[H] got over 1500mhz on their MSI 970 Gaming Card on Stock Volts

[Attic]

Overclocking Results from [H]




Very close to pulling trigger on this card if it comes in stock at newegg.

Very solid review, makes the MSI 970 Gaming card look like one of the best cards to get, new TFV cooler as well, so the fan issues of the TFIV series should be alleviated. 8pin and 6 pin is another nice touch. Fans idle unless temp goes beyond 60c.



Negative side, another 970 OC card benched/reviewed against bone stock competition..... wtf guys? This makes no sense, i'd like to know if this is an nVidia preferred reviewing method or if reviewers are making this call on their own. We've seen tons of 780 OC vs 290 OC reviews, but all of sudden the 970 drops and we seeing highly OC'd 970's (anand used the Evga FTW 970 which is as easy to find in the wild as a unicorn) vs the 947mhz stock 290's and the stock 780's? Why sandbag the competition in the reviews of the 970?

 

[toyota]

Mine will do that too but throttle back and forth a bit a bit under heavy load since my TDP is lower.

and I will just copy and paste what I said there about their review:

Power consumption makes no sense at all. My 780 uses way more power and my MSI 970 uses WAY less than the 970 here. In Metro LL I saw my 780 at stock 1019 boost using a whopping 96 watts more than my 970 at 1290 (out of box clocks). My typical load in demanding games on the 780 was 370 to 390 watts and my 970 only went to just over 300 watts and that was with oced to 1466. Yet here we see the 970 using MORE power than the 780? I have yet to see anything like that anywhere else and it defies logic

 

[iiiankiii]

Wow. That's nuts. Maxwell is a beast. I can't wait until big maxwell hits. Hopefully, it'll be atleast 35% faster than the GTX 980 and with the crazy overclocking headroom. I don't want to deal with multi-gpu setups anymore.

 

[BrentJ]

Overclocking Results from [H]




Very close to pulling trigger on this card if it comes in stock at newegg.

Very solid review, makes the MSI 970 Gaming card look like one of the best cards to get, new TFV cooler as well, so the fan issues of the TFIV series should be alleviated. 8pin and 6 pin is another nice touch. Fans idle unless temp goes beyond 60c.



Negative side, another 970 OC card benched/reviewed against bone stock competition..... wtf guys? This makes no sense, i'd like to know if this is an nVidia preferred reviewing method or if reviewers are making this call on their own. We've seen tons of 780 OC vs 290 OC reviews, but all of sudden the 970 drops and we seeing highly OC'd 970's (anand used the Evga FTW 970 which is as easy to find in the wild as a unicorn) vs the 947mhz stock 290's and the stock 780's? Why sandbag the competition in the reviews of the 970?

I made the call myself. We are already talking a $349 card up against more expensive GPUs, GTX 780 etc... Plus, this is the initial review, always in the initial review I like to get a baseline of performance. We don't know how it's going to perform, so getting a baseline allows us to see how it does up against reference stock clocked cards. As stated on page 2 future reviews may use factory overclocked cards as the comparison as price dictates. We line up our retail reviews by price when we can.

 

[Termie]

...

Power consumption makes no sense at all. My 780 uses way more power and my MSI 970 uses WAY less than the 970 here. In Metro LL I saw my 780 at stock 1019 boost using a whopping 96 watts more than my 970 at 1290 (out of box clocks). My typical load in demanding games on the 780 was 370 to 390 watts and my 970 only went to just over 300 watts and that was with oced to 1466. Yet here we see the 970 using MORE power than the 780? I have yet to see anything like that anywhere else and it defies logic

This MSI Gaming is by far the most power-hungry 970 available. I have no doubt that [H] is completely correct in its power measurements. Same thing was found at other sites. Ex: http://www.bit-tech.net/hardware/graphics/2014/09/19/nvidia-geforce-gtx-970-review/11

The dual PCIe power connectors, one being 8+2-pin, obviously are having an effect here. You don't get extreme stock voltage overclocking without higher-than-average stock volts and high power draw. It's pretty clear that the MSI 970 Gaming uses more power than a 980 while underperforming it, so it won't win any efficiency awards. But it will be a popular card.

 

[toyota]

This MSI Gaming is by far the most power-hungry 970 available. I have no doubt that [H] is completely correct in its power measurements. Same thing was found at other sites. Ex: http://www.bit-tech.net/hardware/graphics/2014/09/19/nvidia-geforce-gtx-970-review/11

The dual PCIe power connectors, one being 8+2-pin, obviously are having an effect here. You don't get extreme stock voltage overclocking without higher-than-average stock volts and high power draw. It's pretty clear that the MSI 970 Gaming uses more power than a 980 while underperforming it, so it won't win any efficiency awards. But it will be a popular card.

OK then Maxwell is a joke if their review is correct. That is a 145 watt gpu pulling more power than the 780. so basically that is a 252 watt 970 all by itself if you subtract the 90 watts they say the system uses without a card. and then another 25 watts for the 12% OC. yeah right...

nearly clock for clock that card is using about 100 watts more than mine. that is laughable and not even possible. again their 780 power usage is low and their 970 power usage is insanely high. this is not the first time they have had screwy power consumption results that did not align with every other review showed not to mention logic.

 

[f1sherman]

OK then Maxwell is a joke if their review is correct. That is a 145 watt gpu pulling more power than the 780. so basically that is a 252 watt 970 all by itself if you subtract the 90 watts they say the system uses without a card. and then another 25 watts for the 12% OC. yeah right...

Nah... that 970 in [H] is 180-ish W.

It looks weird but it's possible. 780 is kinda golden consuming 170-ish.

CPU is clearly not idling - it's around 60W

 

[toyota]

Nah... that 780 in [H] is 180-ish W.

It looks weird but it's possible. 780 is kinda golden consuming 170-ish.

CPU is clearly not idling - it's around 60W

um out of the box factory oced 970 using more power than the 780? please show me another review with such screwy results. again they are showing nearly 100 watts MORE than my 970 at the same clocks.

 

[tviceman]

There are several other Gaming 970 reviews showing it consuming less watts than a 780. [H]'s measurements are an anomaly in that regard.

http://www.guru3d.com/articles_pages/msi_geforce_gtx_970_gaming_review,7.html
Even Termie's link -> http://www.bit-tech.net/hardware/graphics/2014/09/19/nvidia-geforce-gtx-970-review/11 shows it consuming considerably less power than a GTX780.

 

[f1sherman]

That's a VERY rough estimate chart.
Done with old drivers on old cards, and with new drivers on new ones.
Using God know how many different apps (We simply stress the GPU, not the processor.) Dafuq is with everyone hiding their heat app?

Check out GTX 750/Ti , 7950 and R7 265 on that chart.
Using 76 , 93 , 138W and 144W respectively

Nothing against them, they are doing their job fine, but I take G3D noise and power numbers with lots of salt.
Matter of fact, I wouldn't vouch for anyone when it comes to noise and power

Particular review done with such and such cards should be fine by itself, but iterating these results and doing historical comparison is too much for method(s) used.

 

[BrentJ]

For power and temp I use real games to get results, I watch the power meter, note the highest peak power for all cards in the same game, at the same spots in the game so it sampled from the same spot for comparison.

 

[toyota]

For power and temp I use real games to get results, I watch the power meter, note the highest peak power for all cards in the same game, at the same spots in the game so it sampled from the same spot for comparison.

I do too and so do most sites. your numbers are beyond crazy as it would mean Maxwell is hardly any more efficient than the Kepler. in fact any reviewer showing a 145 watt card using more power than the 780 would be questioned. again though you guys have shown some random odd power usage results in the past too.

 

[f1sherman]

For power and temp I use real games to get results, I watch the power meter, note the highest peak power for all cards in the same game, at the same spots in the game so it sampled from the same spot for comparison.

Ouch... peak power ingame usage is something that should correlate with average ingame power consumption, but the two are not the same.

What we are mostly interested is average gaming consumption - preferably done with 100% GPU usage gaming scenario.

 

[toyota]

Ouch... peak power ingame usage is something that should correlate with average ingame power consumption, but the two are not the same.

What we are mostly interested is average gaming consumption - preferably done with 100% GPU usage gaming scenario.

I noted the peak difference when I tested and it was 75 to almost 100 watts more that the 780 used over the 970. it was about that same range for average usage too when I tested. and my 780 was only like 13 mhz over stock where as the 970 was oced pretty high out of the box already exceeding 1300 in most cases.

 

[f1sherman]

So you are willing to use peak power numbers as an indication of power consumption?
Adding yet another systematic error which is perfectly avoidable...
Unless you are really interested in power maximum; im not.

If you are, there you go then - see what we get ^^. I have little doubt [H] numbers are read correctly

 

[Silverforce11]

If MSI choose to pump more volts into by default, then ofc it will OC very well on "stock volts", but at the expensive of much more power use, due to the square nature of vcore.

[H] should definitely do an average load over the game duration as well as peak, because peak by itself doesn't mean much, as we may find its a temporary spike that isn't sustained for long periods.

 

[toyota]

If MSI choose to pump more volts into by default, then ofc it will OC very well on "stock volts", but at the expensive of much more power use, due to the square nature of vcore.

[H] should definitely do an average load over the game duration as well as peak, because peak by itself doesn't mean much, as we may find its a temporary spike that isn't sustained for long periods.

my card uses more volts at 1.225 so if anything the Gaming card uses lower volts than mine.

 

[tviceman]

It's just one review. Like I said, every other review shows the MSI gaming GTX 970 using less watts than the GTX 780.

Here is another one: http://www.kitguru.net/components/graphic-cards/zardon/msi-gtx-970-gaming-4g-review/20/

 

[toyota]

It's just one review. Like I said, every other review shows the MSI gaming GTX 970 using less watts than the GTX 780.

Here is another one: http://www.kitguru.net/components/graphic-cards/zardon/msi-gtx-970-gaming-4g-review/20/

yeah the 780 is using 100 watts more there which makes sense. its probably oced a little but still even stock 780 would still be using WAY more power not less like in the Hard review.

 

[BrentJ]

We use a power meter to look at total system wattage at the wall. The numbers are live, and there is no way to capture all the power numbers from one game to calculate an average.

 

[gamervivek]

um out of the box factory oced 970 using more power than the 780? please show me another review with such screwy results. again they are showing nearly 100 watts MORE than my 970 at the same clocks.

http://www.hardware.fr/articles/928-7/consommation-efficacite-energetique.html

 

[Silverforce11]

We use a power meter to look at total system wattage at the wall. The numbers are live, and there is no way to capture all the power numbers from one game to calculate an average.

Then get a power meter that displays the min, average & max, instead of you eye-balling it for peak (I had no idea this is how you guys did it all this time!) which is not accurate at all to sum up "power use" to inform your readers.

 

[Silverforce11]

http://www.hardware.fr/articles/928-7/consommation-efficacite-energetique.html

It looks like Maxwell's efficiency is part hard/soft-ware as suggested by some review sites, since depending on the load, it could be very efficient or the same as Kepler.

Interesting that Anno loads it like crazy, it must be the DX11 compute they use.

 

[tviceman]

We use a power meter to look at total system wattage at the wall. The numbers are live, and there is no way to capture all the power numbers from one game to calculate an average.

I'm not dismissing your results or saying they're wrong. There are always going to be reviews of hardware that skew from the normal results most other review sites get. It happens all the time. This time, it just happens to be your review's power numbers. Your numbers were probably accurate for the exact samples you tested but, because they are skewed far from the normal results most reviews are showing, are not indicative of most gtx970 samples.

 

[gamervivek]

It looks like Maxwell's efficiency is part hard/soft-ware as suggested by some review sites, since depending on the load, it could be very efficient or the same as Kepler.

Interesting that Anno loads it like crazy, it must be the DX11 compute they use.

http://beyond3d.com/showpost.php?p=1876415&postcount=2363