Slashdot Mirror

Intel built one of these in the research world, except they used flash instead of a drive. It was dubbed the "personal server."

http://www.intel.com/research/exploratory/personal _server.htm

stu =)

But does it run linux?

Yes. (Just search for the "L" word.)

The link in the main article is paraphrased from http://www.intel.com/technology/silicon/45nm_techn ology.htm

Intel64 says:

• 64-bit flat virtual address space
• 64-bit pointers
• 64-bit wide general purpose registers
• 64-bit integer support
• Up to one terabyte (TB) of platform address space <-- It's dangerous!!! It's (32+8)=40-bit !!! It will be tiny like there is not space in 256 times of 4 GiB !!!

Why? Is it the evilness Intel's FUCK YOU to all the world?

TOP SECRET, NON-DISCLOSURE, NDA, WAR, COMMERCIAL BUSSINESS, MILITARY ARMY.

BTW, AMD64 is the official arch name implemented by AMD and Intel, x86-64 has been officially abandonned

Intel64 property is defined only when it is running on an Itanium processor

Is Intel®64 the same technology used in the Itanium® 2 processor?
No. Intel®64 is an extension to Intel's processors based on the IA-32 architecture. The Itanium processor family is based on the EPIC architecture. These are two separate families of processors, based on two different architectures. The Itanium processor family is specifically designed for the most demanding mission-critical applications.

BTW, AMD64 is the official arch name implemented by AMD and Intel, x86-64 has been officially abandonned

Intel64 property is defined only when it is running on an Itanium processor

Is Intel®64 the same technology used in the Itanium® 2 processor?
No. Intel®64 is an extension to Intel's processors based on the IA-32 architecture. The Itanium processor family is based on the EPIC architecture. These are two separate families of processors, based on two different architectures. The Itanium processor family is specifically designed for the most demanding mission-critical applications.

It's Intel 64 to them.

I suspect the real problem is because high end cards are starting to push Shader unification.

From a chipset standpoint, Intel actually makes decent (not spectacular, but better than many) graphics hardware already, they just don't have hardware transformation and lighting (T&L), which gets offloaded to the CPU. That means you can't be throttling your CPU(s)/cores and need a decent pipe between the hardware and memory. Intel said a couple of years back that it's a myth that the bottleneck is usually in T&L and the problem is actually pixel throughput.

As far as I can tell, that means
a) the bottleneck is between geometry (T&L) and shading (pixel hardware), meaning it's because of the software driver.
or
b) the bottleneck is between shading and the display, meaning Intel's hardware is too crappy to push that many pixels.

The first is a meh (no surprise - it's caused by having geometry in software) the second would be a hardware issue Intel needs to resolve to work with larger displays.

Now back to Shader Unification - basically, if companies like nVidia and ATI move to unified shaders they can assign the types they need as needed and not leave many of them idle. Both of those companies have experience in unified shader architectures already (i.e. the Xbox, and GeForce 8 series), so it wouldn't surprise me if this were the trend of the future. Intel needs to move their software T&L into hardware to create a unified architecture - assuming that is the way of the future.

Another issue is that unified architectures are basically high speed generalized floating point units - these have practical uses in other areas besides graphics (physics, supercomputing, even databases - there are even web pages like this one dedicated to it). Intel has to see this as a threat and know that they need a response should their main competitor, AMD (ATI), go in that direction.

Intel have a Game Compatibility List on their website. Not sure what your situation is, but my lappie use i945. DOOM III is fine, but Quake IV isn't. If Intel want to really penetrate the graphics market, obviously their next list will have to be all green spots. "Intel: The way it's meant to be played!" ;-)

http://www.intel.com/support/graphics/intel945gm/s b/CS-021400.htm

Serious gamers bitch that the 9XX series is low end. Maybe it is. But it whips my nVidia 5700 and that's good enough for me!

Here's a list of supported Linuxes. Bizarrely Ubuntu isn't mentioned, but some folks seem to have it running. Mandriva works out of the box. Google is your friend. You may be able to find more up to date information than these: Good Luck!

http://www.intel.com/cd/channel/reseller/asmo-na/e ng/products/linux/feature/279817.htm
http://www.ubuntuforums.org/showthread.php?p=20287 83
http://www.linuxquestions.org/questions/showthread .php?t=435050

Intel have a Game Compatibility List on their website. Not sure what your situation is, but my lappie use i945. DOOM III is fine, but Quake IV isn't. If Intel want to really penetrate the graphics market, obviously their next list will have to be all green spots. "Intel: The way it's meant to be played!" ;-)

http://www.intel.com/support/graphics/intel945gm/s b/CS-021400.htm

Serious gamers bitch that the 9XX series is low end. Maybe it is. But it whips my nVidia 5700 and that's good enough for me!

Here's a list of supported Linuxes. Bizarrely Ubuntu isn't mentioned, but some folks seem to have it running. Mandriva works out of the box. Google is your friend. You may be able to find more up to date information than these: Good Luck!

http://www.intel.com/cd/channel/reseller/asmo-na/e ng/products/linux/feature/279817.htm
http://www.ubuntuforums.org/showthread.php?p=20287 83
http://www.linuxquestions.org/questions/showthread .php?t=435050

The summary doesn't link to the proof, which can be found on the careers section of Intel's website.

I know that on my desktop system, installing the Intel Matrix Storage Manager drivers doubled my disk performance, judging by the "Windows Experience Index." No kidding; apparently the drivers for my chipset that shipped with Vista didn't include any acceleration.

I noticed something similar with this Dell Inspiron E1405 laptop. It's got an Intel ICH7M chipset, and according to the following page, it's supposed to be possible to switch the SATA controller to AHCI mode, rather than legacy mode:

http://www.intel.com/support/chipsets/imst/sb/cs-0 12304.htm

However, there is no option to make this change in the BIOS, like people have reported with similar systems. So in Linux, I'm stuck with the PIIX driver instead of the AHCI driver. I'm not completely sure if there would be any practical benefit to using AHCI mode, but I do know that currently DMA isn't working correctly on the DVD drive, and my options are being artificially limited.

The signalling rates are measured in GT/s not Gbps (correct me if I'm wrong). The new release doubles the current 2.5 GT/s to 5 GT/s. As a comparison, the 2.5 GT/s is about 500 MB/s bandwith per lane thus 16 GB/s in a 32 lane configuration.

I tried to do the math but I just can't get it right with Gbps instead of GT/s.

http://www.intel.com/technology/itj/2005/volume09i ssue01/art02_pcix_mobile/p01_abstract.htm

Intel actually proposed just that. From this website:

Verified by Intel® is a barebone program between Intel, ODM manufacturers and notebook ingredient suppliers that provides interchangeability of key notebook components for simplified sourcing, inventory management and build-to-order flexibility. Unlike most notebooks that use non-standard components, barebones that are Verified by Intel® are designed and tested for interchangeability of common building blocks, including: Hard disk drives, optical disk drives, keyboards, batteries, LCD screens, customizable notebook panels and AC adapters.

Now if VBI took off, I believe that there is a decent argument to abandon desktops. Currently though, there just isn't.

But alas, it probably won't succeed. Even though the major ODMs (who actually design and manufacture the notebooks.. including Quanta, Compal, ASUS, et al) are all for the VBI initiative -- the major OEMs (who manage notebook inventory and customer service.. including Dell, HP, Compaq, et al) are very, very, very much against it. (OEMs have ridiculous margins on custom notebook parts)

You just proved his point about no one noticing. They don't use the "Pentium" brand anymore. The latest generation desktop processors are referred to as "The Intel® Core(TM) processor family".

See?

I think you failed to notice this...

This is the slickest automatic backup system you can possibly imagine for a small Windows or Linux network. Reasonably cheap, RAID 5 if you want it, and it even runs on Linux!

While I don't know what IBM has done specifically for this processor, a catch-phrase that often comes up when talking about solving leakage is "high-K" dielectric. This means that instead of oxidizing the Si surface, they *grow* the dielectric from a different material (often halfnium or zirconium).

From: http://www.intel.com/technology/silicon/high-k.htm

"These new materials, along with the right process recipe, reduce gate leakage more than 100-fold, while delivering record transistor performance... This thicker class of materials, known as "high-k," is likely to replace today's silicon dioxide technology in the next few years."

I'm pretty sure that Intel already "standardized" the computer vision domain (aka what I know about robotics software) with thier OpenCV library. Not to say that it is really any good or particularly noteworthy, just that the two schools that I attended that offered Computer Vision courses, they both used that library.

Intel seems to be doing the most industrial work along side the academics, and reigning them in.

So the OCPN research group here has already gotten our All-Optical packet-routing to work. All optical in that the signals is Never converted from Optical. The switching signals are still electronic, but an integral part of the system is the packet delay (so the signal is delayed while the switches are set).

We, at first, literally used strands of fiber to delay the signal (so a non-variable delay), now we're using the same fiber delay, but between the multiple strands of fiber are the typical 2x2 optical switch (like a Mach-Zender interferometer-based switch), allowing you to switch on/off various delay line segments (thus allowing you to choose the delay, so you can synchronize the incoming signal, etc.). For the next step we'll be integrating this system onto an InP chip (similar to what the article says has been done).
More importantly, what good is IndiumPhoside based technology if everything's made on Silicon??? John Bowers here, recently made the breakthrough in bonding InP to Silicon, paving the way for allowing this technology to actually become useful outside of the long-range communications industry!

Indeed. Intel does the right thing here -- if I'm interpreting their NISPOM-equivalent manual correctly, not only will they not outsource IT, they don't even allow non-Intel employees to have root. The E-ring people definitely need to do something about the contractors who hire employees based solely on clearance and blood pressure, and the poorly-written contracts that allow them to get away with it.

To be fair, Bill Gates denied saying it, and nobody has come up with an original cite or witness to the quote. That doesn't mean that he didn't say something "wrong or stupid" (which he admits to doing on other occasions). Not like he hasn't been wrong in the recent past (SPAM predictions, for example - it's been two years, Bill, and it's getting worse.

While it's possible he's innocent of saying something that funny, Intel for instance has it on their web site in a 2004 article:

http://www.intel.com/pressroom/archive/speeches/ge lsinger20040219.htm

Sounds like the following like may shed some light on this (presuming it's true):

http://www.kuro5hin.org/story/2004/6/18/171425/218

So it possible Bill is innocent. Still it's dang funny to say :-)

So for this significant portion of the notebook market, I think the transition to 64-bit will probably stretch out past Q2 2007. It might go quicker than most, however, because Merom uses the same chipset as Yonah.

http://www.intel.com/products/processor_number/pro c_info_table.pdf

This confirms it. I can't play. But it also lists who can play! Are you on the list? Check it out.

Back in 2000, Intel made an effort to investigate the boot cycle for a PC contrasted with how people actually used their computers. By eliminating things like floppy seeks and redundancies. By streamlining the BIOS, they enabled a POST time of 25 seconds, down from their 60 second benchmark.
http://www.intel.com/technology/magazine/computing /it08001.pdf

Although I can't find any articles right now, I do remember the term "Zero Boot Time PC" from an article, or maybe it was a Comdex buzzword. Bottom line is that PCs still take time to get to the OS, but it's a hell of a lot faster than it used to be. Hibernation doesn't get me up and running any faster, since I'm running 2gb of ram and XP seems to take forever to save/restore this data. The other downside of XP's hibernation is that when you boot the PC, you have to use it within a certain amount of time, or it shuts off again.

Intel's got a Flash RAM solution in the pipes, it seems: http://news.com.com/Intel+cuts+PC+boot+time/2100-1 041_3-5897993.html

As to the reason that the article's poster might have such a long bootup time, I'd chalk it up to SATA, RAM counting/testing and the BIOS in general. After googling "zero boot time" I didn't get any hits for AMD, or off brand chipset makers. That leads me to believe that Intel is probably leading the pack, in terms of making PCs POST/Boot faster.

Yup: http://www.intel.com/technology/efi/

Are you talking about the IP and TCP checksums, or the Ethernet CRC? I just checked the RTL8139B and RTL8139C (you know, the most common Realtek chips out there), and they don't have anything except basic Ethernet framing (including CRC).

On the other hand, Intel's 82559 chips apparently do have it. I have no idea how common those are, though.

I haven't checked the rest.

I asked for datasheets because I've seen a number of people claim to have seen this, throwing around various brand names, without actually being specific enough to allow their claims to be checked.

I suspect that most of today's brand-new on-board Ethernet chips now support TCP, IP, and UDP checksumming, and possibly packet reassembly, but that's a far cry from most Ethernet interfaces that are actually deployed in the field.

On the other hand, obsessive gamers tend to be running next week's hardware anyway.

It is actually sub-$300, better specced than an OLPC, several *gigs* of memory (512M in the OLPC) and a faster processor.

a)"The Classmate PC is also now being labeled as "sub $300," From your link. (My bold)

b)"Although the final price has not yet been determined, Intel says that it will probably start selling them at about $250, and reduce the price to $200 as the volume of orders increases." (My bold) My link

c)It has 256MB RAM as standard and 1GB Flash. Intels product brief [4 page pdf warning]

It looks to me as if they are designed for completely different markets, Intels aim is to get developing nations on the "Upgrade" treadmill, whereas the OLPC project is actually trying to help out in as inexpensive and unobtrusive way as possible. Plus you need mains power for this Classmate PC, whereas the OLPC has a hand charger.

Oh, it does have a faster processor, so give yourself a consolation point.

PDF brochure is here.

On the one hand, not as inexpensive as the OLPC, and it seems to lack some of the features like the mesh networking.

On the other hand, you might actually be able to buy one, and it should be able to play StarCraft.

Intel FTW!

...It's only been 12 months since they changed it to the new one

Actually, they said "chips" not "chip", probably meaning the Centrino platform is made up of a number of ( specified ) chips, and now an 802.11n package is included in the mix. Right now you're still Centrio if you include one of three approved Intel wireless packages... this probably just means they've announced a fourth option. The real question is will OEMs put it in their laptops, will anyone tell buyers that the standard is not approved yet, and how well will it sell... judging by sales of existing "pre-N" stuff, I'm going to guess it's a real standards nightmare already.

Actually, they said "chips" not "chip", probably meaning the Centrino platform is made up of a number of ( specified ) chips, and now an 802.11n package is included in the mix. Right now you're still Centrio if you include one of three approved Intel wireless packages... this probably just means they've announced a fourth option. The real question is will OEMs put it in their laptops, will anyone tell buyers that the standard is not approved yet, and how well will it sell... judging by sales of existing "pre-N" stuff, I'm going to guess it's a real standards nightmare already.

I'd have to agree with the gist of your statement: graphics hardware has definitely not been a high priority for corporate desktops. I wonder if that will start to change as the GUI's grow more resource-hungry.

Some that I've been to (these are all excellent):
Arizona Science Center (Phoenix, AZ)
St Louis Science Center
Tech Museum of Innovation (San Jose, CA)
The Exploratorium (San Francisco)

Some I have yet, even though I live in the area:
Children's Discovery Museum (San Jose, CA)
The Intel Museum (Santa Clara, CA)
Computer History Museum (Mountain View, CA)

If you're looking for geeky museums, the SF Bay Area probably has more of them in a smaller radius than anywhere else in the USA.

Given that Susan Decker from Yahoo! just joined the board of Intel, another company going through a major restructuring - http://www.intel.com/pressroom/archive/releases/20 061116corp_a.htm

"They don't care how much power the chip consumes or how much heat it dissipates" Actually, the Core 2 Duos barely use any more power consumption than their previous single core processors. Why? Simple. They underclocked the processors by 20%, which allowed them to have 87% performance while only using 51% as much energy. Stack two of these on top of each other, and you've got a Dual-Core system that runs 74% faster than a single core, while only using 2% more energy. http://download.intel.com/pressroom/kits/press/cor e2/Rattner_IDFA_Final.pdf

I think there's already been more than one headline on Slashdot previously and on digg that says almost exactly the same thing... IBM is the real winner. But it's not like it's a big story... if it is, then who's the loser? Intel? AMD? Do they really care that they aren't in the consoles? Not as far as I've heard. They're more worried about chasing the living room PC. (Even though I think they'd get into more living rooms with consoles, but I guess it's more work to design a new console processor than it is to make up a silly meaningless standard for "media" PCs... let the marketing staff do the work, instead of paying engineers to make a product.)

It's even hotter. They're all wearing Bunny Suits.

What's wrong with my 10 gigabit ethernet? ( via my Intel PCI Express 10GbE CX4 cards).

Globalization is always to blame when professionals are out of work. Did you know that Intel has offices in Argentina, Australia, Bangladesh, Belgium, Brazil, Cambodia, Canada, Chile, China, Colombia, Costa Rica, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, India, Indonesia, Ireland, Israel, Italy, Japan, Kazakhstan, Lithuania, Malaysia, Mexico, Mongolia, Netherlands, Nigeria, Pakistan, Peru, Philippines, Poland, Portugal, Romania, Russia, Singapore, South Africa, South Korea, Spain, Sri Lanka, Sweden, Switzerland, Taiwan, Thailand, Ukraine, United Kingdom, United States, Venezuela, and finally Vietnam. And you are worried about ONE new office in India (U.S. comparison)?

It amazes me that people think India to be the biggest threat, and that everyone will loose their job to some guy in 'over there' (see: The Outsourcing Bogeyman, Daniel W. Drezner). Well, that 'guy' adapted to his conditions and the conditions of the market, and got a job knowing that he will get paid less than some guy from the U.S. But you know what, he deserves that job more. Even if he hasn't worked harder for it, which I'm not sure if 'they' do or not, he deserves it simply for being who he is and where he is. And you know what? Even I have to compete against him. But I have innovation and adaptability on my side as well, so I really don't care about him. And neither should the average software guy either. We should all worry more about doing our best to adapt, to innovate, and to provide something that someone else doesn't. I shouldn't have to tell you that nearly every large successful company out there survives on this sort of arrangement. And it seems to work. So what do I suggest? Beat them at their own game, if you so wish to compete head to head. Find a way to stay competitive, do something extraordinary with the resources you have, and you will indeed become the one deserving of 'that job'...

Globalization is always to blame when professionals are out of work. Did you know that Intel has offices in Argentina, Australia, Bangladesh, Belgium, Brazil, Cambodia, Canada, Chile, China, Colombia, Costa Rica, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, India, Indonesia, Ireland, Israel, Italy, Japan, Kazakhstan, Lithuania, Malaysia, Mexico, Mongolia, Netherlands, Nigeria, Pakistan, Peru, Philippines, Poland, Portugal, Romania, Russia, Singapore, South Africa, South Korea, Spain, Sri Lanka, Sweden, Switzerland, Taiwan, Thailand, Ukraine, United Kingdom, United States, Venezuela, and finally Vietnam. And you are worried about ONE new office in India (U.S. comparison)?

It amazes me that people think India to be the biggest threat, and that everyone will loose their job to some guy in 'over there' (see: The Outsourcing Bogeyman, Daniel W. Drezner). Well, that 'guy' adapted to his conditions and the conditions of the market, and got a job knowing that he will get paid less than some guy from the U.S. But you know what, he deserves that job more. Even if he hasn't worked harder for it, which I'm not sure if 'they' do or not, he deserves it simply for being who he is and where he is. And you know what? Even I have to compete against him. But I have innovation and adaptability on my side as well, so I really don't care about him. And neither should the average software guy either. We should all worry more about doing our best to adapt, to innovate, and to provide something that someone else doesn't. I shouldn't have to tell you that nearly every large successful company out there survives on this sort of arrangement. And it seems to work. So what do I suggest? Beat them at their own game, if you so wish to compete head to head. Find a way to stay competitive, do something extraordinary with the resources you have, and you will indeed become the one deserving of 'that job'...

You don't specify where you live, but several very large corporations will hire highly motivated, well qualified high school students for internships. For example Intel will offer summer internships to high school students local to one of their major US campuses (Portland, OR; Santa Clara, CA; Chandler, AZ; Folsom, CA; etc.) If you are not local to a tech giant, you will have to beat the pavement and beg a smaller employer.

When I was about your age I wrote an animation program in assembly on my Amiga 500 and showed it at my local Amiga User's Group. This got me noticed by several people who mentored me and got me a job. What stood out then and will stand out now is: can you dig through the documentation and figure it out yourself? You'd be amazed how many professionals there are out there who need substantial hand holding on any new technology. Self-learners are worth their weight in $100 bills. You want to be one of those.

Here's a good self-learner exercise: When I wanted to learn python, I went to an irc Python channel and found The Python Challenge. It is a series of puzzles to solve by writing python programs to figure out the clue to get you to the next puzzle/web page. This worked very well because the puzzles were small enough that you get that "Yes! I did it!" rush every few hours. Don't be afraid to go to the forums for spoilers.

Here's what scares me: The Intelligent Platform Management Interface (IPMI) and the Remote Management and Control Protocol. (RMCP). Many machines in the field implement these protocols in the network controller, independent of the operating system.

These are UDP-based protocols, on port 623. They can be sent from anywhere on the Internet; not just local machines. They provide total power over the target computer. Functions include:

• Change boot device for next boot, including boot from network.
• Turn machine on, off, or reboot.
• Disable keyboard and user on/off switch.

Supposedly machines come out of the factory with an empty set of IPMI remote management passwords in their nonvolatile memory. Supposedly. All it would take would be to slip in a password load somewhere before the machine reaches the customer, and the customer would never notice that they're 0wned. Even a complete reload of the OS won't fix this. You can switch the machine from Windows to Linux and still be 0wned. Or worse, the IPMI hardware could have a built-in password (perhaps for "factory test") that you couldn't even detect unless you knew it. Because all this remote management stuff is already there, it takes a very minor change to make large numbers of machines very vulnerable.

Run IPMItool and find out what machines will talk to you. Try not to reboot your whole server farm by mistake.

http://www.techarp.com/showFreeBOG.aspx?lang=0&bog no=313 and http://linuxmafia.com/faq/Hardware/sata.html as well as http://www.intel.com/support/chipsets/imst/sb/cs-0 15988.htm has some more info.

According to the advertisement of the Core 2 Duo the processor is more suitable for mobile technology because it is more power efficient. I wonder if this also fixes the heating issues, since less power usually implies less heat.

When I saw this headline I immediately thought of this article, an interview with Jen-Hsun Huang (CEO: nVidia) by Wired dated July '02. In it, the intention of overthrowing Intel is made quite clear, and ironically enough they even mention the speculation from a time when it was rumored that nVidia and AMD would merge.

It's actually a very good article for those interested in nVidia's history and Huang's mentality. Paul Otellini ought to be afraid. Very afraid.

Here are details about the new instructions.

RISC is dead. I miss it too.

SiO2 is only used because it is easy to make by oxidizing silicon and it does not react with silicon. For a long time, it was "good enough". Low-k materials would be better, and that is why you hear about them more.

Like the man said, if they want to give Apple an extra million bucks just to install Windows MCE (with BootCamp, of course) and a few extra bits of hardware, why don't they just do it?