so I spent a little time with the reference 980 today...

[toyota]

first off this card is a tank and very heavy. it feels so high quality and it is darn sexy for a reference card for sure. a little bit of a comparison in Metro LL to the 780 and 970 I was using.

Metro LL at 1920x1080 very high settings with SSAA

780 at out of box clocks 1019 boost the whole time
44 fps
384 watts

970 at out of box clocks ranged from 1240 to 1291 boost
46 fps
288 watts

980 with only the power limit raised to 125% ran at 1240 to 1227
52 fps
306 watts

980 with slight bump in voltage oced to 1522/7800 with few dips to 1480 as it hit the power limit
62 fps
372 watts

 

[wilds]

The OC'd 980 really starts to pull ahead! Wish it had unlocked power limit...

 

[toyota]

and for those that think blower cards dont really make a difference to case temps I did a little test in Crysis 3.

with my 780 at stock 1019 boost the I would hit 65 C

with my 970 oced to 1458 boost I would hit 64 C

with the reference 980 oced to 1465 I only hit max temp of 57 C on cpu

now of course that can vary with ambient temp and I know the day I tested the 970 it was a couple degrees warmer than today. when I tested the 780 it was actually cooler than today. so if anything my cpu temps would be another 1 or 2 degrees cooler if a tested on the same day as 780.

 

[SlickR12345]

WOW that is so low. So an overclocked $550 card can barely run Metro at 60fps on 1080p resolution.

Can't believe people payed $750 for the 780ti to run this game at 45fps all this time.

 

[toyota]

WOW that is so low. So an overclocked $550 card can barely run Metro at 60fps on 1080p resolution.

Can't believe people payed $750 for the 780ti to run this game at 45fps all this time.

thats with 4x SSAA. with just 2x SSAA you will never drop below 65 or 70. heck the game looks perfect without SSAA and will have an insanely high framerate all things considered.

 

[Sohaltang]

For reference can you OC the 780 to 1200-1300 and match the tweaked bios 980 memory clock?

I would expect 10+ FPS improvement that would put it in line with the 980.

 

[toyota]

For reference can you OC the 780 to 1200-1300 and match the tweaked bios 980 memory clock?

I would expect 10+ FPS improvement that would put it in line with the 980.

My 780 would not even do but 1163/6800 max even with voltage bump.

with no voltage bump at all I am getting 1465/7400 on the 980 but I have not tried any higher.

 

[Sohaltang]

My 780 would not even do but 1163 max even with voltage bump.

Try with just a 10% OC then? How about jumping in the valley thread? Many people want to see the 980 in relation to other cards not the benchmarks every reviewer used at stock speeds. I see you 770 scores there. No 780.

 

[toyota]

Try with just a 10% OC then? How about jumping in the valley thread? Many people want to see the 980 in relation to other cards not the benchmarks every reviewer used at stock speeds.

I dont have the 780 any longer but I am sure I posted in the valley thread before with it at 1124 or so.

 

[Enigmoid]

How does it do a stock?

 

[toyota]

How does it do a stock?

not sure exactly what you are asking

 

[96Firebird]

Little confused on the 1252 with power limit raised... Did it fluctuate between 1260 and 1252?

 

[Kenmitch]

not sure exactly what you are asking

without raising the power limit.

 

[toyota]

Little confused on the 1252 with power limit raised... Did it fluctuate between 1260 and 1252?

it started at 1265 and then dropped one bin to 1252. it drops one bin no matter what but never drops another if the power limit is not reached. it seems normal as hardwarecanucks.com and some other review I saw noticed the same thing.

EDIT: actually that was 1240 and 1227 it was running out of the box with only power limit raised so I edited the OP too.

 

[Sohaltang]

Thanks for posting valley score. Right in the middle of the 780 pack. Faster than 7, slower than 6. Your 780 @ 1150 should have scored about 3000 or 5% slower than your tweaked 980's 3147. Not sure if its just valley but I have yet to see these cards put any distance on the last generation. Especially a healthy OC'd 780ti.

 

[96Firebird]

it started at 1265 and then dropped one bin to 1252. it drops one bin no matter what but never drops another if the power limit is not reached. it seems normal as hardwarecanucks.com and some other review I saw noticed the same thing.

EDIT: actually that was 1240 and 1227 it was running out of the box with only power limit raised so I edited the OP too.

Ah I see now. I thought the 970 was the 980 at stock, and was wondering how it was performing worse at higher clocks.

 

[toyota]

Thanks for posting valley score. Right in the middle of the 780 pack. Faster than 7, slower than 6. Your 780 @ 1150 should have scored about 3000 or 5% slower than your tweaked 980's 3147. Not sure if its just valley but I have yet to see these cards put any distance on the last generation. Especially a healthy OC'd 780ti.

a 780 ti at 1250 is basically dead even overall with a 980 at 1500.

Nvidia did some crazy marketing and exaggerating as these are cards are no where near 2x perf per watt. really only some of the cards such as the 980 reference even come close to being a 165 watt card. in reality the 980 should have been listed around 185 watts. the nonrefernce cards can be power hogs. I have the Strix 980 sitting here too but I am just not even going to open it. looking at reviews it uses 50-60 watts more out of the box with only a measly 60hz factory oc. and the typical oc on the Strix is no better than the reference card can do.

 

[ams23]

as these are cards are no where near 2x perf per watt.

What they said is 2x improvement in perf. per watt of GTX 980 reference vs. GTX 680 reference, which appears to be pretty accurate (note also that their comparison is GPU-specific power consumption, and not system power consumption, because they are trying to isolate the GPU perf. per watt).

 

[lavaheadache]

Toyota, my reference card seems to be right on par with yours. I need voltage after 1460 on the core. So far I'm up to 7.8 on the memory though.

Performance wise, 980's are no better than last gen if you had a good overclocking Ti. What really separates things is power consumption. Even if only just the reference cards. I barely see over 300 watts. Heck playing borderlands 2 I usually am pulling about 265 watts. It's silly that a highend rig pulls a little more juice than a floor lamp.

 

[toyota]

Toyota, my reference card seems to be right on par with yours. I need voltage after 1460 on the core. So far I'm up to 7.8 on the memory though.

Performance wise, 980's are no better than last gen if you had a good overclocking Ti. What really separates things is power consumption. Even if only just the reference cards. I barely see over 300 watts. Heck playing borderlands 2 I usually am pulling about 265 watts. It's silly that a highend rig pulls a little more juice than a floor lamp.

yes the reference 980 cards seem pretty efficient. mine will run at 1503 on stock voltage which is pretty good. my stock voltage seems to be 1.187. it will always start out at 1.206 then drop down to 1.187 at the same time it drops the clocks 13 mhz. after that it will never go back up or drop if under full load. that drop seems odd but reviews have seen it too plus its consistent in every bench.

 

[KaRLiToS]

WOW that is so low. So an overclocked $550 card can barely run Metro at 60fps on 1080p resolution.

Can't believe people payed $750 for the 780ti to run this game at 45fps all this time.

What do you expect from Metro games with their awsome graphs?

Even 2 x GTX 780ti @ 1400$ would've kept you under 70 fps at 1440p.

Thats why you need to turn down AA in beautiful games like that.


_____________________

Congrats on the cards toyota, enjoy them my friend

 

[toyota]

What they said is 2x improvement in perf. per watt of GTX 980 reference vs. GTX 680 reference, which appears to be pretty accurate (note also that their comparison is GPU-specific power consumption, and not system power consumption, because they are trying to isolate the GPU perf. per watt).

right but that is based on the TDP they gave the 980. in reality its about 50-60% faster than the 680 overall while using basically the same wattage. in many reviews its using just a bit more than the 195 watt 680. now these non reference cards are using 40, 50, and sometimes 60 watts more than the non reference cards. we have even non reference 970 cards using more power than the big old 780. so realistically that 2x perf per watt will only happen with a reference card in the perfect bench. overall its really only about 50% better perf per watt than the 680 which is still great but again thats with the reference cards only.

 

[tviceman]

How is the noise when overclocked? You and I both seem to be noise sensitive, so this is a big question with me. Ryan Smith's small OC investigation revealed a big jump in fan noise, and that scared me from getting a reference version. And what are the ambient temps like in your house when you ran the 980 benchmarks?

 

[toyota]

How is the noise when overclocked? You and I both seem to be noise sensitive, so this is a big question with me. Ryan Smith's small OC investigation revealed a big jump in fan noise, and that scared me from getting a reference version. And what are the ambient temps like in your house when you ran the 980 benchmarks?

well its not bad at all but my cpu cooler is a bit loud so in a real quiet case right by your head it may be a different story. there is tiny bit of coil whine but I can only hear it if I put my head just a few inches from the card. I cant hear it sitting in my chair or from the top of the case.

 

[guskline]

toyota, congrats on the 980! Sounds like a stellar gpu.