evga SC atx2.0 gtx 970 psu requirements

[chrcoluk]

Yes I only got my gtx 760 not that long ago but seems the specs required to play games since next gen launch has skyrocketed so I ordered a SC 970gtx. The finances I think will be hard to justify what I spent this year so will try to recoup some of what I spent by selling the gtx 760.

Question is, is a 600W supply with a 40A 12v rail ok to use on this card?

Only spec's I can find is 500w required.

 

[Suprfire]

Yeah 600W will work with power to spare. what PSU is it exactly?

You could run the system you have in your sig + the GTX 970 on a high quality 400-450W if you wanted to.

 

[n=1]

More than enough.

Just to give you an idea, with my 4930K @ 4.5GHz/1.39V and 2x Gigabyte 970 @ 1506/7700, I'm only pulling 520-550W from the wall when playing Crysis 3. A decent 600W could get you by even with 2 970s in tow.

 

[Armenius]

If you look at the 970 reviews on [H] you'll see their test system is only using 320-350W. I would add ~50W to that for case fans and extra drives. So the 500W power supply requirement from NVIDIA is pretty accurate.

 

[amd7674]

I'm not trying to hijack the thread, and I don't want to open a new one...
I'm looking to upgrade my current GTX650TI GPU on my gaming HTPC.
I have ivy bridge i5 CPU non-oc at stock voltages, one ssd and one laptop HDD in it with 5 fans.
Due to the size of my case HTPC case I cannot fit any cards with heat pipes sticking out, so I'm limited basically to EVGA offerings and Zotac. Currently I'm leaning toward EVGA FTW, due to its fan spindown feature, which would be awesome when watching movies on my HTPC via xbmc.

My current PSU is Antec VP450W (ATX 12V v2.3):
+3.3V@24A, +5V@15A, +12V1@18A, +12V2@18A, [email protected], [email protected]

Would that PSU be enough or it will be pushing it?
Based on PSU calculator site, I should be fine with my current setup... however...

I can swap Corsair VX550W from my other HTPC with Antec, but I would like to avoid it due to the hassle. As you guys know wire management in tight spaces is not fun

The specs for Corsair VX550W (ATX12V V2.2) are:
+3.3V@30A, +5V@20A, +12V@41A, [email protected], +5VSB@3A

would be wise and safer to swap/use Corsair? or can I keep on using my current Antec one? or I should pull trigger on new 600W PSU?

Thanks for your help

 

[chrcoluk]

thanks guys

psu is coolermaster silent pro m 600w

 

[flopticalcube]

I'm not trying to hijack the thread, and I don't want to open a new one...
I'm looking to upgrade my current GTX650TI GPU on my gaming HTPC.
I have ivy bridge i5 CPU non-oc at stock voltages, one ssd and one laptop HDD in it with 5 fans.
Due to the size of my case HTPC case I cannot fit any cards with heat pipes sticking out, so I'm limited basically to EVGA offerings and Zotac. Currently I'm leaning toward EVGA FTW, due to its fan spindown feature, which would be awesome when watching movies on my HTPC via xbmc.

My current PSU is Antec VP450W (ATX 12V v2.3):
+3.3V@24A, +5V@15A, +12V1@18A, +12V2@18A, [email protected], [email protected]

Would that PSU be enough or it will be pushing it?
Based on PSU calculator site, I should be fine with my current setup... however...

I can swap Corsair VX550W from my other HTPC with Antec, but I would like to avoid it due to the hassle. As you guys know wire management in tight spaces is not fun

The specs for Corsair VX550W (ATX12V V2.2) are:
+3.3V@30A, +5V@20A, +12V@41A, [email protected], +5VSB@3A

would be wise and safer to swap/use Corsair? or can I keep on using my current Antec one? or I should pull trigger on new 600W PSU?

Thanks for your help

Yes. The Corsair has a single 12V rail which is what you will need for a 970 otherwise you are nearing the Antec's limit of 196W per 12V rail. The 550W should be plenty.

Any ACX 2.0 cooled EVGA has the spindown at idle feature: 2972 (Stock), 2974 (SC), 2975 (SSC) or 2978 (FTW).

 

[amd7674]

Yes. The Corsair has a single 12V rail which is what you will need for a 970 otherwise you are nearing the Antec's limit of 196W per 12V rail. The 550W should be plenty.

Any ACX 2.0 cooled EVGA has the spindown at idle feature: 2972 (Stock), 2974 (SC), 2975 (SSC) or 2978 (FTW).

Thanks for your help... I guess there is no easy way... LOL

Out of all EVGA offerings and Zotac which one would you recommend? My worry is I don't want hairdryer when playing games in my livingroom. The current Zotac GTX650Ti is whisper quiet. Also I hope my stock ivy bridge i5 won't bottleneck GTX970.

 

[chrcoluk]

I got no opinion on zotac, but I think the evga situation is now no longer as bad as initially made out.

On all ACX 2.0 cards, there is a new firmware that has 0 fan speed at idle. Also it should be not as loud as under load as at time of reviews due to load adjustments as well.

It is also possible the FTW version of the 970 may be quieter than the other EVGA 970s as it has a better heatsink so the fans wont be needed as much.

 

[n=1]

3570K at stock is not going to bottleneck a 970 (or 980), even less so @ 4.5GHz.

Not exactly apples to apples but see this AnandTech review and pay attention to the results with 2x 580 and 2x 7970.

 

[amd7674]

thanks boys I will swap my PSUs and I will go for EVGA FTW.

 

[misterbobby]

3570K at stock is not going to bottleneck a 970 (or 980), even less so @ 4.5GHz.

Not exactly apples to apples but see this AnandTech review and pay attention to the results with 2x 580 and 2x 7970.

Technically it can and will in some games but Crysis 3 is probably the only game where its apparent.

 

[n=1]

CPU bound games (mainly RTS and flight sims) will always be CPU bound.

Crysis 3 is one of the few shooters that benefits from a stronger CPU, but even then the difference between an i5-3470 and i7-3770K is only about 7%. Crank up the settings and make GPU the bottleneck, and there's virtually no difference.

 

[chrcoluk]

amd7674 my FTW will be here in 3-4 hours, if you can wait I will give you noise and temps feedback.

 

[amd7674]

amd7674 my FTW will be here in 3-4 hours, if you can wait I will give you noise and temps feedback.

That would be great... please "beat her up" and post your results

I might getting good deal on Zotac at local store (with easy return policy). It would be around $100 less than EVGA FTW... However if you are happy with it and noise levels at idle/load are excellent, i might be convinced.

Thanks in Advance

 

[chrcoluk]

DPD is delayed, I am watching guy on live tracker after I got delayed email my guess is extra hour to what I quoted, and also I think it will arrive whilst I am watching F1 which may add some time, but I will definitely put it in today.

Aslam is currently making delivery number 28, you are delivery number 41.

Aslam is approximately 2 hours away from you.

You can watch Aslam's progress on our map.

 

[chrcoluk]

its here, but want to do some tests first on my gtx760 today so i have noise and temp baseline to compare against.

After that if there is time before F1 starts I will put in the gtx 970 but may well be after the F1 is done.

 

[amd7674]

its here, but want to do some tests first on my gtx760 today so i have noise and temp baseline to compare against.

After that if there is time before F1 starts I will put in the gtx 970 but may well be after the F1 is done.

No worries, I picked up zotac gtx 970 for over $100 cheaper than EVGA 970 FTW. It is a local store with good 30 days return policy in case of noise issues / coil whine. Unfortunately I won't be getting any free games, which I don't really care about... If you still have time please post your results for others and possibly for me...

Thanks...

BTW... I'm almost done with PSU swapping. It wasn't too bad... Corsair is not modular and it has tons of cables... LOL... it was a lot of fun dealing with wire management...

 

[chrcoluk]

ok I done tests.

gtx 760 temp - fan speed

desktop 27 43%
heaven extreme 70 62%
valley custom 70 60%

gtx 970 temp - fan speed
desktop 31 0%
heaven extreme 73 29%
valley custom 73 29%

notes - the 970 may not drop back to 31 after its been put under load due to idle cpu, temp is still dropping a few mins after I killed bench, and is current 45C, fans start quite and stop early, the threshold for fans seems over 60C.

on noise there is no doubt the gtx 970 is quieter, the fan speed under load is pretty low and less than the 760 is at idle. Overall very happy, if I mute sound and stay in my chair, I still cannot hear anything, but if I put ear to back of case I can hear what seems like a very faint buzzing sound, coil whine? I do not hear it when sat in my chair at front of pc.

Thumbs up from me. I think I will set a 20-30% fan speed tho even when idle.

As I finished writing this up temps is down to 41C.

For the curious my GPU speed hit and was steady at 1531mhz, not sure if thats good or not for these cards. The rated boost is 1367mhz, so I assume I got a good card.

Also temps have now gone down to 33C with 0% fans at time I edited this post.

 

[misterbobby]

CPU bound games (mainly RTS and flight sims) will always be CPU bound.

Crysis 3 is one of the few shooters that benefits from a stronger CPU, but even then the difference between an i5-3470 and i7-3770K is only about 7%. Crank up the settings and make GPU the bottleneck, and there's virtually no difference.

Lol of course if you crank the every settings with a 680 it will become gpu limited. If you have a gtx980 or oced 970 and want 60 fps and have common sense you will use SMAA and that will easily keep you above 60 fps with all other settings maxed if you have 4770k or oced 3770k. If you have an i5 you cant stay above 45 fps even so an i7 makes playable difference for those that use vsync. And your first link is irrelevant as they are using medium settings which also impacts the cpu. I have spent tons of time testing this game on various setups.

 

[n=1]

Umm the whole point of using medium settings is precisely to shift the bottleneck to the CPU as much as possible to tease out how much the CPU impacts FPS. The fact that there was little difference even at medium settings is very compelling evidence as long as you have a true quad core, HT doesn't do much for you in games. And if you crank up settings until you're GPU bound then HT or not it makes no difference as someone has demonstrated already.

If you want something more recent here's one using 780 Ti as the GPU. Going from a dual core i3-4150 to a hex core i7-4960X gave a whooping 15% increase, and between an i5 and i7 there's 1-2 FPS difference. So yes I maintain my position that for a single 970 (or even 980 for that matter), you're gonna have to try very very hard to have that 3570K bottleneck anything. Well assuming you game at 1080p and don't deliberately game at lower resolutions eg 720p to bottleneck your CPU.

 

[misterbobby]

Umm the whole point of using medium settings is precisely to shift the bottleneck to the CPU as much as possible to tease out how much the CPU impacts FPS. The fact that there was little difference even at medium settings is very compelling evidence as long as you have a true quad core, HT doesn't do much for you in games. And if you crank up settings until you're GPU bound then HT or not it makes no difference as someone has demonstrated already.

If you want something more recent here's one using 780 Ti as the GPU. Going from a dual core i3-4150 to a hex core i7-4960X gave a whooping 15% increase, and between an i5 and i7 there's 1-2 FPS difference. So yes I maintain my position that for a single 970 (or even 980 for that matter), you're gonna have to try very very hard to have that 3570K bottleneck anything.

Then you dont understand how things work. Some settings impact the cpu not just the gpu. HT makes a huge difference in this game and that is a FACT. And again an i5 is PLAYABLE limitation in Crysis 3 if you want 60 fps where it is NOT with an i7. But yeah with most other games an i7 means nothing at all but please get your facts straight about Cyrsis 3.

http://maldotex.blogspot.com/2013/02/hyperthreading-and-real-custom-graphics.html

 

[n=1]

And yet the link I posted showed an i5 getting 60 FPS in Crysis 3 with 780 Ti, so yeah.

EDIT: Another chart showing almost no difference between an i5 and i7 at 1080p with GTX 690.

But yes as I have already alluded to if you drop the resolution, or turn settings down to medium and shift the bottleneck back to the CPU, then yeah HT will make a difference. (keep in mind the i5 was a non-K locked at 3GHz and the i7-3770K was OC'd to 4.5GHz, so the differences are larger than what they would be stock for stock) But then in those cases you're >60 FPS anyway so it's very much playable by any definition.

 

[Flogger23m]

Slightly off topic but thought I would ask anyways, although I am sure I am fine. I am looking at getting the MSI Gaming GTX 970 which uses a bit more power than other 970s. Will also plan on doing some mild overclocks (will see how far it goes without upping voltage).

I have an Antec Earthwatts 650 Green PSU:
http://www.newegg.com/Product/Product.aspx?Item=N82E16817371044&cm_re=antec_earthwatts_650-_-17-371-044-_-Product

+3.3V@24A, +5V@24A, +12V1@38A, +12V2@38A, [email protected], [email protected]

4670K @ 4GHZ, SSD, HDD, 2x DIMM RAM sticks (to get two more eventually), DVD drive, 5-6 case fans. Occasionally connect two additional HDDs. I assume my PSU will be plenty?

 

[misterbobby]

And yet the link I posted showed an i5 getting 60 FPS in Crysis 3 with 780 Ti, so yeah.

EDIT: Another chart showing almost no difference between an i5 and i7 at 1080p. But yes as I have already alluded to if you drop the resolution lower, or turn settings down to medium and shift the bottleneck to the CPU, then yes HT will make a difference. But then in those cases you're >60 FPS anyway so it's very much playable by any definition.

Again you have no clue what you are talking about. You must not even own the game and if you actually kept up with the game you would know that there was a patch that came out and the game took more advantage of HT. It also depends on where you are testing and I clearly showed you its a 30% difference on that 2700k when HT was disabled.

And I had a 2500k before this this 4770k so I know damn well the difference. With a 780 I was in the 40s with the 2500k at 4.4 and those same spots with 4770k at 4.4 is over 60 fps. But please continue to argue...

 

[n=1]

Well an i5 choking in certain parts of the game is not the same as it being a global bottleneck, not being able to stay above 45 FPS in general, and thus being unplayable. Your earlier posts seem to imply this (unless I'm reading them wrong), and I have tried to show evidence to the contrary. But alright fair enough.

Anyway, I was simply trying to make the point that if the CPU bottleneck is only apparent in one or a select few games (and only in certain parts on top of that), then it's not really an issue to lose sleep over or make a fuss about, and is simply being blown out of proportion.

EDIT: So I went back and had a look at the first link and tracked down its source. The date of testing was April 23 2014, a full year after the v1.3 patch was released. The results speak for themselves.