Always wondered

[VampD]

so I think I have a good grasp on FSB and multipliers, I always wondered though

how come when theres advertisements for a particular processor they write for example

Intel® Core™ 2 Duo E6750 CPU @ 2.66GHz 1333FSB 4MB L2 Cache 64-bit((Sckt775)Intel® Core™ 2 Duo E6750 CPU @ 2.66GHz 1333FSB 4MB L2 Cache 64-bit)

what does this 1333 FSB mean exactly for this particular processor? does it matter since its lower in the e4500 processors, etc

also same question applies more specifically for mobos, like why do they write

MSI G31M-F Intel G31 Chipset LGA775 FSB1333 DDR2 Mainboard

 

[Sora]

Basically the FSB of chip is actually "quad-pumped" we call it to reach the speed the chip says. For example the FSB of a an E2xxx and E4xxx series is said to be "800" right? However when you go into the BIOS you see that the FSB is actually only 200mhz. The E2200 is 200x11 to reach the specs of 2.2ghz.

1066 FSB is actually just a chip that runs on stock 266 mhz where as 1333 is stock at 333.

FSB is a topic that took me a while to even fully comprehend but let me try to go into further detail to you.

The reason why some chips have a higher FSB than others like the 333 is so they can squeeze a higher FSB with a lower multiplier. An E6600 has a FSB of 266 which is multiplied by 9 as you said you understand this already reaching thus the specified 2.4ghz. The E2200 is also 2.4ghz right however to achieve this it's a 200fsb multiplied by 12.

If both chips were the EXACT same thing aside from the different ways the FSB and multipliers were arranged to reach the specified speeds, the one with the higher FSB would in turn be a slightly faster chip. This is just due to the fact that a higher FSB and lower multiplier is a more efficient way than a low fsb high multiplier. Companies do this for 2 reasons. 1.) To reel their consumers in with a "1333FSB!!!" to make it sound like it's the must have and cream of the crop. Also just in general benchmarks at the same speed are actually slightly higher so it's more efficient.

HOWEVER the problem to us overclockers with a very high native FSB chip is that it's a lot harder to reach a good speed without very good ram...this was a problem over a year and a half ago when the best ram was so expensive and the E6xxx line was ONLY 266 FSB. That means to reach the speeds that these could reach and can reach (which is over 500FSB like mine) you needed ram that could reach DDR2-1000 speeds and a above. Not an easy feat for any ram really that was sub $115 PER GIG so many people only had DDR2-533,677, or value 800. Memory is also "quad-pumped" however in a different way. DDR ram is actually sdram just double pumped and this was what ram in computers from around 2002-2006 on average. Then we developed DDR2 ram which double pumped that. Technically all DDR2-1066 ram is, is just ram running at 266mhz that is quad pumped by technology to run at that frequency. To find what DDR2 ram runs on a 1:1 ratio just take the chip and divide it by half and thats your DDR2-800 speed you need. 1333 FSB chips need at least DDR2-677 ram to let it run on a 1:1.

Motherboards specifiy a FSB becuase they mean they are capable of utilizing chips with a stock fequency of that...pretty much any motherboard can just have the BIOS changed to pick up stock frequencies and thats all they really did when they made the revisions on boards when 1333 fsb chips came out.

So if you see a motherboard that says 1333FSB don't think that your E2200 which runs at only a FSB of 800 won't work becuase it will, they are just saying that the board is programed to work with native 1333 FSB chips thats all.

I hope that answered your question and more as i realize i gave you much more than you asked for

 

[Prototype]

this is good informatio i learned a lot from this cz i was confused also

so now cpu and ddr2 ram are all quad-pumped?

so the lower the ratio like 1:1 it is faster than a 2:1? what order does the ratio go is it RAM:FSB or FSB:RAM?

does the mobo FSB mean when u turn on the computer it will run at 1066 first and it will start overclocking to a higher fsb?

why does mobo show 2 fsb like 1333/1066?

 

[VampD]

Thanks Sora much appreciated, you answered all I needed to know and more! I guess a quick follow up question would be, why is a higher FSB + lower multiplier combo better then the vice versa combo? I thought a higher FSB puts more stress on the system?

 

[Prototype]

Thanks Sora much appreciated, you answered all I needed to know and more! I guess a quick follow up question would be, why is a higher FSB + lower multiplier combo better then the vice versa combo? I thought a higher FSB puts more stress on the system?

it depends

for overclockers it is bad because they cant overclock it very high from stock speed

for non overclockers it is good or if you going to buy from Dell, HP, etc.. that cant be overclocked

 

[Nagasama]

Thanks Sora much appreciated, you answered all I needed to know and more! I guess a quick follow up question would be, why is a higher FSB + lower multiplier combo better then the vice versa combo? I thought a higher FSB puts more stress on the system?

high FSB + lower mulitplier is better because high FSB = more cycles per second (ie 250 fsb performs 50 more cycles per second than 200 FSB) and multiplier = cpu cycles every FSB cycle.
so, 200 x 10 = 2000 or 2ghz
250 x 9 = 2250 or 2.25ghz
lets say for example that your rig tops out at 250fsb with a stock multiplier of 10 (= 2.5ghz).
you would likely be able to drop the multiplier to 9 and get say, 295 FSB (2.65ghz). faster overall, faster computer.
the higher the FSB, the more "valid" the overclock. most serious overclockers use the highest FSB (or external clock for us AMD guys) and the lowest multiplier possible. lol....my unlocked 5000 makes me lazy sometimes though. and yes it will require more voltage and cause more stress to the system. but it is also faster, even at the same speed. ie 2.5ghz high FSB vs 2.5ghz low FSB.

 

[VampD]

high FSB + lower mulitplier is better because high FSB = more cycles per second (ie 250 fsb performs 50 more cycles per second than 200 FSB) and multiplier = cpu cycles every FSB cycle.
so, 200 x 10 = 2000 or 2ghz
250 x 9 = 2250 or 2.25ghz
lets say for example that your rig tops out at 250fsb with a stock multiplier of 10 (= 2.5ghz).
you would likely be able to drop the multiplier to 9 and get say, 295 FSB (2.65ghz). faster overall, faster computer.
the higher the FSB, the more "valid" the overclock. most serious overclockers use the highest FSB (or external clock for us AMD guys) and the lowest multiplier possible. lol....my unlocked 5000 makes me lazy sometimes though. and yes it will require more voltage and cause more stress to the system. but it is also faster, even at the same speed. ie 2.5ghz high FSB vs 2.5ghz low FSB.

hmmm well said...I dont really know why you provided this example though:
so, 200 x 10 = 2000 or 2ghz
250 x 9 = 2250 or 2.25ghz
what if the the fsb was at 225 x10 = 2.25 ghz, would the 250x9 combo be faster even their both at 2.25 ghz? ( according to your last sentence it will be ) but I guess my question is asking...how much faster since their at the same ghz?

 

[alexsabree]

Oh lawd.. i got me a headache

The overall Ghz you get from your chip, and the frequency that your ram runs at will determine the speed of your processor/computer.

At least thats how I understand it.

 

[Nagasama]

hmmm well said...I dont really know why you provided this example though:
so, 200 x 10 = 2000 or 2ghz
250 x 9 = 2250 or 2.25ghz
what if the the fsb was at 225 x10 = 2.25 ghz, would the 250x9 combo be faster even their both at 2.25 ghz? ( according to your last sentence it will be ) but I guess my question is asking...how much faster since their at the same ghz?

i provided that example to illustrate that yes, indeed, the 250 fsb will be faster even though their overall speed is the same. we're talking benchmark and overclocking for the fun of overclocking here. you will not see too much difference in the real world.