Also remember that MP3 has it's roots in VCD technology. MP3 is a quasi-proprietary variant of MP2, which at 224Kb/s is the VCD audio format.
So, it makes sense that ISO VCD's are the next logical step. The quality is decent with a good encoder (320x240 w/NTSC though, NOT DVD-quality 640/720x480), the format is a heckuva lot more open than Real or MS video (the latter being encapsulated MPEG4), and encoders and players are available for Linux. (Some, in low-resolution and low-quality modes, can be run in real time.)
And of course the audio is good, too.:)
- Chad,
(who dosen't trade ISO VCD's though... it just makes sense that ppl _do_)
K6-2's at 350mhz and below are now being fire-saled for under $50 all over the place. The 350 went down first, and then (at least on NECX's components page) everything else below it dropped last week. At those prices, they're one heck of a deal.
(Also note that the Celeron is basically literally a PPro with MMX and less cache, and a smaller socket.;)
Nope... although Intel delayed Coppermine, it's going to make a Katmai based P3/600 sooner. The Coppermine P3/600 will be a lot nicer though (it'll run much cooler, faster, etc.)
The K-7 is MUCH faster at FP, and somewhat faster with integer stuff compared to the Katmai-based P3/600. I wouldn't be suprised to see the Coppermine P3/600 w/256K on-die cache match the K7 on integer though, especially if Intel picks up PC133 SDRAM.
(I wonder how many ppl who screamed 'Cyrix/AMD's FPUs suck!' will scream 'Intel's FPU sux!' after the K7 hits the street...:)
The Celeron is actually somewhat profitable because of it's good yields (which is pretty amazing since it's got 18 million transistors or so because of the cache.) And all of the R&D was paid for quite a while ago. And there's NO WAY they're not making money hand over fist with the P3, even at the 'reduced' pricing. They typically sell a chipset along with the new chips, too.:)
I'd bet the profit margin is MUCH lower on all those K6-2 2 million excess units last quarter didn't help.
So it's rather ambiguous... Intel IS running AMD into the ground with the Celeron, but is playing fair just enough to dodge antitrust issues. Their R&D and production resources allow this.
Also, it has to be noted that Intel provides more of a 'package' to it's OEMs... the 440BX is a tough chipset to compete with, and Intel will happily sell OEMs a variety of motherboards around their chipsets as well. (Including a model with Riva TNT's and SB64PCIs on 'em.) AMD can't provide a whole solution to it's OEMs. (VIA, in theory, will once they absorb Cyrix.)
AMD's best hopes are that the EV6-based motherboards wind up being very good, and that Intel continues to stick with Rambus (which could be the biggest screwup by an industry leader since IBM bungled the marketing of Micro Channel!)
Because... by the end of the year Intel might just have an improved 800mhz Coppermine P3 ready... if the K7 takes off. And AMD can't afford to be foiled again.
A lot of _really_ good stuff is coming from Taiwan... and I wouldn't be suprised to see a couple of the smaller CPU companies get swallowed up by the other bigger companies... (Acer for instance makes or assembles almost everything BUT CPUs, and maybe power supplies. If they had a CPU vendor they'd have more or less the same pieces to play with that Intel does...)
It oughta be interesting to see what happens with Cyrix/VIA now... VIA chipsets have been a bit slower than Intel ones (esp. slot 1), so the infusion of new IP might help. It's also good to see the Cyrix engineers not get laid off either... longer term C/V is probably going to be better off than AMD.
The other thing to look out for is at least one company to start selling a whole 'package' if you will... like Intel does. Some companies can basically buy entire PCs through Intel and slap their name on (or maybe even have Intel put their badge on it in their factories:) AMD does _NOT_ have this, VIA might with the help of some of their subsidiaries like FIC.
I've seen benchmarks where (at the same speed) the slower P2/P3 512K cache pulls ahead of the Celery's 128K cache in SMP, whereas in a uniproc environment at the same speed (and same FSB) they're usually even. Heck, mpg123 is a bit _faster_ (slightly) on a Celery:)
The M2 has the fastest integer performance per MHZ for it's class... if it ran at 400-500mhz it would be competitive. If it had an integrated 128K L2 cache at that speed it would probably eat up Celerys. Alas, the fastest one is only a 250mhz chip, and w/o integrated L2. At least they're cheap.
They _had_ David Cutler until they pushed him out in favor of yes-men during NT4 development because he fought the DirectX stuff (which is WAY too limited in NT4 anyway). Right now he's heading Win64 development.
Whatta shame. He'd never have let the NT5/W2K development situation get so bad... NT3.1 had problems, but they had it pretty well refined in NT3.5x, especially NT3.51.
Most of the Supra stuff I've seen on store shelves are simply branded Rockwell "Win"Modems anyway. Nothing special.:) (If I was to grab a -Modem for a Windozer (95/98 or NT) I'd get a Lucent PCI anyway.)
Something I was thinking about was a machine with a 16MB IDE flash disk, and a compressed ROM file system. This would let you put Linux and the essential stuff on (for something embedded only 2-4MB would actually be needed) and have it boot in a matter of moments. (1min) Also it might be possible to back up the system state to NVRAM or Flash ROM as well, further reducing start up time. But that would require a bit of coding effort.
Are they using *sleeve-bearing* fans or something? Out of anything other than a HUGE sample nothing like that should happen. (Especially if you use Intel Boxed CPUs with Sanyo fans or something like that.) And bad/noisy PS fans are good signs that the whole PS unit is junk, too.
Asus has a TNT board with video in/out, but no driver support outside of Win9X. The video decoder chip does have a Linux driver but the connection to the TNT chip is unknown (i.e. the I2C bus...) so it cannot be made to work.
Another good 2-slot combo (since the V2 takes a PCI slot, it's even) is a TNT(2) or G200 with a bt848/878/879 board like a Hauppauge WinTV, which interoperates with any video card you might grab in the future (it uses PCI and DGA to blast the TV onto the video card). That'll give you a video solution that is fully Free Software.
Thanks to the team working on Video4linux, the bt848 has been well supported for some time (unlike ATI boards), and is in kernel 2.2. And to Matrox for releasing enough of the G200 specs to make a driver, and to Nvidia for making a driver and releasing it with readable source.
What if Richard Stallman decides to change the language of the GPL (or is forced to?) If GPL v2 was challenged in court and found illegal for instance, mewonders if we'll see a GPLv3 that looks suspisiously like the XFree86 license...
Re:...Now we know why the RedHat about face re: KD
on
Red Hat Announces IPO
·
· Score: 1
Also note that the KDE (quite nice) used in RH6 still uses QT 1.44:) I had heard even when 5.9 came out that they'd have to pull KDE if they didn't have the QT2 version ready. Humm...
I lived in one where they wired everything with 10BaseT... through hubs with only five segments for about 150-200 ppl. Sniffers, collisions, and teardrops, oh my.
I remember the admins worried about Quake games. Alas there was no $ to do it the right way with switched ports.
Still, considering how bad the SJSU phone setup is, it was cool.
Slashdot Mirror
oops. I meant some _encoders_ can run in real-time. Any decent decoder will work in real time on a (P/K)6-level machine.
Also remember that MP3 has it's roots in VCD technology. MP3 is a quasi-proprietary variant of MP2, which at 224Kb/s is the VCD audio format.
So, it makes sense that ISO VCD's are the next logical step. The quality is decent with a good encoder (320x240 w/NTSC though, NOT DVD-quality 640/720x480), the format is a heckuva lot more open than Real or MS video (the latter being encapsulated MPEG4), and encoders and players are available for Linux. (Some, in low-resolution and low-quality modes, can be run in real time.)
And of course the audio is good, too. :)
- Chad,
(who dosen't trade ISO VCD's though... it just makes sense that ppl _do_)
K6-2's at 350mhz and below are now being fire-saled for under $50 all over the place. The 350 went down first, and then (at least on NECX's components page) everything else below it dropped last week. At those prices, they're one heck of a deal.
(Also note that the Celeron is basically literally a PPro with MMX and less cache, and a smaller socket. ;)
Nope... although Intel delayed Coppermine, it's going to make a Katmai based P3/600 sooner. The Coppermine P3/600 will be a lot nicer though (it'll run much cooler, faster, etc.)
The Pentium 60 and 66s were 5v space heaters, and barely faster than a DX4/100 at Integer ops. The P 75/90/100 chips were far better.
The K-7 is MUCH faster at FP, and somewhat faster with integer stuff compared to the Katmai-based P3/600. I wouldn't be suprised to see the Coppermine P3/600 w/256K on-die cache match the K7 on integer though, especially if Intel picks up PC133 SDRAM.
(I wonder how many ppl who screamed 'Cyrix/AMD's FPUs suck!' will scream 'Intel's FPU sux!' after the K7 hits the street... :)
The Celeron is actually somewhat profitable because of it's good yields (which is pretty amazing since it's got 18 million transistors or so because of the cache.) And all of the R&D was paid for quite a while ago. And there's NO WAY they're not making money hand over fist with the P3, even at the 'reduced' pricing. They typically sell a chipset along with the new chips, too. :)
I'd bet the profit margin is MUCH lower on all those K6-2 2 million excess units last quarter didn't help.
So it's rather ambiguous... Intel IS running AMD into the ground with the Celeron, but is playing fair just enough to dodge antitrust issues. Their R&D and production resources allow this.
Also, it has to be noted that Intel provides more of a 'package' to it's OEMs... the 440BX is a tough chipset to compete with, and Intel will happily sell OEMs a variety of motherboards around their chipsets as well. (Including a model with Riva TNT's and SB64PCIs on 'em.) AMD can't provide a whole solution to it's OEMs. (VIA, in theory, will once they absorb Cyrix.)
AMD's best hopes are that the EV6-based motherboards wind up being very good, and that Intel continues to stick with Rambus (which could be the biggest screwup by an industry leader since IBM bungled the marketing of Micro Channel!)
Because... by the end of the year Intel might just have an improved 800mhz Coppermine P3 ready... if the K7 takes off. And AMD can't afford to be foiled again.
A lot of _really_ good stuff is coming from Taiwan... and I wouldn't be suprised to see a couple of the smaller CPU companies get swallowed up by the other bigger companies... (Acer for instance makes or assembles almost everything BUT CPUs, and maybe power supplies. If they had a CPU vendor they'd have more or less the same pieces to play with that Intel does...)
It oughta be interesting to see what happens with Cyrix/VIA now... VIA chipsets have been a bit slower than Intel ones (esp. slot 1), so the infusion of new IP might help. It's also good to see the Cyrix engineers not get laid off either... longer term C/V is probably going to be better off than AMD.
The other thing to look out for is at least one company to start selling a whole 'package' if you will... like Intel does. Some companies can basically buy entire PCs through Intel and slap their name on (or maybe even have Intel put their badge on it in their factories :) AMD does _NOT_ have this, VIA might with the help of some of their subsidiaries like FIC.
I've seen benchmarks where (at the same speed) the slower P2/P3 512K cache pulls ahead of the Celery's 128K cache in SMP, whereas in a uniproc environment at the same speed (and same FSB) they're usually even. Heck, mpg123 is a bit _faster_ (slightly) on a Celery :)
The M2 has the fastest integer performance per MHZ for it's class... if it ran at 400-500mhz it would be competitive. If it had an integrated 128K L2 cache at that speed it would probably eat up Celerys. Alas, the fastest one is only a 250mhz chip, and w/o integrated L2. At least they're cheap.
Heck, even server-level SDRAM (ECC, CAS2) is going for ~$1/mb.
They _had_ David Cutler until they pushed him out in favor of yes-men during NT4 development because he fought the DirectX stuff (which is WAY too limited in NT4 anyway). Right now he's heading Win64 development.
Whatta shame. He'd never have let the NT5/W2K development situation get so bad... NT3.1 had problems, but they had it pretty well refined in NT3.5x, especially NT3.51.
Most of the Supra stuff I've seen on store shelves are simply branded Rockwell "Win"Modems anyway. Nothing special. :) (If I was to grab a -Modem for a Windozer (95/98 or NT) I'd get a Lucent PCI anyway.)
Looks like you'd like the XFree license then. Similar to BSD but with no advertising clause.
Something I was thinking about was a machine with a 16MB IDE flash disk, and a compressed ROM file system. This would let you put Linux and the essential stuff on (for something embedded only 2-4MB would actually be needed) and have it boot in a matter of moments. (1min) Also it might be possible to back up the system state to NVRAM or Flash ROM as well, further reducing start up time. But that would require a bit of coding effort.
Are they using *sleeve-bearing* fans or something? Out of anything other than a HUGE sample nothing like that should happen. (Especially if you use Intel Boxed CPUs with Sanyo fans or something like that.) And bad/noisy PS fans are good signs that the whole PS unit is junk, too.
For instance, Debian 2.1 spent 6-months in freeze before release.
Asus has a TNT board with video in/out, but no driver support outside of Win9X. The video decoder chip does have a Linux driver but the connection to the TNT chip is unknown (i.e. the I2C bus...) so it cannot be made to work.
Another good 2-slot combo (since the V2 takes a PCI slot, it's even) is a TNT(2) or G200 with a bt848/878/879 board like a Hauppauge WinTV, which interoperates with any video card you might grab in the future (it uses PCI and DGA to blast the TV onto the video card). That'll give you a video solution that is fully Free Software.
Thanks to the team working on Video4linux, the bt848 has been well supported for some time (unlike ATI boards), and is in kernel 2.2. And to Matrox for releasing enough of the G200 specs to make a driver, and to Nvidia for making a driver and releasing it with readable source.
QT with the old license ships as part of the base RH6.0 set.
What if Richard Stallman decides to change the language of the GPL (or is forced to?) If GPL v2 was challenged in court and found illegal for instance, mewonders if we'll see a GPLv3 that looks suspisiously like the XFree86 license...
Also note that the KDE (quite nice) used in RH6 still uses QT 1.44 :) I had heard even when 5.9 came out that they'd have to pull KDE if they didn't have the QT2 version ready. Humm...
They aren't losing that much money, and I'm sure the sales of $80 RH6 kit will make up for it :)
Seriously, considering how much money their IPO is gonna bring in, they could last for a long time losing $100K/year.
I lived in one where they wired everything with 10BaseT... through hubs with only five segments for about 150-200 ppl. Sniffers, collisions, and teardrops, oh my.
I remember the admins worried about Quake games. Alas there was no $ to do it the right way with switched ports.
Still, considering how bad the SJSU phone setup is, it was cool.
Some ISPs are linking themseleves with ADSL HW providers, so they're in the revenue stream too.