The most time-critical paths in the CPU are also the most important. Those double-pumped ALUs are probably the touchiest components, and also the trace-op caches and branch predicition stuff. Without those, well...
Your statement is so misinformed, I don't even know where to begin. I'm guessing you've never taken a single class in digital electronics, computer architecture, or electrical engineering.
You need a decent heatsink designed for 1U use, and a motherboard with 1U use in mind. Motherboard considerations: The chip socket should be as recessed as possible considering the board standoffs, no large capacitors, as many right-angled cable connectors as possible, etc. Also, the chip sockets have to be oriented so they are not obstructed by board components (such as filled RAM slots). The heatsink has to be full of fins oriented front-to-back. Then you need to install an assload of 60mm fans, and some plastic or carboard airflow guides/hoods to force that air into and out of the CPU heatsink while bypassing everything else. A few of the remaining fans that don't funnel into the CPU cooling "tunnel" will be cooling the power supply on one side, and a few more for general MB/chipset cooling on the other. You will absolutely need fans in the front and back, as close to the front and rear ventilation holes of the case as possible.
Also helpful are HSFs that are horizontally oriented designed for 1U mounting to various socket types... but they are rare.
A heat pipe and a large radiator does passively via convenction what the stirling engine does forcefully for the extra 9W.
The overall cooling effect is still limited by the heat transfer ability of the hot side, which is always some kind of radiator.
I like using heat pipes and a single fan that forces ambient air through multiple radiators (oriented lowest to highest heat load for maximum heat transfer). Cheap and easy.
They just refuse to turn on many optimizations unless they see "GenuineIntel" returned from the CPUID instruction. Thus excluding _everyone_ else. At least they're being fair about it. *eye roll*
The P2 and P3 (and even the Pentium M) are largely unchanged from the original PPro.
P2 = PPro + MMX P3 = P2 + SSE
There are other differences (cache structure and speed, bus signaling, addt. transistors for branch prediction, die shrinks) but it's all the same block diagrams and stuff.
P4 was different... the whole netburst thing. Wow did that suck.
That's not an innovation. Fast-user switching is not allowed on machines on NT domains on purpose. The reason is because the security model gave certain preferences and made certain assumptions about the "console" window session which were not acceptable by beta testing system admins. For example: If a user fast-switches away from a login session you have currently running, and is using it, if you have a remote profile then you are crippled if you try logging in a different machine (as opposed to "taking over" the existing session if you could unlock said machine). A user would then be forced to make that other user get up to log out. There were other problems related to how the new logon screen would attempt to list remote users, handle incorrect password attempts, etc. Basically what people knew about domain administration would be thwarted by the operation of the fast-user switching model which was designed for home users. Windows 2003 handled a few of the problems (but it makes it irrelevant since you can remotely take over console sessions with it)... and Longhorn fixes all of them. But you will probably need Longhorn Server to allow Longhorn clients to get on a domain and fast-user switch.
You don't even need more than one ethernet interface on the system. If you are attaching it directly to a switch you can make it listen on two different addresses and the switch will happily route the traffic accordingly. Of course having two interfaces can increase your throughput...
Pop a win2k server CD (for example, one from MSDN) into your computer. Find the tsoc.in_ file. Expand this (using expand.exe) into a location on your computer. Right click on this and select "install". It should ask you for the windows server 2000 cd (which is in your computer). It will copy the rest of the necessary files, make some registry changes, install the RDP driver and install the TS service and admin tools.
If this doesn't work, you might need NTSwitch to convince the system it's okay to proceed. Alternatively you can change a few lines in the header of the INF file to allow it to work on your particular build/revision of 2000 professional.
You could even install it manually if you know how to kick it with regsrv32, expand.exe and regedt32... the tsoc.inf file is well commented and straightforward (basically you want to look at parts that talk about installing from "nothing" to "50").... But you didn't hear it from me.:-)
it's cool but it's braindead. In Server 2003 (which doesn't cost much more than XP Pro, in fact), you can do 2 remote sessions and a local session without having to log off.
And Windows 2000 can do real-deal terminal services (as many sessions as you want), which can be used for free if you use 2000 or XP as a remote client (and rdesktop with a patch).
I'd rather just add TS to 2000, install the license server locally, and enjoy the flexible goodness.
...it's the firmware Dell shoves on them. It's only designed to hook up with Dell disc arrays and tape drives and everything else can shove it (from their point of view). Do yourself a favor and skip 'em and just by the cards straight from LSI.
All the chips come off the same assembly line. They first test all the chips at a "safe" configuration to check for clearly defective chips (usually about 50% when you talk about 90nm process). Then, they test as many chips as necessary at a speed margin above the rated speed to fill in the "speed bin". As soon as this quota is met, they run the rest through a test at lower rated speed. Since they go from high to low, it's quite conceivable a chip that was not reviewed to fill a higher speed quota will fill the quota of a lower tested speed. Testing takes time, so they don't like to test chips multiple times... they will test them the minimum times necessary to ensure all speed rating bins are filled for a batch of chips. Hence you get chips rated a lower speed which were never tested at a higher speed (wasn't needed to fill predicted demand)... and a good chance of overclockability.
Slashdot Mirror
Someone to give him an elevator summary of what questions he should throw at somebody's kid in college or a starving IT consultant.
The most time-critical paths in the CPU are also the most important. Those double-pumped ALUs are probably the touchiest components, and also the trace-op caches and branch predicition stuff. Without those, well...
Your statement is so misinformed, I don't even know where to begin.
I'm guessing you've never taken a single class in digital electronics, computer architecture, or electrical engineering.
No.
In systems engineering, you focus on how you go about making you kill more of them instead of you.
You need a decent heatsink designed for 1U use, and a motherboard with 1U use in mind.
Motherboard considerations: The chip socket should be as recessed as possible considering the board standoffs, no large capacitors, as many right-angled cable connectors as possible, etc.
Also, the chip sockets have to be oriented so they are not obstructed by board components (such as filled RAM slots).
The heatsink has to be full of fins oriented front-to-back. Then you need to install an assload of 60mm fans, and some plastic or carboard airflow guides/hoods to force that air into and out of the CPU heatsink while bypassing everything else. A few of the remaining fans that don't funnel into the CPU cooling "tunnel" will be cooling the power supply on one side, and a few more for general MB/chipset cooling on the other.
You will absolutely need fans in the front and back, as close to the front and rear ventilation holes of the case as possible.
Also helpful are HSFs that are horizontally oriented designed for 1U mounting to various socket types... but they are rare.
A heat pipe and a large radiator does passively via convenction what the stirling engine does forcefully for the extra 9W.
The overall cooling effect is still limited by the heat transfer ability of the hot side, which is always some kind of radiator.
I like using heat pipes and a single fan that forces ambient air through multiple radiators (oriented lowest to highest heat load for maximum heat transfer). Cheap and easy.
They just refuse to turn on many optimizations unless they see "GenuineIntel" returned from the CPUID instruction. Thus excluding _everyone_ else.
At least they're being fair about it. *eye roll*
...it's just that your system lacks the bandwidth.
I'm guessing you have a 754-pin A64... you've got a crippled bus from your cache to memory.
The P2 and P3 (and even the Pentium M) are largely unchanged from the original PPro.
P2 = PPro + MMX
P3 = P2 + SSE
There are other differences (cache structure and speed, bus signaling, addt. transistors for branch prediction, die shrinks) but it's all the same block diagrams and stuff.
P4 was different... the whole netburst thing. Wow did that suck.
That's not an innovation.
Fast-user switching is not allowed on machines on NT domains on purpose. The reason is because the security model gave certain preferences and made certain assumptions about the "console" window session which were not acceptable by beta testing system admins.
For example: If a user fast-switches away from a login session you have currently running, and is using it, if you have a remote profile then you are crippled if you try logging in a different machine (as opposed to "taking over" the existing session if you could unlock said machine). A user would then be forced to make that other user get up to log out.
There were other problems related to how the new logon screen would attempt to list remote users, handle incorrect password attempts, etc.
Basically what people knew about domain administration would be thwarted by the operation of the fast-user switching model which was designed for home users.
Windows 2003 handled a few of the problems (but it makes it irrelevant since you can remotely take over console sessions with it)... and Longhorn fixes all of them.
But you will probably need Longhorn Server to allow Longhorn clients to get on a domain and fast-user switch.
... heard 'round the world!
You don't even need more than one ethernet interface on the system. If you are attaching it directly to a switch you can make it listen on two different addresses and the switch will happily route the traffic accordingly. Of course having two interfaces can increase your throughput...
then any OS that is capable is fine... BSD, Linux, Solaris, it's all good... provided your comfortable managing the system.
Then the competetion will be tight again.
You don't just hack into it... it's not physically connected.
And that's just for marginally clAssified stuff.
Pop a win2k server CD (for example, one from MSDN) into your computer.
... But you didn't hear it from me. :-)
Find the tsoc.in_ file. Expand this (using expand.exe) into a location on your computer.
Right click on this and select "install". It should ask you for the windows server 2000 cd (which is in your computer). It will copy the rest of the necessary files, make some registry changes, install the RDP driver and install the TS service and admin tools.
If this doesn't work, you might need NTSwitch to convince the system it's okay to proceed. Alternatively you can change a few lines in the header of the INF file to allow it to work on your particular build/revision of 2000 professional.
You could even install it manually if you know how to kick it with regsrv32, expand.exe and regedt32... the tsoc.inf file is well commented and straightforward (basically you want to look at parts that talk about installing from "nothing" to "50").
it's cool but it's braindead.
In Server 2003 (which doesn't cost much more than XP Pro, in fact), you can do 2 remote sessions and a local session without having to log off.
And Windows 2000 can do real-deal terminal services (as many sessions as you want), which can be used for free if you use 2000 or XP as a remote client (and rdesktop with a patch).
I'd rather just add TS to 2000, install the license server locally, and enjoy the flexible goodness.
How does one go about sending SMS messages from a computer (without using provider websites)?
Do you, at decrypt time, specify a source file and a pad file stored on a CD you keep locked in a safe when not in use?
Do you use offsets into a pad and just specify the (secret) offset?
...it's the firmware Dell shoves on them.
It's only designed to hook up with Dell disc arrays and tape drives and everything else can shove it (from their point of view).
Do yourself a favor and skip 'em and just by the cards straight from LSI.
All the chips come off the same assembly line.
They first test all the chips at a "safe" configuration to check for clearly defective chips (usually about 50% when you talk about 90nm process).
Then, they test as many chips as necessary at a speed margin above the rated speed to fill in the "speed bin". As soon as this quota is met, they run the rest through a test at lower rated speed.
Since they go from high to low, it's quite conceivable a chip that was not reviewed to fill a higher speed quota will fill the quota of a lower tested speed.
Testing takes time, so they don't like to test chips multiple times... they will test them the minimum times necessary to ensure all speed rating bins are filled for a batch of chips.
Hence you get chips rated a lower speed which were never tested at a higher speed (wasn't needed to fill predicted demand)... and a good chance of overclockability.
And you can even create a software one built into your timer interrupt... perfect for software-mitigated freezes/crashes
You'd have to inhale _a lot_ of it to get a deadly dose of radiation during that period.
I think they were planning on using the gas form or heavy water, not a tritium-containing compound that will sit in tissues.
I was trying to download some papers from a prof's website at your school at for some reason my connection kept randomly timing out...
:-)
That must be the reason why.