Painting by numbers – Wrong values already ex works?
Before I test more pads in the near future, I also need to briefly touch on the actual business. Companies like Arctic, Gelid, Thermal Grizzly and Alphacool do not produce anything themselves, but import ready-made pads from the big manufacturers or OEMs (or even suppliers). By doing this at scale, they can also buy directly from the source. And this is exactly where the uncertainty begins, because you have to rely on what your trusted supplier writes in the data sheet.
It is really up to the manufacturers to either test these pads themselves or to buy only those products that have already been tried and tested and are also used professionally in many places. Serious Japanese companies can’t afford to misrepresent, but some no-name manufacturers from China can. On the following picture you can see a sample of such a pad with 11 W/(m*K), which in the end was not better than one with 7 W/(m*K). Or you test one with 14 W/(m*K) in the data sheet, which then does not achieve 10 W/(m*K). However, this can only be verified if you really have real comparison objects with correct values and a proper measurement setup.
The next graphic shows how badly such a “replacement” can work. The performance is a full 10 degrees worse, so the price doesn’t really help in the end. If you bring it into the right relation to true performance, then the original pad is hardly more expensive in the purchase. However, the customer doesn’t see it and would happily buy the supposedly better bargain here, even if it’s just a cheat pack in the end. Since he doesn’t know, he actually sleeps better.
In the graph below, we now see the performance of pad #3 from the trials with the OEM replacements (pictured above) and what it is really capable of. No, it’s not really that bad, but it’s clearly less than advertised and thus actually already failed. And I am quite sure that one or the other company sells exactly these pads as such with 14 W/(M*K). It’s in the spec sheet. By the way, we know this game from the thermal paste, where suppliers stubbornly keep the data sheets, although they have already changed the supplier several times and the quality differs extremely. But very few customers will notice it or once really check (be able to check) the content. There’s a lot of money to be made in this business.
In general: The providers of such pads are not primarily responsible for the lack of performance, but they have to ask themselves if they really didn’t know better and if not, why they didn’t test it themselves. With that experience in mind, I will then test incoming pads piece by piece in the near future. Equal right and equal conditions for all, the RTX 3080 I archived along with cooler, spare sleeves and enough thermal paste of the same batch in the box of good deeds.
Bisher keine Kommentare
Kommentar
Lade neue Kommentare
Artikel-Butler
Alle Kommentare lesen unter igor´sLAB Community →