I have a better reason: AMD can fab those CPUS easily and IBM has better things to do than fab chips for Macs. Apple needs to make sure it doesn't have to stop the assembly line for IBM to fab more CPUs. My guess is that Apple will have AMD produce IBM 970 chips alongside IBM. IBM probably doesn't want to be the first in line for Macintosh CPUs, there's not enough money in it for a multi-faceted operation like IBM. AMD can produce ample chips and they might be able to make a profit doing it.
This has nothing to do with Macs and x86, AMD produces a LOT more than just athlon chips, they'll be pumping out AMD-970s with their extra capacity.
That's what we get for hiring the two biggest aviation companies to handle the shuttle for us. They'r more interested in bleeding money from the shuttle program than developing a more efficient way to get the job done. Why would they want to produce a shuttle that let them collect less from NASA?
Anyone looking at the framebuffer device as salvation is barking up the wrong tree. Yes, its direct to hardware, but it's slow as molasses when there's anything happening on the screen. Try running a console-intensive app on a fbconsole and watch the latency of all your apps go through the roof. I can't even play an MP3 when the framebuffer is busy, and I have a top-end system with optimized software.
XFree86 has direct-to-hardware rendering, it has DRI and MITSHM. Granted it uses a few more megs of memory than I'd like it too, it's our best bet for 2D performance out there. It really is quite good, I think a lot of the desktop environments need optimization (KDE is a bad performer all-around, IMO).
The framebuffer exists for three reasons:
1. Simple to program for. 2. Needed for 68k and most PPC machines (no native console mode) 3. Platform Independent.
it doesnt:
1. Perform well enough for casual use. 2. Play well with others. 3. Run X programs without the overhead of X anyway.
Re:LIKE I WOULD TAKE SCSI ADVICE FROM AN AOL USER
on
Linux SMP Round-Up
·
· Score: 1
HAHA! The AOL account is just so my clients can reach me, most of them are AOL users and they have trouble with email outside of AOL (really). I don't have AOL here at home, I 'gave' the account to my dad and switched my apartment over to 100% Linux systems.
The simplicity of having clients reach me @aol.com pays for itself.
Don't forget that there's also an 800MHz G3 chip (based on the 750FX) for those of us who don't need Altivec (server folks, mostly).
PowerLogix has managed to get the latest (last?) of the G3s onto a ZIF for us. I can't wait to get this cool-running power-efficient CPU under the hood of my server. It has an integrated 512K full-speed cache instead of a backside cache (big but slower).
The G3 is a great chip for Linux servers and workstations that don't do video-editing or use AltiVec extensively.
FC and SSA not of value in single-disk systems
on
Linux SMP Round-Up
·
· Score: 1
Is it for workstation use? We're not talking about arrays of disks here, just replacing your IDE drive with a SCSI drive. AFAIK the limiting factor in the disk subsystem is the disk itself.
In a single-disk system it makes (almost) no difference if you use Ultra2Wide, Ultra160, or Ultra320 busses, because the most you're going to get out of the disk is about 60MB/sec. I can't see how SSA or FC would help at all unless you had enough disks RAIDed to hose the bus.
The same is true for IDE busses. UltraATA-66 and UltraATA-133 show less than 10% real-world performance difference when using the same disk, even though the bus is twice as fast with UATA133.
Why go elsewhere? I Install NEC monitors at banks day-in and day-out. One out of ten of the monitors we recieve are DOA and 40% that have been in service over two years are too dim or blurry for professional use. The client wants to keep them because they're so cheap (we get fe750 CRTs for about $120 each).
I personally like Hitachi and Sony monitors. I'd kill for a LaCie ElecronBlue CRT though.
I thought I was the only one. I didn't see it just to spite all the lusers who did.
Check your chipset drivers [offtopic]
on
Linux SMP Round-Up
·
· Score: 2, Informative
I'm running Gentoo Linux 1.4rc1. Everything is built from scratch with optimizations so it's as fast as can be on both machines. I'm running kernel 2.4.20 on both machines.
I should note that the SCSI performance boost is still huge in Windows, but less profound than in Linux due to the way Windows aligns frequently used files on the disk.
As for your performance issues, try updating the drivers for your chipset (Intel INF and Intel Application Accelerator / VIA Hyperion 4-in-1) to make sure you're getting the most from your motherboard.
Your answer is not in the CPU
on
Linux SMP Round-Up
·
· Score: 2, Informative
'Most responsive user experience'? Switch to SCSI. The major bottleneck in any PC is the crappy disk access. I get better app start times on my 400Mhz U2W SCSI system (80MB/sec max) than my Athlon 1400 with ATA-133. The SCSI theoretical speed limit might be lower (in the example above), but real-world performance favors SCSI.
Go get an Adaptec 29160 and a 36GB 10K Cheetah drive for your / and/usr partitions. Put/home on your IDE drive. Get the best of both worlds. When you recover from the investment you can move the whole SCSI deal to the next machine (and it'll STILL kick the next generation PCs ass!).
You probably haven't USED a PPC/Linux box. They make great servers. Here are my reasons:
The G3 and G4 have caches comparable to most Xeon chips, making them a happy medium between consumer CPUs and expensive server CPUs. An Apple blue+white G3 with a SCSI card is blazing-fast for most server purposes, much faster IMO than a comparably clocked PC.
The G3 uses very little power and does not need a fan on its small heatsink, making it ideal for 24/7 uptime. High power-use not only costs money directly, but it costs money to cool off server rooms and offices. Less moving parts (fans) means less downtime too (for both cleaning and repair)
OpenFirmware lets you have quite a bit of control over how your machine boots, your machine doesn't need to bootstrap into protected mode, it's native 32-bit from the ground-up. If there's no monitor connected I can pump my POST screen ove rthe serial port, that's GREAT for network management.
Great compiler support from the GCC team. There's targets for the enire line of PowerPC products.
Straightforward architecture. I'm not a programmer, but it seems to me that x86 has a lot of hacked and bastardized components added to it to make it both fast and backwards-compatable. PowerPC is fast and backward compatable from the get-go, true RISC means that there's very little change in the inner workings of the CPU, there's not a new set of instructions for the compilers and developers to catch up to every year.
PowerPC has it's roots in large high-quality systems (RS/6000 servers). x86 has it's roots in the lowest-end computers of it's day (it was designed for the PC, which is the bottom of the line for computer vendors).
You can FEEL when you're running on a better architecture. The system doesn't ever feel 'laggy' even when it gets bogged down, performance is much better than a similar spec'ed PC, and the cabinet the case is in stays cool all the time.
I use an Athlon for my workstation because it's cheap and fun to hack around with, but my servers are PowerPC and I've never even thought of moving them to x86.
...and ONTO an internal USB bus!
on
Legacy-Free PCs
·
· Score: 1
I support the removal of the floppy from the system, along with the serial and parallel ports. I think that floppy support should be piped over USB, and there oughtta be an internal USB header in every PC. You could get legacy ports/drives on 'legacy-free' PCs with an internal USB bus (should you need them).
How cool would it be if there was a 3.5" USB hot-swap bay on every PC that could take a floppy or flashcard reader? You could stick serial-ata connectors in there too for a general-purpose hot-swap bay instead of a rickety old 1980s floppy device.
Move them off the MOBO please!
on
Legacy-Free PCs
·
· Score: 1
And most people don't need wheelbarrows anymore.
It's time to get rid of the 'geek' connectors on PCs. If you want legacy ports you can get an add-in card with them or a USB shim.
While we're at it, is there a chance we can get a new kind of floppy controller? I think it's lame how my Athlon slows to a crawl when I'm formatting a floppy. Maybe serializing floppy access and adding a 32K buffer would fix things.
Right, now instead of paying for a $1000 roll of real film you'll have to pay MS $995 for the right to use their 'WM9 theater encoder deluxe'. (prices are completely made up)
I think the major hurdle for the independent film industry is the vast droves of morons who would rather pay $9 to see a bunch of computer-rendered explosions than $5 to watch a well thought-out film.
That's exactly my problem. I've become too used to firing up three or for apps at once and letting the kernel sort out the resource management. I'm not firing up more apps than I need, I can USE all those apps. Something about making myself home in UNIX-land has made me content to have 10-15 open windows on different desktops waiting for input. Why not have OpenOffice, XMMS, Mozilla, Konqueror and WindowMaker all fire up when I log in? That way I don't have to individually open something and close it; it's all open, all the time!
The trouble is when I try to do the same thing in Windows, I have to stretch out the taskbar to three rows and when I push the system too far it kills all the unsaved data in ALL of those windows! I don't know if I'll ever be able to go back.
What you say is only MOSTLY true. I know I break any win9x box within 5 minutes of logging in just because I'm too speedy for it, the 9x kernel and I do NOT get along. WinNT/2K/XP seem to be alright with me, and the Linux Kernel LOVES me, I've only caused Linux to become unresponsive twice in my life (and both were because I was TRYING to).
Mac OS 9 is also crashy for me, I kill it right off the bat (even a clean install). My dad, who has a more linear approach to using the computer, can go weeks without a single crash on the same Mac.
I find myself more and more of a liability on older systems, I just make them crash too much, does anybody else have this problem?
What irks me is that there's not a -march=c3 target in recent GCC releases. The C3 currently works best when you use '-m486 -m3dnow -mmmx' which is nasty. VIA needs to kick a GCC developer a few thousand and a few books so GCC can get a proper target for VIA's products. Until there's proper scheduling and cache-management for this processor (on the compiler end) everything is gonna feel REAL SLOW on it.
That's what I'm looking for. There's got to be a way to mount that thing in my G3 B&W, maybe with a ZIF adapter card. I just want a 750 with a giant or superfast cache, I'll underclock it if I have to. If I can't get that I'll wait for the Pegasos PPC board and go Apple-free.
I had too many headaches getting something like that working with Mozilla, the file system differences really nail you, paths are all different and stuff.
What I'd like to see is an option to add 'windows' and 'unix' options to the mail paths in prefs.js so when I run in linux the profile comes from '~/.mozilla...' and when I run in Windows it picks up from 'U:\.mozilla...' even though those are really the same folder.
What I wouldn't give to have a recent-design G3 in my old blue-n-white.
I wish someone would produce new G3s with giant integral and backside caches, they would absolutely fly on low-end and midrange servers where altivec is nothing more than extra heat to dissipate.
Anyone know of a way to get one of those swanky new 750CXe G3s from the latest iBook into an older blue-and-white? They're different pinouts and voltages from my POV.
I thought the G4 FP improvements were because it intrinsically routed some of those ops through the AltiVec units.
AFAIK, the G4 was a G3 with SMP circuitry and AltiVec units, at least in it's first incarnation. It's been a while since it debuted though, and the evidence that the G4 has become more streamlined since branching from the G3 is good (if they are the same 'core' the G3 would be able to clock much higher since it's smaller).
That's what I thought the situation was. I know a lot of printed graphics folks who jacked up their working resolution to 6000 DPI just to feel important and to justify their own time expenditure. I tried explaining once that going above the printer's resolution was a waste of time and I was quickly hushed and told that that's what the new PCs were for.
Slashdot Mirror
I have a better reason: AMD can fab those CPUS easily and IBM has better things to do than fab chips for Macs. Apple needs to make sure it doesn't have to stop the assembly line for IBM to fab more CPUs. My guess is that Apple will have AMD produce IBM 970 chips alongside IBM. IBM probably doesn't want to be the first in line for Macintosh CPUs, there's not enough money in it for a multi-faceted operation like IBM. AMD can produce ample chips and they might be able to make a profit doing it.
This has nothing to do with Macs and x86, AMD produces a LOT more than just athlon chips, they'll be pumping out AMD-970s with their extra capacity.
That's what we get for hiring the two biggest aviation companies to handle the shuttle for us. They'r more interested in bleeding money from the shuttle program than developing a more efficient way to get the job done. Why would they want to produce a shuttle that let them collect less from NASA?
Anyone looking at the framebuffer device as salvation is barking up the wrong tree. Yes, its direct to hardware, but it's slow as molasses when there's anything happening on the screen. Try running a console-intensive app on a fbconsole and watch the latency of all your apps go through the roof. I can't even play an MP3 when the framebuffer is busy, and I have a top-end system with optimized software.
XFree86 has direct-to-hardware rendering, it has DRI and MITSHM. Granted it uses a few more megs of memory than I'd like it too, it's our best bet for 2D performance out there. It really is quite good, I think a lot of the desktop environments need optimization (KDE is a bad performer all-around, IMO).
The framebuffer exists for three reasons:
1. Simple to program for.
2. Needed for 68k and most PPC machines (no native console mode)
3. Platform Independent.
it doesnt:
1. Perform well enough for casual use.
2. Play well with others.
3. Run X programs without the overhead of X anyway.
HAHA! The AOL account is just so my clients can reach me, most of them are AOL users and they have trouble with email outside of AOL (really). I don't have AOL here at home, I 'gave' the account to my dad and switched my apartment over to 100% Linux systems.
The simplicity of having clients reach me @aol.com pays for itself.
I'm interested in embedding WhiteTrash into Firebird. I bet it'll go real fast. Hope it doesn't crash much.
Don't forget that there's also an 800MHz G3 chip (based on the 750FX) for those of us who don't need Altivec (server folks, mostly).
PowerLogix has managed to get the latest (last?) of the G3s onto a ZIF for us. I can't wait to get this cool-running power-efficient CPU under the hood of my server. It has an integrated 512K full-speed cache instead of a backside cache (big but slower).
The G3 is a great chip for Linux servers and workstations that don't do video-editing or use AltiVec extensively.
Is it for workstation use? We're not talking about arrays of disks here, just replacing your IDE drive with a SCSI drive. AFAIK the limiting factor in the disk subsystem is the disk itself.
In a single-disk system it makes (almost) no difference if you use Ultra2Wide, Ultra160, or Ultra320 busses, because the most you're going to get out of the disk is about 60MB/sec. I can't see how SSA or FC would help at all unless you had enough disks RAIDed to hose the bus.
The same is true for IDE busses. UltraATA-66 and UltraATA-133 show less than 10% real-world performance difference when using the same disk, even though the bus is twice as fast with UATA133.
Why go elsewhere? I Install NEC monitors at banks day-in and day-out. One out of ten of the monitors we recieve are DOA and 40% that have been in service over two years are too dim or blurry for professional use. The client wants to keep them because they're so cheap (we get fe750 CRTs for about $120 each).
I personally like Hitachi and Sony monitors. I'd kill for a LaCie ElecronBlue CRT though.
I thought I was the only one. I didn't see it just to spite all the lusers who did.
I'm running Gentoo Linux 1.4rc1. Everything is built from scratch with optimizations so it's as fast as can be on both machines. I'm running kernel 2.4.20 on both machines.
I should note that the SCSI performance boost is still huge in Windows, but less profound than in Linux due to the way Windows aligns frequently used files on the disk.
As for your performance issues, try updating the drivers for your chipset (Intel INF and Intel Application Accelerator / VIA Hyperion 4-in-1) to make sure you're getting the most from your motherboard.
Intel Chipset Driver Matrix
VIA Hyperion Downloads
'Most responsive user experience'? Switch to SCSI. The major bottleneck in any PC is the crappy disk access. I get better app start times on my 400Mhz U2W SCSI system (80MB/sec max) than my Athlon 1400 with ATA-133. The SCSI theoretical speed limit might be lower (in the example above), but real-world performance favors SCSI.
/usr partitions. Put /home on your IDE drive. Get the best of both worlds. When you recover from the investment you can move the whole SCSI deal to the next machine (and it'll STILL kick the next generation PCs ass!).
Go get an Adaptec 29160 and a 36GB 10K Cheetah drive for your / and
For young folks with good eyes I'd say to go for the 12" screen. AFAIK, The resolution is the same on the 12" and 14" inch screens.
The 12" ibook is smaller and lighter, and IMO it looks better (the 14" model seems dwarfed by it's own screen).
You probably haven't USED a PPC/Linux box. They make great servers. Here are my reasons:
The G3 and G4 have caches comparable to most Xeon chips, making them a happy medium between consumer CPUs and expensive server CPUs. An Apple blue+white G3 with a SCSI card is blazing-fast for most server purposes, much faster IMO than a comparably clocked PC.
The G3 uses very little power and does not need a fan on its small heatsink, making it ideal for 24/7 uptime. High power-use not only costs money directly, but it costs money to cool off server rooms and offices. Less moving parts (fans) means less downtime too (for both cleaning and repair)
OpenFirmware lets you have quite a bit of control over how your machine boots, your machine doesn't need to bootstrap into protected mode, it's native 32-bit from the ground-up. If there's no monitor connected I can pump my POST screen ove rthe serial port, that's GREAT for network management.
Great compiler support from the GCC team. There's targets for the enire line of PowerPC products.
Straightforward architecture. I'm not a programmer, but it seems to me that x86 has a lot of hacked and bastardized components added to it to make it both fast and backwards-compatable. PowerPC is fast and backward compatable from the get-go, true RISC means that there's very little change in the inner workings of the CPU, there's not a new set of instructions for the compilers and developers to catch up to every year.
PowerPC has it's roots in large high-quality systems (RS/6000 servers). x86 has it's roots in the lowest-end computers of it's day (it was designed for the PC, which is the bottom of the line for computer vendors).
You can FEEL when you're running on a better architecture. The system doesn't ever feel 'laggy' even when it gets bogged down, performance is much better than a similar spec'ed PC, and the cabinet the case is in stays cool all the time.
I use an Athlon for my workstation because it's cheap and fun to hack around with, but my servers are PowerPC and I've never even thought of moving them to x86.
I support the removal of the floppy from the system, along with the serial and parallel ports. I think that floppy support should be piped over USB, and there oughtta be an internal USB header in every PC. You could get legacy ports/drives on 'legacy-free' PCs with an internal USB bus (should you need them).
How cool would it be if there was a 3.5" USB hot-swap bay on every PC that could take a floppy or flashcard reader? You could stick serial-ata connectors in there too for a general-purpose hot-swap bay instead of a rickety old 1980s floppy device.
No problems here. Consider a distribution with better package management. All I did was:
/etc/init.d/samba restart
emerge sync
emerge -u world
The clients even stayed connected, not a hiccup.
And most people don't need wheelbarrows anymore.
It's time to get rid of the 'geek' connectors on PCs. If you want legacy ports you can get an add-in card with them or a USB shim.
While we're at it, is there a chance we can get a new kind of floppy controller? I think it's lame how my Athlon slows to a crawl when I'm formatting a floppy. Maybe serializing floppy access and adding a 32K buffer would fix things.
Right, now instead of paying for a $1000 roll of real film you'll have to pay MS $995 for the right to use their 'WM9 theater encoder deluxe'. (prices are completely made up)
I think the major hurdle for the independent film industry is the vast droves of morons who would rather pay $9 to see a bunch of computer-rendered explosions than $5 to watch a well thought-out film.
That's exactly my problem. I've become too used to firing up three or for apps at once and letting the kernel sort out the resource management. I'm not firing up more apps than I need, I can USE all those apps. Something about making myself home in UNIX-land has made me content to have 10-15 open windows on different desktops waiting for input. Why not have OpenOffice, XMMS, Mozilla, Konqueror and WindowMaker all fire up when I log in? That way I don't have to individually open something and close it; it's all open, all the time!
The trouble is when I try to do the same thing in Windows, I have to stretch out the taskbar to three rows and when I push the system too far it kills all the unsaved data in ALL of those windows! I don't know if I'll ever be able to go back.
What you say is only MOSTLY true. I know I break any win9x box within 5 minutes of logging in just because I'm too speedy for it, the 9x kernel and I do NOT get along. WinNT/2K/XP seem to be alright with me, and the Linux Kernel LOVES me, I've only caused Linux to become unresponsive twice in my life (and both were because I was TRYING to).
Mac OS 9 is also crashy for me, I kill it right off the bat (even a clean install). My dad, who has a more linear approach to using the computer, can go weeks without a single crash on the same Mac.
I find myself more and more of a liability on older systems, I just make them crash too much, does anybody else have this problem?
What irks me is that there's not a -march=c3 target in recent GCC releases. The C3 currently works best when you use '-m486 -m3dnow -mmmx' which is nasty. VIA needs to kick a GCC developer a few thousand and a few books so GCC can get a proper target for VIA's products. Until there's proper scheduling and cache-management for this processor (on the compiler end) everything is gonna feel REAL SLOW on it.
That's what I'm looking for. There's got to be a way to mount that thing in my G3 B&W, maybe with a ZIF adapter card. I just want a 750 with a giant or superfast cache, I'll underclock it if I have to. If I can't get that I'll wait for the Pegasos PPC board and go Apple-free.
I had too many headaches getting something like that working with Mozilla, the file system differences really nail you, paths are all different and stuff.
What I'd like to see is an option to add 'windows' and 'unix' options to the mail paths in prefs.js so when I run in linux the profile comes from '~/.mozilla...' and when I run in Windows it picks up from 'U:\.mozilla...' even though those are really the same folder.
I guess I should go enter this in bugzilla!
What I wouldn't give to have a recent-design G3 in my old blue-n-white.
I wish someone would produce new G3s with giant integral and backside caches, they would absolutely fly on low-end and midrange servers where altivec is nothing more than extra heat to dissipate.
Anyone know of a way to get one of those swanky new 750CXe G3s from the latest iBook into an older blue-and-white? They're different pinouts and voltages from my POV.
I thought the G4 FP improvements were because it intrinsically routed some of those ops through the AltiVec units.
AFAIK, the G4 was a G3 with SMP circuitry and AltiVec units, at least in it's first incarnation. It's been a while since it debuted though, and the evidence that the G4 has become more streamlined since branching from the G3 is good (if they are the same 'core' the G3 would be able to clock much higher since it's smaller).
That's what I thought the situation was. I know a lot of printed graphics folks who jacked up their working resolution to 6000 DPI just to feel important and to justify their own time expenditure. I tried explaining once that going above the printer's resolution was a waste of time and I was quickly hushed and told that that's what the new PCs were for.