Do I make a good low-overclock?

[jordibax]

I read somewhere that an effective overclock is done is done everytime you reach up a certain number of mhz, like cycles, depending of the card architecture. This means that you can increase 5 mhz on the core or ram gpu and don´t get any increase performance untill you get exactly 6 more mhz, it`s called somewhat i don´t remember and you must calculate this formula for each you are dealing with. I left this issue until i feel like to spend time on it, meanwhile, i decided to make an easy overclock that consists in see what is my gpu core and overclock it at the same mhz that the fastest card with the same core is clocked. For example, There is one Asus x800 gto that has R48O clocked at 400 mhz, in this case, i would clock at 540 mhz since x850xt PE pci-e has the same core and theorically this r480 chip is supposed to deal with this clocks easily. Then I see if they have the same mem modules, and if it is so, I clock the same way too. If the highest card had some extra clooling system, then I try to add something similar. Is this low-overclocking system totally safe or i am missing something?so far, i have not fried any card...

[phranq]

I read somewhere that an effective overclock is done is done everytime you reach up a certain number of mhz, like cycles, depending of the card architecture. This means that you can increase 5 mhz on the core or ram gpu and don´t get any increase performance untill you get exactly 6 more mhz, it`s called somewhat i don´t remember and you must calculate this formula for each you are dealing with. I left this issue until i feel like to spend time on it, meanwhile, i decided to make an easy overclock that consists in see what is my gpu core and overclock it at the same mhz that the fastest card with the same core is clocked. For example, There is one Asus x800 gto that has R48O clocked at 400 mhz, in this case, i would clock at 540 mhz since x850xt PE pci-e has the same core and theorically this r480 chip is supposed to deal with this clocks easily. Then I see if they have the same mem modules, and if it is so, I clock the same way too. If the highest card had some extra clooling system, then I try to add something similar. Is this low-overclocking system totally safe or i am missing something?so far, i have not fried any card...


No overclocking is "totally safe". Sometimes the company uses the cores that failed as say an x850xt and make an x800gto out of it. So if I'm understanding you right, no, the only guarantee is that it will run at the specs it was built to run at, anything higher is a bonus and will never be "totally safe". It's still fun to do it though. :)

[jordibax]

No overclocking is "totally safe". Sometimes the company uses the cores that failed as say an x850xt and make an x800gto out of it. So if I'm understanding you right, no, the only guarantee is that it will run at the specs it was built to run at, anything higher is a bonus and will never be "totally safe". It's still fun to do it though. :)
Really?So if i got you, if they make a chip that is supposed to run at X mhz, and some part of the total production do not resist that clock, they use these chips to release a lower card range with that chips on them running at a lower clocks? then, from now on, I would not surprise if even using that not too agressive method i fry some card. znx

[phranq]

Really?So if i got you, if they make a chip that is supposed to run at X mhz, and some part of the total production do not resist that clock, they use these chips to release a lower card range with that chips on them running at a lower clocks? then, from now on, I would not surprise if even using that not too agressive method i fry some card. znx


It really depends on the card, it varies from card to card. Sometimes you will get a card that will work just fine, but often times a card doesn't meet the standards, and gets down clocked.