How to run a Pentium IV processor at 5GHz speed

hul2000

Member
Oct 17, 2006
5,050
2
0
39
Gallifrey
Anusha said:
Because if your claims are correct, people can run faster in the winter season :rofl:

The main flow in Pentium IV architecture (a.k.a. NetBurst architecture) is high power consumption and producing more heat than any previous processors.

At the launch of the P4, Intel stated NetBurst was expected to scale to 10 GHz (over several fabrication process generations). However, the NetBurst architecture ultimately hit a frequency ceiling far below expectation—the fastest retail Pentium 4 never exceeded 4 GHz. Intel had not anticipated a rapid upward scaling of transistor power leakage that began to occur as the chip reached the 90 nm process node and smaller. This new power leakage phenomenon, along with the standard thermal output, created cooling and clock scaling problems as clock speeds increased. Reacting to these unexpected obstacles, Intel attempted several core redesigns ("Prescott" most notably) and explored new manufacturing technologies. Nothing solved their problems though and in 2005-6 Intel shifted development away from NetBurst to focus on the cooler running Pentium M architecture. In March 2006, Intel announced the Intel Core microarchitecture, which puts greater emphasis on energy efficiency and performance per clock. The final NetBurst-derived products were released in 2006, with all subsequent product families switching exclusively to the Intel Core microarchitecture.

The NetBurst differed from the preceding Intel P6 - of CPUs branded Pentium III, II, etc. - by featuring a very deep instruction pipeline to achieve very high clock speeds[4] (up to 4 GHz) limited only by max. power consumption (TDP) reaching up to 115 W in 3.6–3.8 GHz Prescotts and Prescotts 2M[5] (a high TDP requires an additional cooling that can be noisy or expensive).

So basically higher the the clock speed the higher the heat. thus the cooler the processor gets the more the clock speed (OK it's not that simple, but anyway)
 

Anusha

Member
Jun 13, 2006
25,864
5
0
hul2000 said:
The main flow in Pentium IV architecture (a.k.a. NetBurst architecture) is high power consumption and producing more heat than any previous processors.





So basically higher the the clock speed the higher the heat. thus the cooler the processor gets the more the clock speed (OK it's not that simple, but anyway)
Another n00b statement! Cooler the CPU, you can achieve higher overclock :P
Clock speed doesn't go up automatically just because you cool it down LOL
Then how can people be running 3GHz P4s at 40C and 60C at the same time? (40C person should be running it at 3.2GHz, right? :P)
 

hul2000

Member
Oct 17, 2006
5,050
2
0
39
Gallifrey
Anusha said:
Another n00b statement! Cooler the CPU, you can achieve higher overclock :P
Clock speed doesn't go up automatically just because you cool it down LOL
Then how can people be running 3GHz P4s at 40C and 60C at the same time? (40C person should be running it at 3.2GHz, right? :P)

Speed to Temperature does not change linearly, it changes exponentially.
 

Anusha

Member
Jun 13, 2006
25,864
5
0
When you run the CPU at a lower temperature (anything that works with clock, for that matter), the resistances inside the chip gets lower and the rise times/switch times/delays get lower. So you can increase the clock speed up to the level which it still works at the correct delays. (lower temperature gives more headroom to overclock more)

Clock speed is an external input. It is generated by a crystal (well, it is more complex...it is generated by a Phase Locked Loop actually) on the motherboard. Then the motherboard has many multipliers to give the PCI clock, PCI-E clock, FSB and all the other devices.

The CPU will multiply the FSB (which is a multiplied value of the original clock generated by the PLL) to the correct value specified by the manufacturer using the multiplier (eg: 3GHz P4E has a multiplier of 15). Multiplier is a fixed value (generally).

Buy increasing the FSB, you are actually increasing the PLL clock (that corresponds to the FSB). New mobo's can change the clocks (FSB, PCI etc.) independently.

DO YOU THINK THE MOTHERBOARD KNOWS THAT THE CPU IS RUNNING VERY COOL AND THAT IT CAN INCREASE THE CLOCK SPEED???
 
Last edited:

Anusha

Member
Jun 13, 2006
25,864
5
0
hul2000 said:
Speed to Temperature does not change linearly, it changes exponentially.
Dammit! I was just taking an example!

Speed to temperature doesn't necessarily increase exponentially.

I guess if the E6300 (1.86GHz) CPU runs at 50C at full load, the E6850 runs at around 80C at full load :P (that's even when you consider it liearly). According to you, it must be running at well over 100C :rofl:

Just think for a second before you open your mouth!!!
 

hul2000

Member
Oct 17, 2006
5,050
2
0
39
Gallifrey
Anusha said:
When you run the CPU at a lower temperature (anything that works with clock, for that matter), the resistances inside the chip gets lower and the rise times/switch times/delays get lower. So you can increase the clock speed up to the level which it still works at the correct delays. (lower temperature gives more headroom to overclock more)

Clock speed is an external input. It is generated by a crystal (well, it is more complex...it is generated by a Phase Locked Loop actually) on the motherboard. Then the motherboard has many multipliers to give the PCI clock, PCI-E clock, FSB and all the other devices.

The CPU will multiply the FSB (which is a multiplied value of the original clock generated by the PLL) to the correct value specified by the manufacturer using the multiplier (eg: 3GHz P4E has a multiplier of 15). Multiplier is a fixed value (generally).

Buy increasing the FSB, you are actually increasing the PLL clock (that corresponds to the FSB). New mobo's can change the clocks (FSB, PCI etc.) independently.

DO YOU THINK THE MOTHERBOARD KNOWS THAT THE CPU IS RUNNING VERY COOL AND THAT IT CAN INCREASE THE CLOCK SPEED???

Well, that's why they cool the mother board too. the piezoelectricity of the crystal varies on temperature (as much as 4% if you cool from 85 C to -55 C).
 

Anusha

Member
Jun 13, 2006
25,864
5
0
hul2000 said:
Well, that's why they cool the mother board too. the piezoelectricity of the crystal varies on temperature (as much as 4% if you cool from 85 C to -55 C).
well, it is NOT a Crystal!!! PLLs are much much much more stable!