[VC]NVIDIA GeForce GTX 980, GTX 980 SLI, GTX 970, 3DMark performance

Fastx

Senior member
Dec 18, 2008
780
0
0
First of all, we have GeForce GTX 980 scores with three different clock speeds. We were unable to confirm which clock is is the actual reference clock, so we decided to post all three. However, GTX 980 running at 1127 MHz was our point of reference. I think it’s quite clear now, that GeForce GTX 980 has: 4GB GDDR5 memory, 256 bit bus and 7 GHz memory clock. Today we are adding more info about the core clock. In our chart you can see three speeds: 1127, 1178 and 1190 MHz. It doesn’t really matter which clock you choose, because 1127 MHz were already enough to outperform all current single-gpu high-end cards. Of course, once you add more juice to 780 Ti or 290X you can actually reach GTX 980 performance. However, GTX 980 at 1.19 GHz will not be easy to beat.

Just remember, this is just synthetic scenario, 4k gaming will surely show different results.

GeForce GTX 980 SLI

We also have GTX 980 SLI score. To understand how efficiently SLI connection is operating in high performance Maxwell GPUs, we will compare parts running at the same clock speeds in 1 and 2-GPU configuration. That said, GTX980 SLI at 1178 MHz offers 59.5% scaling, which is average at best. Radeon R9 295X is still faster.

GeForce GTX 970 4GB


According to our data, GTX 970 has a core clock of 1051 MHz. So technically, it shouldn’t be a problem to add extra speed, as we already know that GM204 can easily operate at 1150-1190 frequencies. For such reason, likelihood of GTX 970 beating GTX TITAN or 290X is relatively high.



http://videocardz.com/52166/nvidia-geforce-gtx-980-gtx-970-gtx-980m-gtx-970m-3dmark-performance
 
Last edited:

mindbomb

Senior member
May 30, 2013
363
0
0
the 970 might be a 200 watt part, hence it would be almost directly comparable to the r9 285...which it would crush if this is true.
 

Wild Thing

Member
Apr 9, 2014
155
0
0
Not just "here"...XDMA crossfire is superior to SLI across the range of Radeons that feature it in both scaling and frame latency.
 

jpiniero

Lifer
Oct 1, 2010
14,841
5,456
136
the 970 might be a 200 watt part, hence it would be almost directly comparable to the r9 285...which it would crush if this is true.

Hmm, so the 970 might be 1600 cores after all. That'd be more around 150 W range with the 1920 core 980 around 170-180 W. The 980 SLI is likely suffering from bandwidth problems.

How popular do you think something like the 980M would be? Look at that score, it's way faster than the 880M.
 

JDG1980

Golden Member
Jul 18, 2013
1,663
570
136
If accurate, those figures are very impressive. Keep in mind that this is on the same node as before, and that GTX 980 will almost certainly have a TDP of 200W or below. Compare that to 250W for GTX 780 (and R9 280X on the AMD side). You may not care about power consumption, but many people do, since lower TDP means quieter and more lightweight cooling solutions, and easier compatibility with lower-wattage power supplies and smaller cases. Not everyone wants a giant fire-breathing case with a 1000W+ PSU.

The next generation of Mac Pro should be something to behold: Haswell-EP with up to 18 CPU cores, and dual Maxwell Quadro cards with superior performance and lower TDP.

GTX980 SLI slower than R295 X2 ?


Probably because of the narrower memory bus.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,108
1,260
126
I hope this bench is false, because this is actually atrocious. I know the graph shows the GTX 980 as slower than a 780ti with both at stock, but for all intents and purposes, that is basically parity between them with those scores.

Also, given that the gtx 980 will have a 256 bit bus with less available memory bandwidth than the 780, titan cards and 780ti, it is not going to scale as well with overclocking like we see comparing a 680/770 to a 780/780ti/titan. So it would almost certainly be slower at anything over 1080p and quite likely even at 1080p with overclocking accounted for.

Wondering what the plan would be there. You can't market on better perf/w without more performance as well. Just offering that well it's the same performance but uses less power is not inspiring upgrades. Perhaps the review kit will have a caveat that it's not intended to replace gk110 cards as an upgrade path or intended for high resolutions. Push it more as a pair of more affordable cards for 1080p. The 970 looks about the same as an R9 290, may be a decent mid-range card, you'd guess with that performance it wouldn't cost more than $350-$400 given the 290's price.
 

JDG1980

Golden Member
Jul 18, 2013
1,663
570
136
Also, given that the gtx 980 will have a 256 bit bus with less available memory bandwidth than the 780, titan cards and 780ti, it is not going to scale as well with overclocking like we see comparing a 680/770 to a 780/780ti/titan. So it would almost certainly be slower at anything over 1080p and quite likely even at 1080p with overclocking accounted for.

I would be very surprised if GTX 980 wasn't at least sufficient for 2560x1440, given the current popularity of that resolution. For 4K, no single-card solution will be sufficient for high-framerate "AAA" gaming until 16nm FinFET becomes available (assuming TSMC ever manages to figure it out).

It's quite possible that Maxwell (at least GM2xx) has some sort of texture compression, like AMD's GCN 1.2, to increase memory bandwidth utilization. I doubt that they would release a new $499 card which is bottlenecked by too narrow a memory bus at ordinary resolutions.
 
Feb 19, 2009
10,457
10
76
Looks good for a mid-range Maxwell, it shows amazing IPC gains since they are on the same node. If the die size is ~300mm2 like Gk104, AMD is going to be in a world of hurt.

Just think about that a bit, a ~300mm2 die on the same node performing the same as the massive GK110 die, with much less power use.

It means big Maxwell is going to pwn hard.
 

iiiankiii

Senior member
Apr 4, 2008
759
47
91
Weak sauce, if the benchmark is true. I'm assuming this is targeted towards gtx 680 generation of users. Side-grade for 780/780ti. 28nm is really holding back the performance of these cards.
 
Feb 19, 2009
10,457
10
76
Weak sauce, if the benchmark is true. I'm assuming this is targeted towards gtx 680 generation of users. Side-grade for 780/780ti. 28nm is really holding back the performance of these cards.

You assume correct. It's a MID-RANGE Maxwell, GM104.

It replaces 780/ti and serves as an attractive upgrade for users on weaker GPUs.
 

iiiankiii

Senior member
Apr 4, 2008
759
47
91
Looks good for a mid-range Maxwell, it shows amazing IPC gains since they are on the same node. If the die size is ~300mm2 like Gk104, AMD is going to be in a world of hurt.

Just think about that a bit, a ~300mm2 die on the same node performing the same as the massive GK110 die, with much less power use.

It means big Maxwell is going to pwn hard.
Yeah, Big Maxwell on 16/20nm is shaping up to be the next major upgrade. Looks good from that perspective.
 

Attic

Diamond Member
Jan 9, 2010
4,282
2
76
I'm expecting $400 and $500 for the 970 and 980.

Performance increase is meh, obviously. Priced right these could be great though.

Maybe 980ti comes out later and does more for performance.
 

tviceman

Diamond Member
Mar 25, 2008
6,734
514
126
www.facebook.com
10% faster than 780 TI at GTX 770 power levels. Looks really solid. Everyone here complaining is doing so based on name only. If it was named GTX Z85 then no one would have any room to complain because there would be no naming precedent set (same story with Tonga's R9 285 name). AMD has absolutely no answer to this to speak of on the horizon. Faster, smaller, and way more power efficient than Tahiti.

If it's priced at $499 it will sell great, despite all the complaining going on in this thread. If the GTX 970 is priced at $399, it'll crater AMD's entire product pricing. R290x will have to drop $379 to look competitive and r9 290's to $299. More than likely though, I bet Nvidia prices GTX 980 at $549, GTX 970 at $449, and GTX 960 TI at $349. GK104 will likely get rebadged to GTX 960 at $249 to take on R9 285.
 
Last edited:

Zanovar

Diamond Member
Jan 21, 2011
3,446
232
106
10% faster than 780 TI at GTX 770 power levels. Looks really solid. Everyone here complaining is doing so based on name only. If it was named GTX Z85 then no one would have any room to complain because there would be no naming precedent set (same story with Tonga's R9 285 name). AMD has absolutely no answer to this to speak of on the horizon. Faster, smaller, and way more power efficient than Tahiti.

If it's priced at $499 it will sell great, despite all the complaining going on in this thread. If the GTX 970 is priced at $399, it'll crater AMD's entire product pricing. R290x will have to drop $379 to look competitive and r9 290's to $299. More than likely though, I bet Nvidia prices GTX 980 at $549, GTX 970 at $449, and GTX 960 TI at $349. GK104 will likely get rebadged to GTX 960 at $249 to take on R9 285.


Jesus calm down tv,im sure it will still sell.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,108
1,260
126
10% faster ? With the 780t running at boost clocks the 980 is a bit slower or are you basing that on the 780ti running at the base clock ? Heck, my 780tis boost to 1200 out of the box.

Performance is terrible for a new card. Nothing new brought to the table. The price will likely be it's saving grace.

I am still doubtful these are accurate though as this is abysmal for a new X80 flagship, not even faster... Seems implausibly bad.
 

96Firebird

Diamond Member
Nov 8, 2010
5,712
316
126
I think you're reading the clockspeeds wrong. In fact, we have no idea what the clockspeeds mean. The author doesn't even know which clockspeed is default. Are they using the stock 780Ti base speed, no matter the boost? And the other results are overclocked?

As for the SLI performance, I'm sure driver improvements will come since this is their first time with Maxwell SLI.
 

BFG10K

Lifer
Aug 14, 2000
22,709
2,980
126
As others have pointed out, if this is a mid-range part, the performance is excellent and bodes well for the high-end part(s).

Still, 3DMark never meant squat for real-world performance anyway. Let's wait for real game benchmarks before passing judgement.
 

HurleyBird

Platinum Member
Apr 22, 2003
2,726
1,342
136
Looks impressive for what should be a smaller more power efficient chip on the same node, but disappointing for a generational leap. That said, grain of salt and all that.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |