8 replies to this topic

[Nemo]

OCC has just published a review on the MSI N275GTX Twin Frozer OC - http://www.overclock...gtx_twinfrozer/

[Silverfox]

Looks good Frank!

[PingoPongo]

Wanted to get this card... but its too expensive compared to the HD 4890 here.
I mean it does outperform the HD 4890 in many cases but paying an extra $65 (720 EEK) just isnt worth it...

Anyway a very nice review! :thumbs-up: The card is also nice

[Ziphead]

OK... I need some help here.

This can't be right.... can it?

Temps on the MSI GTX 275 Frozr

Here is a quote from the overclocking section of the review.


"With the maximum overclock and 100% fan speed the N275GTX idled at 31 Celsius and loaded at 51 Celsius after looping 3dmark06 game test 3 for an hour. With the fan speed controlled by the drivers running at the stock clock speeds the card idled at 31 Celsius and loaded at 54 Celsius. Pretty decent numbers when all things are considered."


Pretty decent numbers???? I would say those are unbelievable numbers!

How can the numbers be the same stock as with the max overclock? I know 1 is with the fans at 100% and the other with the fans controlled by the drivers... but what are the chances that they would come out the same?

and here's an image that I found that does not agree.

[ccokeman]

OK... I need some help here.

This can't be right.... can it?

Temps on the MSI GTX 275 Frozr

Here is a quote from the overclocking section of the review.


"With the maximum overclock and 100% fan speed the N275GTX idled at 31 Celsius and loaded at 51 Celsius after looping 3dmark06 game test 3 for an hour. With the fan speed controlled by the drivers running at the stock clock speeds the card idled at 31 Celsius and loaded at 54 Celsius. Pretty decent numbers when all things are considered."


Pretty decent numbers???? I would say those are unbelievable numbers!

How can the numbers be the same stock as with the max overclock? I know 1 is with the fans at 100% and the other with the fans controlled by the drivers... but what are the chances that they would come out the same?

and here's an image that I found that does not agree.

The chances are pretty good that they come out the same. Idle temperatures are Idle temperatures. If the voltage to the card does not go up then the heat generated is not going to go up when you are idling with 2d clock speeds. Fair enough reasoning on that part considering my i7 920 idles at close to the same temperature at 2.6 and at 4.1GHz. Under load the drivers just don't hammer the fans until you really hot (above around 80 C). The coolers just do their job that well not to mention I use a well ventilated case so the ambient inside my case is the same as outside it. Then you increase fan speed without increasing voltage to the GPU and even with the clock speed increase the voltage set with the 3d clocks is not going to change so the only increase in heat comes from the increased frequency of the core not an increase in voltage . Increased fan speed no jump in voltage and if the cooler works well enough you can have lower temperatures under load overclocked with the fan speed at 100%.

The review you used to get that image used Furmark to generate the load tempoeratures. Furmark generates load temperatures far higher than you will see in a game much the same way LinPak will hammer your CPU temps higher than F@H ,Prime 95 or OCCT will. It is a benchmark and stability test. I use it to test stability and then pull temperatures with something realistic. 2560x1600 4xAA 16xAF Loop Game tests 3and 4 for an hour while monitoring the temps. Not to mention the test area was 82/83F on the image you linked and mine is down at 74/75 f so I hope this helps with your questions.

[RJR]

ccokeman, Very nice review on the MSI GTX275.

I noticed you did a little folding, post OC (to verify stability), and was just a little curious of the PPD you achieved??

[ccokeman]

I did not check because the variability of the wus makes a consistent PPD count not really feasible but it did give me a boost to my average production.

[Ziphead]

The chances are pretty good that they come out the same. Idle temperatures are Idle temperatures. If the voltage to the card does not go up then the heat generated is not going to go up when you are idling with 2d clock speeds. Fair enough reasoning on that part considering my i7 920 idles at close to the same temperature at 2.6 and at 4.1GHz. Under load the drivers just don't hammer the fans until you really hot (above around 80 C). The coolers just do their job that well not to mention I use a well ventilated case so the ambient inside my case is the same as outside it. Then you increase fan speed without increasing voltage to the GPU and even with the clock speed increase the voltage set with the 3d clocks is not going to change so the only increase in heat comes from the increased frequency of the core not an increase in voltage . Increased fan speed no jump in voltage and if the cooler works well enough you can have lower temperatures under load overclocked with the fan speed at 100%.

The review you used to get that image used Furmark to generate the load tempoeratures. Furmark generates load temperatures far higher than you will see in a game much the same way LinPak will hammer your CPU temps higher than F@H ,Prime 95 or OCCT will. It is a benchmark and stability test. I use it to test stability and then pull temperatures with something realistic. 2560x1600 4xAA 16xAF Loop Game tests 3and 4 for an hour while monitoring the temps. Not to mention the test area was 82/83F on the image you linked and mine is down at 74/75 f so I hope this helps with your questions.

Thanks... yes that helps a lot. Thanks for the extensive explanation.

Great review... I can't wait for my MSI GTX 275 Frozr to show up from Newegg.

This card has to be the best bang for the buck out there... especially when you take into account the OC headroom.

[ccokeman]

Glad I could clear it up! Enjoy the card