Strange, the "Linux on Large Systems Foundry" link doesn't seem to indicate any problems. Hemos, what work is still to be done? The fact that there is continuing development doesn't surprise me, but don't make it sounds like "Linux isn't ready for the mainframe". The only posting in the scalability forum is "asdfasdf".
On the other hand, this looks like a great portal for Linux on mainframe users, with news and a 'library' of information/links on high-availability, parallel programming, shared memory, and SMP, among other topics.
In the article they mention that one of the $400,000 servers can replace 'hundreds of servers' and that their $400k is comparable to an average mainframe cost of $750k.
Hmm. If you have a voice modem it wouldn't be too hard to just call 911 and play an audio file "This computer was stolen, blah blah blah" through the modem. Don't forget the AT command to turn off the speaker first.
Agurkan was saying (I think) that if you design a cluster with a certain Gflop budget, you can achieve the same net performance with approximately 10% less P4 nodes if those P4 nodes are 10% faster.
So maybe there's other specialty cluster hardware that drives up the per-box cost, but if you start talking about dual-cpu boxes, it gets even worse...
Ten nodes with $600 (2 x Athlon 2000+) and $1500 Gb NIC = $21,000
Nine nodes with $1200 (2 x P4 2.2 GHz) and $1500 Gb NIC = $24,300
for comparable net performance.
I guess if you're trying to break a speed record, a 10% speed gain might be important even if the price doubles, but if you have to be on the 'bleeding edge' then you're probably not concerned with $$$ anyway.
That's pretty good! I'd take out the 'hard-coded' reference to an arbitrary US Dollar amount, though. $100 million might sound about right for a really punishing fine today, but maybe $100 sounded reasonable for a fine back in the 19th century. These documents have staying power.
Q: Won't this protection interfere with the concept of "fair use"?
A: We are committed to protecting the rights of our artists and copyright holders.
Q: Yeah, but with official Phillips CD's I can make legal copies for personal use. It's convenient. Why can't I do that with your new discs?
A: We are committed to protecting the rights of our artists and copyright holders...
Well if I got anything out of reading the benchmarks, it's that AMD is dead on with their "Athlon XP 2000+" marketing game. I wasn't sure I liked the idea, after reading benchmarks where the 1.67 GHz AMD chip ran neck-and-neck with the 2.0 GHz Northwood, I have to concede that it seems accurate.
Bottom line? This might be healthy competition between AMD and Intel. Let's hope these two continue to push each other higher than they might climb on their own.
Well, with sufficiently small characters, it should look fine. In fact, as you approach a 1-pixel font it should look quite nice. Add color and you're there. Oh, wait...
"(the operating system) will be QNX, a real time OS that supports massive parallelism and has very low overhead. QNX is fast! QNX is also Posix compliant, so there is lots of software that almost works under it."
If you're looking for software that almost works, I know of an OS that might fit your needs. You're not going to hook this thing up to the Internet, though, are you?
Give me some credit. But to be 'anthropomorphosized' as you put it, there has to be a perceivable reaction to a stimulus. For fear of light, I suppose a photodiode could turn on the 'turning' circuit or the 'reverse' circuit. I would guess that most of the smarts are probably sensors that turn on/off mechanical drive sections in the toy.
Well, that's true, but I think they're getting at the component count/complexity rather than the technicality of what makes a computer.
I think I saw a commercial for these bugs, and IIRC they would 'seek out others of their kind' and were generally portrayed as being intelligent. My first thought was that there's a computer inside. I'd like to know how he could have programmed responses like "fear of light" with a handful of discrete transistors.
From the "Making the electronics" section (emphasis mine, of course):
Solder remaining parts into the transmitter. Put the three 74HC04's in stack (like they are fucking), and solder pins of equal numbers together. The schematic follows.
at PC AV Tech website. A SB16 with 75 dB S/N and 74 dB dynamic range isn't acceptable to me played though a consumer stereo amplifier and decent speakers. It's probably good enough for the tinny 10-watt speakers that come with most computers these days, and I'm sure that's what Creative was targeting.
But there's some middle ground between 'fanless net-booting computers' and wanting to hear decent sound.
Re:MP 1900+ same as XP 1900+
on
Athlon MP Reviewed
·
· Score: 2, Interesting
I've heard there are 760MPX boards coming out (with 66 MHz support for the 64-bit PCI slots; is there anything else?) but I haven't seen any yet. Anyone see any non-sketchy details anywhere? The timing I've read about was mid-November. Anand Tech probably had the best article with pictures of some boards. Here's a link to his Preview from Comdex.
Ok, I'll back down a notch and admit it's listed as "(EXPERIMENTAL)" in the kernel configuration dialogs. But I still don't think that's reason enough to break something that once worked.
Have you tried the FireWire stuff? I see "Firewire updates" in the ChangeLog, but then again I've seen ieee1394 updates since 2.4.6 that didn't work. Of course I'll find out for myself soon enough.
Not to flame (maybe I'm just cranky today) but I thought 2.4 was the stable version tree. I've been stuck back on 2.4.6 because every later version has locked up (can't even telnet in) when I try to use the ieee1394 system.
I know what I'll hear, 'fix it yourself, it's open source' etc. But seriously, how does non-working code get released as a 2.4.x series kernel?
What if you bought an un-approved motherboard and found out that it in fact does hide the MHz on boot-up?
Point being, there could be other reasons why you'd want to avoid an unapproved motherboard. Perhaps motherboard manufacturers that fail AMD qualification will simply display MHz on boot-up and let people make their inferences...
Very interesting, and I'll probably try it if I can ever find the time, but one thing struck me as odd. The project is called "Linux from scratch" yet one of the first things in the LFS book is "We are going to build the LFS system by using an already installed Linux distribution such as Debian, SuSe, Slackware, Mandrake, RedHat, etc."
Yeah, I see where they're going, this is just to create the filesystem and kickstart everything, but still, I thought they were promoting the 'from scratch' concept.:)
I'm not sure they were marketing to audiophiles. I think it was more like the mp3 crowd. And anyone who hates lugging CD's around. I don't change the CD's in my trunk-mounted 6-disc changer nearly as often as I'd like to because of the inconvenience.
And while I don't claim to be an audio engineer or anything, I thought that the MP3 format was tested with audio experts versus CD in double-blind tests on systems far better most of us could afford, and they couldn't tell the difference. So your statement that mp3's sound like crap seems naive.
I mean, sure, MP3's sound like crap played through a $20 sound card on tinny speakers, but then so would a 16-bit 44.1 kHz.wav file. If CD audio sounds better on a system, it's probably because the CDROM's audio output bypasses a cheesy D/A converter on the sound card.
I'll bet very few of us could reliably tell the difference between a good MP3 encoding and 16-bit 44.1 kHz in a blind test.
I remember signing up and then getting an e-mail when they were ready. Signing up was a cool idea, but when push came to shove I didn't want to blow that much money on a car stereo. And it wasn't even that it was so expensive, it was that I thought the units would get much cheaper (probably the week after I bought mine). Looks like they didn't... yet.;)
Slashdot Mirror
Strange, the "Linux on Large Systems Foundry" link doesn't seem to indicate any problems. Hemos, what work is still to be done? The fact that there is continuing development doesn't surprise me, but don't make it sounds like "Linux isn't ready for the mainframe". The only posting in the scalability forum is "asdfasdf".
On the other hand, this looks like a great portal for Linux on mainframe users, with news and a 'library' of information/links on high-availability, parallel programming, shared memory, and SMP, among other topics.
In the article they mention that one of the $400,000 servers can replace 'hundreds of servers' and that their $400k is comparable to an average mainframe cost of $750k.
Hmm. If you have a voice modem it wouldn't be too hard to just call 911 and play an audio file "This computer was stolen, blah blah blah" through the modem. Don't forget the AT command to turn off the speaker first.
Agurkan was saying (I think) that if you design a cluster with a certain Gflop budget, you can achieve the same net performance with approximately 10% less P4 nodes if those P4 nodes are 10% faster.
So maybe there's other specialty cluster hardware that drives up the per-box cost, but if you start talking about dual-cpu boxes, it gets even worse...
Ten nodes with $600 (2 x Athlon 2000+) and $1500 Gb NIC = $21,000
Nine nodes with $1200 (2 x P4 2.2 GHz) and $1500 Gb NIC = $24,300
for comparable net performance.
I guess if you're trying to break a speed record, a 10% speed gain might be important even if the price doubles, but if you have to be on the 'bleeding edge' then you're probably not concerned with $$$ anyway.
Think about that.
Ten nodes with $300 Athlon 2000+ and $1500 Gb NIC = $18,000.
Nine nodes with $600 P4 2.2 GHz and $1500 Gb NIC = $18,900.
Where is that savings?
That's pretty good! I'd take out the 'hard-coded' reference to an arbitrary US Dollar amount, though. $100 million might sound about right for a really punishing fine today, but maybe $100 sounded reasonable for a fine back in the 19th century. These documents have staying power.
Q: Won't this protection interfere with the concept of "fair use"?
A: We are committed to protecting the rights of our artists and copyright holders.
Q: Yeah, but with official Phillips CD's I can make legal copies for personal use. It's convenient. Why can't I do that with your new discs?
A: We are committed to protecting the rights of our artists and copyright holders...
Yeah, I read your response.
Interesting that in your opinion, "rude is rude," but we should be forgiven your IRC "chat antics" where you "blow off steam."
I smell a double standard.
Well if I got anything out of reading the benchmarks, it's that AMD is dead on with their "Athlon XP 2000+" marketing game. I wasn't sure I liked the idea, after reading benchmarks where the 1.67 GHz AMD chip ran neck-and-neck with the 2.0 GHz Northwood, I have to concede that it seems accurate.
Bottom line? This might be healthy competition between AMD and Intel. Let's hope these two continue to push each other higher than they might climb on their own.
Well, with sufficiently small characters, it should look fine. In fact, as you approach a 1-pixel font it should look quite nice. Add color and you're there. Oh, wait...
"(the operating system) will be QNX, a real time OS that supports massive parallelism and has very low overhead. QNX is fast! QNX is also Posix compliant, so there is lots of software that almost works under it."
If you're looking for software that almost works, I know of an OS that might fit your needs. You're not going to hook this thing up to the Internet, though, are you?
Give me some credit. But to be 'anthropomorphosized' as you put it, there has to be a perceivable reaction to a stimulus. For fear of light, I suppose a photodiode could turn on the 'turning' circuit or the 'reverse' circuit. I would guess that most of the smarts are probably sensors that turn on/off mechanical drive sections in the toy.
Well, that's true, but I think they're getting at the component count/complexity rather than the technicality of what makes a computer.
I think I saw a commercial for these bugs, and IIRC they would 'seek out others of their kind' and were generally portrayed as being intelligent. My first thought was that there's a computer inside. I'd like to know how he could have programmed responses like "fear of light" with a handful of discrete transistors.
From the "Making the electronics" section (emphasis mine, of course):
Solder remaining parts into the transmitter. Put the three 74HC04's in stack (like they are fucking), and solder pins of equal numbers together. The schematic follows.
Nice.
at PC AV Tech website. A SB16 with 75 dB S/N and 74 dB dynamic range isn't acceptable to me played though a consumer stereo amplifier and decent speakers. It's probably good enough for the tinny 10-watt speakers that come with most computers these days, and I'm sure that's what Creative was targeting.
But there's some middle ground between 'fanless net-booting computers' and wanting to hear decent sound.
I've heard there are 760MPX boards coming out (with 66 MHz support for the 64-bit PCI slots; is there anything else?) but I haven't seen any yet. Anyone see any non-sketchy details anywhere? The timing I've read about was mid-November. Anand Tech probably had the best article with pictures of some boards. Here's a link to his Preview from Comdex.
Maybe it's a new Google site:
http://pRon.google.com!!
Ok, I'll back down a notch and admit it's listed as "(EXPERIMENTAL)" in the kernel configuration dialogs. But I still don't think that's reason enough to break something that once worked.
Sigh, all in the name of progress...
Have you tried the FireWire stuff? I see "Firewire updates" in the ChangeLog, but then again I've seen ieee1394 updates since 2.4.6 that didn't work. Of course I'll find out for myself soon enough.
Not to flame (maybe I'm just cranky today) but I thought 2.4 was the stable version tree. I've been stuck back on 2.4.6 because every later version has locked up (can't even telnet in) when I try to use the ieee1394 system.
I know what I'll hear, 'fix it yourself, it's open source' etc. But seriously, how does non-working code get released as a 2.4.x series kernel?
What if you bought an un-approved motherboard and found out that it in fact does hide the MHz on boot-up?
Point being, there could be other reasons why you'd want to avoid an unapproved motherboard. Perhaps motherboard manufacturers that fail AMD qualification will simply display MHz on boot-up and let people make their inferences...
Very interesting, and I'll probably try it if I can ever find the time, but one thing struck me as odd. The project is called "Linux from scratch" yet one of the first things in the LFS book is
:)
"We are going to build the LFS system by using an already installed Linux distribution such as Debian, SuSe, Slackware, Mandrake, RedHat, etc."
Yeah, I see where they're going, this is just to create the filesystem and kickstart everything, but still, I thought they were promoting the 'from scratch' concept.
I think it's worth pointing out that the linked article reads:
...a discovery that might lead to semiconductor-based quantum computers.
Emphasis mine.
I'm not sure they were marketing to audiophiles. I think it was more like the mp3 crowd. And anyone who hates lugging CD's around. I don't change the CD's in my trunk-mounted 6-disc changer nearly as often as I'd like to because of the inconvenience.
.wav file. If CD audio sounds better on a system, it's probably because the CDROM's audio output bypasses a cheesy D/A converter on the sound card.
And while I don't claim to be an audio engineer or anything, I thought that the MP3 format was tested with audio experts versus CD in double-blind tests on systems far better most of us could afford, and they couldn't tell the difference. So your statement that mp3's sound like crap seems naive.
I mean, sure, MP3's sound like crap played through a $20 sound card on tinny speakers, but then so would a 16-bit 44.1 kHz
I'll bet very few of us could reliably tell the difference between a good MP3 encoding and 16-bit 44.1 kHz in a blind test.
I remember signing up and then getting an e-mail when they were ready. Signing up was a cool idea, but when push came to shove I didn't want to blow that much money on a car stereo. And it wasn't even that it was so expensive, it was that I thought the units would get much cheaper (probably the week after I bought mine). Looks like they didn't... yet. ;)
Shouldn't 1394b ("double FireWire") be FireFireWireWire and not just FireFire?