GTX 970 vs. R9 280

[AMDRazer]

Hello all,

I want to know which is better; the 970 or 280. What are the differences and similarities?

Thanks, AMDRazer

 

[ko888]

The Radeon R9 280 is just a rebadged Radeon HD 7950.

The Radeon R9 285 is now the replacement for the Radeon R9 280.

I guess that would be you that said it is the replacement. And just so we are clear the 285 is not a replacement for the 280 either.

http://www.tomshardware.com/reviews/amd-radeon-r9-285-tonga,3925.html
Note at the top af the article in bold:
On paper, the new Tonga-based R9 285 looks to be slightly slower than the R9 280 it is intended to replace, but there's more than meets the eye.

http://www.anandtech.com/show/8460/amd-radeon-r9-285-review
In the first paragraph:
"In the process the R9 285 would be a partial refresh of their R9 280 series lineup, supplying it with a new part that would serve to replace their nearly 3 year old Tahiti GPU."

http://www.hardocp.com/article/2014/10/14/xfx_r9_280_double_dissipation_3gb_video_card_review
In the third paragraph:
"The R9 280 is basically being replaced by the AMD Radeon R9 285."

http://www.eteknix.com/sapphire-dual-x-amd-r9-285-tonga-2gb-graphics-card-review/
In the fourth paragraph:
"we can now confirm the R9 285 makes the R9 280 End-Of-Life (EOL)."

There are many more articles that point it out specifically. You can keep denying it all you want.

 

[RCFProd]

Sometimes we, as a community, really have to realise that VRAM isn't everything.

I see this all the time. The R9 280 demolishes the GTX 960! Whilst really, the 960 gets the best performance at 1080p. Yes, maybe at higher resolutions you might have call on this but the discussions are way too much biased on VRAM at times.

 

[IamTimTech]

whatever you say buddy, whatever you say. How many people do you see asking about the 285? That's what I thought.

 

[RCFProd]

I prefer the GTX 960 over the R9 285. That's why you won't see me name the R9 285 in builds. Not bad at all though.

 

[ko888]

As the R9 280 disappears from the market then you will be hearing more about the R9 285.

 

[AMDRazer]

As the R9 280 disappears from the market then you will be hearing more about the R9 285.

Reply almost a year later:

I still have an R9 280 in my build, lol. I can't find listings for it anywhere. The last time i saw it was at least 6 months ago for 170.

Such a shame that its replacement is the R9 380

 

[RCFProd]

As the R9 280 disappears from the market then you will be hearing more about the R9 285.

Reply almost a year later:

I still have an R9 280 in my build, lol. I can't find listings for it anywhere. The last time i saw it was at least 6 months ago for 170.

Such a shame that its replacement is the R9 380

R9 285 = R9 380. They're the same.

 

[JackNaylorPE]

Why don't you use a chart that is closer to the OP's display resolution instead of one that is well beyond it?

Nothing even close to a 25 - 50% difference anywhere.

Try and follow the conversation .... the reason I didn't use a chart for 1050 1) because, as clearly stated stated, I wasn't comparing two cards but the same 970 card at different resolutions and b) because I was responding to a quote about 3360 resolution which I indicated was close to 1440p and therefore a significant 25-50& drop in performane could be expected compared to 1080p

And yes, the math was sitting right there for you, if you tell me where you got lost I can explain it to you

At 3360, you are talking 1.7 times more pixels than 1080p or approximately the same as 1440p

That is **not** going to be a small hit, more like 25 - 50% than 5%.....on the MSI 970 for example

Crysis 3 drops from 42 fps to 25.4
BF4 drops from 77.9 fps to 50.2
Far Cry 3 drops from 62.6 fps to 39.3
Tomb Raider drops from 59.5 fps to 35.0

Ok, lets do the math ... Please explain how any of these comparisons fall into the 5% range ? The way I was taught math ....

25.4 / 42 = 60.5% aka 39.5% drop
50.2 / 77.9 = 64.4% aka 35.6% drop
39.3 / 62.6 = 62.8% aka 37.2% drop
35.0 / 59.5 = 58.8% aka 41.2% drop

Again, all those numbers fit into the 25% - 50% performance drop that one would expect moving from 1080p to 3350 x 1050 or 1440p.


Since you seem to be a stickler in insisting that one only address the OPs 1050 resolution instead of the 3360 I was actually responding to, and respomd to the OP's stated 1050 resolution, I find it ironic that you don't to the same.

The OP is talking about the 280 vs 970, and what do your own references show using the faster 280x?

Overall performnce = 100% / 69% = 45% advantage

 

[ko888]

If you do decide to get a Radeon R9 290X make sure your power supply unit has sufficient power capacity and sufficient number and type(s) of PCI Express supplementary power connectors for that specific graphics card.

Okay, I'm sorry. The graph shows the 970 above 290x, and 290x is better? I'm confused.

What display resolution are you planning to use with the graphics card? If it's above 1920x1080 then the R9 290X will be the better performing card.

If your display resolution is 1920x1080 then get the GeForce GTX 970.

I will be using two 1680x1050 monitors, so 3360x1050. But I will not use both for gaming.

Why don't you use a chart that is closer to the OP's display resolution instead of one that is well beyond it?

Nothing even close to a 25 - 50% difference anywhere.

Try and follow the conversation .... the reason I didn't use a chart for 1050 1) because, as clearly stated stated, I wasn't comparing two cards but the same 970 card at different resolutions and b) because I was responding to a quote about 3360 resolution which I indicated was close to 1440p and therefore a significant 25-50& drop in performane could be expected compared to 1080p

And yes, the math was sitting right there for you, if you tell me where you got lost I can explain it to you

At 3360, you are talking 1.7 times more pixels than 1080p or approximately the same as 1440p

That is **not** going to be a small hit, more like 25 - 50% than 5%.....on the MSI 970 for example

Crysis 3 drops from 42 fps to 25.4
BF4 drops from 77.9 fps to 50.2
Far Cry 3 drops from 62.6 fps to 39.3
Tomb Raider drops from 59.5 fps to 35.0

Ok, lets do the math ... Please explain how any of these comparisons fall into the 5% range ? The way I was taught math ....

25.4 / 42 = 60.5% aka 39.5% drop
50.2 / 77.9 = 64.4% aka 35.6% drop
39.3 / 62.6 = 62.8% aka 37.2% drop
35.0 / 59.5 = 58.8% aka 41.2% drop

Again, all those numbers fit into the 25% - 50% performance drop that one would expect moving from 1080p to 3350 x 1050 or 1440p.

You keep ignoring the part where @AMDRazer stated "But I will not use both for gaming."

 

[JackNaylorPE]

Doesn't change anything..... the only relevance when picking the cards is when the resolution being used chnages what card has the advantage.

The 970 is still faster outta the box (overall based upon TPU testing) than the 280, 280x, 290, 290, 290x AT ALL RESOLUTIONS up to 1080p.... With all cards overclocked, the 970 is faster than 280, 280x, 290, 290, 290x, 390, 390x AT ALL RESOLUTIONS up to 1440p. The 9xx series cards overclock from 17% (970) - 31% (980 Ti).... the R9 300 series only 6 - 8%

 

[RCFProd]

It is true that with aircooling the GTX 970 is much better at overclocking, latest benchmarks for the following titles Rise of the Tomb Raider, Star Wars: Battlefront, The Division and Hitman: Absolution all show higher FPS out of the box for the R9 290X and R9 390 though, important mention!

 

[JackNaylorPE]

Air cooling is a non factor on 9xx series card... speaking from experience, you get no higher performance from water cooling. Twin 970s for example top out at 68C, well below the 80C throttling point. Look at any TPU 900 series non-reference card review ... even the G1 980 TI which can draw as much as 360 watts at extreme OCs, doesn't get close to the 85C throttling point.

 

[RCFProd]

I meant to say that the R9 290X/R9 390 is not that good for OC'ing until you get it watercooled, until then it stands no chance to the GTX 970 when overclocking.

 

[JackNaylorPE]

Agreed, WC'ing will help but as the 3xx are basically tweaked 2xx designs there's not all that much headroom left because they have been so aggressively clocked in the box. And, since CLC type cooling doesn't help all that much, it's not a big return. The FuryX for example, only brought home 5%

https://www.techpowerup.com/reviews/AMD/R9_Fury_X/34.html