Now, this is a question that has been grating against me for some time now.
And I am seriously thinking about building a new rig either for myself or someone.
So I thought I would ask in this forum of knowledgeable gentlemen.
Many companies(including Overclockers, Scan and many more) sell pre-overclocked bundles and rigs.
To do this, they claim to 'cherry pick' components to use in these rigs. To define 'cherry pick', my understanding of the term is that: company has a selection of components in stock, in their warehouse or wherever. Some of these components will exceed the manufacturers specifications and will readily overclock, sometimes very much so. Others will only work at the maximum rating as set by the manufacturer. From this set of stock they select components, test them on the rig they want to overclock, and from those that readily do overclock they choose to go into the rig.(In practice it will probably be a more sophisticated process with components being separated into tiers; I digress).
The upshot is that there are certain components, which, when tested, are shown not to be able to overclock past the manufacturer's specifications.
The question is : what happens to these components?
I would guess, like most commercial enterprises, these companies seek to get the maximum profit, so in due course sell these components(which they know cannot be overclocked) back to their customers at the standard stock price.
That would mean that buying a stock(non cherry picked, guaranteed overclockable) component from such company, vastly decreases the odds of 'winning the silicon lottery'(which I define as: purchasing a readily overclockable component from a random pool of components for sale). Given the increased(sometimes most likely vastly so) pool of low qualify components in the general component pool this would mean that for some companies, (I would guess) that the odds of buying a stock component which overclocks well is close to zero because of this practice.
CPUs and motherboards are likely to be most affected.
Is my model correct ? Am I missing some pertinent information ?
My purchasing pattern(and more importantly, the recommendations I give to others, is likely going to be affected, and I really would like to know if the model as I have above surmised, is accurate).
My thanks to anyone who takes the time to answer.
Gavin786
And I am seriously thinking about building a new rig either for myself or someone.
So I thought I would ask in this forum of knowledgeable gentlemen.
Many companies(including Overclockers, Scan and many more) sell pre-overclocked bundles and rigs.
To do this, they claim to 'cherry pick' components to use in these rigs. To define 'cherry pick', my understanding of the term is that: company has a selection of components in stock, in their warehouse or wherever. Some of these components will exceed the manufacturers specifications and will readily overclock, sometimes very much so. Others will only work at the maximum rating as set by the manufacturer. From this set of stock they select components, test them on the rig they want to overclock, and from those that readily do overclock they choose to go into the rig.(In practice it will probably be a more sophisticated process with components being separated into tiers; I digress).
The upshot is that there are certain components, which, when tested, are shown not to be able to overclock past the manufacturer's specifications.
The question is : what happens to these components?
I would guess, like most commercial enterprises, these companies seek to get the maximum profit, so in due course sell these components(which they know cannot be overclocked) back to their customers at the standard stock price.
That would mean that buying a stock(non cherry picked, guaranteed overclockable) component from such company, vastly decreases the odds of 'winning the silicon lottery'(which I define as: purchasing a readily overclockable component from a random pool of components for sale). Given the increased(sometimes most likely vastly so) pool of low qualify components in the general component pool this would mean that for some companies, (I would guess) that the odds of buying a stock component which overclocks well is close to zero because of this practice.
CPUs and motherboards are likely to be most affected.
Is my model correct ? Am I missing some pertinent information ?
My purchasing pattern(and more importantly, the recommendations I give to others, is likely going to be affected, and I really would like to know if the model as I have above surmised, is accurate).
My thanks to anyone who takes the time to answer.
Gavin786
