cancel
Showing results for 
Search instead for 
Did you mean: 

why there are no any cooling for the vram and vrm on rog strix 1050ti

ir3555
Level 7
what a shame, charging premium price for inadequate cooling as a result in poor overclocking....
asus even prohibits users to mod additional heatsinks on them.... stupid protective sticker.
15,109 Views
21 REPLIES 21

Korth
Level 14
Reference GTX1050Ti is rated for 75W/300W TDP and max operating temp of 97C. If your card exceeds rated temp (97C) then the cooing is inadequate. And the Strix 1050Ti has DirectCU II plus a pair of ASUS 0dB Wing-Blade fans.

But instead of paying "premium price" for the Strix 1050Ti you could pay "normal price" for reference 1050Ti ... and get a card with substantially inferior cooling, lol.

So why do you say it's inadequate cooling? What are your maximum observed temps under sustained peak load?

Few people bother to mod fans and heatsinks onto GPU cards. If they upgrade the cooling at all, it'll almost always be to waterblocks.
"All opinions are not equal. Some are a very great deal more robust, sophisticated and well supported in logic and argument than others." - Douglas Adams

[/Korth]

Korth wrote:
Reference GTX1050Ti is rated for 75W/300W TDP and max operating temp of 97C. If your card exceeds rated temp (97C) then the cooing is inadequate. And the Strix 1050Ti has DirectCU II plus a pair of ASUS 0dB Wing-Blade fans.

But instead of paying "premium price" for the Strix 1050Ti you could pay "normal price" for reference 1050Ti ... and get a card with substantially inferior cooling, lol.

So why do you say it's inadequate cooling? What are your maximum observed temps under sustained peak load?

Few people bother to mod fans and heatsinks onto GPU cards. If they upgrade the cooling at all, it'll almost always be to waterblocks.


i am not talking about gpu die cooling, but the cooling of vram and vrm on the card.
this strix 1050ti has 100% 75w max stock power limit settings and it can be raised to 200% 150w max.


maybe you should go check some 1050ti from other brands like galax oc which has active vrm cooling from the main heatsink.
or gigabyte 1050ti g1 which even has thermal pad on the back of the pcb serving for the REAL cooling function of the backplate.

Menthol
Level 14
Like Korth says, it's additional cooling for Vram and VRM is just not needed on a 1050ti, the cooling supplied by the factory heatsink is enough

xeromist
Moderator
Interesting you should mention backplate cooling. It doesn't make a tangible difference. None. It's just an aesthetic accessory that provides a bit of extra support. A PCB is not as good a heat conductor as the heat spreaders on the front of the chips so even if you can get the back a few degrees cooler it does not make the chips any cooler.
Relevent test: https://www.youtube.com/watch?v=OBsxuNXyDJk

So what Korth and Menthol are getting at is whether your VRM is actually overheating or if you are just assuming it will.

What are your actual temps?
A bus station is where a bus stops. A train station is where a train stops. On my desk, I have a work station…

xeromist wrote:
Interesting you should mention backplate cooling. It doesn't make a tangible difference. None. It's just an aesthetic accessory that provides a bit of extra support. A PCB is not as good a heat conductor as the heat spreaders on the front of the chips so even if you can get the back a few degrees cooler it does not make the chips any cooler.
Relevent test: https://www.youtube.com/watch?v=OBsxuNXyDJk

So what Korth and Menthol are getting at is whether your VRM is actually overheating or if you are just assuming it will.

What are your actual temps?


still, i am not talking about the gpu die temp.
vrm cooling and vram cooling should never be cheap out.
four phases vcore for 75w may be okay. but for 150w max, i wont take that risk.
unless, the additional 6pin connector and the 150w max power limit settings on the bios are just for "aesthetic".


fyi
http://www.pcdiy.com.tw/assets/images/768/54a2da157c6153e93c3a30783e6a154f.jpeg
http://www.pcdiy.com.tw/assets/images/768/8686e445c925cb6de031cadcc6ece88e.jpeg

if you want to know more about how thermal pads on the backplate help reducing vram temp and vrm temp, you can go check gamersnexus reviews of previous evga disaster and the latest zotac 1080ti amp extreme review showing how bad backplate is trapping the vrm heat.

ir3555 wrote:
you can go check gamersnexus reviews of previous evga disaster and the latest zotac 1080ti amp extreme review showing how bad backplate is trapping the vrm heat.



not trying to dispute the importance of VRM thermal pads at all... but rather, only that specific sentence seems queer, almost oxymoron-ish. so i'm counter-inquiring you on that.


didn't that very same video mention that the ASUS Strix cards being simply the best? in their conclusion, zotac's 1080ti overdesigned but all of that didn't do nutthin.

just doubly making sure that i wasn't hearing things, and that i pride on my own information accuracy, here's the time where he said it: https://youtu.be/iiLSxhyiZss?t=843

i'm wondering if the same logic can be applied on ASUS' 1050Ti, and questioning whether the pads are really that necessary.
heck, even if i'm wrong in my suspicion, that's fine too. imma just take the damn card, stick some pads there, and job done.
no siggy, saw stuff that made me sad.

ir3555 wrote:
if you want to know more about how thermal pads on the backplate help reducing vram temp and vrm temp, you can go check gamersnexus reviews of previous evga disaster and the latest zotac 1080ti amp extreme review showing how bad backplate is trapping the vrm heat.


Sure, a high end card with a bad backplate design could trap heat. However I don't know if the Strix 1050ti has a bad backplate design and it's definitely not a high end card. Do you have any evidence for _this card_ that the cooling is inadequate or that the backplate is causing problems? It seems like you are assuming based on issues with a different class of card from a different company.
A bus station is where a bus stops. A train station is where a train stops. On my desk, I have a work station…

Menthol
Level 14
We understand the difference between GPU die and VRM, VRAM and what you are talking about, we just don't see the concern
Do you own this card, are you actually having any issues with the card
ASUS must have tested and decided that the air passing through the heatsink provides adequate cooling to these components
You have the option to purchase from another manufacturer

Korth
Level 14
My (anecdotal) experience is that backplates do improve temps but a tiny margin, <2C at best. On higher-powered cards which run hot, a good thermal-pad-and-backplate arrangement does pull a little heat off the PCB. Lesser cards don't benefit as much from backplates because the front-side cooling is already enough to keep them fairly cool, less excess heat on the PCB means less heat can transfer to the backplate. And, to be fair, most backplates are primarily cosmetic, many don't mate well with thermal hotspots or even bother to mount thermally conductive pads/fillers, very few backplates (that I've seen) are designed with functionally efficient heat transfer or heat radiation in mind. Yes, there are some bad (even very bad) backplate designs out there which will actually increase real temps on the GPU card.

1050Ti is one of the weakest of the NVIDIA GeForce 10 Series cards - 1030, 1050, 1050Ti, 1060, 1070, 1080, 1080Ti, Titan X (GP102), Titan Xp.
It's roughly one-fifth as powerful as a 1080Ti. It has 4GB GDDR5 - comparable to a GTX980. Yet reference 1050Ti VRMs are much more efficient than those seen on the very fastest of the fastest factory-overclocked GTX980 cards (Galax HOF and EVGA Kingpin), and those volt-modded cards pulled far more than 150W.

Reference 1050Ti cooling is a standard active heatsink arrangement with one centered fan. Many 1050Ti cards are actually "half-sized" versions with stripped-down VRMs and cooling, yet they overclock fairly well. The Strix 1050Ti is a "full-sized" card with DirectCU II cooling (which has direct contact only with the GPU but also covers the VRM and VRAM components), it has two better-and-larger-than-reference fans, it has a backplate, it follows reference PCB and VRM layout (for compatibility, I think) but it uses far better quality components with substantially higher performance specs.

So I'm of the opinion that a Strix 1050Ti should easily overclock well with the ASUS-provided cooling.

You do understand that the VRM and VRAM hardware is immediately adjacent to the main GPU ASIC - as physcially and electrically close as possible in circuit to maximize signal integrity and minimize response latency - so it's also covered by the DirectCU II heatsink (which is itself all covered by forced airflow)? This is already "active" VRM cooling, quite comparable to what's on the Galax card although perhaps not as hyped in the marketing, it just doesn't get more "active" without putting (small) direct-contact heatpipes onto the (small) VRM chips themselves - and that would consequently leave less space to put (big) heatpipes onto the (big) GPU ASIC. And - assuming the Strix 1050Ti lacks the thermal pads seen on the Gigabyte card - you can purchase enough of the very-highest-efficiency thermally conductive pads/strips to cover the entire card for a little less than five bucks.

Although I'm with the others in this thread - I ask whether you've actually observed "unacceptable" high temps on an overclocked Strix 1050Ti or if you're just shopping for the very best of the best overclocking 1050Ti card you can find? If it's the latter then I'd even go so far as to say don't waste your time, the Strix is perhaps the best and certainly among the best at overclocking and cooling, but it's still a waste of time compared to a GTX1060 or better, lol.
"All opinions are not equal. Some are a very great deal more robust, sophisticated and well supported in logic and argument than others." - Douglas Adams

[/Korth]