Palit GTX 750 Ti KalmX temperature

[QORD]

Hi,

I've read quite a few reviews about this card and checked the forum but it seems no one is having the same problem
with this gfx. The problem is the temperature and I am not sure if I should return the gfx or if it's working as intended.
My Palit GTX 750 Ti KalmX runs on 95°C running FurMark in an open case without any additional fans. CPU temperature
is around 40°C so that can't be the reason for such temperatures. In game the temperature reaches around 80°C
which is what I would expect by the FurMark stress tests compared to all the reviews out on the internet.

Oh the ambient temperature was 21°C. I wrote an email to Palit and reply is kinda neutral:

The passive heatsink will work better with some air flow inside the PC case from the system fan.
Otherwise please contact the store where you bought the graphic card to have further check or claim warranty.

Cheers,
Blaž

[quest_for_silence]

Hi,

I've read quite a few reviews about this card and checked the forum but it seems no one is having the same problem
with this gfx. The problem is the temperature and I am not sure if I should return the gfx or if it's working as intended.
My Palit GTX 750 Ti KalmX runs on 95°C running FurMark in an open case without any additional fans. CPU temperature
is around 40°C so that can't be the reason for such temperatures. In game the temperature reaches around 80°C
which is what I would expect by the FurMark stress tests compared to all the reviews out on the internet.

Oh the ambient temperature was 21°C. I wrote an email to Palit and reply is kinda neutral:

The passive heatsink will work better with some air flow inside the PC case from the system fan.
Otherwise please contact the store where you bought the graphic card to have further check or claim warranty.

Cheers,
Blaž

Given that it's out of any enclosure, put a 120mm fan in front of the Palit card, spin it at a paltry 600rpm (or similar pace): does anything change in temp?

By the way, broadly speaking Palit answer sound right and in good accordance with some respectable web review (not to mention that your S3 works a lot better than the Palit stock chunk of metal).

[QORD]

Given that it's out of any enclosure, put a 120mm fan in front of the Palit card, spin it at a paltry 600rpm (or similar pace): does anything change in temp?

By the way, broadly speaking Palit answer sound right and in good accordance with some respectable web review (not to mention that your S3 works a lot better than the Palit stock chunk of metal).

OK, I will do a quick fan test but that isn't the aim of a passive product if I'm not mistaken. Anything I should keep my
eye on? I might test it with S3 which is laying around but that costs me of the warranty on a brand new gfx plus
suddenly a gfx costs 40eur more!

So no one with same testing environment around the forum?

[quest_for_silence]

OK, I will do a quick fan test but that isn't the aim of a passive product if I'm not mistaken.

I'm not sure whether you're mistaken: broadly speaking, any fanless cooler rely upon case airflow (unless the cooler is the case itself), and whether there's no airflow their performance is often rather sub-par.
So eventually, if you think you can use a passive VGA without any airflow, and still getting low temps, I guess you're wrong.

Anything I should keep my eye on?

Using some benchmark like either Heaven or Valley, you should just check whether with a minimal airflow the GPU temp goes down, and whether increasing that airflow the temp will decline more sharply: those could be two symptoms that everthing is working as it should.

With reference to your reported temps, they seem compatible with TechPowerUp! findings, so that apparently they're not pointing out any underlying thermal issue.

I might test it with S3 which is laying around but that costs me of the warranty on a brand new gfx plus
suddenly a gfx costs 40eur more!

I just told you that the S3 is a more effective cooler, not to use it to test the card: it would work better than the KalmX one, if you want to stick with that card.

Even re-applying the thermal paste (something like AC MX-2 is the bare minimum) often gives some benefit.

[QORD]

I'm not sure whether you're mistaken: broadly speaking, any fanless cooler rely upon case airflow (unless the cooler is the case itself), and whether there's no airflow their performance is often rather sub-par.
So eventually, if you think you can use a passive VGA without any airflow, and still getting low temps, I guess you're wrong.

I didn't and I'm not expecting low temps, but 95°C is to high, that's all I'm saying. I will report my results and I will try my luck
with another gfx, maybe anothter Palit or maybe Zotac.

I just told you that the S3 is a more effective cooler, not to use it to test the card: it would work better than the KalmX one, if you want to stick with that card.

Even re-applying the thermal paste (something like AC MX-2 is the bare minimum) often gives some benefit.

Do you have any reviews in mind?

Btw this review talks about 73°C without any case air flow other then PSU. Or did google translate mess the translations?

[CA_Steve]

There's always this summary...

Palit GTX 750 Ti KalmX 2GB - Passive card
Techpowerup - card reaches 83C in test, where Nvidia starts to reduce Boost clocking...so, may require directed air for max performance. YMMV.
Kitguru - card reaches 56C gaming and 64C in Furmark in a well ventilated case with water cooled CPU.
Computer Base - worth reading for thermal analysis
Guru3D - 80C load temp; 55C with ventilation

Cards above 50W TDP typically need some airflow across the fins in order to keep below throttling.

[xan_user]

If you are comfortable with the 80c at gaming, why worry?
- or do you plan on running furmark all the time?!?

my fanless asus gtx750 got very hot running furmark/gaming and throttled in my open air rig, so i ducted my cpu fan towards it.

have you tried yours in a case, with some airflow?

[quest_for_silence]

I will try my luck with another gfx, maybe anothter Palit or maybe Zotac.

There's more than a chance that it couldn't make any substantial difference, but whether an RMA process won't cause any loss for you, why not...

Do you have any reviews in mind?

Are you really asking me for reviews?

I think I already gave you the link to one of the more well regarded (TechPowerUp!, 83°C), but you can check also Guru3D (74-75°C with minimal ventilation).

Only german sites credit the KalmX of lower temp: the one you linked (CB, 64°C) clearly used active case airflow to cool the card (with a 200mm fan), and with reference to fully passive operation they stated (to be fair, that's the Google translation): - "A special case is the operation in a "OpenBench Case", ie without any housing. In this case, the cooling concept no longer works and the card is even warmer. But the natural convection on the heat sink then is no longer sufficient for the heat exchange.". From what I was able to understand even within their enclosure (CM 690 III and R4) they said that the card throttled: "...reaches 73 degrees Celsius, which leads in conjunction with the Power Target, that the graphics card clocks down. The target temperature of 80 ° C is not achieved in this case... In Fractal Design R4... Without chassis fan hardware is then too warm and working at Target Temperature". Obviously I may miss something in that translation, but my own experience with passive cooling tell me that donkeys don't fly and that KalmX is a donkey (even on this, I may be wrong).

On the other hand, HardwareLuxx (65°C) didn't say whether they rely upon case airflow or not, but as they thanked Thermaltake I guess there's a chance they are using one of their enclosure.

[QORD]

Sorry, with reviews I had S3&Palit in my mind, not the Palit gfx itself.

[quest_for_silence]

Sorry, with reviews I had S3&Palit in my mind, not the Palit gfx itself.

There should be a couple of user reviews in the GTX 750Ti about some co-forumers who installed the S1 on it.

[QORD]

If you are comfortable with the 80c at gaming, why worry?
- or do you plan on running furmark all the time?!?

Good point Well I just want to get ready for the summer as much as possible, nothing else.

I've attached R5 default fan directly on a side case and the temp went down from 95°C to 83°C
so you guys were right, seems there are no thermal issues.

[quest_for_silence]

so you guys were right, seems there are no thermal issues.

If you don't need the money you can make of, you may slap the S3 on that card (and report here the relevant temps, please).

[QORD]

If I do so I will most definitely report to you guys. I must order better thermal paste and use it on CPU also
- Phobya LM or do you recommend anything else? Shall I avoid LM and just go for Thermal Grizzly Kryonaut?
I'm sorry, didn't do any deep researches on that topic yet.

[quest_for_silence]

If I do so I will most definitely report to you guys. I must order better thermal paste and use it on CPU also
- Phobya LM or do you recommend anything else? Shall I avoid LM and just go for Thermal Grizzly Kryonaut?
I'm sorry, didn't do any deep researches on that topic yet.

Liquid metal thermal compounds have often turned out (as well as rather expensive) of somewhat difficult application/removing, and reported as corrosive for aluminium and alloy heatsinks: from the few things I read on Phobya's one, it wouldn't seem that good (though thermal pastes reviews are often sort of black magic).

By the way, I'm expecting a syringe of Kryonaut these days, I will report you whether it's that effective against more proven, less expensive pastes (mainly AC MX4 and AS5, but I've also some more else) in real life.

[QORD]

Cool, I will wait with the purchase till you get back to me with the results, thanks!

[QORD]

Hey, just a quick update on S3 & Palit 750 Ti .. just installed it and the improvement on completely passive
configuration is not that remarkable. During gaming the temp raises a lot slower compared to stock Palit
cooler and it stops around 79°C while the ambient temp is 23°C. However FurMark results are a lot better, 83°C.

I will also test it with same case fan as before and compare the results with the previous ones (83°C).

I still have few small modifications to make (gluing one GPU screw since its completely loss) and maybe
adjust thermal pads a bit better. Overall im pleased with "final" result, shame I wasted additional money for
the passive Palit version but maybe other 750 ti wouldn't perform as good with S3, I for sure don't want to know

Oh, I used the Thermal grizzly kryonaut thermal paste on GPU and CPU. CPU temps aren't any lower, would
probably be a lot better if I had any airflow.

[quest_for_silence]

Oh, I used the Thermal grizzly kryonaut thermal paste on GPU and CPU. CPU temps aren't any lower, would
probably be a lot better if I had any airflow.

I completed my testing today, these are my findings (I'd stress these are just my findings, not a general rule):
FIRST SESSION

Code: Select all

╔══════╦══════════════════════════════╦═══════════╦═══════════╦══════════╦══════════╗
║ Rank ║             TIM              ║ LOAD (°C) ║ IDLE (°C) ║ MAX (°C) ║ MIN (°C) ║
╠══════╬══════════════════════════════╬═══════════╬═══════════╬══════════╬══════════╣
║    1 ║ Arctic Cooling MX4           ║ 45.5      ║ 34.5      ║ 47       ║ 34       ║
║    2 ║ Thermal Grizzly Kryonaut     ║ 46.5      ║ 34.5      ║ 47       ║ 31       ║
║    3 ║ Thermalright Chill Factor 3  ║ 46.5      ║ 33.5      ║ 52       ║ 31       ║
║    4 ║ Prolimatech PK1              ║ 46.5      ║ 34        ║ 48       ║ 33       ║
║    5 ║ Arctic Silver 5              ║ 47.5      ║ 34.5      ║ 48       ║ 34       ║
║    6 ║ IC Diamond Carat 7           ║ 47.5      ║ 34.5      ║ 48       ║ 34       ║
║    7 ║ Noctua NT-H1                 ║ 47.5      ║ 35        ║ 49       ║ 32       ║
║    8 ║ CM Thermal Compound Kit      ║ 48        ║ 35        ║ 49       ║ 31       ║
║    9 ║ Shin Etsu X-23-7783D         ║ 48        ║ 34.5      ║ 49       ║ 31       ║
║   10 ║ Arctic Ceramique 2           ║ 48.5      ║ 35        ║ 50       ║ 33       ║
╚══════╩══════════════════════════════╩═══════════╩═══════════╩══════════╩══════════╝

SECOND SESSION

Code: Select all

╔══════╦══════════════════════════════╦═══════════╦═══════════╦══════════╦══════════╗
║ Rank ║             TIM              ║ LOAD (°C) ║ IDLE (°C) ║ MAX (°C) ║ MIN (°C) ║
╠══════╬══════════════════════════════╬═══════════╬═══════════╬══════════╬══════════╣
║    1 ║ Thermal Grizzly Kryonaut     ║ 45.5      ║ 33.5      ║ 47       ║ 31       ║
║    2 ║ Thermalright Chill Factor 3  ║ 46        ║ 33.5      ║ 47       ║ 32       ║
║    3 ║ Prolimatech PK1              ║ 46.5      ║ 34.5      ║ 47       ║ 33       ║
║    4 ║ Arctic Cooling MX4           ║ 47        ║ 34.5      ║ 48       ║ 32       ║
║    5 ║ IC Diamond Carat 7           ║ 47        ║ 34.5      ║ 48       ║  0       ║
║    6 ║ Noctua NT-H1                 ║ 47        ║ 34        ║ 48       ║ 32       ║
║    7 ║ Arctic Silver 5              ║ 47.5      ║ 34.5      ║ 49       ║ 34       ║
║    8 ║ CM Thermal Compound Kit      ║ 47.5      ║ 35        ║ 49       ║ 31       ║
║    9 ║ Shin Etsu X-23-7783D         ║ 48.5      ║ 35        ║ 50       ║ 32       ║
║   10 ║ Arctic Ceramique 2           ║ 48.5      ║ 35        ║ 49       ║ 33       ║
╚══════╩══════════════════════════════╩═══════════╩═══════════╩══════════╩══════════╝

THIRD SESSION

Code: Select all

╔══════╦══════════════════════════════╦═══════════╦═══════════╦══════════╦══════════╗
║ Rank ║             TIM              ║ LOAD (°C) ║ IDLE (°C) ║ MAX (°C) ║ MIN (°C) ║
╠══════╬══════════════════════════════╬═══════════╬═══════════╬══════════╬══════════╣
║    1 ║ Thermalright Chill Factor 3  ║ 45.5      ║ 32.5      ║ 47       ║ 30       ║
║    2 ║ Prolimatech PK1              ║ 46.5      ║ 33.5      ║ 47       ║ 30       ║
║    3 ║ Thermal Grizzly Kryonaut     ║ 46.5      ║ 33        ║ 47       ║ 31       ║
║    4 ║ Arctic Cooling MX4           ║ 46.5      ║ 34        ║ 47       ║ 30       ║
╚══════╩══════════════════════════════╩═══════════╩═══════════╩══════════╩══════════╝

MEAN OF THE THREE SESSIONS

Code: Select all

╔══════╦══════════════════════════════╦═══════════╦═══════════╦══════════╦══════════╗
║ Rank ║             TIM              ║ LOAD (°C) ║ IDLE (°C) ║ MAX (°C) ║ MIN (°C) ║
╠══════╬══════════════════════════════╬═══════════╬═══════════╬══════════╬══════════╣
║    1 ║ Thermalright Chill Factor 3  ║ 46.0      ║ 33.2      ║ 48.7     ║ 31       ║
║    2 ║ Thermal Grizzly Kryonaut     ║ 46.3      ║ 33.7      ║ 47.0     ║ 31       ║
║    3 ║ Prolimatech PK1              ║ 46.3      ║ 34.0      ║ 47.3     ║ 32       ║
║    4 ║ Arctic Cooling MX4           ║ 46.5      ║ 34.3      ║ 47.3     ║ 32       ║
╚══════╩══════════════════════════════╩═══════════╩═══════════╩══════════╩══════════╝

As you can see the Thermal Grizzly Kryonaut performed unremarkably well, and IMHO the relevant cost (here in Italy) is apparently unjustified for air cooling (at least, that's my conclusion).

[CA_Steve]

Luca, what were the test conditions?
- processor
- cooler
- load app
- curing times ( do any of these benefit from curing time?)

[quest_for_silence]

Luca, what were the test conditions?
- processor
- cooler
- load app
- curing times ( do any of these benefit from curing time?)

Here is the testbed:

TIM_test_SPCR_2.jpg

Here we refer to such a system as a "muletto" (in english it would sound as "forklift" or "spare car", but I guess the exact correspondence would be lost in translation), it's the secondary rig the young son of my brother uses to study and game, when he go visit his grandparents.

Therefore specs are pretty outdated, but enough to let him play with some ease from League of Legends to CS:GO on a 1680x1050 display.
It's an AMD Phenom II X4 940 Black Edition (140W TDP) cooled by a Cooler Master Hyper 412S (the ASUS in background is a GTX660), it was on my bench because my nephew messed up with the operating system and it didn't boot into Windows 8.1 anymore.

To load the system I used Prime95 SmallFFT, each test lasted approximately about 2 hours (for each TIM half an hour of idling, one hour Prime, half an hour idling): curing time was minimal, since Kryonaut is supposed to not need it, mostly the first half hour.
One partial exception is the second session with Arctic Silver 5: given I was not satisfied of its first outcome, I left the system running a 1 thread Prime95 Blend for 24hrs before the relevant second session (eventually AS5 scored quite the same).

Are you thinking to change your favourite TIM, Steve? AC MX2/MX4 look promising, money and performance wise.

[CA_Steve]

MX4 looks pretty good.

[QORD]

Hey, here are my latest temps with S3 properly mounted on Palit 750 Ti Kalmx:
- ingame 75-76°C
- furmark 79-80°C

Room temp was 22°C. I will test it with a system fan and keep you guys updated.

Well for zero airflow Kryonaut seems a waste at least in my case

[xan_user]

im always curious in these test what would happen if each paste was tested multiple times. my bet is the same paste could have a few degrees difference each time it was applied and mounted.

[CA_Steve]

The paste is just one part of the equation...and a lesser part at that.

[leon.]

im always curious in these test what would happen if each paste was tested multiple times. my bet is the same paste could have a few degrees difference each time it was applied and mounted.

Have you seen skinneelabs.com? I was kind of surprised.