Big, (fault tolerant or not) single-image Symmetric Multi Processing are certainly NOT, on a $/cpu basis cheaper than a cluster of the same number of cpus. Vendors not only make lower margins on clustered systems, these absolute amount of the cost is lower. Clustering with small-way (typically 2-way) commmodity systems will always be cheaper than big SMP - whether you get real (or in the case of some clusters, effective virtual) fault tolerance or not. The sensible rule-of-thumb today is that if your application lends it self to be parallelized easily, you cluster - if it absolutely requires a single MP image of x cpus, you have to do SMP. And most things in the real world fall in between. So, if you CAN cluster effecively, you SHOULD cluster. Next question.
Re:Exactly what *is* the Dell aversion to AMD?
on
Dell Dumping Itanium
·
· Score: 1
If a company made any effort to use AMD's products, some companies have reported that Intel would "suddenly" run out of important server-class chips to ship out to them, and marketing incentive payments would dry up and not be paid.
Funny, HP has a very broad line of Opteron systems, but, as far as I can see, has no shortage of Intel processors and is probaly getting getting their incentives pretty much as usual because HP sells a lot more Xeon-systems than Opteron-systems and neither HP nor Intel would want to damage that mutually beneficial realationship. The real reason Dell uses Intel, is probably that AMD can't deliver the quantitiy of product Dell needs to introduce a server model as a sucessful product. Other vendors can have niche products which they don't expect to sell as much of, but that doesn't work for Dell. Itanium is actually growing in market share year on year in the big system space - BUT it still wasn't generating the volume of business that works with Dell's model.
Exactly right. If you create a game where mugging is possible, then joining that game will potentially expose you to this violence. If the victum, assumed, based on docs and other profers that no such thing could occur, perhaps they have a grounds for a civil suit against the game company- but he would probably have to prove they knew this was possible to be negligent, I think.
It's widely known that Intel was going to bring an evolved Pentium M-like architecture to desktop and and server. Better perf/watt. What Apple's gonna buy.
$300 is cheap for a Linux HPC cluster node....
on
Xbox 360 for $300
·
· Score: 1
...when the mod hacks make this possible. The interesting question is what has Ms done to try to stop this from happening...
| Assert your right to clean your teeth with a garden hoe and other non-standard uses for things you've bought and paid for! |
Read your "Eats, Shoots and Leaves". He may very well have meant to say "try and minimize" if he meant to suggest that two things be done: 1) 'try' the bugs (test them, put them under trial) AND 2) 'minimize' the bugs (attempt to reduce the number of the bugs which you have tested). So, while I agree that it's statistically likely that he intended to say what you think he meant - his sentence is still grammatically correct if you interpret it the way I've suggested. Cheers!
The best way to convince him to do it for free is to take a linksys or other generic access point, configure it as wide open (leave the name "linksys") and plug it into your inverter attached to the cigarette lighter in your car. Park outside the cafe in question, turn on you access point and go inside with your WiFi laptop. Show him how pointless it is to try to charge for wireless when there's a least one other free acess point already in his neighborhood. Seriously, it's only a matter of time until this is legitimately true - he'll only be deluding himself, and hurting his business plan if he counts on a wireless revenue stream.
I've had the Mosberger book for years. Seminal. Yes, HP has by far the most Itanium OSes: (besides Windows and Linux) HP-UX, OpenVMS and now the NS OS. What was trying to point out was that I think many may think that Linux was going to be supported ON the NS platform - I can't imagine that would ever happen - but since the currently supported development environment is Windows with a cross-compiler that integrates into Visual Studio, Linux might be able to play the same role if a Linux native NS cross-compiler was available.
But is now running on Itanium processors (was MIPS). I suspect the Linux connection is that you may be able to now use Linux-based tools for development and the cross-compiler -now you have to use Windows and Visual Studio IDE.
1) "It's an Intel processor, so it has to have great x86 compatibility and performance. If it doesn't run that well, it's a failure." Why doesn't anybody rap an IBM Power processor (or a HP PA or Sun SPARC or...) for having NO (not just slow) x86 compatibility? Itanium is the only non-x86 processor (one with a different native instruction set) that does ANY native emulation of x86 -so does this win praise from people who like Power but wish it supported x86 as well as it's native instruction set? STRANGELY, NO. 2) "It uses an instruction set (EPIC) that only Intel makes so no competing processors can use it. This is very bad from a company with as big a market share as Intel." Well, Itanium is not (not yet anyway) a big market share leader in the big enterprise space that it's targeted at...like IBM Power. Of course IBM would sue your ass off if you tried to independently build and market a Power-compatible processor. The truth is that in the high end of the enterprise computing market, proprietary is the norm and Intel is no worse or better than any of the other players in this space. If you don't think Intel SHOULD NOT be trying to make processors that compete at the high end, fine, but that's what IS where Intel aims this processor at. (Anybody think that AMD WOULDN'T do their OWN exclusive high end processor and unique instruction set if they thought they could sell enough of it to make it work?). 3) "Itanium is hard to work with, being so different with the VLIW instructions, so dependent on compilers and so darned parallel the compilers excrete binaries in structures called 'bundles'. Why didn't Intel just build a better x86 instruction set processor?" It always facinates me that you can sometimes hear this comment voiced in the same conversation that somebody else claims there are no major innovations happening in computer architecture. For better for worse, Intel stuck it's neck out and did a very elegant highly parallel processor design that is not CISC or RISC. And does Intel get credit for spending a ton of money to bring to market something that ISN'T just a tarted up decendent of the 4004 cpu from the 70s? You answer. (There's lot's more to say here, but I'm out of time...)
Re:SGI's Linux is for Itanium not MIPs
on
SGI Faces Bankruptcy
·
· Score: 2, Insightful
While Linux is available for practically anything, including old SGI MIPs hardware, SGI never suggested people use anything other than IRIX on MIPs.
If you want to bitch to SGI about how well Linux runs on platfroms they don't support it for, while we're at it, let's give Microsoft a hard time about what a pain it is to run Linux on the xBox.
SGI's change to Linux is to support SGI's Altix line of Itanium based systems which inlcude the fastest commercially available supercomputer in the world (Number 2 on Top500 list - #1 one is a specialized IBM design that's not based a commercially available product like the SGI Altix)
Also, there are many spook agencies all over the world using SGI gear that you don't get very much publicity about. While these, unfortunately, are not changing the bottom line for SGI, I doubt that certain gov'ts - esp the US - will let SGI go into bankruptcy.
Does your firm have any problem with you partially basing your purchasing decisions on personal political beliefs? Why do I suspect you haven't asked your manager if this boycott is ok? BTW, while the typical AMD system does NOT have Intel ethernet silcon in it (which I'm sure you're happy about) I regret to inform you that it very likely has Broadcom ethernet components and since Broadcom just purchased an Israeli company, I guess that's out too. You might find some old DEC 'Tulip' chipset-based network cards available... but -oops! that semiconductor business is owned by Intel. Don't worry, I'm sure your fellow workers will have no problem supporting your principles by using dial-up to get to the internet AND your internal servers.
1) DRM is just a capability in the hardware which can be used for good or evil -like a lot of other things. 2)You can choose to execute OS and software stacks that take advantage of DRM or not. (I really doubt the open source community, for one, will adopt this uniformly) 3) Before you assume that all DRM is intended to enable some Orwellian conspiracy to smoke out you and your illegal MP3s and warez, consider that the many of the messes we're in for everything from stolen credit cards and identity theft to spam and other similar issues are the consequence of not being able to positively and authoratively authenticate users and systems and application objects across the internet. DRM is an unfortunate name for the ability to affirm authenication -since asserting IP ownership, whatever you think about the value of this, is just one of the lesser uses of what this technology actually is. Making authenication sloppy and untrustworthy is no solution to ensuring civil liberties.
I assume you assumed when I wrote "Apple does have legal issues with Appl" (sorry for truncation) that the first Apple I mentioned where the computer people. Can see why that was warranted....
Also because the 'modular' model assumes that on board integrated video, sound, etc is not good enough for you that you want to use advanced 3rd party cards (as the original poster pointed out, except in the exceptional case like you that already has a set of discrete adapters - integrated is cheaper). So these modular boards assume you want the fastest cpus, most RAM/card slots, and other high end fundimental features to support you attempt to build a high-end system. So, the 'simple' motherboards are not just more expensive because they're rarer, but because they're build with high-end goodies for the basic infrastucture.
In fact, Apple got in trouble in the 70s when they first used the word and had to agree that Apple (computer company) would not get into the business of Apple ( music company). Of course, Apple (music company) cried foul recently over iTunes.... So, you can see cross-industry trademarks DO have to be negotiated.
Slashdot Mirror
...watch for the "Dempsey" in Q1 and the "Woodcrest" low power, high thoughput in Q2.
The Paxville DC is a stopgap.
Big, (fault tolerant or not) single-image Symmetric Multi Processing are certainly NOT, on a $/cpu basis cheaper than a cluster of the same number of cpus. Vendors not only make lower margins on clustered systems, these absolute amount of the cost is lower. Clustering with small-way (typically 2-way) commmodity systems will always be cheaper than big SMP - whether you get real (or in the case of some clusters, effective virtual) fault tolerance or not. The sensible rule-of-thumb today is that if your application lends it self to be parallelized easily, you cluster - if it absolutely requires a single MP image of x cpus, you have to do SMP. And most things in the real world fall in between. So, if you CAN cluster effecively, you SHOULD cluster.
Next question.
If a company made any effort to use AMD's products, some companies have reported that Intel would "suddenly" run out of important server-class chips to ship out to them, and marketing incentive payments would dry up and not be paid.
Funny, HP has a very broad line of Opteron systems, but, as far as I can see, has no shortage of Intel processors and is probaly getting getting their incentives pretty much as usual because HP sells a lot more Xeon-systems than Opteron-systems and neither HP nor Intel would want to damage that mutually beneficial realationship.
The real reason Dell uses Intel, is probably that AMD can't deliver the quantitiy of product Dell needs to introduce a server model as a sucessful product. Other vendors can have niche products which they don't expect to sell as much of, but that doesn't work for Dell. Itanium is actually growing in market share year on year in the big system space - BUT it still wasn't generating the volume of business that works with Dell's model.
...you can all go ahead and write games about fighting evil girls, though.
...since they've never allowed any kind of overclocking.
The conservative choice.
Or, try sending mail to addresses at
their site
Exactly right. If you create a game where mugging is possible, then joining that game will potentially expose you to this violence. If the victum, assumed, based on docs and other profers that no such thing could occur, perhaps they have a grounds for a civil suit against the game company- but he would probably have to prove they knew this was possible to be negligent, I think.
It's widely known that Intel was going to bring an evolved Pentium M-like architecture to desktop and and server. Better perf/watt. What Apple's gonna buy.
...when the mod hacks make this possible. The interesting question is what has Ms done to try to stop this from happening...
| Assert your right to clean your teeth with a garden hoe and other non-standard uses for things you've bought and paid for! |
...of course. Otherwise through torture, threat of prison (or worse, threat of killing future book deals) they might squeal.:-)
Read your "Eats, Shoots and Leaves". He may very well have meant to say "try and minimize" if he meant to suggest that two things be done:
1) 'try' the bugs (test them, put them under trial)
AND
2) 'minimize' the bugs (attempt to reduce the number of the bugs which you have tested).
So, while I agree that it's statistically likely that he intended to say what you think he meant - his sentence is still grammatically correct if you interpret it the way I've suggested.
Cheers!
...chips for a while. There's a ton of them listed here . Look for the ones called EM64T.
And Dell has been selling these 64-bit chips for long time too.
The best way to convince him to do it for free is to take a linksys or other generic access point, configure it as wide open (leave the name "linksys") and plug it into your inverter attached to the cigarette lighter in your car. Park outside the cafe in question, turn on you access point and go inside with your WiFi laptop. Show him how pointless it is to try to charge for wireless when there's a least one other free acess point already in his neighborhood.
Seriously, it's only a matter of time until this is legitimately true - he'll only be deluding himself, and hurting his business plan if he counts on a wireless revenue stream.
I've had the Mosberger book for years. Seminal.
Yes, HP has by far the most Itanium OSes: (besides Windows and Linux) HP-UX, OpenVMS and now the NS OS.
What was trying to point out was that I think many may think that Linux was going to be supported ON the NS platform - I can't imagine that would ever happen - but since the currently supported development environment is Windows with a cross-compiler that integrates into Visual Studio, Linux might be able to play the same role if a Linux native NS cross-compiler was available.
Wow. The making nice is the first step that leads us to Linus being hired by Microsoft to put Windows on Linux
Gary Wolf is a prophet.
But is now running on Itanium processors (was MIPS). I suspect the Linux connection is that you may be able to now use Linux-based tools for development and the cross-compiler -now you have to use Windows and Visual Studio IDE.
Check out
1) Windows Services for UNIX
2) Windows Scripting Host
The folks that normally crow about Intel being a technological follower of AMD should graciously acknowledge that Intel was ahead of AMD on DDR2...
1) "It's an Intel processor, so it has to have great x86 compatibility and performance. If it doesn't run that well, it's a failure."
Why doesn't anybody rap an IBM Power processor (or a HP PA or Sun SPARC or...) for having NO (not just slow) x86 compatibility? Itanium is the only non-x86 processor (one with a different native instruction set) that does ANY native emulation of x86 -so does this win praise from people who like Power but wish it supported x86 as well as it's native instruction set? STRANGELY, NO.
2) "It uses an instruction set (EPIC) that only Intel makes so no competing processors can use it. This is very bad from a company with as big a market share as Intel."
Well, Itanium is not (not yet anyway) a big market share leader in the big enterprise space that it's targeted at...like IBM Power. Of course IBM would sue your ass off if you tried to independently build and market a Power-compatible processor. The truth is that in the high end of the enterprise computing market, proprietary is the norm and Intel is no worse or better than any of the other players in this space. If you don't think Intel SHOULD NOT be trying to make processors that compete at the high end, fine, but that's what IS where Intel aims this processor at. (Anybody think that AMD WOULDN'T do their OWN exclusive high end processor and unique instruction set if they thought they could sell enough of it to make it work?).
3) "Itanium is hard to work with, being so different with the VLIW instructions, so dependent on compilers and so darned parallel the compilers excrete binaries in structures called 'bundles'. Why didn't Intel just build a better x86 instruction set processor?"
It always facinates me that you can sometimes hear this comment voiced in the same conversation that somebody else claims there are no major innovations happening in computer architecture. For better for worse, Intel stuck it's neck out and did a very elegant highly parallel processor design that is not CISC or RISC. And does Intel get credit for spending a ton of money to bring to market something that ISN'T just a tarted up decendent of the 4004 cpu from the 70s? You answer.
(There's lot's more to say here, but I'm out of time...)
While Linux is available for practically anything, including old SGI MIPs hardware, SGI never suggested people use anything other than IRIX on MIPs.
If you want to bitch to SGI about how well Linux runs on platfroms they don't support it for, while we're at it, let's give Microsoft a hard time about what a pain it is to run Linux on the xBox.
SGI's change to Linux is to support SGI's Altix line of Itanium based systems which inlcude the fastest commercially available supercomputer in the world (Number 2 on Top500 list - #1 one is a specialized IBM design that's not based a commercially available product like the SGI Altix)
Also, there are many spook agencies all over the world using SGI gear that you don't get very much publicity about. While these, unfortunately, are not changing the bottom line for SGI, I doubt that certain gov'ts - esp the US - will let SGI go into bankruptcy.
Does your firm have any problem with you partially basing your purchasing decisions on personal political beliefs? Why do I suspect you haven't asked your manager if this boycott is ok?
BTW, while the typical AMD system does NOT have Intel ethernet silcon in it (which I'm sure you're happy about) I regret to inform you that it very likely has Broadcom ethernet components and since Broadcom just purchased an Israeli company, I guess that's out too. You might find some old DEC 'Tulip' chipset-based network cards available... but -oops! that semiconductor business is owned by Intel.
Don't worry, I'm sure your fellow workers will have no problem supporting your principles by using dial-up to get to the internet AND your internal servers.
1) DRM is just a capability in the hardware which can be used for good or evil -like a lot of other things.
2)You can choose to execute OS and software stacks that take advantage of DRM or not. (I really doubt the open source community, for one, will adopt this uniformly)
3) Before you assume that all DRM is intended to enable some Orwellian conspiracy to smoke out you and your illegal MP3s and warez, consider that the many of the messes we're in for everything from stolen credit cards and identity theft to spam and other similar issues are the consequence of not being able to positively and authoratively authenticate users and systems and application objects across the internet. DRM is an unfortunate name for the ability to affirm authenication -since asserting IP ownership, whatever you think about the value of this, is just one of the lesser uses of what this technology actually is. Making authenication sloppy and untrustworthy is no solution to ensuring civil liberties.
I assume you assumed when I wrote "Apple does have legal issues with Appl" (sorry for truncation) that the first Apple I mentioned where the computer people. Can see why that was warranted....
Also because the 'modular' model assumes that on board integrated video, sound, etc is not good enough for you that you want to use advanced 3rd party cards (as the original poster pointed out, except in the exceptional case like you that already has a set of discrete adapters - integrated is cheaper). So these modular boards assume you want the fastest cpus, most RAM/card slots, and other high end fundimental features to support you attempt to build a high-end system. So, the 'simple' motherboards are not just more expensive because they're rarer, but because they're build with high-end goodies for the basic infrastucture.
In fact, Apple got in trouble in the 70s when they first used the word and had to agree that Apple (computer company) would not get into the business of Apple ( music company). Of course, Apple (music company) cried foul recently over iTunes....
So, you can see cross-industry trademarks DO have to be negotiated.