• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Which 970?

Which 970?

  • Total voters
    47
I like the look of the MSI badge when it lights up and the cooler looks nicer.
 
the gigabyte g1 is the best card available, not sure why that's not on your list. the msi is second best, beating the strix at its own gimmick (0 rpm fans at idle) by a significant margin. there's no reason at all to even consider the evga.
 
Odellus said:
the gigabyte g1 is the best card available, not sure why that's not on your list. the msi is second best, beating the strix at its own gimmick (0 rpm fans at idle) by a significant margin. there's no reason at all to even consider the evga.
Click to expand...

Maybe its too big for his case? That was the situation with me so I went MSI. The MSI barely fit on mine with about a centimeter of space left due to my hard drive rack.
 
So the MSI...

The Gigabyte G1 wont work because it is a long card, and I will be using 2 Way SLI.
 
The only bad thing about the MSI is it's hard to find, I'm waiting to hear from Amazon when they are getting them back in stock.
 
Scatcats22 said:
It is $200 more for only 10-15% more performance.

I also found a Sapphire Vapor-X R9 290 for $320. I heard that is the best cooler...
Click to expand...

i thought you already decided, based on your other thread, that you had settled on the 970. there's absolutely no reason to consider a 290. a 290X, ok, but not a 290.
 
polonyc2 said:
but performance wise (cooling, overclocking etc) it makes no difference?
Click to expand...

Backplates...

1. Give the video card extra support, adding rigidity to help with sagging, warping, and bending.
2. Help protect critical components from being damaged.
3. Help absorb heat (similar to a heat sync) and/or redirect airflow (Gigabyte Windforce) for possible better cooling and overclocking.
4. They look kewl, brah!

For the 980, NVIDIA started putting them on the reference cards for the first time, so there must be something to it.

GTX-980-123-9.jpg
 
Last edited:
I'm partial to EVGA and try to always buy cards from them cause they're good quality with excellent customer service. That said you can't really go wrong with Asus or MSI either but all things being equal I'd go EVGA even if it's a few bucks more.
 
The only reason to go with EVGA this time round would be for the support. The physical hardware is inferior to the competition (Asus, MSI, Gigabyte) whether in terms of performance or cooling.
 
n=1 said:
The only reason to go with EVGA this time round would be for the support. The physical hardware is inferior to the competition (Asus, MSI, Gigabyte) whether in terms of performance or cooling.
Click to expand...


Except where it is shown to be superior in both performance and cooling... like the FTW.
 
I've bought EVGA for years, but all I've read lately about the 970's they arent the best to go with other than service if something goes wrong. Between coil whine and overall build quality, it doesnt reassure me.

I'd avoid ASUS like the plague after dealing with them in the past and trying to get support... nightmare of epic proportions.

I'm thinking of going MSI this time around for a change. Still not 100% on that though. Would prefer a blower style cooler, which, atm, is onlyon EVGA it seems, though I guess there are others in the pipe coming out soon.
 
flopticalcube said:
It has the fastest clocks out of the box and is amongst the quietest of cards under at idle or load, equalling the Windforce which is much bigger.
Click to expand...

stock clocks mean absolutely nothing.

once again, where? because i am definitely not seeing what you're seeing.
 
I take the view that stock clocks are all that matters because in the OC lottery, they are the only guarantee.

anandtech FTW review said at 42db it was "one of the quietest midrange enthusiast cards we have ever reviewed."
 
flopticalcube said:
It has the fastest clocks out of the box and is amongst the quietest of cards under at idle or load, equalling the Windforce which is much bigger.
Click to expand...

Except the card is basically clocked to its limits already and overclocks like an absolute dud. For comparison purposes I can do 1506 boost 7800 mem on my Gigabyte 970s without even touching voltage (1.2V for one card, 1.206V for the other). And if it weren't for borked drivers not being able to compensate for ASIC differences, my boost clock would be 1524.

The only reason it would be amongst the quietest would be for the 0db passive mode, so not even a fair comparison to the WindForce that doesn't have the feature.

Even with the new BIOS on the FTW it is still louder than either the Asus, Gigabyte, or MSI, while still running 4C hotter at 73C compared to the hottest running Asus 970 (69C).

And to top it off, the FTW costs as much as the Gigabyte while only using 4 phase power for the core, while the other 3 all use 6 phases, with Asus and MSI being cheaper as well.

So I am not wrong when I say if you buy the FTW you'd buying it purely for EVGA's support.
 
Last edited:
n=1 said:
Except the card is basically clocked to its limits already and overclocks like an absolute dud.

The only reason it would be amongst the quietest would be for the 0db passive mode, so not even a fair comparison to the WindForce that doesn't have the feature.

Even with the new BIOS on the FTW it is still louder than either the Asus, Gigabyte, or MSI, while still running 4C hotter at 73C compared to the hottest running Asus 970 (69C).
Click to expand...

OC lottery means a single sample can't be used to know how a particular card will OC in general. That remains to be seen.

Its not the 0 db mode that gives the card its quiet rating under load, its the less aggressive fan profile.

Finally, you cannot compare sound levels unless the tests are set up to be identical. I doubt guru3d and anandtech share sound measurement set ups.

As seen in the OCN 970 owners thread, adding more power does not necessarily lead to higher OCs in the 970s. They reach their frequency threshold long before their power/temp threshold and that is down to how lucky you are in the chip you get.
 
And less aggressive fan profile means higher temperatures...

Have a look at the OC thread here and count how many people are able to push 1500+ boost and 7600+ memory.

flopticalcube said:
It has the fastest clocks out of the box and is amongst the quietest of cards under at idle or load, equalling the Windforce which is much bigger.
Click to expand...

flopticalcube said:
I take the view that stock clocks are all that matters because in the OC lottery, they are the only guarantee.

anandtech FTW review said at 42db it was "one of the quietest midrange enthusiast cards we have ever reviewed."
Click to expand...

I'm simply going by what I have, if you would like to provide additional data please do so. And at least I'm backing up my claims with data, whereas you seem to be basing your conclusions on an unqualified statement made by Anandtech. And by your logic, we may as well not compare anything because different setups, silicon lottery and all that. :rolleyes:
 
Last edited:
Sad, only have the choice of doing an EVGA ACX 2.0 or a Zotac via special order at Staples. Going to go with the Zotac.
 
I originally wanted the EVGA, but now looking at that MSI. The Gigabyte Gaming G1 is $20 more, and I can overclock the MSI/EVGA for just as good performance.
 
n=1 said:
And less aggressive fan profile means higher temperatures...

Have a look at the OC thread here and count how many people are able to push 1500+ boost and 7600+ memory.





I'm simply going by what I have, if you would like to provide additional data please do so. And at least I'm backing up my claims with data, whereas you seem to be basing your conclusions on an unqualified statement made by Anandtech. And by your logic, we may as well not compare anything because different setups, silicon lottery and all that. :rolleyes:
Click to expand...

Clearly not. To measure sound levels properly requires and anechoic chamber and calibrated equipment like SPCR does. In lieu of that, a repeatable setup can measure sound levels to within an acceptable precision.

If you don't understand what the silicon lottery is and how it contributes to OC ability, you can't understand how two identical cards OC so differently. To eliminate this factor we need a broad sample of each manufacturers cards and see what the limits are. This will eliminate the variability in the chip's ability to OC. Until then, the only guarantee you have are the manufacturer's stock clocks.

As far as chip temps go, like I said, 970 hits the speed wall long before temps get too toasty. This isn't Fermi.
 
So why did we vote for and have a poll then? That made no sense what so ever to do that and then do it how you did. Anyway I hope you are happy with what brand you buy. Most in here seem to be loyal to one brand and only one brand. It doesn't matter if they have a bad gpu come out they will still buy it to support their company. Again which is fine.

I'm just the type of person that needs the best bang for buck gpu though. Great customer service is indeed a great thing. But the bottom line is everyone here will not have that gpu for a long time. So they will never have to use that part to begin with. But it is nice to have that though.

Update the thread on the model you get and post some benchmarks, overall boost and overclocking with the ASIC score also. It will be interesting to see what you get and the model. Good luck on the purchase.

* Please note the review samples that are sent out for reviews are hand picked also. Not to mention the controlled climate they are kept in. Makes a big difference on the ending results. *
 
flopticalcube said:
Clearly not. To measure sound levels properly requires and anechoic chamber and calibrated equipment like SPCR does. In lieu of that, a repeatable setup can measure sound levels to within an acceptable precision.
Click to expand...

So how do you justify your claime that the FTW is "amongst the quietest of cards under at idle or load" then, given what you just said?

If you don't understand what the silicon lottery is and how it contributes to OC ability, you can't understand how two identical cards OC so differently. To eliminate this factor we need a broad sample of each manufacturers cards and see what the limits are. This will eliminate the variability in the chip's ability to OC. Until then, the only guarantee you have are the manufacturer's stock clocks.
Click to expand...

GPU boost adds considerable complexity to what your actual gaming boost clock would be, and that "out the box" number can be completely off the mark. For example the Gigabyte 970 is officially rated at 1329 boost, but mine runs 1366 out of the box without me touching a single thing. If you're going by the "max boost clock", then explain to me how it is that of all the games AnandTech tested, the FTW did not come anywhere near the 1418 MHz figure in one single game? The highest it achieved was 1373 MHz in BF4. If I had to guess, it's simply running into the TDP wall. More on that later.

With that being said I had a look at the gaming clocks FTW achieved out the box, and it's nothing spectacular. My Gigabyte 970 runs 1366 MHz out the box, and does not fluctuate regardless of the games I play (Crysis 1/2/3, Far Cry 3, Watch Dogs to name a few). Since Crysis 3 is the only common game I have with AnandTech, I'm not going to draw any conclusions apart from saying that lone data point does not give the FTW any clock advantage over my Gigabyte.

Back to TDP. The EVGA 970 FTW, like the Asus Strix 970, has the same amount of deliverable power. FTW uses 2x6 pins, Strix uses 1x8 pin, for a total of 225W. Is it any coincidence then, that the Asus Strix 970 consistently failed to reach 1500 boost across all five different reviews?

There is a fundamental power delivery limitation on the FTW that simply cannot be overcome, much like the Asus Strix 970. And before you tell me n=5 is still too small of a sample size, yes I fully understand that, and I'm well aware of what the silicon lottery is. The point being while n=1 might be an outlier, n=5 starts to paint a reasonable trend, and that trend simply points towards that overclocking is likely going to be limited by how much power the card is able to obtain. Both the FTW and Asus Strix lag behind the curve at 225W compared to MSI and Gigabyte, both of which allow a maximum of 300W deliverable power (6+8 pins for both).

As far as chip temps go, like I said, 970 hits the speed wall long before temps get too toasty. This isn't Fermi.
Click to expand...

So when I pointed out that the FTW is not "...superior in both performance and cooling", temps all of sudden become irrelevant? Keep shifting the goalposts chief.
 
Last edited:
n=1 said:
So how do you justify your claime that the FTW is "amongst the quietest of cards under at idle or load" then, given what you just said?
Click to expand...
anandtech's words, not mine. I would have expected them to keep testing conditions similar so that results can be compared.

GPU boost adds considerable complexity to what your actual gaming boost clock would be, and that "out the box" number can be completely off the mark. For example the Gigabyte 970 is officially rated at 1329 boost, but mine runs 1366 out of the box without me touching a single thing. If you're going by the "max boost clock", then explain to me how it is that of all the games AnandTech tested, the FTW did not come anywhere near the 1418 MHz figure in one single game? The highest it achieved was 1373 MHz in BF4. If I had to guess, it's simply running into the TDP wall. More on that later.

With that being said I had a look at the gaming clocks FTW achieved out the box, and it's nothing spectacular. My Gigabyte 970 runs 1366 MHz out the box, and does not fluctuate regardless of the games I play (Crysis 1/2/3, Far Cry 3, Watch Dogs to name a few). Since Crysis 3 is the only common game I have with AnandTech, I'm not going to draw any conclusions, but suffice to say that lone data point does not give the FTW any clock advantage over my Gigabyte.



So when I pointed out that the FTW is not "...superior in both performance and cooling", temps all of sudden become irrelevant? Keep shifting the goalposts chief.
Click to expand...
No temps are not relevant as long as they are under the thermal limits. No goal post shifting at all.

Again, you can't compare single samples. You need a larger pool of samples to eliminate variability... the silicon lottery.
 
You must log in or register to reply here.
Back
Top