Let's see some decent Linux drivers from ATi and a benchmark showdown at Linuxhardware.org. Till then, it might be wise not to make such sweeping remarks to the Slashdot crowd.
I don't see why this particular set of benchmarks is special enough to deserve attention on Slashdot. Okay, so they run through a nice variety of benchmarks and they're a fairly credible source. But it has nothing to do with Linux, nothing to do with Apple, and not much to do with OpenGL. The benchmarks are all done in Windows XP with DirectX 9. Even the UT2003 benchmarks were done in DX9.
About the only thing I can tell about this set of benchmarks is that OpenGL and Linux are ignored completely. At least most other reviewers benchmark Quake 3, Serious Sam 2, or some synthetic OpenGL benchmark.
I don't really see what's wrong with this approach. Using desktop managers as an example, some people simply want a full-featured (Or bloated, if you prefer) window manager like KDE so they have similar functionality to commercial OSes. Others may want a smaller one like WindowMaker. Either project would concievably be further along if the developers from one abandon their own projects and joined with the other.
That's assuming there wasn't a lot of internal bickering going on as to how things should be done, however. I think internal strife is far more dangerous and inhibiting than forking a project. The only way to make dozens, hundreds, or thousands of developers set their attitudes and egos aside for the sake of reaching a common goal is to offer them loads of cash.
I certainly like the idea of using effective air cooling like that rather than extreme / expensive methods like dipping the thing in fluorinert.
The only concern I have is servicing the boards. I've never worked with one of those racks before, but I imagine Linux Networx did a good job at making the boards fairly easy to swap out in case of node failure.
Here are some pics of the Evolocity 2 nodes in action for everyone to check out.
In many clusters today, each node relies on a hard disk to hold a distribution. This is a real big problem because hard disks generate a lot of heat. This is especially true with 10-15K RPM SCSI drives.
To add to that, most cluster node manufacturers that I know of design their 1U cases so that the hard disks are mounted on the front of the case for hot-swapping (Which in itself should suggest that people have problems with drives failing). The fans lie somewhere behind the hard disks. For example, this is a picture of an Appro 1122H dual Opteron server. Some cool air (But not a lot) is drawn from the vents on the front-right of the case, but a lot of hot air gets sucked away from the hard disks and blown over the CPUs. A better solution is just to take out the hot swap bays, make sure there's no SCSI backplane in the way, and let the cool air be drawn in from all across the front.
There are some manufacturers who will put a cardboard heat tunnel around the intake vent, lead it over some blowers, and eventually have to run over the CPUs. That's good design for keeping the CPUs cool, but it basically leaves the hard disks and other components such as north bridge chips and memory modules to fend for themselves when it comes to fighting thermal death. Often times they fail.
I have yet to see a hard disk that was smart enough to spin down to a lower speed when it reached a certain temperature. Even if they were that smart, who wants an HDD performing poorly when you've already spent ungodly amounts of money for bleeding edge 15K ultra-320 drives? As far as I'm concerned, they're practically built to destroy themselves. This is why diskless cluster nodes are so appealing to me.
Sorry for my little rant there. I hope it helped to clarify what I meant when I was emphasizing the advantages of having fewer moving parts.
Pink is the largest LinuxBIOS cluster in the world to date. The only moving parts in each node are cooling fans. This translates into a small savings on hardware, but more importantly means the nodes are by design more reliable thanks to fewer moving parts. Each node is powered by two 2.4GHz Xeon processors with four gigs of RAM and Myrinet 2000 interconnects.
The raw hardware power, while impressive, isn't what makes this cluster unique. The kicker is in the software, more specifically Clustermatic 3 featuring LinuxBIOS. Stuff happens and nodes fail, but thanks to LinuxBIOS they can be back up in a matter of seconds, not minutes.
Additional tools for the frontend node from Linux Networx makes updating nodes super-easy. You can flash each node's BIOS with a single command all in a matter of seconds. BProc allows you to run basic shell commands on any node without even installing a distribution on those nodes. w00t!
What we see here is a big shift away from expensive hardware and proprietary software. The software powering this cluster is 100% GPL, so users save a fortune in software licensing costs alone. And while these P4 nodes in particular aren't exactly cheap, they provide pretty darn good power and are far less expensive than Alpha servers. Also, using the x86 architecture means that consumer boards are not far behind in clustering. In fact, you can check out the LinuxBIOS homepage and see some pretty cheap boards that are supported already. So if you have some spare cash lying around and a couple weekends to kill, you can pick up a cheap board + cpu + memory combo and set it up as a slave node for your desktop machine with the same software these guys use to power this huge cluster.
I'd want Redhat and Mandrake to dispapear from that picture entirely. RPMs are simply beyond redemption at this point for me. Give me Portage or give me death!
It's one of the only times a company has made an in-house port of the game. Not just another 3D Tetris or something, but a real gamer's game.
Tribes and Heavy Gear certainly were great thanks to the efforts of Loki, don't get me wrong. And we have SDL, Loki_update, and some other great tools thanks to Loki.
However, when Loki went out of business, support for most of their games got dropped as well. You could not call up a company like Sierra and ask for Tribes 2 support for Linux. Patches were also discontinued. In the case of Tribes 2, it was just luck that Sam Lantinga was working for Sierra (Blizzard, to be more specific) and was allowed to patch the Linux version as well.
Bioware has made a commitment. They will always support Linux. They will, not a third party who may go out of business at some point while the game is still being developed. In-house porting is what makes the Linux port of NWN so signifficant.
Re:Easy way to get a cluster up and running?
on
Linux Clustering
·
· Score: 2, Informative
(Shameless plug)
Head over to LinuxNetworx now for a LinuxBIOS-ready Evolocity II cluster!
Note: I'm not an employee of LinuxNetworx, but they still kick ass.
Nothing productive socially? I suppose I could do what most people my age do, much of which involves substance abuse, destroying property, etc, but instead I find joy in things that don't have (real world) death in the equation. Putting my enthusiasm for technology to use and gaming on-line w/ friends seems like a good choice.
I will agree with you if that person seriously does not have connections with anyone in his real life community. However, I do feel there are a signifficant number of people like myself who see video gaming as just another group activity to kill time like watching a movie, sports, etc. If you're going to just sit around and chat with friends, why not do it while conquering enemy territory and slaying opponents?
Define "convince." Anand's review shows that the FX5900 is indeed a worthy competitor for the latest ATi offering, and the superior Linux drivers put nVidia over the top for Linux users.
If you're concerned about nVidia's ethics, perhaps you should check out ATi's background.
Ah, that puts it in perspective. I wish the headline included that rather than overly generalizing it by saying stiff like "Help us shake it down in preparation for Mozilla 1.4 final."
Hardly. Below are two links that have video coverage of Zeta at CeBit2003: http://ddanneels.free.fr/Zeta-CeBIT200 3.avi http://gravity24hr.com/mirror/zeta/BeOS-Zet a-Prese ntation-CeBIT2003.avi
I really wish Slashdot would stop covering incremental releases of software. While I'm not a Mozilla user, I'm guessing that Mozilla will go through at least two more release candidates before 1.4 final is out, which I do think should be covered. It's a bit silly to follow any project through all the release candidates.
What??? I thought MS products were affordable for everyone ($199 for WinXP Home, $479 for Office XP standard edition))! That's why they own 93+ percent of the desktop market, because people buy their products for their value, not because they're forced to in any way!
/* End sarcasm */
Someone in another message pointed out that a billion dollars in software is after the price markup. One billion divided by the total cost of Office XP and Windows XP home retail, $678, comes in at just under 1.5 million bundles. If it costs MS $.05 to stamp out the bundles, and I would consider even that a conservative estimate for 2-4 CDs in mass production, that comes in at around $75,000. So they're spending about $75K so they can claim a billion dollars in donations. Wow! Take that 75K and divide it by 1 billion to get the percentage MS is spending compared to what they're claiming to give away:.000075 percent. Now how's that for a return on an investment!
Perhaps Redhat/Mandrake/SuSE/any distro with some cash should make a special edition of their Linux distro with an important sounding name, set the MSRP at $1000 per copy, stamp out tens of thousands of CDs and mail 'em out like AOL CDs to charities and claim to have given up tens or hundreds of millions.
I'm having trouble believing my own numbers, so if I made a math error please feel free to correct me.
>> What's worse than that, though, is that they are still trying to pretend that it's not the case.
Since when? Jen-Hsun Huang admits defeat (But promises a comeback):
"Tiger Woods doesn't win every day. We don't deny that ATI has a wonderful product and it took the performance lead from us. But if they think they're going to hold onto it, they're smoking something hallucinogenic."
Let's not jump on nVidia too harshly for this. Sure, this spectacle seems to have gained a lot more publicity than ATi's own cheating ( linklinklink ). At least when nVidia cheated in 3DMark, they publically denounced synthetic benchmarks.
"The R200 path has a slight speed advantage over the ARB2 path on the R300, but only by a small margin, so it defaults to using the ARB2 path for the quality improvements. The NV30 runs the ARB2 path MUCH slower than the NV30 path. Half the speed at the moment. This is unfortunate, because when you do an exact, apples-to-apples comparison using exactly the same API, the R300 looks twice as fast, but when you use the vendor-specific paths, the NV30 wins.
The reason for this is that ATI does everything at high precision all the time, while Nvidia internally supports three different precisions with different performances. To make it even more complicated, the exact precision that ATI uses is in between the floating point precisions offered by Nvidia, so when Nvidia runs fragment programs, they are at a higher precision than ATI's, which is some justification for the slower speed."
>>My question is why a closed, not free product gets a plug every time they put out a new release?
My question is why every time there's a new MacOS version people hail it as the evolution of Unix and such while other closed, not free software gets shunned.
A lot of people have come down pretty hard on Robertson for allowing people to use root by default. Here's an idea: Offer solutions instead of just bitching about the problem.
The root issue is of serious concern since not only are files exposed to damage, but services can be run that can compromise security as well. Here's my idea:
1. Classify accounts by access level and make a graphical interface during bootup to set up user accounts. Don't give everyone root right off the bat, but give them something like "Administrator" or "poweruser." Make "root" a checkbox that's unchecked by default, label it with a warning and maybe draw some cute little skulls near by with a can of rat poison when you move your cursor over the checkbox. Have a "read more" option that explains, in gory detail, the risks of creating a user with root privileges and why it should only be used only for (insert creative name for) system tasks. If the user hasn't quit due to boredom after the first sentence and given up on "root" then at least he/she might understand why the root account is so special and tread with caution.
Anyway, back to the watered down root privs. Give these access levels high privileges that will be required for installing software later on (See part 2), but not enough to say, start a telnet server, ping flood someon, or rm -rf / . This will be the default permission. Also, place all users in the same group. Give them suggestions in a drop-down menu like "family" to make the process more recognizable.
Obviously, users aren't going to want to keep track of multiple passwords. And I just know I'm gonna get roasted for this, but make the root password the same as the password entered in for the first user. The idea here is to make the user think twice about an operation that he/she is going to do. When a user does something that requires root privs, ask for confirmation with a password. Maybe even have a little "help" button that explains that when the box pops up asking for confirmation, they're about to perform a task that could be potentially harmful to the system.
Of course this is less secure than it should be and having the same root and user password won't help much if a hacker manages to get in or something. However, for an OS with Micheal's intentions we have to cut the user a lot of slack, dreary as it may sound to all of us.
2. I know this would be enormously tedious and require tremendous effort, but I see it as part of the process of creating a distribution from scratch with Micheal's objective in mind. "Standardize" directories where stuff should go. Yep, that's it. Keep directories like/usr/X11R6 under lock 'n key for "root" only, where modification will require a password, but chmod +777/usr/local/ (Hopefully all users will be in the same group by default) and other such directories commonly used to install apps. Make sure that all packages in Click 'n Run do not try to write to directories that users are not given permission to.
Unfortunately, my suggestions are crude at best and go off a few assumptions. For one, I assume that people will not try to upgrade extremely important things like KDE without reading some documentation first (How about some kinder, gentler documents that assume no Linux knowledge available from Lindows?). And like Micheal, I also assume that people aren't going to be using Lindows to make a network filesever or something delicate like that.
That's all I have for now. Feel free to butcher / add / mod me up or down, but don't give me any "STFU N00B" responses for sympathizing with Micheal's cause. Apologies in advance if I misspelled some easy words, I'm typing this up on an old laptop with a small screen and crappy keyboard.
Slashdot Mirror
Pretty much any nVidia card since the TNT.
Let's see some decent Linux drivers from ATi and a benchmark showdown at Linuxhardware.org. Till then, it might be wise not to make such sweeping remarks to the Slashdot crowd.
If you want to run Linux, go with the GeForce.
I don't see why this particular set of benchmarks is special enough to deserve attention on Slashdot. Okay, so they run through a nice variety of benchmarks and they're a fairly credible source. But it has nothing to do with Linux, nothing to do with Apple, and not much to do with OpenGL. The benchmarks are all done in Windows XP with DirectX 9. Even the UT2003 benchmarks were done in DX9.
About the only thing I can tell about this set of benchmarks is that OpenGL and Linux are ignored completely. At least most other reviewers benchmark Quake 3, Serious Sam 2, or some synthetic OpenGL benchmark.
Hahahaha! Mod the parent up :)
I don't really see what's wrong with this approach. Using desktop managers as an example, some people simply want a full-featured (Or bloated, if you prefer) window manager like KDE so they have similar functionality to commercial OSes. Others may want a smaller one like WindowMaker. Either project would concievably be further along if the developers from one abandon their own projects and joined with the other.
That's assuming there wasn't a lot of internal bickering going on as to how things should be done, however. I think internal strife is far more dangerous and inhibiting than forking a project. The only way to make dozens, hundreds, or thousands of developers set their attitudes and egos aside for the sake of reaching a common goal is to offer them loads of cash.
I certainly like the idea of using effective air cooling like that rather than extreme / expensive methods like dipping the thing in fluorinert.
The only concern I have is servicing the boards. I've never worked with one of those racks before, but I imagine Linux Networx did a good job at making the boards fairly easy to swap out in case of node failure.
Here are some pics of the Evolocity 2 nodes in action for everyone to check out.
In many clusters today, each node relies on a hard disk to hold a distribution. This is a real big problem because hard disks generate a lot of heat. This is especially true with 10-15K RPM SCSI drives.
To add to that, most cluster node manufacturers that I know of design their 1U cases so that the hard disks are mounted on the front of the case for hot-swapping (Which in itself should suggest that people have problems with drives failing). The fans lie somewhere behind the hard disks. For example, this is a picture of an Appro 1122H dual Opteron server. Some cool air (But not a lot) is drawn from the vents on the front-right of the case, but a lot of hot air gets sucked away from the hard disks and blown over the CPUs. A better solution is just to take out the hot swap bays, make sure there's no SCSI backplane in the way, and let the cool air be drawn in from all across the front.
There are some manufacturers who will put a cardboard heat tunnel around the intake vent, lead it over some blowers, and eventually have to run over the CPUs. That's good design for keeping the CPUs cool, but it basically leaves the hard disks and other components such as north bridge chips and memory modules to fend for themselves when it comes to fighting thermal death. Often times they fail.
I have yet to see a hard disk that was smart enough to spin down to a lower speed when it reached a certain temperature. Even if they were that smart, who wants an HDD performing poorly when you've already spent ungodly amounts of money for bleeding edge 15K ultra-320 drives? As far as I'm concerned, they're practically built to destroy themselves. This is why diskless cluster nodes are so appealing to me.
Sorry for my little rant there. I hope it helped to clarify what I meant when I was emphasizing the advantages of having fewer moving parts.
Pink is the largest LinuxBIOS cluster in the world to date. The only moving parts in each node are cooling fans. This translates into a small savings on hardware, but more importantly means the nodes are by design more reliable thanks to fewer moving parts. Each node is powered by two 2.4GHz Xeon processors with four gigs of RAM and Myrinet 2000 interconnects.
The raw hardware power, while impressive, isn't what makes this cluster unique. The kicker is in the software, more specifically Clustermatic 3 featuring LinuxBIOS. Stuff happens and nodes fail, but thanks to LinuxBIOS they can be back up in a matter of seconds, not minutes.
Additional tools for the frontend node from Linux Networx makes updating nodes super-easy. You can flash each node's BIOS with a single command all in a matter of seconds. BProc allows you to run basic shell commands on any node without even installing a distribution on those nodes. w00t!
What we see here is a big shift away from expensive hardware and proprietary software. The software powering this cluster is 100% GPL, so users save a fortune in software licensing costs alone. And while these P4 nodes in particular aren't exactly cheap, they provide pretty darn good power and are far less expensive than Alpha servers. Also, using the x86 architecture means that consumer boards are not far behind in clustering. In fact, you can check out the LinuxBIOS homepage and see some pretty cheap boards that are supported already. So if you have some spare cash lying around and a couple weekends to kill, you can pick up a cheap board + cpu + memory combo and set it up as a slave node for your desktop machine with the same software these guys use to power this huge cluster.
I'd want Redhat and Mandrake to dispapear from that picture entirely. RPMs are simply beyond redemption at this point for me. Give me Portage or give me death!
It's one of the only times a company has made an in-house port of the game. Not just another 3D Tetris or something, but a real gamer's game.
Tribes and Heavy Gear certainly were great thanks to the efforts of Loki, don't get me wrong. And we have SDL, Loki_update, and some other great tools thanks to Loki.
However, when Loki went out of business, support for most of their games got dropped as well. You could not call up a company like Sierra and ask for Tribes 2 support for Linux. Patches were also discontinued. In the case of Tribes 2, it was just luck that Sam Lantinga was working for Sierra (Blizzard, to be more specific) and was allowed to patch the Linux version as well.
Bioware has made a commitment. They will always support Linux. They will, not a third party who may go out of business at some point while the game is still being developed. In-house porting is what makes the Linux port of NWN so signifficant.
(Shameless plug)
Head over to LinuxNetworx now for a LinuxBIOS-ready Evolocity II cluster!
Note: I'm not an employee of LinuxNetworx, but they still kick ass.
Any news on supporting those damned Broadcom 802.11 chipsets for wireless? Prism seems to have been phased out already :(
Nothing productive socially? I suppose I could do what most people my age do, much of which involves substance abuse, destroying property, etc, but instead I find joy in things that don't have (real world) death in the equation. Putting my enthusiasm for technology to use and gaming on-line w/ friends seems like a good choice.
I will agree with you if that person seriously does not have connections with anyone in his real life community. However, I do feel there are a signifficant number of people like myself who see video gaming as just another group activity to kill time like watching a movie, sports, etc. If you're going to just sit around and chat with friends, why not do it while conquering enemy territory and slaying opponents?
Define "convince." Anand's review shows that the FX5900 is indeed a worthy competitor for the latest ATi offering, and the superior Linux drivers put nVidia over the top for Linux users.
If you're concerned about nVidia's ethics, perhaps you should check out ATi's background.
Ah, that puts it in perspective. I wish the headline included that rather than overly generalizing it by saying stiff like "Help us shake it down in preparation for Mozilla 1.4 final."
Hardly. Below are two links that have video coverage of Zeta at CeBit2003:0 3.avit a-Prese ntation-CeBIT2003.avi
http://ddanneels.free.fr/Zeta-CeBIT20
http://gravity24hr.com/mirror/zeta/BeOS-Ze
I really wish Slashdot would stop covering incremental releases of software. While I'm not a Mozilla user, I'm guessing that Mozilla will go through at least two more release candidates before 1.4 final is out, which I do think should be covered. It's a bit silly to follow any project through all the release candidates.
Someone in another message pointed out that a billion dollars in software is after the price markup. One billion divided by the total cost of Office XP and Windows XP home retail, $678, comes in at just under 1.5 million bundles. If it costs MS $.05 to stamp out the bundles, and I would consider even that a conservative estimate for 2-4 CDs in mass production, that comes in at around $75,000. So they're spending about $75K so they can claim a billion dollars in donations. Wow! Take that 75K and divide it by 1 billion to get the percentage MS is spending compared to what they're claiming to give away:
Perhaps Redhat/Mandrake/SuSE/any distro with some cash should make a special edition of their Linux distro with an important sounding name, set the MSRP at $1000 per copy, stamp out tens of thousands of CDs and mail 'em out like AOL CDs to charities and claim to have given up tens or hundreds of millions.
I'm having trouble believing my own numbers, so if I made a math error please feel free to correct me.
Closed source drivers or not, nVidia's Linux performance rocks.
Of course, that's only among those that actually reviewed Linux support...
>> What's worse than that, though, is that they are still trying to pretend that it's not the case.
Since when? Jen-Hsun Huang admits defeat (But promises a comeback):
"Tiger Woods doesn't win every day. We don't deny that ATI has a wonderful product and it took the performance lead from us. But if they think they're going to hold onto it, they're smoking something hallucinogenic."
Let's not jump on nVidia too harshly for this. Sure, this spectacle seems to have gained a lot more publicity than ATi's own cheating ( link link link ). At least when nVidia cheated in 3DMark, they publically denounced synthetic benchmarks.
I think this might be what you're referring to:
"The R200 path has a slight speed advantage over the ARB2 path on the R300, but only by a small margin, so it defaults to using the ARB2 path for the quality improvements. The NV30 runs the ARB2 path MUCH slower than the NV30 path. Half the speed at the moment. This is unfortunate, because when you do an exact, apples-to-apples comparison using exactly the same API, the R300 looks twice as fast, but when you use the vendor-specific paths, the NV30 wins.
The reason for this is that ATI does everything at high precision all the time, while Nvidia internally supports three different precisions with different performances. To make it even more complicated, the exact precision that ATI uses is in between the floating point precisions offered by Nvidia, so when Nvidia runs fragment programs, they are at a higher precision than ATI's, which is some justification for the slower speed."
-John Carmack, Jan 29 2003
>>My question is why a closed, not free product gets a plug every time they put out a new release?
My question is why every time there's a new MacOS version people hail it as the evolution of Unix and such while other closed, not free software gets shunned.
A lot of people have come down pretty hard on Robertson for allowing people to use root by default. Here's an idea: Offer solutions instead of just bitching about the problem.
/usr/X11R6 under lock 'n key for "root" only, where modification will require a password, but chmod +777 /usr/local/ (Hopefully all users will be in the same group by default) and other such directories commonly used to install apps. Make sure that all packages in Click 'n Run do not try to write to directories that users are not given permission to.
The root issue is of serious concern since not only are files exposed to damage, but services can be run that can compromise security as well. Here's my idea:
1. Classify accounts by access level and make a graphical interface during bootup to set up user accounts. Don't give everyone root right off the bat, but give them something like "Administrator" or "poweruser." Make "root" a checkbox that's unchecked by default, label it with a warning and maybe draw some cute little skulls near by with a can of rat poison when you move your cursor over the checkbox. Have a "read more" option that explains, in gory detail, the risks of creating a user with root privileges and why it should only be used only for (insert creative name for) system tasks. If the user hasn't quit due to boredom after the first sentence and given up on "root" then at least he/she might understand why the root account is so special and tread with caution.
Anyway, back to the watered down root privs. Give these access levels high privileges that will be required for installing software later on (See part 2), but not enough to say, start a telnet server, ping flood someon, or rm -rf / . This will be the default permission. Also, place all users in the same group. Give them suggestions in a drop-down menu like "family" to make the process more recognizable.
Obviously, users aren't going to want to keep track of multiple passwords. And I just know I'm gonna get roasted for this, but make the root password the same as the password entered in for the first user. The idea here is to make the user think twice about an operation that he/she is going to do. When a user does something that requires root privs, ask for confirmation with a password. Maybe even have a little "help" button that explains that when the box pops up asking for confirmation, they're about to perform a task that could be potentially harmful to the system.
Of course this is less secure than it should be and having the same root and user password won't help much if a hacker manages to get in or something. However, for an OS with Micheal's intentions we have to cut the user a lot of slack, dreary as it may sound to all of us.
2. I know this would be enormously tedious and require tremendous effort, but I see it as part of the process of creating a distribution from scratch with Micheal's objective in mind. "Standardize" directories where stuff should go. Yep, that's it. Keep directories like
Unfortunately, my suggestions are crude at best and go off a few assumptions. For one, I assume that people will not try to upgrade extremely important things like KDE without reading some documentation first (How about some kinder, gentler documents that assume no Linux knowledge available from Lindows?). And like Micheal, I also assume that people aren't going to be using Lindows to make a network filesever or something delicate like that.
That's all I have for now. Feel free to butcher / add / mod me up or down, but don't give me any "STFU N00B" responses for sympathizing with Micheal's cause. Apologies in advance if I misspelled some easy words, I'm typing this up on an old laptop with a small screen and crappy keyboard.