. The same is true for Cell phones, which are also loss-leaders, subsidized by the various phone companies, nevertheless, TI and Qualcomm have done quite well selling chips for phones.

As for Power vs. PowerPC, the original discussion was RISC vs. CISC and my point has always been that RISC was technologically superior and thus has been very successful, which is true. I said nothing about PowerPC. My point was that CISC did not beat out RISC, rather it morphed itself into pseudo-RISC.

Finally, as for building a faster 68K, Motorola already did that. The 88K line was a flop! In 1991, x86 was deeply entrenched as the defacto standard of the PC market. Only a revolutionary change like the PowerPC had any hope of overcoming the x86/intel/microsoft juggernaut. In fact, the PowerPC did succeed in gaining some market share for the Mac. Ultimately, it failed to displace x86, but the reasons for that were mostly non-technological (eg. Apple pulling the rug out from under the Mac Clone business, especially Motorola’s Clone business, Motorola’s other business troubles siphoning away resources, cut-throat competition between IBM and Motorola for the small Mac business and Apple’s manipulation of both sides, FUD from Microsoft and Intel, etc…)

Having said all of that, it should also be recognized that Motorola and later Freescale leveraged PowerPC into a very profitable and successful embedded business. I don’t know if IBM lost money on PowerPC, but it should be recognized that their highly profitable POWER business is relatively low-volume, but requires them to operate a state-of-the-art High-Volume semiconductor Fab. Keeping such a fab loaded, even with only marginally profitable parts, such as PowerPC or Cell (for Video Games) is enough to make the overall operation profitable.

Game consoles are sold as loss-leaders. I bought a PS3 when it first came out for less than the cost of a standalone Blu-Ray player.

IBM’s Power was around long before PowerPC. That is a separate business and is not part of the $5 billion. The server numbers you quoted are Power, not PowerPC. Since PowerPC appeared in 1991, x86 sales have been measured in the 100s of billions of dollars.

If IBM wanted Mac business, they could have much more cheaply collaborated with Motorola to build a faster 68K. It took Apple years to get their entire OS ported to PowerPC, and then they switched to x86.

I remember Motorola long-term sales projections for PowerPC from around 1992, and they overestimated by several orders of magnitude.

I don’t think you can argue that IBM threw $5 billion “down the drain” chasing after RISC, when they have made many billions selling these same chips for Macs (now defunct), Video game consoles, and High-end Servers. In fact their domination of the high-end computing market (210 of the top 500 supercomputers are IBM machines) is a major reason why IBM is currently one of the most successful IT companies on the planet.

RISC was the right direction for IBM, not a pipe-dream lead by starry-eyed academics. Sure they over-hyped the doom-and-gloom scenario for CISC in their bid to wrestle the PC away from the dominant market player. But that was just good old fashioned FUD marketing, not bad business. Unfortunately, their efforts put them squarely in the cross-hairs of Microsoft, the grand masters of FUD and the rest is history.

Michael D Smith (04:06:27) :

The fact is, anon, that real scientists ARE being dragged through the mud over AGW. The shocking thing is, that too few scientists have the courage to speak up, counter the political hype and stop the madness…

Go back read all the posts. Scientists (Im one) who may disagree have 2 choices, 1) do nothing/say nothing, 2) get wrapped into the lump with the lot who wrote these posts.
guess what someone who doesnt want to attempt carear suicide is going to do.

Some more NASA highlights:

It was a science fiction fantasy come true: Ten years ago this summer, NASA announced the discovery of life on Mars.
http://www.cbsnews.com/stories/2006/08/05/ap/tech/mainD8JAGIEG0.shtml

New evidence of life on Mars spotted by NASA
http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2009/01/15/MNV715B3HE.DTL

And, although Intel now has an SOC chip based around the Pentium M core, I believe, ARM and the 604 architectures still get most of the embedded design wins, and MIPS is still around in things like the Cavium processors and so forth.

The lack of self-criticism in the world of climate science is a primary reason why sites like WUWT and CA exist. Academics can’t bludgeon the energy basis of the world economy and expect that all 7 billion people on earth will go quietly – no matter how clueless the politicians may be.

Your argument about “design principles” is correct, which is exactly the point – the nearly unanimous “consensus” of academics were wrong. There was never any need to change instruction set architectures for personal computers. IBM threw $5 billion down the drain because of out of control academics, who took one piece of correct information and blew it up into a nonexistent problem and solution. Sound familiar?

The PC marketplace in the future will be dominated by x86 + Larrabee, Nvidia or AMD GPUs. ARM will be used a coprocessor for lightweight tasks.
http://news.cnet.com/8301-1001_3-10162275-92.html

I think you’re misreading those comments. “Scientists” has quotation marks, indicating that a particular person who uses inflammatory language like ‘coal trains of death’ is abusing their status as a scientist.

All real scientists are skeptics. But they are also human, and some of them are bought by money, status or ideology. That’s what President Eisenhower was warning us against in his farewell speech.

The two examples above refer to Michael Mann and James Hansen. They have sold out for personal gain and status. But they are not the typical scientist, who cringes at their perversion of the scientific method.

Rather than getting upset at someone who refers to others making outlandish statements like ‘coal trains of death,’ or to someone who uses a known false algorithm to promote their “hockey stick” conclusion, your proper target of scorn should be those corrupt scientists who make the great preponderance of honest scientists look devious and corrupt.

The public tends to generalize, and Mann and Hansen are the prime contributors to the public’s view of scientists. They are always in the news. Those two, more than anyone else, have contributed to the false impression that all scientists are corrupt and political.

This site points out the truth of the matter. You can see that a lot of serious thought goes into most of these posts. And if you think about it, you will see that the sooner that the false prophets of doom are marginalized, the better it will be for mainstream science and the taxpaying public.

“…seriously incomplete. When the “scientists” give us the debunked “hockey stick” and talk of “coal trains of death”, and when the policy makers…”

a couple of scientists do shoddy work, it catches on politically…now all scientiest are worthless??

use your head

I hate to throw cold water on a good debate, but the real reason that RISC was never able to beat out x86 was that NexGen and later AMD and Intel figured out that they could adopt RISC design principles into x86 designs. Basically, they used decoder stages to convert CISC instructions into RISC-like micro-ops, which can run on a RISC-like x86 core. Such a design may take a small performance hit relative to native RISC (eg. 5-10%), but otherwise provides RISC-like scalability. The small performance hit is in the noise when you consider variations in process capabilities (eg. raw transistor performance) and advantages of CISC for certain applications. Add in the backward compatibility of x86 for a large installed base and there was no way RISC could displace x86.

Having said that, don’t count RISC as dead just because it’s not competitive on the desktop. ARM processors pretty much dominate the low-cost, high-performance market for embedded processors (think iPod’s and SmartPhones). Freescale’s PowerPC chips drive most Cisco routers and thus form the backbone of the internet. And finally, IBM’s Cell architecture drives all three leading game controllers (Wii, Xbox, PlayStation3) and it’s POWER Architecture fuels four of the world’s ten fastest supercomputers.

As a scientist, science teacher and physician, your glib sarcasm about the solution is to cut out government funding altogether actually gets at the real problem: nearly every scientist in the nation is dependent on the government for funding and position and credibility, and if they step very far off from the political ‘consensus’ for certain issues, a scientist will lose all three. Indeed, most all scientists are dependent on the scientific-technological elite and its relationship with the political class for training and acceptance into the community, essentially training and admitting only those to the academy who are blindered by the ‘consensus.’