Well it isn't like that anymore let me assure you!
On the Intel 945 chipset, Windows installs fine if SATA is set to IDE mode which is the default. If you set it to AHCI *OR* RAID mode, then it requires drivers from Intel http://www.intel.com/design/chipsets/matrixstorage _sb.htm. This was with a clean XP SP2, and I'd imagine it'd be the same for Win2k also.
Generally SATA controllers do require drivers, unless they are in an IDE compatabile mode. It appears that the Intel 945 has a newer ICH which is likely the reason why this is the case. Going forward, I can't see the situation changing.
RAID mode isn't the issue, it is the IDE compatability mode.
To create custom MSTs for Acrobat, which you can use to disable all of the annoying crap. Well, apart from the Yahoo search! I suggest also http://www.appdeploy.com/ can be useful for finding ways to disable stuff in installers.
Sure you can get it for some PC servers, often via an expensive add-on card, but every Sparc machine has this built-in from the desktops to the servers. Until PC servers break from the legacy BIOS, and add features like this as standard equipment they will just be PCs that happen to be running a server OS.
See but every Sparc machine has this built-in, after saying that only some PC servers have it. That isn't really fair, as I think these days you'll struggle to find a server from a tier 1 vendor that doesn't have some sorts of lightouts management.
Also there is confusion. Sure, every Sun I can remember also console redirection to keyboard that it is NOT lightsout management. You can't power on a Sun from it's keyboard console. This was a useful feature in it's day, but a LOM is far more useful and I agree with the GP, is essential these days.
As I point out, the crufty BIOS isn't a limitation. I can use lightsout management on a Sunfire v20z just fine thanks!
It's standard too on all HP servers these days too. I believe Dell have a RAS console as standard, although I'm not big on Dell servers so I could be wrong.
This isn't a SPARC thing though, even the Sunfire v20z has it as well and that is Opteron. Of course, technically it was made by Newisys and not Sun. I've not seen the newer Sun Opteron server's LOM yet, but from the specs it looked the same.
I think you are missing the point. I too had an Athlon-XP mobile, but my desktop motherboard (and all others from what I could tell) did not support the lower voltages the XP-M can use, nor do they support clock scaling.
So basically, your XP-M is running at the same voltage as a stock part and can't be clock scaled. So it is essentially acting as a desktop part. The advantage of the XP-M was better overclocking for the most part, and some motherboards did provide a slightly lower core voltage. The XP-M was designed run at 1.45v, the desktop XP was designed to run as v1.65. My ASUS board would go to 1.575v which is what I ran it at.
As to why a 2600+ would even overheat, well, I suspect your heatsink is either badly attached (lack of heatsink compound?) of the fan isn't running too well. There is no way it should overheat if it is fitted properly and everything is working.
Generally you are correct, 64-bit code is generally slower as it requires more cache & bandwidth than 32-bit code. The only real advantage is in encryption, databases, etc which requires the large word widths where it can provide benefits (beyond more memory available).
Of course, we are specfically talking about x86 where the above isn't true.
In normal 32/64-bit modes (e.g. SPARC, POWER, etc) the modes are pretty much the same. For x86 it isn't, you get double the amount of registers (8 vs 16) which can provide LARGE performance benefits for a given task. Also, you can built your x86-64 binary and assume you've got SSE2, MMX, etc and any other instructions that have been added since the i386 came about. Basically, you can build a better binary.
Ok, I don't think this is right. I watched part of the keynote the next day and when Jobs was talking about performance he said both chips were using the "best possible compiler" and these were identified as Intel's compiler for x86 and the IBM compiler for the PPC.
Although using Xcode, yes, they use GCC. I think at this point they were trying to get the best number possible.:)
Of course, when the G5 came out they used GCC when comparing it against the Pentium 4 as this was "fair". More likely, it was due to the fact the Pentium 4's architecture didn't show good performance when running under GCC due to it's long pipeline and SIMD dependenicies. They didn't show how an Athlon64 did either, which I suspect would have been close.
Not that I'm saying the G5 is a bad chip, but it is interesting how they used GCC then and now use Intel's compiler for benchmarks!
To be fair, most of the current sucess is due to AMD's Opteron. To say Sun's v40z is quicker than a Dell system isn't really giving Sun any credit, especially as the v40z was designed by Newisys and is just badged by Sun.
Also, Intel's 64-bit x86 implementation is a bag o crap as they only did it to keep up with AMD. They'd have rather you use Itanium still.
It'll be useless for any sort of gaming, as it has 8 integer cores and 1 FPU core between the lot of them. So any heavy FPU stuff is going to be really slow.... it should also be said that FPU cores are complicated which is why Sun was able to do this system.
The integer performances of their cores isn't going to be great either. For web servering, I suspect it'll be a good box (although I'd like to see some non-Sun benchmarks/SPEC results for this). Comparing it to a Dell system isn't fair either, as the Intel processors are junk. How it compares to Sun's own AMD based systems (which are faster and cooler) is more useful - especially as they are already dual core (2 FPUs too....) and are heading 4 cores.
Sure, ICANN does like to involve other countries but it's ultimate boss is the US government. I don't want it to be under UN control, but I don't want it to be under the US government control.
Clinton agreed to make it independent, but that changed:
The US government, which funded the development of the internet in the 60s, said in June it intended to retain its role overseeing Icann, reneging on a pledge made during Bill Clinton's presidency. Since Icann was created, the US commerce department has not once interfered with its decisions.
After all, ICANN voted and agreed to use the.xxx domain but that offended some conservative types and their protests to the US government got it stopped. So now, they have interfered.
Not really... all they actually said there would be Intel Macs shipping by June 2006 and the cutover would be done by June 2007. There wasn't a great deal of detail of exactly what they had in mind here.
They didn't say they would *start* shipping Intel Macs in June, just that they would be available. There is a difference.
DDR-2 does have lower power requirements, which is a win for laptops. More battery, and less heat. But I do agree it is likely, if anything, to degrade performance.
What Apple should have done is put in the 7448 core from Freescale. It is pin compatabile with the current 7447a and has a faster FSB (200MHz) as well as more cache and better power management. Along with other general core improvements.
Of Unix server sales? Linux has 0% because it is an OS not a tier 1 hardware provider!
As is also clearly stated in the above link, Linux has a $1.4 billion market and is separate from the Unix market (which I presume means Solaris, AIX, HPUX, etc).
Plus did I mention that sun has about 40% of the Unix server market, which if I'm not mistaken is about four times the size of the Linux server market?
The Linux market was $1.4 billion. Sun have 29.5% of a $4.3 billion market, which if my maths serves me correct is less than the Linux market.
As for why switch from SPARC? Price/performance is the main thing. Being able to use any x86 manufactuer with Linux is more flexible than Sun (or Fuji at a push) as a SPARC supplier. Niagara would be good for quiet web servers, but for any sort of real performance it isn't going to be there yet. Rock maybe more interesting, but that is sometime away yet. Sun's sales are currently declining too (and have done for a few years), so it isn't like SPARC is selling that well.
As a general comment, "64bit program will be bigger and probably slower" is generally true. Things that benefit from 64-bit (databases, encryption) will speed up but everything else will slow down. Of course, when you speak x86-64 that isn't true at all. You get several advanatages in the design (more reigsters, flat memory, assume high level processor (no compat building for a Pentium 1)) which do speed it up.
In EDA tools, I've seen SPARC 64-bit binaries go slower and the x86-64 ones go quicker. That isn't a critcism of SPARC, it is just proof of the benefits of the x86-64 mode.
Anyway in THIS case I don't see any benefits of having a 64-bit OOo as such - the problem is if OOo is 32-bit, then it needs a 32-bit userland. So on my x86-64 boxes I've got a load of duplicated libraries (GTK etc) in 32-bit mode just to support OOo. If OOo was 64-bit, I simply would need less stuff. That has to be good right?
HP have had "integrated lightsout" for years which does exactly that. I've got P3 based boxes I can do just about anything to remotely.
This isn't really anything special. Although it is a very nice solution, so I'm glad as many servers as possible have it as it does make your life much easier!
How can they differentiate themselves? Opterons have on-board memory controllers, so as long they are all connected then there is little else you can do.
Read the v20z and X4100 specs. Oh look, same memory controller, same chipset, etc. There is nothing they can change to make any particular difference.
The Sun v40z will scale well, and is a dual core x86-64 server not made by Sun. I'm sure the DL585 from HP will do the same job because they use exactly the same memory technology.
The only magic you can do with Opterons is when you go above 8 sockets, and you need some logic (which Newisys is making) to get above that.
Lessons:
- Sun hardware is too reliable: the machines will be technologically obsolete before they fail. Sun can save costs there, because nobody appreciates it anyway unless they are building a spacecraft or nuclear power plant.
I'd like to put comment on that - from what you've said, 1991 Sun h/w is very reliable. These days it just isn't the same. It doesn't appear to be any better than the HP/Dell servers we've got. Their level of support isn't what it was, but maybe that is just me.
I'm not saying they are unreliable, but there don't appear anything special in build quality/etc than the other servers we've got. The desktop ultra 5 was a piece of junk though, we had a load and they essentially all failed whilst are our slightly older Compaq desktops didn't fail at all..... I may not trust the newer HP desktops so much though.
Not really.... the Pentium 4 is heavily dependent on vectorisation to get any sort of good performance. Such auto-vectorisation only appered in gcc 4.0, and more work on it is going into gcc 4.1.
If they had used ICC, then they'd have had this and the results would have been very different. Don't believe me? Go look at the SpecINT/FP scores for the P4 & 970 (IBM have posted some). The P4 is much much higher.
Now, if they'd used GCC and pitted it against an Opteron then they would have also lost as the Opteron is fine running "normal" code.
There are plenty of benchmarks out there if you google for them. Anandtech certainly has a few.
You'll find Opterons are better performing, and scale better. They scale better because each Opteron has it's own memory where as dual/quad Xeon has to fight for the same amount of bandwidth. Opteron's onboard memory controller also gives very low latency to main memory, which helps a great deal.
You'd probably find a dual Opteron would easily beat a dual Xeon is almost all tasks. Most likely exception is streaming & encoding video, there it will just be "close".
This is a rather silly statement given the number of ATX motherboards out there. Whilst ATX has it's problems and BTX had some advantages, just remember one thing - there are no BTX motherboards available for AMD based systems.
With the layout of the BTX boards, it makes it very difficult for AMD to use it as their onboard memory controller is too far from the memory IIRC. Of course for Intel, they only have to route to the northbridge which then has to route to the memory so they can get it to work.
As AMD's processors run cooler, there is less need for a BTX design also currently.
Slashdot Mirror
Well it isn't like that anymore let me assure you!
e _sb.htm. This was with a clean XP SP2, and I'd imagine it'd be the same for Win2k also.
On the Intel 945 chipset, Windows installs fine if SATA is set to IDE mode which is the default. If you set it to AHCI *OR* RAID mode, then it requires drivers from Intel http://www.intel.com/design/chipsets/matrixstorag
Generally SATA controllers do require drivers, unless they are in an IDE compatabile mode. It appears that the Intel 945 has a newer ICH which is likely the reason why this is the case. Going forward, I can't see the situation changing.
RAID mode isn't the issue, it is the IDE compatability mode.
You can use this piece of Adobe software:
? ftpID=2709
http://www.adobe.com/support/downloads/detail.jsp
To create custom MSTs for Acrobat, which you can use to disable all of the annoying crap. Well, apart from the Yahoo search! I suggest also http://www.appdeploy.com/ can be useful for finding ways to disable stuff in installers.
See but every Sparc machine has this built-in, after saying that only some PC servers have it. That isn't really fair, as I think these days you'll struggle to find a server from a tier 1 vendor that doesn't have some sorts of lightouts management.
Also there is confusion. Sure, every Sun I can remember also console redirection to keyboard that it is NOT lightsout management. You can't power on a Sun from it's keyboard console. This was a useful feature in it's day, but a LOM is far more useful and I agree with the GP, is essential these days.
As I point out, the crufty BIOS isn't a limitation. I can use lightsout management on a Sunfire v20z just fine thanks!
It's standard too on all HP servers these days too. I believe Dell have a RAS console as standard, although I'm not big on Dell servers so I could be wrong.
This isn't a SPARC thing though, even the Sunfire v20z has it as well and that is Opteron. Of course, technically it was made by Newisys and not Sun. I've not seen the newer Sun Opteron server's LOM yet, but from the specs it looked the same.
I think you are missing the point. I too had an Athlon-XP mobile, but my desktop motherboard (and all others from what I could tell) did not support the lower voltages the XP-M can use, nor do they support clock scaling.
So basically, your XP-M is running at the same voltage as a stock part and can't be clock scaled. So it is essentially acting as a desktop part. The advantage of the XP-M was better overclocking for the most part, and some motherboards did provide a slightly lower core voltage. The XP-M was designed run at 1.45v, the desktop XP was designed to run as v1.65. My ASUS board would go to 1.575v which is what I ran it at.
As to why a 2600+ would even overheat, well, I suspect your heatsink is either badly attached (lack of heatsink compound?) of the fan isn't running too well. There is no way it should overheat if it is fitted properly and everything is working.
One thing to note, the X1900XT has already been released so the X1800XT is already obsolete!
Newegg search for X1900XT.
The X1900XT is a very good chip it seems from reviews so far.
Generally you are correct, 64-bit code is generally slower as it requires more cache & bandwidth than 32-bit code. The only real advantage is in encryption, databases, etc which requires the large word widths where it can provide benefits (beyond more memory available).
Of course, we are specfically talking about x86 where the above isn't true.
In normal 32/64-bit modes (e.g. SPARC, POWER, etc) the modes are pretty much the same. For x86 it isn't, you get double the amount of registers (8 vs 16) which can provide LARGE performance benefits for a given task. Also, you can built your x86-64 binary and assume you've got SSE2, MMX, etc and any other instructions that have been added since the i386 came about. Basically, you can build a better binary.
Wikipedia is useful as always.
Ok, I don't think this is right. I watched part of the keynote the next day and when Jobs was talking about performance he said both chips were using the "best possible compiler" and these were identified as Intel's compiler for x86 and the IBM compiler for the PPC.
:)
Although using Xcode, yes, they use GCC. I think at this point they were trying to get the best number possible.
Of course, when the G5 came out they used GCC when comparing it against the Pentium 4 as this was "fair". More likely, it was due to the fact the Pentium 4's architecture didn't show good performance when running under GCC due to it's long pipeline and SIMD dependenicies. They didn't show how an Athlon64 did either, which I suspect would have been close.
Not that I'm saying the G5 is a bad chip, but it is interesting how they used GCC then and now use Intel's compiler for benchmarks!
So have some other games. Like the new Call Of Duty 2 patch:
Call Of Duty 2 1.01 patch review.
Two things:
1. I don't trust SPEC figures as they are normally cooked by the companies involved (see Sun and SpecFP).
2. The Xeon systems would be outpaced by Opteron systems, especially the dual core systems.
I'm not saying the T2000 won't be good at web serving, but I would like some 3rd party evidence of it.
To be fair, most of the current sucess is due to AMD's Opteron. To say Sun's v40z is quicker than a Dell system isn't really giving Sun any credit, especially as the v40z was designed by Newisys and is just badged by Sun.
Also, Intel's 64-bit x86 implementation is a bag o crap as they only did it to keep up with AMD. They'd have rather you use Itanium still.
It'll be useless for any sort of gaming, as it has 8 integer cores and 1 FPU core between the lot of them. So any heavy FPU stuff is going to be really slow.... it should also be said that FPU cores are complicated which is why Sun was able to do this system.
The integer performances of their cores isn't going to be great either. For web servering, I suspect it'll be a good box (although I'd like to see some non-Sun benchmarks/SPEC results for this). Comparing it to a Dell system isn't fair either, as the Intel processors are junk. How it compares to Sun's own AMD based systems (which are faster and cooler) is more useful - especially as they are already dual core (2 FPUs too....) and are heading 4 cores.
Clinton agreed to make it independent, but that changed:
From The Guardian.
After all, ICANN voted and agreed to use the .xxx domain but that offended some conservative types and their protests to the US government got it stopped. So now, they have interfered.
Not really... all they actually said there would be Intel Macs shipping by June 2006 and the cutover would be done by June 2007. There wasn't a great deal of detail of exactly what they had in mind here.
They didn't say they would *start* shipping Intel Macs in June, just that they would be available. There is a difference.
DDR-2 does have lower power requirements, which is a win for laptops. More battery, and less heat. But I do agree it is likely, if anything, to degrade performance.
What Apple should have done is put in the 7448 core from Freescale. It is pin compatabile with the current 7447a and has a faster FSB (200MHz) as well as more cache and better power management. Along with other general core improvements.
Of Unix server sales? Linux has 0% because it is an OS not a tier 1 hardware provider!
As is also clearly stated in the above link, Linux has a $1.4 billion market and is separate from the Unix market (which I presume means Solaris, AIX, HPUX, etc).
Perhaps you should re-read it?
Then you need to put down the kool-aid....
IBM was the leader in worldwide Unix server revenue with 31% share, while HP and Sun were statistically tied for the number two position, with 30.0% and 29.5% share, respectively.
The Linux market was $1.4 billion. Sun have 29.5% of a $4.3 billion market, which if my maths serves me correct is less than the Linux market.
As for why switch from SPARC? Price/performance is the main thing. Being able to use any x86 manufactuer with Linux is more flexible than Sun (or Fuji at a push) as a SPARC supplier. Niagara would be good for quiet web servers, but for any sort of real performance it isn't going to be there yet. Rock maybe more interesting, but that is sometime away yet. Sun's sales are currently declining too (and have done for a few years), so it isn't like SPARC is selling that well.
I've had that happen on 8.02 on Gentoo Linux also.....
As a general comment, "64bit program will be bigger and probably slower" is generally true. Things that benefit from 64-bit (databases, encryption) will speed up but everything else will slow down. Of course, when you speak x86-64 that isn't true at all. You get several advanatages in the design (more reigsters, flat memory, assume high level processor (no compat building for a Pentium 1)) which do speed it up.
In EDA tools, I've seen SPARC 64-bit binaries go slower and the x86-64 ones go quicker. That isn't a critcism of SPARC, it is just proof of the benefits of the x86-64 mode.
Anyway in THIS case I don't see any benefits of having a 64-bit OOo as such - the problem is if OOo is 32-bit, then it needs a 32-bit userland. So on my x86-64 boxes I've got a load of duplicated libraries (GTK etc) in 32-bit mode just to support OOo. If OOo was 64-bit, I simply would need less stuff. That has to be good right?
HP have had "integrated lightsout" for years which does exactly that. I've got P3 based boxes I can do just about anything to remotely.
This isn't really anything special. Although it is a very nice solution, so I'm glad as many servers as possible have it as it does make your life much easier!
How can they differentiate themselves? Opterons have on-board memory controllers, so as long they are all connected then there is little else you can do.
Read the v20z and X4100 specs. Oh look, same memory controller, same chipset, etc. There is nothing they can change to make any particular difference.
The Sun v40z will scale well, and is a dual core x86-64 server not made by Sun. I'm sure the DL585 from HP will do the same job because they use exactly the same memory technology.
The only magic you can do with Opterons is when you go above 8 sockets, and you need some logic (which Newisys is making) to get above that.
I'd like to put comment on that - from what you've said, 1991 Sun h/w is very reliable. These days it just isn't the same. It doesn't appear to be any better than the HP/Dell servers we've got. Their level of support isn't what it was, but maybe that is just me.
I'm not saying they are unreliable, but there don't appear anything special in build quality/etc than the other servers we've got. The desktop ultra 5 was a piece of junk though, we had a load and they essentially all failed whilst are our slightly older Compaq desktops didn't fail at all..... I may not trust the newer HP desktops so much though.
Not really.... the Pentium 4 is heavily dependent on vectorisation to get any sort of good performance. Such auto-vectorisation only appered in gcc 4.0, and more work on it is going into gcc 4.1.
If they had used ICC, then they'd have had this and the results would have been very different. Don't believe me? Go look at the SpecINT/FP scores for the P4 & 970 (IBM have posted some). The P4 is much much higher.
Now, if they'd used GCC and pitted it against an Opteron then they would have also lost as the Opteron is fine running "normal" code.
There are plenty of benchmarks out there if you google for them. Anandtech certainly has a few.
You'll find Opterons are better performing, and scale better. They scale better because each Opteron has it's own memory where as dual/quad Xeon has to fight for the same amount of bandwidth. Opteron's onboard memory controller also gives very low latency to main memory, which helps a great deal.
You'd probably find a dual Opteron would easily beat a dual Xeon is almost all tasks. Most likely exception is streaming & encoding video, there it will just be "close".
This is a rather silly statement given the number of ATX motherboards out there. Whilst ATX has it's problems and BTX had some advantages, just remember one thing - there are no BTX motherboards available for AMD based systems.
With the layout of the BTX boards, it makes it very difficult for AMD to use it as their onboard memory controller is too far from the memory IIRC. Of course for Intel, they only have to route to the northbridge which then has to route to the memory so they can get it to work.
As AMD's processors run cooler, there is less need for a BTX design also currently.