Why is the same GPU faster?
If you compare the RX 6650XT with two fans to the one with three fans, you might notice that the one with more fans is a few megahertz faster in both base clock and turbo clock. There are also differences in overclocking and GPU performance.
Why is that? Doesn't AMD simply hand over the chip to the companies and they just take care of the chassis? To what extent will the differences be noticeable?
The chip manufacturer specifies what it considers to be optimal and safe, what the card manufacturer does, but remains its thing.
It doesn’t do much if it’s a lot, it might be 5% more performance. Instead of 60 FPS 63, it would not be worth the speech.
Better cooling allows a higher overclocking, therefore the differences.
Hello,
Both are not necessarily related to the number of fans. Of course, models with three fans are often better cooled and accordingly have more possibilities for overclocking etc.
Yeah, right. There are some key data and that was
Hardly. Most likely in cooling. You can see the volume and the temperatures. From the performance, the cards are almost identical.
With the connections, features and design you can also distinguish the cards
LG
Yes AMD and NVIDIA let the chips finish and then hand them over to the card manufacturers.
Although AMD and NVIDIA make a few specifications such as these chips to be operated nominally. However, the card manufacturers can overclock the chips as long as they do not exceed the specified power limit and the temperatures are not too high.
For better cooling solutions, the chips can therefore usually also be clocked somewhat higher.
However, AMD and NVIDIA sometimes also set limits on overclocking, so that a cheaper chip does not make a more expensive competition (but it is rarely the case since the chips usually differ from the performance).
Overall, however, the performance gain by the overclocking is quite manageable, so that usually the very expensive maps are hardly worth it if these are already close to the next better.
Here are some examples:
https://www.mindfactory.de/shopping_cart.php/basket_action/load_basket_extern/id/f24e36221b61b24350ebc9daad046ffc49b66ea975737cb9cd0
It is worth buying a cheap card with better chip than an expensive card with worse chip.
More fan = better cooling (at least in normal case) = allows higher clock = more power 🤷 ♂️
And the graphics cards with custom cooling are also sold overclocked by manufacturers.
The graphics cards are simply overclocked. Yes AMD gives the various custom design manufacturers of graphics cards only the chip, the manufacturers like Sapphire, XFX etc. but not only build a cooler on it, which can also adjust the bios of the Gpu for themselves and increase the clock speeds of the graphics chip and the Vrams. Higher clocks usually mean higher voltage and higher voltage more waste heat, so the cards that are overclocked are also those that usually have larger coolers and more airs.
In principle, you can also overclock your graphics card yourself, then you do nothing other than these manufacturers if you manually increase the frequencies for the chip and memory and raise the voltage easily.
However, how far and how well these overclocked is always a question of silicone lottery and that can not affect the manufacturers either, as they get the chips from AMD.
The better the cooling, the more clock can have the graphics card. The difference is minimal, but more cooling is always good.
Because 3 fans can dissipate more heat generated by overclocking.