Thank you for the advice Th0nt and Mikehhhhhhh,
I´m not trying to go that way, I just using that to fully load my cards ("Brute force"), since loading some games (GTA-Shadow of the Tomb Raider) not always do.
My simulations takes some time loading the gpu, so the software used by this guys clearly forces my card in no time (I think even better than many benmarchs stress test). I need that info to get as stable as I can when my real load start (Scheduled on may).
In fact, I hated those guys (Minners), they make these cards tooo expensive.
But now, they have a lot information (and software) to help me prepare my computer for the heavy load.
Love and hate, hate and love.
Going to minning the days the computer is not making simulation?
I don´t know yet, but I don´t think so, not for now.
For now, I got some results, but can´t find how to check mem temp on HiveOs, because is a linux based os.
On stock, no extra fan or heatsink, the cards got 106 MH, with 70º on GPU one card and the other can´t remember, but much lower, something like 62º (No memory temp). Within minutes one card got down to 80 MH and the lower card on 90 MH.
30 minutes later they still got this results but the room was as hot as a mini sauna.
I saw a lot of red alerts and numbers on the HiveOs interface.
I stopped all (I think that was dangerous for my card)
I set up all the fans again. Tune to 90% speed the side fans, added my extra heatsinks (As you saw on the pictures), then change the memory to 1200, the power to 285 for each card, the core -300 and fan 95% all the time and the results changed:
Started equal, but in minutes the cards changed to this:
Upper card: 101 MH, 60º no idea memory temp (That is drive me crazy).
Lower card: 107 MH, 51º
I left it all day, and the result was the same for almost 9 hours.
I think the the 50º celsius is a good threshold to start but... I want to be sure because that is the GPU temp, not memory temp.
I want to know the memory temp as I see in windows (Junction). Anyone have an idea on how to do so on HiveOs?