Why do so many overclock by using LOW mutiplers?

[Twista]

For instance i seen someone using 9.5 x 200 or 10.5 x 190 why do they just use a high mutilpler as in 13.5 x 140 or 14x149.


Thanks for your replies.

 

[moonshinemadness]

You begin to get bottle necks in the PCI cards and such like because they have the original bus speed, so although the CPU is running faster the other components arent...I think anyway....correct me if im wrong.

 

[Corporate Thug]

the higher the FSB the better the performance. lets take for example a 1700+ B core. mine is doin 2200 Mhz right now but i get better performance if it's at 2200mhz@ 220*10 rather than 200*11

 

[bgeh]

and the higher the FSB, the higher the bandwidth. so you also get better performance

 

[Necrolezbeast]

It's like a Corvette, or say another car with a beefy engine and only having a .25in exhaust...thoughput is greatly compromised and performance is lost

 

[Alphanos]

Wow, there are some pretty bad replies here so far.

The CPU clockspeed is determined by the multiplier and the FSB. The FSB also controls the speed of the RAM in the system. Thus using a higher multiplier and lower FSB will not give performance as high as a low multipliers with a high FSB, since a high FSB is also increasing the speed of the memory. The memory speed increases will often improve performance even more than the clockspeed increases (that's why Intel and AMD are using new processors & chipsets now that support a higher FSB to allow faster communication with the memory).

 

[ChefJoe]

Sigh... 12*200 vs 12*201... difference of 12 MHz 24*100 vs 24*101... difference of 24 MHz. It's about the resolution. With the lower multipliers, the CPU can be overclocked with the FSB (which I've only seen in MHz increments) and you may be able to get an extra few MHz in the region of max stability.

Now, the other thing is that low multipiers often mean the "lower end" products at a relatively new/high FSB. Lower cost, clearer skies overhead (1.6a's were being made from the same silicon as the 2.4 GHz chips - the gamble was if it left the factory with a fault that earned it the lower speed or if it was the fact that people wanted to buy 1.6 GHz chips and not the expensive 2.4 GHz chip), and higher FSB for the same CPU MHz is usu. a good thing.

 

[sodcha0s]

Sigh... 12*200 vs 12*201... difference of 12 MHz 24*100 vs 24*101... difference of 24 MHz. It's about the resolution. With the lower multipliers, the CPU can be overclocked with the FSB (which I've only seen in MHz increments) and you may be able to get an extra few MHz in the region of max stability.

Now, the other thing is that low multipiers often mean the "lower end" products at a relatively new/high FSB. Lower cost, clearer skies overhead (1.6a's were being made from the same silicon as the 2.4 GHz chips - the gamble was if it left the factory with a fault that earned it the lower speed or if it was the fact that people wanted to buy 1.6 GHz chips and not the expensive 2.4 GHz chip), and higher FSB for the same CPU MHz is usu. a good thing.

OK, can somebody translate that gibberish? I mean, I KINDA get what he's saying, but it really doesn't make much sense....

 

[ChefJoe]

In the world of locked multipiers, overclocking is done by increasing the front speed bus (FSB). You can only bump that bus in MHz increments. If your CPU maxes out at 2110 MHz and your 24X multipiers mean you have to choose between 2090 and 2114 (for example) you'd be at a slight loss. With a 12X multiplier, you could hit 2102 and be closer to the max for that cpu. While this is a part of overclocking, the economics are probably a bigger part.

Now, there's economical concerns as well. Lower multipliers are going to be found in the slower products at a set FSB. When a (for arguement's sake, Intel) CPU moves to a 200 MHz FSB and there's 2.4C (12X multiplier) and 3.0C (15X multiplier), both CPUs are from the same material but one costs much less. Same material may mean that Intel had to sell these pieces as 2.4Cs because that's what people were buying, but this CPU might have been able to run as a 3.0C if people wanted to pay the money for it. The other possibility is that the processor had a defect that made it only stable up to 2.4 GHz and you're not going to be able to get beyond that.

Earlier posts were talking about higher FSBs. In general, yes, higher FSBs mean more bandwidth available to feed the CPU... while a good thing in itself, I think higher default FSBs (in the case of overclocking CPUs) tends to signal a new core and a new range of CPUs with the appropriate low and high speeds/multipliers. Alphanos's description of why they develop to a higher FSB is a good one, but I wouldn't consider that an explanation of overclocking. The memory bus and the FSB can run independently, but matched is the ideal.

As the original poster hasn't checked in or responded, I think it's safe to let this one drop.

 

[4x4expy]

Originally posted by: sodcha0s

OK, can somebody translate that gibberish? I mean, I KINDA get what he's saying, but it really doesn't make much sense....

I think maybe he accidentally posted in the wrong thread.

 

[soja]

lol @ 4x4expy.

Just keep it simple for the guy. He asked why people use higher fsb, all that jibber jabber (but good info) is only going to confuse him.

 

[4x4expy]

The question was accurately answered by Alphanos already.