GPUBoss Review Our evaluation of R9 380 vs 970 among Desktop GPUs


Battlefield 3, Battlefield 4, Bioshock Infinite and 21 more


T-Rex, Manhattan, Cloud Gate Factor, Sky Diver Factor and 1 more


Face Detection, Ocean Surface Simulation and 3 more

Performance per Watt

Battlefield 3, Battlefield 4, Bioshock Infinite and 32 more


Battlefield 3, Battlefield 4, Bioshock Infinite and 32 more

Noise and Power

TDP, Idle Power Consumption, Load Power Consumption and 2 more


Overall Score

Nvidia GeForce GTX 970 

GPUBoss recommends the Nvidia GeForce GTX 970  based on its benchmarks, compute performance and noise and power.

See full details

Cast your vote Do you agree or disagree with GPUBoss?

Thanks for adding your opinion. Follow us on Facebook to stay up to date with the latest news!

Differences What are the advantages of each

Front view of Radeon R9 380

Reasons to consider the
AMD Radeon R9 380

Report a correction
Slightly more shading units 1,792 vs 1,664 128 more shading units
Slightly more texture mapping units 112 vs 104 8 more texture mapping units
Front view of GeForce GTX 970

Reasons to consider the
Nvidia GeForce GTX 970

Report a correction
Better PassMark score 8,563 vs 6,049 More than 40% better PassMark score
Significantly higher effective memory clock speed 7,012 MHz vs 5,500 MHz More than 25% higher effective memory clock speed
Higher memory bandwidth 224.4 GB/s vs 176 GB/s Around 30% higher memory bandwidth
Better 3DMark06 score 15,071 vs 12,191 Around 25% better 3DMark06 score
Higher clock speed 1,050 MHz vs 918 MHz Around 15% higher clock speed
Higher pixel rate 58.8 GPixel/s vs 29.38 GPixel/s More than 2x higher pixel rate
Significantly more render output processors 56 vs 32 24 more render output processors
Better fire strike factor score 72.7 vs 57.51 More than 25% better fire strike factor score
Significantly higher memory clock speed 1,753 MHz vs 1,375 MHz More than 25% higher memory clock speed
Higher BioShock infinite framerate 107.6 fps vs 80.8 fps Around 35% higher BioShock infinite framerate
Better PassMark direct compute score 4,375 vs 2,938 Around 50% better PassMark direct compute score
Better T-Rex score 8.69 frames/s vs 7.07 frames/s Around 25% better T-Rex score
Higher crysis 3 framerate 50.6 fps vs 39.4 fps Around 30% higher crysis 3 framerate
Higher battlefield 4 framerate 88.6 vs 59.5 Around 50% higher battlefield 4 framerate
Lower TDP 148W vs 190W More than 20% lower TDP

Benchmarks Real world tests of Radeon R9 380 vs GeForce GTX 970

Bitcoin mining Data courtesy CompuBench

Radeon R9 380
410.27 mHash/s
GeForce GTX 970
212.39 mHash/s

Face detection Data courtesy CompuBench

Radeon R9 380
84.59 mPixels/s
GeForce GTX 970
105.04 mPixels/s

T-Rex (GFXBench 3.0) Data courtesy CompuBench

Radeon R9 380
GeForce GTX 970

Manhattan (GFXBench 3.0) Data courtesy CompuBench

Radeon R9 380
GeForce GTX 970

Fire Strike Factor Data courtesy FutureMark

Sky Diver Factor Data courtesy FutureMark

Battlefield 4


Showing 15 comments.
The two cards were released "years" in one another. Why fight over the two? A year can make a difference. But still the GTX 970 with 2gb plays more. For me that is. I've used the R9 380, I built another one for my li'l brother and we play similar games. Like warthunder. Hell I even have a GTX 480 that runs warthunder on FULL PRESET OF EVERY SETTING and it runs ON 60FPS! Cheese, I like AMD but NVIDIA cards plays more. AMD HAS GOOD CARDS FOR A REASONABLE PRICE, CARDS THAT CAN PERFORM BUT THEIR IS A REASON WHY NVDIA CARDS ARE FREAKING EXPENSIVE, BECAUSE IT WILL GIVE YOU MORE.
Ehmmm... You''re just nvidia fanboy.Don't get me wrong but you aren't right with 380 is about 5-10 fps more than 960 even at ""Nvidia Gameworks Titles"" and really mostly people think about performance , not about TDP (In my opinion) so yeah....
Lol don't even know which side u r weighing down on
Lol, wtf resolution u playing at to get that kind of power gap between the two? Why if it is rly high are you playing with 1080p cards, and why if it is rly low, are you rocking 4gb vram? Again, look at benchmarks from Linustechtips, jayztwocents, awesomesauce network, Paul's hardware, whatever, and get your damn story straight. At 1080p, both cards max games at 60+fps. This is what they were designed to do, and this is what they do very well. Up the resolution to higher, and the gap only widens in AMDs favour, due to higher bandwidth between the gpu and vram, as well as more of the aforementioned vram. Before posting out of ur ass, check at least one benchmark site to corroborate what u r abt to say.
Lol, the 380 IS the competition to the 960, occupying the Same place in the market and the gpu lineups of both companies, do you even know how economics works? And I'll enjoy my 280 and the £100 worth of games or a phone or whatever I want to buy whilst still getting 1080p60 in every game today maxed (except for crysis, but even the Titan x can't really run crysis)
Lol 'blows everything but fury X out of the water' huh??? What about the r9 390x? Or the fury, or the nano? Have you even looked at the damn benchmarks in the real world??? Do you even have two brain cells for connections to form between? And to mention the drivers is asking for trouble, nvidia with game works is a piece of shite and you know it, crimson for AMD has more than closed the software gap between the two. Even a bloody 290 is beating out a 970 in dx12 benchmarks! This should not happen, but it is. A £100 less expensive card beating or matching a 970 in real world tests??? True, the 970 beats this card, but with a £100 delta, what did you actually expect?
hope those amd fan boy benchtest keep you hooked! cause i aint fucking fooled.
and $100 cheaper then the r9 380. 15% better for $100 cheaper. but yah the r9 380 is just betta, cause some amd fan boys said so. DO NOT LISTEN TO ANYONE ON THIS SITE, its biased both sides. for sure. but having used both cards gtx 960 and the r9 380, i found that my gtx 960 2gb, was superior. in all my games, in all my video's in every fucking thing i do, it is superior. oh and it uses way less power. 15% better $100 cheaper, and uses 60 less watts. you need to be smoking jib and injecting heroin to think that a r9 380 is better. lol fallout 4 ultra EVGA gtx 960 2gb runs at a rock solid 60fps ULTRA fallout 4 high Gigabyte r9 380 4gb runs at 1080p at 57 fps, on high. NOT ULTRA. wait what? 2 less gigs, 60 less watts, less shading units less everything, BUT PUMPS OUT HIGHER FRAMES ON A HIGHER SETTING? gg night amd fan boys
my solid facts, ive never used amd, and the one time i did, i couldnt believe that a 4gb gigabyte r9 380, couldnt run fallout 4 on ultra without lag. my 2 gb gtx 960 runs it on ultra, without lag. what more needs to be said? powerful junk vs engineered cards. no competition. Never has been, and most likely never will be. sorry amd fan boys. your shit sucks. hahahahahahahaha. oh and everyone comparing a gtx 960 to a r9 380 is a fucking retard, since a r9 380's competition is the gtx 970. and the 970 shits all over this, day and night, every second of every game, the gtx 970 will laugh at everything but a fury x.\ enjoy. im so glad i bought nvidia, cause this card is amazing. about 15%-20% better then my r9 380. maybe a well clocked or stable card, i dont know, i dont care, its better. lol.
uhm, gtx 970 blows everything but a fury x out of the water. lol. thats not biased, thats just facts. amd drivers dont function 100% nvidia is known for having the best functioning drivers. I WILL CHOOSE THE PROPER FUNCTIONING CARD 99/100 times. and that 1 time i get amd, ill be so disappointed, that ill fucking return it for nvidia... LOL LIKE ALWAYS. LIKE ALWAYS, LIKE ALWAYS, returned my r9 3804gb for a gtx 960 2gb, and the gtx 960 2gb runs everything higher, smoother, AND FUCKING BETTER. I DONT CARE ABOUT NUMBERS AND FAGGOTS BENCHTEST, id rather see for myself, and find out that nvidia sells cheaper better cards. lol my 960 was $230 my r9 380 was $330 MY GTX IS BETTER. LOLLLOLLOLOLOLOLLOOL LOLLOLOLOLOLOLLOLLOLOLOLOL :LP:P:P:P::P::PPOLOLOLO<LOLOLO
do you think i dont know that the memory speed differences and core clocks mean nothing? it is very misleading to represent it like this when the two different architectures work in two completely different ways (look at CUDA vs openCL for reference,) so representing these numbers in these ways can cause misunderstanding and spreads misinformation. That is how the results are biased, and that is why you should really look at in game benchmarks or things like firestrike to get a real idea of how these cards perform
check out for example the video from jayztwocents on this topic also, gpuboss has a bad reputation of biasing their results pro nvidia, check out for example, the bias is real and is noticeable
How are they bias? Its complete fact.
wow, more bias for nvidia from gpuboss
comments powered by Disqus