F5F Stay Refreshed Power Users Overclocking Overclocking may be damaging the GTXZ 1070.

Overclocking may be damaging the GTXZ 1070.

Overclocking may be damaging the GTXZ 1070.

H
Hecchicero
Member
171
02-27-2016, 09:06 AM
#1
evening all, just a quick question.
I had put on an overclock of +300 on the Core and + 300 on the memory, when i tried firestrike the whole program froze. ive now found a decent setting and the card seem to be running stable and cool.
When firestrike crashed when i did the +300 on the core and memory, would this of damaged the card? (damage that hasnt showen its ugly head yet?). or does the card have fail safes? ive got my power target set to 112%, is this a potential safety feuture?
H
Hecchicero
02-27-2016, 09:06 AM #1

evening all, just a quick question.
I had put on an overclock of +300 on the Core and + 300 on the memory, when i tried firestrike the whole program froze. ive now found a decent setting and the card seem to be running stable and cool.
When firestrike crashed when i did the +300 on the core and memory, would this of damaged the card? (damage that hasnt showen its ugly head yet?). or does the card have fail safes? ive got my power target set to 112%, is this a potential safety feuture?

S
sheepbottle5
Junior Member
44
03-05-2016, 08:43 AM
#2
Your card is in good condition. To determine its limits, you need to identify where it might fail. The cards include built-in safeguards. Heat is the main issue since prolonged exposure to high temperatures can harm a card. However, the cards reduce speed (throttle) when temperatures reach 85°C for NVidia. This slows them down until the heat subsides. It doesn't matter what your clock settings are; it automatically adjusts.
S
sheepbottle5
03-05-2016, 08:43 AM #2

Your card is in good condition. To determine its limits, you need to identify where it might fail. The cards include built-in safeguards. Heat is the main issue since prolonged exposure to high temperatures can harm a card. However, the cards reduce speed (throttle) when temperatures reach 85°C for NVidia. This slows them down until the heat subsides. It doesn't matter what your clock settings are; it automatically adjusts.

A
alexfreedom123
Junior Member
19
03-09-2016, 01:25 PM
#3
Sorry about the error in the title! GTX 1070*
A
alexfreedom123
03-09-2016, 01:25 PM #3

Sorry about the error in the title! GTX 1070*

L
Lord_Thurlvin
Junior Member
10
03-16-2016, 12:01 PM
#4
Your card is in good condition. To determine its limits, you need to identify where it might fail. The cards include built-in safeguards. Heat is the main issue since prolonged exposure to high temperatures can harm a card. However, the cards reduce speed (throttle) when temperatures reach 85°C for NVidia. This slows them down until the heat subsides. It doesn't matter what your clock settings are; it automatically adjusts.
L
Lord_Thurlvin
03-16-2016, 12:01 PM #4

Your card is in good condition. To determine its limits, you need to identify where it might fail. The cards include built-in safeguards. Heat is the main issue since prolonged exposure to high temperatures can harm a card. However, the cards reduce speed (throttle) when temperatures reach 85°C for NVidia. This slows them down until the heat subsides. It doesn't matter what your clock settings are; it automatically adjusts.

M
MC_Alfin
Junior Member
37
03-16-2016, 12:17 PM
#5
Thanks so much! I feel relieved now.
M
MC_Alfin
03-16-2016, 12:17 PM #5

Thanks so much! I feel relieved now.

G
Goldensoul133
Member
129
03-17-2016, 05:53 PM
#6
The memory requirement of 300 should be manageable, but a high core count will cause issues. Begin with short intervals of 5-10 and perform a windowed stress test to observe both the OC program and the test together. Gradually increase the clock by 5-10 every few minutes until it crashes, which will indicate your maximum limit.

OC memory behaves slightly differently, but the changes are minor. Execute the stress test identically, adjusting memory in the same way, but this time focusing on detecting artifacts during the test. When they emerge, you'll have reached your memory threshold.

Set your power limit to 120%.
G
Goldensoul133
03-17-2016, 05:53 PM #6

The memory requirement of 300 should be manageable, but a high core count will cause issues. Begin with short intervals of 5-10 and perform a windowed stress test to observe both the OC program and the test together. Gradually increase the clock by 5-10 every few minutes until it crashes, which will indicate your maximum limit.

OC memory behaves slightly differently, but the changes are minor. Execute the stress test identically, adjusting memory in the same way, but this time focusing on detecting artifacts during the test. When they emerge, you'll have reached your memory threshold.

Set your power limit to 120%.