The CPUs that AMD or Intel makes are basically all the same with just different FSB speeds and/or different multiplier limits, right? I've heard that after they manufacture a bunch of CPUs, they test them and set the FSB/multi/cachesize accordingly for an individual chip's optimal performance and stability. Does anybody [i]know[/i] this to be true? Or do you think it's just a pricing scheme created by profit-hungry anti-Logans?
I don't know much about processor chips, so any insight as to what would cause this chaos factor in the manufacturing process is appreciated.
So the C2D E8200 and C2D E8600, which are both made from the Wolfdale chip, with the same FSB/L2cache but with either a 8x or a 10x multiplier... have different architectures?
each cpu family is very different, not only in the frequency or the stuff that says on the box. There are much more things that aren't mentioned anywhere that make a cpu fast. Sometimes , cpus that are supposed to be slower tend to be faster. You have to treat each one like it's unique and always , always look at the benchmarks. Intel and amd can make much faster cpus if they want. The whole faster and faster thing is a lie. Everything is prearranged . but if you want to get the best cpu out there , always buy the one with the smallest architecture , 45 nm , 32 (in some time).When a cpu series is launched what intel and amd do is that they usually make 3 models with different frequencies and the equivalent prices. Your best bet is to buy the slowest one and overclock it more than the best of the line for example e8400 vs e8500 the same thing with different clocks. and yes they lock the multiplier , for example the black edition of phenom. They can and they do it. in your face
for example e8400 vs e8500 the same thing with different clocks. and yes they lock the multiplier , for example the black edition of phenom. They can and they do it. in your face
So they don't lock the multiplier at different level to ensure best performance and stability? Because that is what I've heard, and that's what I want to know.
if they want to do that for your own good , why don't they set it to the "best" option by default and leave it unlocked so that you can tweak it as much as you want ?
if they want to do that for your own good , why don't they set it to the "best" option by default and leave it unlocked so that you can tweak it as much as you want ?
And then they give you a replacement per the warranty when you don't listen to them and fry your chip?(Rhetorical question not based on my own views as I don't have any yet)
but not that i am thinking about it again , that would require the motherboard to set the value to a memory in the cpu , which would require new motherboard standards hmmm
but not that i am thinking about it again , that would require the motherboard to set the value to a memory in the cpu , which would require new motherboard standards hmmm
It would also require the additional installation of non-volatile memory onto the cpu, so I think it's safe to say that they can't read the last value of the cpu multiplier.
no , it could be read via th pins on the cpu . In the same way your bios reads the information about your cpu model, thus i don't think that would even require a change in the sockets structure
this guy has written something interesting "The multiplier lock is used by CPU manufacturers to prevent consumers and dealers from overclocking the CPU. When overclocking became mainstream, profit margins for CPU manufacturers lowered because users wouldn't need to upgrade to a faster processor."
no , it could be read via th pins on the cpu . In the same way your bios reads the information about your cpu model, thus i don't think that would even require a change in the sockets structure
I think your wrong and that any multiplier to use, other than the stock/lock, is stored somewhere on the mobo. But it doesn't really matter much cuz even if they couldn't read it, it really isn't much proof that they lock the multiplier for our own good.
no , it could be read via th pins on the cpu . In the same way your bios reads the information about your cpu model, thus i don't think that would even require a change in the sockets structure
I think your wrong and that any multiplier to use, other than the stock/lock, is stored somewhere on the mobo. But it doesn't really matter much cuz even if they couldn't read it, it really isn't much proof that they lock the multiplier for our own good.
what i meant is that if they wanted to make the info stored on the cpu they could read it via the pins . i didn't actually say that they can do it.
what i meant is that if they wanted to make the info stored on the cpu they could read it via the pins
I think any info stored on the cpu isn't rewritable as it would need to be to store the value of the last multiplier. Again though, even if reading from that data from the cpu is impractical or even hypothetically impossible, it just means that locking the multiplier to avoid paying out warranties is a valid argument, but a phenom black edition with a 3 year warranty is valid evidence that this isn't the case.
As far as I know, it might still be the case, but it probably isn't. I checked newegg, and Intel also includes a 3 year warranty on their extremes, but in that case, I'd suggest that the huge pricetag would overcome all the extra replacements that they might have to issue. I'm still researching, but I think that it's just a way to sell the consumer a product that's not as good as it should be so they can charge much more for the less intentionally-degraded ones. Since both Intel and AMD lock multipliers and charge accordingly, that would seem to be collusion within a duopoly, however since this is most often seen as illegal, the fact that they've been able to do it for so long without much controversy makes me think that there might me a more reasonable explanation.