Slashdot Mirror
Intel 32/64-bit Nocona CPU
OCGeek writes "A picture of the upcoming Nocona processor of the Xeon family that has 64-bit
extensions known as Intel EM64T has appeared on
VR-Zone website. Nocona will have
604 pins and supports HyperThreading, SSE3, PCI Express, DDR2, Vanderpool
technology."
They made the mistake to have not one, but two featured stories on Slashdot today. No wonder their site is down, LOL
Here is the picture:
+--+
| |
+--+
Just for starters, notice that all the hardware sites get their test units from the manufacturers. In other words, they call the manu and say 'please send me a free hard drive to test for a review'. The manu then tries out 5 units to find the one that works best and sends it.
Consumers Reports, on the other hand, goes to the store and buys a random unit, same as you or I might.
Personally, I trust www.storagereview.com, but they do the same thing.
Perhaps you missed the whole DeCSS issue? "Without licensed DVD players for Linux and other operating systems, an entire class of computer users is completely cut off from viewing DVDs."
Phwoar! I'll have some of that.
Slashdot. Pornography for nerds.
EM64T
Remember, it's spelled x86-64.
Save your wrists today - switch to Dvorak
... would they call it a "Sno-Cona"?
First time I've heard of it. I know about all the other stuff mentioned but not this. And now It's slashdotted on top of all.
Anyone know?
In anticipation of intel's move away from MHz numbers and confusing names, I predict the nonoca will adopt the name "Intel Xeon Championship Edition."
Why can't I get this to run on my WXP machine? I have XP Pro installed....
You linux geeks get all the good toyz!!
Darn you, Darn you to Redmond!
What do I get?
Well.. I guess I do get all the neat patches.
Here's the skinny on vanderpool.9 94215
http://www.newscientist.com/news/news.jsp?id=ns99
It actually seems a really interesting technology. The CPU itself can generate virtual machines that can run different OS's simultaneously. Kinda like hyperthreading but on a much lower level.
Hmmm, these kind of sites are becoming a nuisance.
Sorry, that website uses broken embed tags and Windows-specific registry CLSIDs to point to quicktime player. I don't have a "registry" or a "quick time" player. For those of us who choose our own browser helper applications (instead of it being decided by a "registry") here is the relevant link.
Intel talked about this at the last developers conference. Its the ability to run OSes and applications in partitions that are protected from trashing each other. Here's a blurb from one of the keynote addresses (about halfway down):
You may remember at the last IDF, Paul Otellini in his keynote did a demonstration and introduced a new technology, a new star "T" called Vanderpool Technology or VT. In that demo, he was in a home environment where he demonstrated by creating different stations in a virtualized station. You are able to run your PVR in one partition and the games in another partition without interfering with each other.
VT has applications not just in the digital home but also in the digital office. What are some of these usage models? Let's take a look. VT, likewise, can be used in business computers to create different partitions, to provide an IT partition where the IT mission-critical applications are well protected and not compromised by the user. At the same time, it can create partitions that can provide legacy support. In other words, applications that may not run under the new operating system.
Now, this is the kind of thing that's actually fairly common encountered in both large enterprises as well as more medium business.
An example we see in accounting software or asset tracking software, they're written and validated on an old operating system that have not been reported or validated.
As an example, my sister is a dentist and she has a billing system on her computer. She wouldn't dare to upgrade it because there's no support of porting that billing system to a new OS. And as a result, she continues to run on old hardware, old OSs, that expose herself to productivity and security issues. Not a good situation.
So let's take a look at how this actually works. I'd like to invite Jason Davidson out here to show us how VT benefits the enterprise.
(Demo begins and ends.)
BILL SIU: So in the coming several years, we'll be working with many of our business colleagues, many of you present here, to develop this capability and bring this kind of improvement to the enterprise. We think this is of just great value to manageability, providing both end user benefits as well as IT value.
One assumes the demo shows them crashing an application yet the other application keeps on working.
to the Engineer, the glass is neither half full nor half empty. Its just two times too big.
Quick screen capture of the entire (tiny) article here
The GDM link points to here (japanese)
The Xeon roadmap link points to another vr-zone article here (probably also slashdotted)
Kinda contentless, apart from the pic.
These sites are almost always reviewing products that haven't hit the market yet. They can't just go out and buy a retail unit if there aren't any available yet.
This is also how they can get away with paper launches... Make a few samples available to the reviewers to make it seem like the processor is available. In these cases, usually the review sample is such an early revision that anything a consumer touches probably works better.
This was demonstrated at the fall 2003 Intel Developer Forum. They operated two virtual machines, one running linux and one running windows, and rebooted one of the machines with the other unaffected.
I'm not sure which one they rebooted but I have a pretty good guess.
...on portugal and brasil... just google for cona and you'll see what i mean :-)
Support for PCI Express and DDR2 are dependent on the chipset, not the processor, in Intel CPUs. So saying that the Nocona processors support PCI Express and DDR2 is pretty stupid... Any Intel processor could use them so long as they were running on a chipset that did.
Of course, Intel normally releases new chipsets with a new revision of a processor family, but that is another matter entirely. Since the site is down, I have no idea if this is discussed at all.
If this could be done efficiently, and in a way which allowed users to easily switch between the two OSes, one could run linux and windows simultaneously. Then, instead of having to use a second rate application for those apps which haven't been replicated in the linux world, one could easily switch back to windows for those few necessary apps which were holding one back from trying out linux.
Linux adoption would go up as people find it easier to try it out without abandoning their familiar windows apps, which leads to more linux development, which results in more replacement of those windows apps(since there is still the cost benefit to switching to linux).
I was really worried until the end of the snippet when Intel mentioned Dance Dance Revolution 2 support...
- A software solution like VPC that can be easily added to existing hardware (at a large performance cost)
- IBM's
?That should be, IBM's hardware abstraction technology
Just for starters, notice that all the hardware sites get their test units from the manufacturers. In other words, they call the manu and say 'please send me a free hard drive to test for a review'. The manu then tries out 5 units to find the one that works best and sends it.
...that there's such things as rated speeds. For a CPU that would be something like "This CPU is rated at 3.0 GHz, but it'll overclock to 3.6 GHz". Maybe the average consumer CPU won't overclock to that. But it's a pretty sure thing it *will* work at 3.0GHz, and that's the benchmarks I read.
As for harddisks, I imagine they find one with no remapped sectors (a "perfect" disk) but otherwise, I doubt they can do much either without rigging the specs. There's simply not much room for variability these days. Maybe they have a perfectly balanced/aligned disk that could do more than 7200rpm, but that's a different story.
Kjella
Live today, because you never know what tomorrow brings
Doh. Isn't it obvious that he's talking about rollbacks?
/*
The ones in common use don't support rollbacks.
e.g. you don't get to do stuff like:
begin;
rm -rf somedir
(ooops! should have been somedir/* )
rollback;
I'll gladly be corrected if I'm wrong.
...it appears that Intel is a gnat's eyebrow away from having to use liquid cooling on that motha! 150 watts! OMG!
"...Well, there's egg and bacon; egg sausage and bacon; egg and spam; egg bacon and spam; egg bacon sausage and spam..."
Ah, SIE (Start Interpretive Execution)! I wondered when that mainframe functionality would make to PC silicon. It's about as old as Linux, so I guess it's about time.
Can You Say Linux? I Knew That You Could.
My PC has a Pentium III 866 MHz, and it supports a port of DDR (not just 2nd Mix but all the way to 8th Mix through bumper packs) just fine through the StepMania simulator. If you want to contribute AMD64 builds of StepMania, go right ahead; StepMania is free software.
(FICTIVE STORY)
Well actually this happened:
First engineer: Look, a new processor
Second engineer: Hey, it's a male
Bypassing Portuguese: ??? Nocona?
Signature Pro version 1.13.2-3 release 83.5 beta3try7 after-breakfast edition
Ummm.... isn't this like MacOS running VirtualPC? I mean, really, what's the big difference?
If I didn't have absolutely NOTHING to do, I wouldn't be here.
Sat on a shelf next to me is an Amiga 4000/040, made in 1993. It has a 68040 CPU in it, which is fully virtualizable (which means that you can run an OS in userspace by pretending to the OS that it's in kernelspace). Most other non-toy CPU architectures are fully virtualizable (e.g. SPARC, PPC, and various mainframes). The notable exception? x86.
It's nice to see that Intel has finally been dragged kicking and screaming into the 1970s. Give them another 20 years, and they might "invent" RISC.
Soon they won't need the actual x86 instruction set at all!
Buy Opterons .... They scale better. Dual systems.. its about neck and neck with Xeons but go to quads and the Opterons eat Xeons for Lunch. Oh .. and the Opterons are cheaper too. It's a no brainer folks. It wall take alot more then copying AI64 from AMD to put the Xeons on top. Indeed soon with how the Opterons scale they will eat up the Itantics too.
Depends on the tasks set for the CPU's. For some tasks, dual-cpu's are the sweet spot for performance and cost. And if you're running renders, physics calcs etc a lot, the Xeons are the way to go, and for databases etc, the Opterons are the way to go. And besides, the dual Xeons have beaten dual Opterons, despite the Opterons running in 64-bit mode, with all those extra registers.... Now just think about what the Xeons will be able to do when they also get to play with all those extra registers.
I don't think you can just boot a stock copy of Solaris on any SPARC machine without a thin layer of virtualization glue.
In any case, this technology doesn't remove that need (they mention the need for a "Hyper-OS" and small modifications to the host operating system)... it just pushes a lot of the common stuff (simulating interrupts, catching exceptions) into the hardware.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
1) The chip interface to the northrbridge has been improved and will allow it to go "Really Fast".
2) The chip has an intergrated memory controller and/or PCI express bridge/controller ala Opteron.
THIS THING CAN TURN ON A DIME, MACROSSZERO STYLE ALSO FUCK BETA, ~NYORON
is it a 64bit CPU that can act like a 32 bit CPU or it's 16bit predecessor (which is, itself based on an 8 bit design).?
I can understand why Intel wanted to go to a clean 64bit CPU implementation, but It's a bit late in the game for them.
Free Software: Like love, it grows best when given away.
Vanderpool is basically a re-optimization of priorities and costs. Read this for more:
Intel won't say shit about it if you ask, I have several times. I was at both the IDF demos on it, and they said all of nothing technical. I found out anyway. :)
http://www.theinquirer.net/?article=14289
That said, what is Pellston and Foxton? I know one of them.....
-Charlie
That New Scientist is basically repeating the Intel line on the Tech, which is to say nothing. You notice the entire article is basically saying that it is about virtualization, no specifics. It is pretty close to word for word what the Intel PR people will tell you. Journalism indeed.
I was at both IDF keynotes, and they gave demos, and did nothing much. I asked, they told me squat. I found out and wrote it up, I posted a link to my story above, I won't re-post the link.
It annoys me when Intel announces a tech, tells you it will be the greatest thing since sliced bread, but then won't talk about it if you ask. Kind of like a certain Unix vendor we all know....
-Charlie
Isn't that what the recycle bin does?
*ducks*
Be wary of any facts that confirm your opinion.
I wonder how long it will take Intel to move 64 bit technology into the P4 line?
By adding it in the Xeon, they legitimize the technology. But, they don't put it in the consumer chips. So, this makes the Athlon 64 a lot more attractive.. Compared to the Intel chips, the A64 has high end technology in a low cost chip.
If AMD ever completes their unfortunate socket shuffle, the A64 could really take off.
but does it have the proverbial kitchen sink?
:)
Or is the heat sink merely that heavy?
Juuussst kidding.
Don't get me wrong, I love AMD and only buy
AMD for myself. If Vanderpool works the way
I'm hearing it's supposed to..... I have a lot
of customers who can use that technology YESTERDAY!!!
Last year even!
Please bring this about in an AMD-64 Version Pleaaaaaaase!!!!
No .. its the HT links and on board Memory Controls that help the Opterons win. The beat them 32bit or 64bit. I have not seen any test that show Dual Xeons doing anything better then the Opterons with the exception of puting out more heat. Please show us a Link to said test.
Sounds sort of like Xen's approach, but with hardware support.
pb Reply or e-mail; don't vaguely moderate.
uhm, the 586 and up have been internally RISC since the early 90s. The Itanium's VLIW is sort of an even more extreme design than RISC was.
Jeremy
Fade to intro: granny is sipping tea while working on the computer. She's checking some new recipes online and sending an e-mail to grandson jimmy. She hits the 'Send' button in Outlook and WHA-ZAM! that email is sent so fast by her new Intel 64-bit Nacona that it's almost illegal.. Wowza.
I prefer to get Opterons over "Xeopterons", if for no other reason than because Intel blatantly ripped off the 64-bit extensions from AMD, and didn't even bother mentioning them in the "ia32e" specification documentation.
Granted, AMD is making designs based on Intel's ancient and decrepit architecture, but at least they acknowledge this and give Intel credit where credit is due. Many of AMD's AMD64 technology papers are published as the differences between Intel's IA32 papers and their design.
Of course, the fact that Opterons scale better due to not sharing all memory bandwidth between CPUs, using HyperTransport for interCPU communication, and having a dedicated and integrated low-latency memory controller for each individual CPU helps in the Opteron-vs-Xeopteron choice as well....
Computer Science is no more about computers than astronomy is about telescopes. --E. W. Dijkstra
talking about some crappy Intel CPU because the AS/400 (iSeries) has done this for quite some time and the RS/6000 (pSeries) will be able to do this before the end of this year with the announcement of Power5.
It's not just the mainframes. The iSeries (AS/400) can do this, including sub-processor partitioning, as well as their pSeries (RS/6000, i.e. UNIX) line. With the release of Power5 this year, the pSeries line will get virtual I/O and sub-processor partitioning.
Maybe you meant one of these when you said mainframe.
For a troll, that shit was fuckin funny... congrats!
Never mind. I wish at times like these that Slashdot let you nuke your posts.
Opteron has coolness factor built in... but in the end Intel wins... sad but true. It's strictly up to Intel to mess this one up.
I understand that at the time, it was in AMD's best interest to sign the cross-licensing agreements... but here is a clear case where it will hurt AMD. AMD better have a rabbit in their hat, or it's "show's over".
The 40 million lines of code is for the whole system. That would be like the kernel, xfree86, KDE, Mozilla, and all the system tools and utilities.
"Taligent is still pure vapor. Maybe they'll be the last who jumps up on Openstep... "
Search google for pages in Portugese with the word "cona", then "translate this page" on the first link and you get this perversion:
" MY CURIOUS TOUPEIREX:
ALREADY WE CAME BACK Of the MISSION THAT In them TOOK the LISBON... or EITHER TO FOLLOW MY BROTHER-in-law CELESTINO To SUCH NECESSITY GAY TO TAKE OFF CLEAN IF IT WOULD BE GAY OR NOT. I, the CELESTINO, COUSIN ILDA, NEIGHBOR ARMANDINA And the SISTER Of It, the CONSTANCY PREPARED A GOOD MERENDA And Set It WAY Of the CAPARICA."
Worse than goatse!
--
E_NOSIG