Slashdot Mirror

I had a couple of 'why' questions and found a possible answer.

This Intel forum:
http://software.intel.com/en-us/forums/showthread.php?t=67843

Contains many complaints about the performance of the Intel compiler under WinCE (as well as Intel selling it all off to Marvel).

If you ask me, MS wants a chip that they can optimize for their OS. Seems liek this will lead it down a proprietary hole, not unlike Apple.

So, is that the real story, "MicroSoft, now more like Apple!"

Intel already licenses the ARM architecture and uses it in their xScale line of processors

Currently Intel supplies compilers, but this Intel forum contains many complaints about performance under WinCE.

http://software.intel.com/en-us/forums/showthread.php?t=67843

Seems, like MicroSoft decided to take things into their own hands

> Most games, with a few exceptions, are single-threaded applications. Gamers are much better off with a higher clocked dual core system than a slower-clocked, 6 core system.

False, for games written after 2006 - 2008. If a game is cross platform such that it is meant to run on PS3 or XBox360, then it is pretty much garanteed to be multi-threaded - you'll never get great performance if your game is single threaded on those consoles, especially on the PS3 where you have 6 SPUs.

References:
* http://www.gamasutra.com/features/20060906/monkkonen_01.shtml
* http://techreport.com/articles.x/11237
* http://software.intel.com/file/1478
* http://en.wikipedia.org/wiki/Multi-core_processor#Disadvantages
* http://scrawlfx.com/2008/06/killzone-2-uses-4-12-of-ps3s-6-cell-chip-cores

You have access to this thing called the Internet. Go to ark.intel.com - it's been there a long long time. It tells you *all* the details of every chip they sell.

They do sell a huge array of chips with many subtle differences - it's not like the good old days when there were 4 speeds of 486 to choose between - bus speeds, chip speeds, integrated GPU, process node, cache sizes, VT-x extensions, SSE4.2, AES, power draw, number of cores, Hyper-threading, form factor, socket, lead-free, PCIe lanes, ebmbedded...

Many, many people don't care, and buy their PCs to a price, and based on how it looks. For those of us that do care, or need a particular feature, it's *really* easy to find out what's what.

I thought Intel's website had a pretty good "joe consumer" cpu choosing wizard to help them pick out a CPU.

Except thats not what he said - what he described would be to play the game on the PC in your bed-room/office using the TV in your living room as the video output without having to physically move the bed-room PC. Basically sounds like what Intel's WiDi (wireless display) that have had ads on OTA TV lately is claiming to achieve.

The thing is, what if that fails, then you have to wait 10 other attempts for somebody to come out with a common display adapter that actually sticks in the market?

And let the battle for a new standard begin.

I had thought Light Peak was the likely replacement technology.

10Gbps and backward compatible with USB.

"At 10Gb/s, you could transfer a full-length Blu-Ray movie in less than 30 seconds. Optical technology also allows for smaller connectors and longer, thinner, and more flexible cables than currently possible. Light Peak also has the ability to run multiple protocols simultaneously over a single cable, enabling the technology to connect devices such as peripherals, displays, disk drives, docking stations, and more."

http://techresearch.intel.com/articles/None/1813.htm

I wasn't too happy that Intel axed the parallel port, but I could get cards/USB adapters for that. Now they axe PCI?

One example of a new model with a parallel port. If you want an upcoming 6-series just use a PCIe card.

answered my own question. I looked up the z530 Atom cpu on: http://ark.intel.com/Product.aspx?id=35463 and it DOES support Intel VT technology as a 32 bit processor.

Negative: http://ark.intel.com/Product.aspx?id=35641&code=atom+330

As for 64-bit, they may well have gotten custom Z530s that can do 64-bit - make a large enough order, and Intel will do custom stuff.

"The ATOM doesn't support virtualisation in hardware"

Incorrect.

http://www.seamicro.com/?q=node/38
"Processor Specification Intel Z530: 1.6GHz, Single Core, Dual Thread x86 Processor"

http://ark.intel.com/Product.aspx?id=35463
"Intel® Virtualization Technology (VT-x) - Yes"

Not entirely true. Some of the Atoms have VT support and some have AMD64 support, but I don't think any have both.

http://ark.intel.com/ProductCollection.aspx?familyId=29035 (Doesn't show AMD64 support on the main page, but you can drill down to find 'em).

I just downloaded Google Chrome 3.0.192.0 for Mac and it crashed before I could even open a page. There is no excuse for this; my Mac Pro is perfect in every way with eight 2.93 GHz cores, 32 GB RAM, and a fresh install of Mac OS X Leopard v10.5.7. Ergo any crashing Google Chrome does is Google Chrome's own fault!

Why is it that Apple and Mozilla can do this but Google can't? I ran Internet Explorer 8 for months before its final release, Firefox 3.5 since its 3.1 days, and found Safari 4 Developer Preview more stable than Safari 3. In fact, even WebKit is more stable than Chrome.

What really baffles me, however, isn't the instability I've come to expect from Google, but that Google has the audacity to ask for personal user info to improve its browser. Is the search engine maker datamonger really so desperate for my private information that it's stooped to the level of Trojan horses to get it?

They should ask me that when it doesn't crash on launch.

Everything Google does is just another way to sieve personal data away for targeting ads. This kind of Big Brother crap is more repulsive than the fat programmers that make it possible. Google, with its deep pockets and doctoral scholars, thinks that by holding user data hostage it can maneuver around Apple and Microsoft. While this may be true, I'm not willing to be a part of it.

In using Google's search, Gmail, Chrome or whatever else the faceless robot of a company invents, the user is surrendering their personal information to a giant hivemind. No longer are their personal preferences some choice they make; they're a string of data processed by a Google algorithm: Google dehumanizes its users!

So while Google is arrogant enough to paint spyware shiny so it can parse our browsing habits, the least they could do is make sure it doesn't crash. If Apple, Microsoft, and Mozilla can get their preview releases right, why can't Google? And now they're making their own operating systems?

Get real, Google! I'll use your crashing codebloat when my Mac is cold and dead and I'm looking for handouts. Until then, quit mining my personal data!

Here's the Intel one I've been using for some info, since it's basically their standard, created for the Itanium: http://download.intel.com/technology/efi/SF09_EFIS001_UEFI_PI_TCG_White_Paper.pdf

HW Manufacturers will require EFI firmware to be signed in order to install it. See "executable verification" in this PDF

I just downloaded Google Chrome 3.0.192.0 for Mac and it crashed before I could even open a page. There is no excuse for this; my Mac Pro is perfect in every way with eight 2.93 GHz cores, 32 GB RAM, and a fresh install of Mac OS X Leopard v10.5.7. Ergo any crashing Google Chrome does is Google Chrome's own fault!

Why is it that Apple and Mozilla can do this but Google can't? I ran Internet Explorer 8 for months before its final release, Firefox 3.5 since its 3.1 days, and found Safari 4 Developer Preview more stable than Safari 3. In fact, even WebKit is more stable than Chrome.

What really baffles me, however, isn't the instability I've come to expect from Google, but that Google has the audacity to ask for personal user info to improve its browser. Is the search engine maker datamonger really so desperate for my private information that it's stooped to the level of Trojan horses to get it?

They should ask me that when it doesn't crash on launch.

Everything Google does is just another way to sieve personal data away for targeting ads. This kind of Big Brother crap is more repulsive than the fat programmers that make it possible. Google, with its deep pockets and doctoral scholars, thinks that by holding user data hostage it can maneuver around Apple and Microsoft. While this may be true, I'm not willing to be a part of it.

In using Google's search, Gmail, Chrome or whatever else the faceless robot of a company invents, the user is surrendering their personal information to a giant hivemind. No longer are their personal preferences some choice they make; they're a string of data processed by a Google algorithm: Google dehumanizes its users!

So while Google is arrogant enough to paint spyware shiny so it can parse our browsing habits, the least they could do is make sure it doesn't crash. If Apple, Microsoft, and Mozilla can get their preview releases right, why can't Google? And now they're making their own operating systems?

Get real, Google! I'll use your crashing codebloat when my Mac is cold and dead and I'm looking for handouts. Until then, quit mining my personal data!

I just downloaded Google Chrome 3.0.192.0 for Mac and it crashed before I could even open a page. There is no excuse for this; my Mac Pro is perfect in every way with eight 2.93 GHz cores, 32 GB RAM, and a fresh install of Mac OS X Leopard v10.5.7. Ergo any crashing Google Chrome does is Google Chrome's own fault!

Why is it that Apple and Mozilla can do this but Google can't? I ran Internet Explorer 8 for months before its final release, Firefox 3.5 since its 3.1 days, and found Safari 4 Developer Preview more stable than Safari 3. In fact, even WebKit is more stable than Chrome.

What really baffles me, however, isn't the instability I've come to expect from Google, but that Google has the audacity to ask for personal user info to improve its browser. Is the search engine maker datamonger really so desperate for my private information that it's stooped to the level of Trojan horses to get it?

They should ask me that when it doesn't crash on launch.

Everything Google does is just another way to sieve personal data away for targeting ads. This kind of Big Brother crap is more repulsive than the fat programmers that make it possible. Google, with its deep pockets and doctoral scholars, thinks that by holding user data hostage it can maneuver around Apple and Microsoft. While this may be true, I'm not willing to be a part of it.

In using Google's search, Gmail, Chrome or whatever else the faceless robot of a company invents, the user is surrendering their personal information to a giant hivemind. No longer are their personal preferences some choice they make; they're a string of data processed by a Google algorithm: Google dehumanizes its users!

So while Google is arrogant enough to paint spyware shiny so it can parse our browsing habits, the least they could do is make sure it doesn't crash. If Apple, Microsoft, and Mozilla can get their preview releases right, why can't Google? And now they're making their own operating systems?

Get real, Google! I'll use your crashing codebloat when my Mac is cold and dead and I'm looking for handouts. Until then, quit mining my personal data!

What? The cheapest LGA1156 CPU listed on Newegg, the Pentium G6950 for $98, supports VT. http://ark.intel.com/Product.aspx?id=43230 So does the cheapest i3 they offer, the $115 530. http://ark.intel.com/Product.aspx?id=46472

The older LGA775 chips did use VT as a selling point, but I believe that most current Intel CPUs support it. Since VT was required for Win7's VirtualPC XP Mode, it became a more standard option recently.

What? The cheapest LGA1156 CPU listed on Newegg, the Pentium G6950 for $98, supports VT. http://ark.intel.com/Product.aspx?id=43230 So does the cheapest i3 they offer, the $115 530. http://ark.intel.com/Product.aspx?id=46472

The older LGA775 chips did use VT as a selling point, but I believe that most current Intel CPUs support it. Since VT was required for Win7's VirtualPC XP Mode, it became a more standard option recently.

How you managed to read that I was suggesting that there weren't any and I was somehow lamenting that fact is beyond me. How you turned my post questioning the choice of location into "telling others how to conduct their business" is mystifying. You must be great fun in person when someone asks "why?"

To summarize my post and possibly increase your reading comprehension: I questioned why anyone would build a data center in California. I provided an example of an alternative that I was familiar with.

Since you seem to think that I was suggesting that there aren't any big data center projects here, I'll do the name-drops: Google, Amazon, Facebook, Intel, and Mission West all have or are constructing data centers here. There are more, of course, but those are well-known.

Intel's Extreme line, for one. The X25-E goes up to 64GB. It's a 2.5" form factor, but it's a SATA drive and you can use a 3.5" bay with mounting rails to put it in a desktop.

GP is right about being expensive...expect to pay over $600 for the 64GB model.

The CPU in my Acer Aspire 5930 is one example: http://ark.intel.com/Product.aspx?id=36750

Actually Intel had a radical way to handle this - Larrabee. It was going to be 48 in order processors on a die with Larrabee new instructions. There was a Siggraph paper with very impressive scalability figures for a bunch of games running DirectX in software - they captured the DirectX calls from a machine with a conventional CPU and GPU and injected them into a Larrabee simulator.

This was going to be a very interesting machine - you'd have a machine with good but not great gaming performance and killer server performance - servers are naturally "embarrassingly parallel" because you can have one thread per client. A sort of x86 take on Sun's Niagra.

Of course there are problems with this sort of approach. Most current games are not very well threaded - they have a small number of threads that will run poorly on an in order CPU. So if the only chip you had was a Larrabee and it was both a CPU and a GPU the GPU part would be well balanced across multiple cores. The CPU part would likely not. You have to wonder about memory bandwidth too.

Larrabee was switched to be a GPU only and then canned.

Of course as a pure GPU it is a bit of a poor design. Real GPUs don't drag in x86 compatibility - they can implement whatever instruction set is best and nothing else. The instruction set is not publicly exposed and can change from generation to generation. You can cram a lot more than 48 cores onto a GPU and the peak performance is higher. Power consumption is lower too.

Still a modern gaming GPU is huge - there's no way you're going to cram it and a modern GPU onto a die and get something affordable. Then again CPUGPU chips are probably not aimed at gamers - there's an argument for having a CPU and a stripped down integrated GPU on one chip for netbooks like the latest Atoms do.

You could cram in a chipset too to reduce the price on netbooks.

Sigh, I know *I'm* the one actually feeding the troll here, but: http://www.intel.com/consumer/products/technology/graphics.htm

The page for the GMA 950 even has this hilarious tidbit:
"With a powerful 400MHz core and DirectX* 9 3D hardware acceleration, Intel® GMA 950 graphics provides performance on par with mainstream graphics card solutions that would typically cost significantly more."
Whoever wrote that line must have been borrowing Steve's Reality Distortion Field.

Sigh, I know *I'm* the one actually feeding the troll here, but: http://www.intel.com/consumer/products/technology/graphics.htm

The page for the GMA 950 even has this hilarious tidbit:
"With a powerful 400MHz core and DirectX* 9 3D hardware acceleration, Intel® GMA 950 graphics provides performance on par with mainstream graphics card solutions that would typically cost significantly more."
Whoever wrote that line must have been borrowing Steve's Reality Distortion Field.

It was indeed a mere observation of conjuncture. That said, it has been an extraordinarily useful one in the form of a challenge to humankind. Without it we would not have progressed the way we have. Intel is using Moore's law as a road map, forcing other companies *coughAMDcough* to innovate just to keep up. And that is why we have the enormous speeds available today. So we have a prediction that shaped the future. Why bother? Because our dreams shape our world.

You might be able to simulate it, but the real advance is that Intel developed high-k (very low leakage) transistors. A transistor that's "OFF" still has some amount of leakage current, thus power consumption and heat emission. In "older" transistors (and thus older processors), that leakage current was significant, at least when multiplied over millions of transistors. So even if you deactivate cores on the fly, you won't see nearly the power savings because your deactivated cores are still consuming a significant portion of your power (or thermal) budget. Likewise, if you were to overclock on the fly, you still wouldn't be able to achieve nearly the performance gains (not to mention that overclocking in the CPU by changing the clock multiplier is usually preferable to merely increasing the clock speed, which is the typical method these days).

Some more info is available here:
http://www.intel.com/technology/silicon/high-k.htm
http://download.intel.com/design/processor/applnots/320354.pdf?iid=tech_tb+paper (White Paper, PDF).

You might be able to simulate it, but the real advance is that Intel developed high-k (very low leakage) transistors. A transistor that's "OFF" still has some amount of leakage current, thus power consumption and heat emission. In "older" transistors (and thus older processors), that leakage current was significant, at least when multiplied over millions of transistors. So even if you deactivate cores on the fly, you won't see nearly the power savings because your deactivated cores are still consuming a significant portion of your power (or thermal) budget. Likewise, if you were to overclock on the fly, you still wouldn't be able to achieve nearly the performance gains (not to mention that overclocking in the CPU by changing the clock multiplier is usually preferable to merely increasing the clock speed, which is the typical method these days).

Some more info is available here:
http://www.intel.com/technology/silicon/high-k.htm
http://download.intel.com/design/processor/applnots/320354.pdf?iid=tech_tb+paper (White Paper, PDF).

" So obviously this whole concept of parallel computing and multi-core processors has just whizzed right past you, huh?
that has absolutely NOTHING to do with Moore's law.

"eah ok maybe you can't fit more transistors on a 5mm x 5mm chip"
That IS Moore's law. You stated right there why the poster is correct.

gah.

" but you can fit a LOT of chips in a mini tower case...

AND? Are you saying making bigger cases means Means moore's law isn't going away?

At least make an attempt to upstand it by reading this:

ftp://download.intel.com/research/silicon/moorespaper.pdf

What he basically did was putting haar-like features on the face. Haar-like feature are derived from Haar wavelet.

Sadly, nowadays this is a myth. Current MLC and SLC SSD's have (on average) 10,000 and 100,000 writes (respectively) before any bitwear will occur. While this number is small, remember that all modern mainstream SSD's have wear leveling algorithms built into the controller. Intel rates their drives' minimum useful life at 5 years [pdf link - page 10], with an estimated life of 20 years. Note that this number is based on 20GB of writes per day, every day. SSD's nowadays will have no problems with acting as a cache for the system.

Performance is not what moores law is about.
http://en.wikipedia.org/wiki/Moore's_law
ftp://download.intel.com/museum/Moores_Law/Articles-Press_Releases/Gordon_Moore_1965_Article.pdf

Today they can today make 15nm cpus. They however are not cost effective. Moores law is about getting rid of packages which cost money. Transistor doubling is how they are doing it in a smaller surface area. System on a chip is the logical conclusion of Moores law. Bringing in the north and south bridges into the CPU. The GPU is next.

15 years ago we had a dedicated ISA card for IDE. These days it is a few pins out of the CPU for SATA.

But like you said too bad their GPU is not exactly 'decent'. It is ok for light tasks. But ask it to do anything more and it fails quite badly. It is about where nVidia and ATI were in 2004-2005 (even that pushes it a bit).

More cache doesn't get you much after a certain point it doesn't change your hit ratio. So it is either the whole memory subsystem or the GPU next. The GPU is the smaller of the two. So it is the logical next candidate to be integrated. Decode accelerators will probably be the next big thing, then memory with a removing of L2/L3 cache.

There will be a day you buy your CPU and it pretty much is 'almost' a computer. You will plug it into a small breakout board. You will decide do you want 'only 16 gig or 32 gig of memory?'

The other possibility is hundreds of CPUs on one chip. But for desktop use anything past about 3 is 'gravy' right now. I have a 4way right now with 4 hyper threads. I really need to go out of my way to tax it.

It will be interesting.

You clearly haven't bothered to look at any die photos. Check out the Atom. Less than half of the main area is L2. You can see the smaller L1 blocks and other caches, but the vast majority of the rest is logic. That's at least 40-50% logic, more if you count the FSB IO areas. larger CPUs aren't much different.

Then you're also missing the fact that the caches themselves are more complicated (and larger per size) on x86, due to the aforementioned extra snooping required.

To be fair, I don't know of a better way to test, and I'd love to see a discussion of better utilities. If I tried this I'd probably do mprime and keep an eye out for MCE's in the system logs, but don't delude yourself into thinking that core is error free because you ran prime95.

There are quite a few tools, mainly found in the overclocking communities. OCCT, Linx and Intel Overburn just to name some.

This is actually what he *needs*:

http://www.intel.com/cd/business/enterprise/emea/eng/189154.htm
http://www.amd.com/gb-uk/Processors/ProductInformation/0,,30_118_14397,00.html

No one at the company had the forethought to do something like this. The same could also be accomplished merely by purchasing 20 identical Dells or HPs. Every serious vendor offers business platforms for a 2-3 year period, to solve or prevent this exact problem. The problem was purchasing dissimilar hardware to begin with. Now you're paying the price. *Next time* do everything you can to push $employer to buy identical machines. Explain the reasons for doing so. Use the data at the Intel and AMD sites above to make your case.

Actually, the i5's in question also support hyperthreading. The only difference is a tiny speed and cache boost in the i7. http://ark.intel.com/Compare.aspx?ids=47341,43560,43544

Also, they should have named the mobile ones m5 and m7 (or i5m and i7m if BMW has a problem with them using M5).

and SuSe lists none of their repositories in their installer

Software update ... starts the package manager.

Click on "Repositories" menu (duh!)

Click on "add" button.

Click on the "community repositories" radio button, and/or choose "scan with slp"

Right now my main desktop has more than 360,000 files reported by apt as being installed.

... and I've got between 2 and 3x that. Pick the defaults and it works. I get conflicts because I override the defaults, but I also know how to fix that with a quick click on the second choice instead of the first.

So what cpu and motherboard are you running on? It can't be Intel or ARM or AMD

And which of these companies claim to be devoted to open source? Not a one

Both AMD and Intel are into open source. Do your research. They're both members of the Eclipse Foundation, AMD has contributed hardware to developers, as well as the servers running groklaw, etc., and Intel is active with all sorts of stuff, and ARM sponsored the code to get debian to run on ARM.

So again, since all these companies have put money where their mouth is with respect to open source - Intel, for example, contributed 4.1% of the kernel code

Intel's position as the fourth-biggest corporate source of contributions "happened by virtue of the work done by four of the top-20 developers. ... Intel has a lot of people working on the kernel, many of whom spend little time in the limelight," Corbet wrote in the same posting.

So, since Chipzilla also works with the Beast from Redmond, my question stands - what cpu and motherboard do you use? Because that's no reason not to use OpenSUSE. Novell has done a LOT for the community.

The aspire is not a netbook. It's a core 2, and it's battery life need not try to compare. (which, btw, it's a core SOLO, not a dual core...) ...

Depends on the model. I got a brand new AS1410-2920 a couple of months ago with a SU2300 processor for the price I mentioned ($450 USD).

Like I said in another post at these prices iPad isn't competing with netbooks. It is competing with laptops. There are a lot of cheap Intel CULV laptops, or AMD NeoX2 laptops. Netbooks using Atom are even cheaper. You can get an Acer Aspire One for $300 USD.

As for H.264 decoding, I could do it with a single core 1.4 GHz Athlon XP using CoreAVC, so I doubt I cannot do it in a dual core SU2300 1.2 GHz. However since the Intel GMA 4500 MHD chipset does it in hardware the point is moot.

Adequate vitamin D (the sunlight vitamin) helps prevent pneumonia:
    http://www.google.com/search?hl=en&q=pneumonia+vitamin+d

At the end of the winter, Ed Roberts' vitamin D supplies would have been depleted.

The right amount of vitamin D also helps prevent influenza, cancer, heart disease, and a variety of other illnesses:
    http://www.vitamindcouncil.org/treatment.shtml

All computers should come with a warning label about this, IMHO. :-)
    http://blogs.intel.com/csr/2010/02/with_all_of_the_debate.php

I'd suggest it is possible that vitamin D deficiency is the leading cause of death of computer users including most slashdotters. See also:
    "A Decade Of Vitamin D Supplementation Would Save $4.4 Trillion Over A Decade; Would Save $1346 Per Person Per Annum"
    http://www.lewrockwell.com/sardi/sardi111.html

White macbook MSRP? no. But MacMall and bestbuy will both sell one for that price, as will apple if you ask nicely. Education price is $899, but you do not need to verify to anyone you actually qualify, just ask them to beat that price and throw in a few extras like mac Mall does.

For the Atom ability to decode H.1264, maybe you should read this:
http://software.intel.com/en-us/articles/quick-reference-guide-to-intel-integrated-graphics/

H.264 is only supported on GMA-HD (and then only partially), and requires an iseries or higher CPU to perform 720p.

There are numerous maches advertised as netbooks over $500, but i agree, only a moron buys one. That said, none UNDER $500 can match the iPads specs, period. Closest competitor is a $629 Acer if you include HTML5 video full screen as an expectation.

I have an eee. Rarely use it. It's good for a basic document edit, or research, but as a companion device for traveling, it's lack of any gaming or video performance makes it less than useless to me. Given the boot time, I'd rather pull out a real 13" notebook any time. I keep it around for other people only, and bring it to the beach so i don't have to worry about damaging an expensive machine.

while D3D kept some stuff on the CPU.

That is a good thing if the GPU is your bottleneck.

Hate to be a prick, but there's no "8 core" i7... I suspect you meant 8mb cache, or perhaps 8 processing threads?...

reference: http://www.intel.com/products/processor/corei7/specifications.htm

No; but the micro-ITX doesn't need to connect to a computer to do its computing...

Unless Microsoft changes the licensing on home versions of Windows to allow multiple concurrent terminal sessions(currently, even Pro versions are only 1 session at a time, if somebody logs in via RDP, the local desktop locks, and if somebody logs in locally, the RDP session dies. You have to step up to Server to get concurrent access, and deal with CALs), and allows joe and jane sixpack to use devices like this on the computers they are actually going to buy(which they do in a limited sense already. Media center extenders don't count as terminal sessions, though I don't know how many are supported), the economics of thin-clients in the home are going to be pretty dubious(anybody who is currently rocking a vintage tektronix xterm just fine thanks very much needn't flame me, I know it works for you).

Most home broadband connections are a bit on the slow side for anything more than light work(word processing via ICA or NX? no problem. Video/gaming? Bigger issue), and paying some vendor a monthly fee for your "cloud desktop" isn't all that thrilling a prospect, compared to just buying a cheap PC and running Google Docs if you want them, and desktop apps if you don't. If you already need a full computer to support your thin client, the temptation to just skip the thin client and move the computer is fairly strong(especially when the full computer is increasingly likely to be a laptop).

Only in the most fanatically noise sensitive environments is it worthwhile to have a thin client setup just for silence purposes. Decent PCs these days are actually pretty quiet, home theatre needs can generally be satisfied by fully solid state networked media streamer boxes, and thin clients good enough for things like pro audio work are Pricey(the software based ones are the cheap seats, $100-$200 for the thin client box, runs over normal ethernet, speaks RDP, or X11, or NX, or ICA. If you want fully software transparent ones, that won't make a hash of 3d graphics or multiple high speed peripherals on the terminal end, or support multiple high resolution monitors, you can end up spending real money for something whose thin client costs more than a normal PC, and that demands a fiber run between the thin client end and the rackmount or blade end).

The viability of something resembling thin-client computing in the home isn't impossible; but, for wide acceptance, it would pretty much require Microsoft(whose incentive involves selling Windows licences) to suddenly dedicate itself to making thin client access work, without charging CAL fees that strangle the whole enterprise. It would be pretty handy to, for instance, have a single program on your computer, or one of your computers, show up as an icon on your TV, activate it, and away you go. Issue is, given the complexity of the software/licensing situation, if that single program is something generic like a web browser or messaging client, it will probably be cheaper, and certainly be simpler, to just embed a $25-$50 ARM SBC running linux or WinCE to do the job. If it is something particular, for which an embedded board won't do, there remains the fact that you can get a full, albeit atom powered, PC for $200 that is close to silent, and easy to connect to your TV(plus, since none of the thin client access protocols are involved, things like 3d graphics and video work seamlessly). That establishes a very challenging price point under which a hypothetical thin client arrangement must slip.

Then, of course, there is intel's new "Wireless display" stuff, coming soon to virtually all but the cheapest and most awful intel based notebooks. Unless they've been hiding some showstopper of a bug during the tech demos and early reviews, that setup will mean "game over" for any small time attempts at bringing thin clients home.

get a notebook (not net , note) with the new intel core i7 640M (U or L - http://ark.intel.com/Compare.aspx?ids=43563,47700,). The U model at 18W TDP gives the atom chips a run for their money (well, not really in same category, but much better performance/wattage than anything before mobile nehalem came along). Lenovo has the x201s, but I'd take a look at the fujitsu P770 series - http://store.shopfujitsu.com/fpc/Ecommerce/buildseriesbean.do?series=P770. Not really in the netbooks price range, and not featuring ION (which is a true shame) but more than adequate from power consumption versus performance standpoint. If you want a portable gaming platform, either alienware or one of the eurocom models ( http://www.eurocom.com/ ) might suit you better.

You'll be glad to hear that in the SFF (2.5") form factor consumer SSDs have now reached density parity (512GB) with enterprise SAS spinning discs, and passed FC SFF SSDs in price parity ($4k/TB). That's still spendy for consumer gear, but for enterprise drives it's cheaper and the 50x IOPs make it better. Intel and Micron's 25nm technology has not yet hit the market but when it does density will double and performance will quadruple. These individual chips deliver 200MB/s, and individual devices will be able to deliver 2GB/s and/or insane IOPs. Large (128MB or more) RAM cache on individual drives is now in the consumer market, and enterprise drives include supercapacitors to ensure writes for data integrity. The advent of these technologies will push down prices of currently available gear at the (as you noted) accellerated rate. The conversion will now begin. These flash drives have a better MTBF, and a better failure mode (fail on write rather than fail on read). TRIM support is looking good in W7 and Linux, but array controller firmwares need to be updated to support it, and they're working on that at a furious pace.

TRIM drivers for XP and Server 2003 would be the courteous thing for Microsoft to do for its committed Software Assurance customers who just can't migrate yet because they bought into that whole IE6/iis/.NET V1 thing - but I don't see Microsoft doing that because they want to shuffle those customers on to W7 and Server 2008 whether they're ready or not. What are they going to do if Microsoft doesn't enable TRIM support on legacy platforms - migrate to Linux? Not likely: they're Microsoft shops and they'll take what Redmond gives them.

We need a new interconnect to pull this all together because bandwidth and IOPs are getting out of hand. lightpeak looks like it if it has the right features. We've passed the performance abilities of copper, so something optical will be the order of the day. If LightPeak doesn't cut it, we can consider PCIe-F (PCI Express over Fibre) or some new thing.

The times, they are a-changing. Spinning disk is the new tape. Tape? Hopefully the kids coming up today will be challenged by the question "what was data tape?" With luck they will be confused by the ambiguity between punched paper and magnetic tape and not pursue that shameful episode further.

Sometimes I find [RMS and the FSF] as annoying as the beggars that shake the cup of coins under your nose to make you give them something. No fucking way.

Really? Stallman asked you for money? Funny, because I never heard about him asking for anything in return for GCC and GDB. Intel, on the other hand...

Intel® Compiler Suite Professional Edition for Linux: $1,349

Whoa!

As FlyingBishop said here before me, quid pro quo. A lot of people owe RMS and the FSF a lot.

Intel (the company) is interested in real robotics. Yes, it may ultimately be a way to sell more processors, but they do serious research into robotics. See this page for example, or this page. They have also been heavily involved in image processing (for robotics and other things) for many years, for example with OpenCV.

Who knows what the marketing department is really interested in besides making Intel look cool.

The project is actually very impressive. The some of the technical details are here including how the learning algorithm was implemented. Reinforcement learning, I knew it!

You might expect a Power6 to be at a bandwidth disadvantage when compared to the x5560 but when you look at the memory bandwidth specs they're surprisingly close. Once you move up to Power7 you begin to see a big difference.

Xeon x5560 memory bandwidth/chip: 32 GB/s [1]
Power6 memory bandwidth/chip: 32 GB/s (??? This may be wrong. I see references to 50GB/s but I believe that's peak)
Power7 memory bandwidth/chip: 100 GB/s [3]

It's not as far fetched as you think.
Intel has been developing a "manageability engine" that lives in the ICH and MCH. Now while this is on the up and up, the tech is obviously there.
http://www.intel.com/technology/platform-technology/intel-amt/
-nB