Tech Support banner

Status
Not open for further replies.
1 - 7 of 7 Posts

·
Registered
Joined
·
507 Posts
Discussion Starter #1
Hi,

I recently flashed a custom VBIOS which pushed my GPU Voltage from 1.25v to 1.275v when under load, for my MSI GTX 970.

I am a little nervous you could say when it comes to the temperatures of how hot the GPU Memory(VRAM) is as well as the VRMs as I can't really see what temperatures they are running at (no sensors on Hwinfo64 or GPUz), I only see the actual GPU temp. When under load my GPU temp is never going past 66 degrees (aggressive fan profile).

I was hoping if you guys could tell me whether I should be concerned with how hot my GDDR5 Samsung GPU memory is getting as well as the VRM temperatures.
I was purchased these-
PcCooler RHS-03 Copper DDR DDR2 DDR3 Memory Cooler Ram Heatsink (Pack of 8) | eBay

to place ontop of the Memory (at the back/rear of the card, as it has no back plate) to help with the temperatures slightly.
I touched the Memory when it was under full load and could keep my finger on it for 1-2 seconds before I have to take my finger off because of the heat (silly to touch the GPU when powered, i know :eek:, but I didn't touch the VRM part) So I think the temperature is still okay but I'm still a bit concerned. I don't want to go and touch the back of the VRMs on the backplate as they could be hot and they're really just little pins poking out the back.


What I am trying to ask is if I should attach these tiny heatsinks to possibly help with lowering the temperatures or if I should not worry about the VRM and Memory and enjoy the power that my card can use - stable at 1540MHz Core and 4001MHz Memory with the higher voltage vbios. (Previously was only able to reach 1470ish core on stock bios)

If you were to recommend attaching these heatsinks to put on the back of the card, should I only put them on the 4 Memory chips or should I put a few along the VRM line too as many as I can fit in a line going down as shown in the image below.

https://gyazo.com/9cb2a51d70d4f76e6d81f5f74ee56a80.png


Please correct me if I've labelled this wrong.

I've purchased a Non-Contact IR heat gun which should be coming soon so I will be able to provide more accurate temp readings next week.
https://www.amazon.co.uk/gp/product/B01AT9TON0/ref=oh_aui_detailpage_o00_s01?ie=UTF8&psc=1


Thank you in advance for your time! :smile:
 

·
Registered
Joined
·
507 Posts
Discussion Starter #2
The temperatures for the VRM are showing a max of 85c, and the VRAM temps I've recorded a max of 65c.


I'm assuming these temps are perfectly fine/safe?
 

·
Team Manager - Networking, Moderator - Hardware
Joined
·
19,836 Posts
Don't know that anyone will have information that specific. Use whatever software was provided by the gpu maker and use that to monitor the temps. Whatever they are monitoring in the software will be the numbers everyone (including the manufacturer) reference to determine proper operating temp.
 

·
Registered
Joined
·
2,779 Posts
The VRM chokes can run around 120c so the temps you are seeing is quite normal.
Use Thermal pads if you must attach heatsinks, to prevent shorting the card out.
 

·
Registered
Joined
·
507 Posts
Discussion Starter #5

·
Registered
Joined
·
507 Posts
Discussion Starter #7
Helped slightly :D dropped temperatures by around 4-5 degrees.

I think I am still within safe temps for my memory, even without the heatsinks so probably getting no benefit from adding them on
 
1 - 7 of 7 Posts
Status
Not open for further replies.
Top