P4 3010... the ultimate rip off...

[Raystonn]

Only until recently could even attempt to overclock an intel system without overclocking the memory.

Not quite true. I am running a Celeron 300A at 450MHz. I am not overclocking my memory, but I am overclocking the processor via the FSB. How was this done? By using PC100 memory. (The Celeron 300A was designed for use with PC66 memory.) If PC1066 memory were being sold then I would recommend that instead of Samsung PC800. However, it is not available and I have tested the Samsung RDRAM with success.

At the time of my earlier posts when I was providing recommendations on the overclocking of memory, I had thus far not encountered any memory that actually would safely overclock in a consistent manner. Each module had to be tested individually to see if it would work. If not tested absolutely thoroughly it might appear to work fine, while corrupting data here and there without your knowledge. This was not desirable. I gave my best analysis at the time. However, at this date things have changed. Due to Rambus's stringent quality-control and timing definitions, coupled with Samsung's extremely good yields, I have encountered RAM that overclocks with a near 100% success rate by 33%. These are the Samsung PC800 RDRAM modules. Take this information and do with it what you like. I am not going to force anyone to overclock their memory. But if you do, you should be careful to thoroughly test your system afterward.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Ncogneto]

so you went from a celeron 300a to a p4 setup with nothing in between?

It's not what they tell you, its what they don't tell you!

 

[Raystonn]

I also have a Pentium III system. The processor is overclocked via the FSB as well, but the memory itself is running within its own specifications. Overclocking the FSB has long been a good method of overclocking the processor. Overclocking the memory has never been very accepted as usually there is already memory available (pretty cheap as well) that goes beyond the specifications for your system. This allows you to overclock without pushing your RAM.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Ncogneto]

only when done in increments that allow our pci and AGP to stay in spec as well. Not easily done on a p3 system.

It's not what they tell you, its what they don't tell you!

 

[Raystonn]

Keeping the AGP and PCI in spec was a difficulty. I agree. This is why I usually keep to the clockspeeds that allow me to use the correct dividers. For example, I could have overclocked my Celeron 300A past 450MHz. Many people were able to achieve over 500MHz. A few people reported speeds of around 600MHz with super cooling. (I do not know if those reports were true.) But I like to ensure that the rest of my system remains within specifications. I stuck with 450MHz, using PC100 memory, and a 2/3 PCI/AGP divider. Many people search out components that will work on overclocked buses. This requires more time invested than I wish to give, and limits the hardware you are able to use. I do not suggest people travel that route.

-Raystonn


= The views stated herein are my personal views, and not necessarily the views of my employer. =

 

[Ncogneto]

The processor is overclocked via the FSB as well, but the memory itself is running within its own specifications.

Only possible with a select few p3's in a given range ( ie usually p3 700 100fsb) on a I815 board running at 133 fsb. Bx boards would be running the agp and pci out of spec. Most 100 fsb p3's will not run at 133 fsb ( exceptions being the 700's).

You always pull out the exception that proves the rule don't you? Thus my dancing in the grey area remark.

It's not what they tell you, its what they don't tell you!

 

[MadCat]

Only possible with a select few p3's in a given range ( ie usually p3 700 100fsb) on a I815 board running at 133 fsb. <b>Bx boards would be running the agp and pci out of spec.</b>

There were some BX boards (i.e., the ABIT BX6r2) that had 1/4 PCI multipliers that kicked in at 124MHz or higher. In this case, only the AGP bus would be running out of spec.

Most 100 fsb p3's will not run at 133 fsb ( exceptions being the 700's).

Many P3 550Es could also run at 133MHz FSB.