Slashdot Mirror

As much as I love AMD, I would like to know what the submitter was thinking calling it a battleground. It's only a fair fight for AMD so long as Intel's not interested - AMD (and their manufacturing partner née subsidiary) can't match Intel's manufacturing abilities. AMD doesn't have an Ultra Low Voltage chip; Intel has a 1.6GHz Core 2 Duo that runs at 10W, meanwhile it's murky at best for AMD's competing chip line, the Neo. The only specs given out to the press for the new fual core version, which is 18W for the 1.6GHz version, with the chip still being built on the 65nm process which hobbles AMD from the start. Bear in mind that the Neo is Athlon 64 based, which means that it's not clock-for-clock competitive with the Core 2 Duo (you'd need a Phenom II-based core for that). In other words, the Intel chip eats less power and gets more performance at the same time.

Mod parent troll. Do you work for intel or something? http://www.amd.com/us-en/ConnectivitySolutions/ProductInformation/0,,50_2330_9863_9864,00.html
There's a link for the AMD geode processor that pulls 1.1W.

So if Intel's serious about this, it's only a battle so long as they don't decide to crush AMD with products and pricing. Intel is light years ahead of AMD in the mobile space due to their process technology advantage. Even TFA points out that they expect 8 hours out of the Intel CPUs, but only 5 hours out of the AMD CPUs. It's entirely lop-sided in Intel's favor.

WTF? This argument only makes sense in the fantasy land where companies don't care about making money. Intel is doing everything they can in every space they compete in, including this one.

Now TFA does mention AMD will have Congo later this year, but even if that's 45nm (AMD has not commented on that matter), it's unlikely that they'd be able to meet Intel's power envelope. When you look at the desktop chips this stuff is derived from, the Phenom II takes more transistors and as a result power than the Core 2 Duo, and that's only to reach a clock-for-clock parity. Congo wouldn't change this.

Citation needed?

As much as I love AMD, I would like to know what the submitter was thinking calling it a battleground. It's only a fair fight for AMD so long as Intel's not interested - AMD (and their manufacturing partner née subsidiary) can't match Intel's manufacturing abilities. AMD doesn't have an Ultra Low Voltage chip; Intel has a 1.6GHz Core 2 Duo that runs at 10W, meanwhile it's murky at best for AMD's competing chip line, the Neo. The only specs given out to the press for the new fual core version, which is 18W for the 1.6GHz version, with the chip still being built on the 65nm process which hobbles AMD from the start. Bear in mind that the Neo is Athlon 64 based, which means that it's not clock-for-clock competitive with the Core 2 Duo (you'd need a Phenom II-based core for that). In other words, the Intel chip eats less power and gets more performance at the same time.

So if Intel's serious about this, it's only a battle so long as they don't decide to crush AMD with products and pricing. Intel is light years ahead of AMD in the mobile space due to their process technology advantage. Even TFA points out that they expect 8 hours out of the Intel CPUs, but only 5 hours out of the AMD CPUs. It's entirely lop-sided in Intel's favor.

Now TFA does mention AMD will have Congo later this year, but even if that's 45nm (AMD has not commented on that matter), it's unlikely that they'd be able to meet Intel's power envelope. When you look at the desktop chips this stuff is derived from, the Phenom II takes more transistors and as a result power than the Core 2 Duo, and that's only to reach a clock-for-clock parity. Congo wouldn't change this.

Sure, the atom is a tad underpowered, but they should be able to run half-life.

I just bought an Intel Atom 330 for a lightweight file server, and I have to say I'm thoroughly impressed. 64-bit, dual-core, virtualization extensions, and low-power to boot for around $80 which includes the motherboard. Simply unbeatable.

I'm not sure I agree. Fry's regularly advertises low power $60-80 AMD CPU and motherboard combinations that have much better performance than the in-order Intel Atom processor. I have one of their cheap Sempron systems that idles at 35 watts.

By the way, the Atom 330 doesn't have virtualization extensions. (athlons do)

From the description, it sounds like this is found in TBB and research done for DARPA in PGAS. GCN had a blog post, "Does parallel processing require new languages?", about this the other day.

I don't think so - embedded OS' just don't sell in the kind of numbers you'd need.

This site might provide some insight as to why.

Pare away the heat sink and all that junk, add super small RAM and flash storage and ... hand held computers (like the article notes from Toshiba). Microsoft better not be resting on its laurels and should either be beefing up Windows Mobile or porting Windows 7 to ARM ... or they're going to miss out big time again.

People keep saying this sort of thing, but I really don't see it being viable. A Snapdragon is probably going to end up being at best the same speed as an Atom for native code. Windows 7 is probably quite portable and from the tests I've done on the Beta on Atoms might run quite okish on an 1Ghz Snapdragon if it were ported. Even there we're talking about a 1Ghz in order core with a memory controller designed for cellphone SDRAM. High performance desktop memory is really different to the stuff used in cellphones - the buses are narrower and slower. Here's are the details for an Atom

http://processorfinder.intel.com/details.aspx?sSpec=SLB73

Note the bus speed, 533Mhz and the cache size, 512KB. By desktop standards the Atom is slow. Most Arm systems run memory much slower than this and have less cache. Look at the Snapdragon based Toshiba L01

http://pdadb.net/index.php?m=specs&id=1855&view=1&c=toshiba_l01

It uses "mobile DDR SDRAM". I don't know the clock speed, but look at this

http://www.eetasia.com/ART_8800457078_499486_NP_197bb814.HTM
Hynix claims 'fastest' 512Mbit mobile DDR SDRAM with a 185Mhz clock speed.

See the thing is that the sort of memory you get in a cellphone is a lot slower than the stuff you get in a desktop because the power budget is so much less. If you want to run desktop applications or emulate an x86 that will really bite you.

Once you get past the OS it gets worse. Office is probably less portable than Windows and Office 200x runs terribly on an Atom and would be worse on Snapdragon given the lower performance memory. Most Windows applications will not be ported and will run even worse in emulation - a Snapdragon emulating x86 will be unusably slow.

Of course maybe ARM will do a Jazelle style extension where common x86 instructions are turned into ARM ones via an extra pipeline stage. I think that would mean a Snapdragon chip would run x86 code say 90% as fast as an Atom at the same clockspeed. Still a 1Ghz Atom is not a quick chip.

The difference between a good ratio and a bad ratio between vm's and hardware is often due to the need in some combinations to run every vm's IO state through the BIOS in order to complete an IO. That's a lot of interrupt state passing and subsequent process rescheduling. You get a multiplier when multiple vm's are all competing for the same trap completion and queues grow as a result. I know that Intel at least has a chip set that optimises this (they call the feature VT) . Ring 0 instruction completion has a huge multiplying effect on virtualisation efficiency. Right chip set = good, wrong chip set = sorta.

Ahem!:
 

Intel Atom Processor 330
...

Supported Features:

• Dual Core
• Hyper-Threading Technology 1
• Intel EM64T 2
• Execute Disable Bit 3

The push to 64-bit would probably go faster if Intel's Atom processor were to up and die. Almost the entire product line of Atoms are 32-bit only.

Point well taken. Yeah do like Ireland did! Whoring themselves out to multinationals has worked out *perfectly* for them.

Multinational corporations are quick to use the competitive advantages to use as a bargaining chip. As in my earlier example, Intel.

http://www.intel.com/community/ireland/index.htm Intel does have a manufacturing plant there.

Along those lines, many people missed the Redmond giant is building a new headquarters.. If taxes get bad, they are not locked into the US tax the evil corporations.
http://www.highbeam.com/doc/1G1-132007757.html They also see the lets get the evil corporations and their overseas tax havens. I hope the current administration gets a clue before world economic reality hits them in the forehead after they close the US operations.

According to intel it's about 0.04.

Put Chandler on the map? A stop in Chandler means Intel employees can commute from Mexico - or Flagstaff.

It's true that TRIM is very much in it's infancy, but the clouds aren't as dark in SSD's future as they once were (even as little as a month ago). Many, many, many companies see SSD's as the future of storage and I'm inclined to agree with them. With that kind of muscle propelling development and increased consumer interest fuelling funding, the landscape is and will continue to change very rapidly.

My own take on things, FWIW, is that tapes will go the way of the floppy and spinning disks will become near-term storage for most enterprises. Look for SSD's to become mainstream SAN devices, especially as hardware manufacturers remove the driver/OS hurdle and present SAN devices as "just another disk". And as SSD's mature, I expect the performance gap between DRAM and SSD's to shrink quite a lot.

And there happens to be one set of T5500's that have vmx, so you are probably one of the lucky ones. http://software.intel.com/en-us/forums/watercooler-catchall/topic/54562/

I went to Intel's Core 2 Duo list of chips and counted. 18/60 chips don't have the VT extension. That's 30%. I have no idea on the relative popularity of these chips but if 30% of the chips out there can't run the XP emulation mode, I'd say that's pretty significant, wouldn't you? It's not like these chips are ancient or anything either, I'm guessing their Core 2 Duo is their most popular chip for desktops right now and maybe even laptops.

Regarding the limit of C: using the third dimension can help minimize the distance between components. A different layout of transistor communications paths (such as a linear pipeline) can help with bandwidth of processing, though not latency, by signals not needing to traverse the entire chip in a single clock cycle.

Regarding density: it's not the density that doubles, but transistor count (number of components per integrated circuit.)

Actually.. what's concerning to me is that David Stewart hasn't put out a new OpenSolaris on Intel video in a couple of months.

Moore's law is fundamentally flawed in that it predicts a never ending exponential (linear in the log domain) progression. It is bound to slow down and eventually stop, yet it fails entirely to take that into account.

That said, Intel still takes the idea deadly seriously when it comes to their marketing and future plans.

Think of it a self prophetic goal:

http://www.intel.com/technology/mooreslaw/

See http://blogs.intel.com/research/2008/10/rattner_the_promise_of_wireles.php . Tesla was, obviouly, much omre ambitious.

Hey, sorry mate but I forgot to get you some info you requested ages ago regarding Vectorization.

Check out:
http://www.intel.com/intelpress/sum_vmmx.htm

It's a little dated but the concepts and programming techniques are still relevant. It does a pretty thorough job.

Apologies for hijacking.

Say we have the following PHYSICAL machines in the server rack:

[VM-Host 1]

You rent space in a virtual machine on VM-HOST1.
This virtual machine is named [srealm.VM].

Who on earth is going to base their rack-mounted server on a micro-ATX motherboard designed for low-end workstations?

Except that the exploit depends on very specific architecture. http://www.intel.com/products/desktop/motherboards/DQ35JO/DQ35JO-overview.htm

I may be missing something here, but the hardware doesn't appear to be the first choice of people building server farms.

Since Oracle likes primarily using "their own thing", my guess is they'll move to Solaris, and their Linux distro will take a bow, since it's based off of someone elses work, that they've not yet acquired.

Solaris used to be the primary development environment and when Oracle switched to Linux the developers seemed to miss DTrace.

In the past, Solaris was the best platform to deploy Oracle on. That may still be true today, even with all the support Oracle has put into Linux. Oracle has kept up with Solaris/Sparc but lagged releases for Solaris/x86. Hopefully that changes now.

As much as I like Linux, I still prefer Solaris, especially since Solaris 10.

Sun's hardware works best (faster doesn't mean better) with Solaris, so I can't see Oracle dropping Solaris. I agree that it wouldn't be surprising to see Oracle moving more towards Solaris.

Intel X25-E, 2.6 watts, 3300 Write IOPS, 35000 read IOPS*. So only one or two orders of magnitude more efficient...

And though no prices are given in the article for the FAWN, at $800 for the X25-E it's probably less expensive too. Particularly if you include setup and administration costs.

Not a bad idea in general, and not a bad idea in specific for 5 years ago, but pathetically outclassed in every area by a high end modern SSD.

* http://download.intel.com/design/flash/nand/extreme/extreme-sata-ssd-datasheet.pdf

> I've seen programmers who grew up with Pascal carefully turning multiplies into sequences of adds and shifts. Great, except that something like the Athlon can do two multiplies in parallel, but only one shift at a time (because most code contains a lot more multiplies than shifts), stalling the pipeline.

That strength reduction of multiplication to bit shifting might not be valuable anymore for modern CPUs surprised me, so I looked up the instruction latency and throughput for the Core 2. To see for yourself, download http://download.intel.com/design/processor/manuals/248966.pdf and look at Appendix C.

The timings show that integer multiply, IMUL, is still quite a bit slower from the bit shift instructions (SHL/SHR). So, turning multiplies into adds and shifts is still a Good Thing.

However, these days, a compiler can do this type of thing for you much of the time, so doing the transformation by hand is usually not necessary. Also, since optimizing the source code in this way does take time and decreases source readability, the only cases where people should really still be doing manual strength reduction are in performance-critical code where the compiler, for whatever reason, is missing the optimization opportunity.

---linuxrocks123

If by recently you mean in the past 6 months, I am terribly puzzled by this, as all currently shipping Precisions use the Intel 82567LM chipset, which has a direct download from Intel's website here.

I'll give you two good reasons we're not ever going to see this.

There's a bigger reason Lake Erie water won't be pumped to the Southwest, there's a compact or agreement between the US and Canada.

NOTHING in the southwest is an important center for commerce. Los Angeles is a major port city, a major tourism destination, and a home for many large businesses. Phoenix?

Intel is in Phoenix, including fabs. From 2005, "Intel To Build New 300 mm Wafer Factory In Arizona".

Can dry-up and wither in the desert heat, for all the world cares.

Yea, for all I care. I oppose any more building and growth there. That applies to LA and Imperial Valley too though. If not for the water pumped from the Colorado River the farms in Imperial Valley would not even exist.

Falcon

I'll give you two good reasons we're not ever going to see this.

There's a bigger reason Lake Erie water won't be pumped to the Southwest, there's a compact or agreement between the US and Canada.

NOTHING in the southwest is an important center for commerce. Los Angeles is a major port city, a major tourism destination, and a home for many large businesses. Phoenix?

Intel is in Phoenix, including fabs. From 2005, "Intel To Build New 300 mm Wafer Factory In Arizona".

Can dry-up and wither in the desert heat, for all the world cares.

Yea, for all I care. I oppose any more building and growth there. That applies to LA and Imperial Valley too though. If not for the water pumped from the Colorado River the farms in Imperial Valley would not even exist.

Falcon

I agree. PostgreSQL is much closer to Oracle than MySQL is. Anyone that thinks MySQL is the best replacement for Oracle likely doesn't know much about Oracle.

It seems that sun has done a bit with PostgreSQL as well. Too bad they bought MySQL. They should have instead invested in making PostgreSQL better, at least developing better replication and clustering. That way, PostgreSQL would have been an even stronger alternative to Oracle.

Oracle used to have Solaris/SPARC as their main development platform, then they switched to Linux. That seems to have been a big blow to Sun. While Oracle still releases Oracle for Solaris/Sparc along with Linux, but the Solaris/x86 versions are always slow. I don't 11g has been released for Solaris/x86 yet.

If I was Jonathan Schwartz, I would have rather put the $1bln they spent on MySQL on PostgreSQL. I don't think it would have even really taken that much either. I'm still just baffled over spending $1bln on a company that I think made $50mln in it's best year!?!?!

Anyway... Oracle developers might not have been too happy about moving away from Solaris because they'd lose DTrace.

I thought I heard something about there being some bad blood between Ellison and Sun but I don't know what that was about.

I still think Cisco should be more interested.

Well, no not really, even the atom with 47 million transistors would not be able to fit into any reasonable amount of space if the transistors had not shrunk down to the current sub micron sizes (40 nm now, 30 nm next year?) therefore would have to run at an incredibly slow clock rate, literally due to the speed of light limitations.

Light travels ballpark about 1 ns per foot,
so if your CPU is a hundred feet across, guess what your clock rate has to be less than?

The biggest number I found here:

http://www.intel.com/pressroom/kits/quickreffam.htm

was 820 million transistors.

You really want those to be small, so size is a BIG part of the cost of an IC.

The basic unit cost of Silicon is the wafer. The more parts you can jam on a wafer, the cheaper each individual part is.

Interestingly enough, these days, the transistors are practically free, and software development costs are becoming the bottleneck.

From the C++ prototype guide, which is just the ISA made into a terribly complex C++ wrapper, they support these transcendental functions in the ISA:
EXP2_PS - Exponential Base-2 of Float32 Vector
LOG2_PS - Logarithm Base-2 of Float32 Vector
RECIP_PS - Reciprocal of a Float32 Vector
RSQRT_PS - Reciprocal of the Square Root of a Float32 Vector

They also provide library functions that implement everything else you'd want (sin, cos, etc) in software, I assume using Newton-Raphson iteration.

Wrong on sooo many counts.

The Dell doesn't have Bluetooth, but it can. The Wireless is an INTEGRATED card (does not consume a slot at all).

The Dell has that you claim it doesn't:

• 4 USB 2.0 ports
• Gigabit Ethernet
• Optical Audio
• Bluetooth (option)

The Dell has and the Mac doesn't have:

• Blu-Ray drive (option)
• 8-in-1 card reader
• Internal EVDO/HSDPA modem
• Logitech GamePanel LCD screen

One of the reasons I like PostgreSQL is it is more like Oracle.

If you're doing a migration from Oracle, especially one that has a lot of pl/sql functions. Here's some good advice for converting pl/sql to pl/pgsql.

Also, going from PostgreSQL to Oracle seems easier as well. With PostgreSQL you can use more Oracle like features so if you need to move to Oracle, you can take advantage of some of it's advanced features instead of migrating simple tables and sql statements.

Sun was actively involved with PostgreSQL before they bought MySQL. I was really dissapointed with their decision, especially at the price.

My guess is they weren't really buying MySQL for the technology, they were buying it for the community. Overnight, a ton of MySQL users and developers were part of Sun's open source community. Building communities takes time and Sun was having a hard time doing it with some of their projects.

All it seems they did though is fund MySQL forks. Kinda messed up for the MySQL developers to do that but we don't know all the details.

$1 billion dollars could have funded a lot of improvements to PostgreSQL better clustering, replication, visual tools, and more. A better PostgreSQL could hurt Oracle more than buying MySQL. After Ellison announced he'd be moving his developers from Solaris workstations to Linux workstations, it could have been a nice comeback.

It also seems that the switch form solaris to linux might not have been developer driven. Even MS knows you have to keep your developers, developers, developers happy.

I don't have to. Intel already has:

http://download.intel.com/support/motherboards/server/sb/enterprise_class_versus_desktop_class_hard_drives_.pdf

In fact, here is a link to the S5000 boards describing using the EFI shell to update the LSI 1068E sas firmware:

http://downloadcenter.intel.com/Detail_Desc.aspx?ProductID=2305&DwnldID=15521&lang=eng

It's not just ray tracing

This Intel paper on Larrabee (pdf) shows pretty good scaling (90% of linear with 48 processors, linear with 32) with shipping DirectX 9 games like Gears of War, F.E.A.R., and Half Life 2 Episode 2.

What's impressive is that the games are unmodified - the only change is in the graphics driver which tiles the rendering and allocates one core per tile.

You could easily imagine a CPU/GPU hybrid that would do this. And you could use the processing power in a server too - imagine a thread pool servicing requests, it's not just for games. Larrabee is x86, but not apparently PC compatible, but you could probably get x86 Windows to run on it with the right HAL.

It's a similar idea (up to six blades sharing up to 14 SAS drives and up to 2 switches with a web interface to control the whole thing) and it's been available for over a year.

I, for one, welcome our new strong ARM'd overlords.

You'll be happy to know that Intel can provide you with those chips too.

I wouldn't be so sure about that.

You might be interested in this whitepaper from Intel. What they find is that the Windows CIFS client write pattern creates serious fragmentation problems for ext3. The problems are mitigated (though probably not completely solved) in XFS precisely by what you mention - delayed allocation.

Intel and Lenovo have partnered to launch the New Frontiers Sweepstakes as a promotion for the upcoming STAR TREK movie directed by JJ Abrams: http://boldlygo.intel.com/newfrontiers From now through May 8 you can enter the sweepstakes on the web. Prizes include: Zero Gravity Flight â" Win a trip for four to experience zero gravity the only way possible without actually going into space. Star Trek Premiere in Hollywood â" Win a trip for four to the starâstudded red carpet event. http://boldlygo.intel.com/newfrontiers

Intel and Lenovo have partnered to launch the New Frontiers Sweepstakes as a promotion for the upcoming STAR TREK movie directed by JJ Abrams: http://boldlygo.intel.com/newfrontiers From now through May 8 you can enter the sweepstakes on the web. Prizes include: Zero Gravity Flight â" Win a trip for four to experience zero gravity the only way possible without actually going into space. Star Trek Premiere in Hollywood â" Win a trip for four to the starâstudded red carpet event. http://boldlygo.intel.com/newfrontiers

Before we start, let me make a prediction: You never asked about the MTBF of your hard disk, right...?

http://www.intel.com/design/flash/NAND/mainstream/

a) When Intel says "new level of ... reliability", maybe it means they thought about this problem when they designed the drive.

b) When they say "NAND flash", maybe it means they're not using the cheapest MLC memory as mentioned in that scary wikipedia article.

c) When their datasheet says "Minimum useful life of five years, assuming 20Gb/day of writing", maybe they got those numbers from real engineers, with degrees.

d) When their datasheet also says, "Should the host system attempt to exceed 20 GB writes per day by a large margin for an extended period, the drive will enable the endurance management feature to adjust write performance, this feature enables the device to have, at a minimum, a five year useful life", maybe they were really really paranoid about saying 'five years' because they know people will start class-action lawsuits if it doesn't work out.

So, um, how this even got greenlighted in 2009 is beyond me. It's like 1999 called wanting its flash-myths thread back.

Well, according to Intel, the i7 is a Nehalem processor:
http://www.intel.com/technology/architecture-silicon/next-gen/index.htm?iid=tech_micro+nehalem

Dell just happens to be selling a Studio XPS 435 with:
Intel Core i7-920 processor(8MB L2 Cache, 2.66GHz)
Genuine Windows Vista® Home Premium Edition SP1, 64-Bit
Dell 24 inch S2409W Widescreen Flat Panel
Single Drive: 16X CD/DVD burner (DVD+/-RW) w/double layer write capability
6GB Tri-Channel DDR3 SDRAM at 1066MHz - 6 DIMMs
750GB 7200 RPM SATA Hard Drive
ATI Radeon HD 4670 512MB1
Integrated 7.1 Channel Audio

for... $1,549

I am not shouting and I hope nobody believes you. Guess a Mac Pro isn't the only way to get a Nehalem processor. Whats up with mac fanboys?

I find it hard to believe that the Linux kernel developers never heard of ICC. Or, to take another example, never used Codewarrior or XL C (IBM's PPC compiler, especially good for POWER5 and Cell) or DIAB (or Wind River Compiler or whatever they call it now). Or even Visual C++. Personally I've had the pleasure of using them all.. they all do things differently, but when you have a development team which is using more than one.. I once worked on a team where most of the developers had DIAB, but they didn't want to pay for licenses for EVERYONE, so it was just for the team leaders and release engineering guys, so we all got GCC instead. We had to be mindful not to break the release builds.. and the work ethic meant everything went pretty much fine all round.

All of them have at one time or still today produce much better code and have much better profiling than GCC and are used a lot in industry. If the commercial compiler doesn't do what you want or is too expensive, GCC is your fallback. Linux turns this on it's head because it "wants" to use as much free, GNU software, but I don't think the development process should be so inhibited as to ignore other compilers - especially considering they are generally always far better optimized for an architecture.

As a side note, it's well known that gcc 2.95.3 generates much better code on a lot of platforms, but some apps out there are refusing to compile with gcc 2.x (I'm looking at rtorrent here.. mainly because it's C++ and gcc 2.x C++ support sucks. This is another reason why commercial compilers are still popular :) and some only build with other versions of gcc, patches flying around to make sure it builds with the vast majority, significant amounts of development time is already "wasted" on compiler differences even on the SAME compiler, so putting ICC or XCC support in there shouldn't be too much of a chore, especially since they are broadly GCC compatible anyway.

Like the article said, most of the problem, and the reason they have the wrapper, is to nuke certain gcc-specific and arch-specific arguments to the compiler, and the internal code is mostly making sure Linux has those differences implemented. There is a decent white-paper on it here. The notes about ICC being stricter in syntax checking are enlightening. If you write some really slack code, ICC will balk. GCC will happily chug along generating whatever code it likes. It's probably better all round (and might even improve code quality generated by GCC, note the quote about GCC "occasionally" doing the "right" thing when certain keywords are missing) if Linux developers are mindful of these warnings, but as I've said somewhere in this thread, Linux developers need some serious convincing on moving away from GCC (I've even heard a few say "well, you should fix GCC instead", rather than take a patch to fix their code to work in ICC)

Stick to the fractal art.

Thanks for noticing. It has done pretty well for me lately.

Wake me when Microsoft gots somethin that runs on stuff somebody wants. Even I know Vista is the suckage.

No, don't. I really could pass a purple twinkie about what Microsoft thinks is good stuff even if they buy adds in my mags that say it's good enough. If you want to get on my stuff then wise up.

Thankfully, Intel is hearing me, yo. Otherwise I'd be waiting like until Jasmine textes me back, which is like for ev-er.

If normal digital projection is 4,096 x 2,160 pixels and this projects at 2,048 x 1,080, I would guess this is a typical anaglyph method using half the pixels for each eye. In other words, normal digital projection.

It could be that they are using two different projectors using polarized light, and that this system uses lower resolution to lower costs. But I would expect a polarized solution to simply use two standard projectors at 4K. Of course the reporter apparently doesn't care which method is being used. It could very well be something gimmicky a crappy like Intel's Intru3D used for the superbowl, or similar technologies which use two colored lenses but try very hard to say "but we're not anaglyph".

BTW my TV is not properly color-adjusted, so the 3D super bowl stuff actually looked like hell to me. I assume a cinema would bother to fix this for a 2-hour feature running continuously, whereas I don't give a crap for a 15 minute commercial.

http://scoop.intel.com/2009/02/did-you-catch-the-3d-super-bowl-commercials.php

There's another one that uses yellow and some other color, might be an offshoot of Intel but using a slightly different second color for the other eye.

Intel had a Fourth-Quarter Revenue of $10.7 Billion , so it isn't quite an insignificant amount, but if it were to completely disappear it wouldn't be a catastrophic problem.

Look at http://software.intel.com/sites/oss/