[Rumor (Various)] AMD R7/9 3xx / Fiji / Fury

Page 78 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

exar333

Diamond Member
Feb 7, 2004
8,518
8
91
http://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_980_Ti/32.html

The 970 has about 60% better perf/watt than the 290. If we take AMD's claim at face value the Nano's perf/watt will be significantly better than the 970.

If true, that would be pretty nice. I wonder what these will cost...if they are just downclocked Pros, why buy the pro at all if you could get this and OC it? Maybe more cut-down than the pro + lower clocks and voltage?

Interestly enough, at the 390x prices, I wonder if its worth a gamble to get the nano and just see how high the clocks can go, voltage put aside? Similar to a low-voltage CPU SKU you can get today. Interesting.
 

Grooveriding

Diamond Member
Dec 25, 2008
9,147
1,329
126
You're right. But the 780 didn't exactly equal Titan performance. The 980 Ti is so close to Titan X as to make no real difference.

It's worse than that. The custom 980ti cards are better than the TX right out of the box. The original Titan retained relevancy until the 780ti. This time the TX is a full chip from the get go and they configured GM200 in 980ti to be almost the same in performance. So when you add a custom board, vs the always reference TX, Titan X becomes slower, hence pointless.

They killed the TX two months after releasing it. It was obvious something was up. Dropping the 980ti made good sense in the hope of locking down as many buyers as possible for that price point before they could see Fury X and make a comparison based buy.
 

sm625

Diamond Member
May 6, 2011
8,172
137
106
perfwatt_3840.gif


http://www.techpowerup.com/reviews/Gigabyte/GTX_980_Ti_G1_Gaming/31.html


A 50% increase in perf/watt really doesnt do anything for the 290X. It would still be 10-20% below the 980ti. Which means the Fury X would need to draw 300+W to match a 260W 980ti. AMD basically openly admitted Fury cannot compete on perf/watt. Since they cant do that, there is no realistic way they can win an overclocking contest. Look at how the big maxwell perf/w scales with clockspeed. 15% performance gained for only a 4% loss in efficiency.
 
Last edited:

geoxile

Senior member
Sep 23, 2014
327
25
91
If true, that would be pretty nice. I wonder what these will cost...if they are just downclocked Pros, why buy the pro at all if you could get this and OC it? Maybe more cut-down than the pro + lower clocks and voltage?

Interestly enough, at the 390x prices, I wonder if its worth a gamble to get the nano and just see how high the clocks can go, voltage put aside? Similar to a low-voltage CPU SKU you can get today. Interesting.

IIRC Raja mentioned something about optimized power on the boards. So perhaps the Nano's tiny board has something to do with the power consumption. Fiji may not overclock well.
 

IEC

Elite Member
Super Moderator
Jun 10, 2004
14,597
6,076
136

exar333

Diamond Member
Feb 7, 2004
8,518
8
91
I thought you were talking about Fury Nano which is air cooled?

Interesting if true. Some other sites have speculated that the Nano would be CLC as well, but based on the picture in the blog/AMD event it appears air-cooled? Watercooled would likely help TDP even more, so I wonder if that will be available too?
 

monstercameron

Diamond Member
Feb 12, 2013
3,818
1
0
cant wait to see review not mention perf/watt if amd has a lead there! this will be the most hilarious and informative round of reviews yet!

so:
perf/watt
perf/dollar
smoothness
software features
drivers

let us see which review ignores all these to spin AMD in a negative light.
My bet is on pcper...
 

AnandThenMan

Diamond Member
Nov 11, 2004
3,991
626
126
Interesting if true. Some other sites have speculated that the Nano would be CLC as well, but based on the picture in the blog/AMD event it appears air-cooled? Watercooled would likely help TDP even more, so I wonder if that will be available too?
If you had a chance to watch the stream you would have seen Fury Nano, tiny card that is air cooled. Really awesome looking I must say.
 

exar333

Diamond Member
Feb 7, 2004
8,518
8
91
If you had a chance to watch the stream you would have seen Fury Nano, tiny card that is air cooled. Really awesome looking I must say.

Yeah, couldn't watch the event, unfortunately. Just able to see the live blog.

I wonder if custom Nanos will be released? That could be a much better option for max performance vs. the 390x. Can't wait to see review. The 2x efficiency is a little out there, but amazing if true.
 

Enigmoid

Platinum Member
Sep 27, 2012
2,907
31
91
We'll see if gamers really care that much about performance per watt as a metric soon enough. A Fiji-based Nano at 2x perf/watt vs R9 290X would be the #1 card in that chart. And significantly better than a GTX 970.

Just like the 980/970 with 2x efficiency over the 770?

I will not take anything at face value unless confirmed by a third party.
 

beginner99

Diamond Member
Jun 2, 2009
5,315
1,760
136
Mark Huddy was not impressive at all. The only parts of the presentation that were interesting were the Fiji discussion (Thanks Lisa Su) and when game developers took over.

He should be let go. This is a crucial release for AMD and his portion of the presentation was depressing and underwhelming. The puns were awful. Also, "Chief Gaming Scientist"... give me a break.

I'm really excited for the Fiji Nano. I hope the pricing makes sense.

Yeah, he was terrible. I can't believe AMD does not have a guy who is a better presenter. In fact most of them were bad to mediocre presenters, also Lisa Su. I mean shouldn't such a company have at least one guy how is a top notch presenter exactly for events like this?
 

Stuka87

Diamond Member
Dec 10, 2010
6,240
2,559
136
Just like the 980/970 with 2x efficiency over the 770?

I will not take anything at face value unless confirmed by a third party.

Efficiency does not mean less power draw. It means better performance for a given amount of power usage. If you take a card that draws 100W and gets 50fps, and another card that draws 100W but gets 100fps, the second card has 2x the efficiency.
 

exar333

Diamond Member
Feb 7, 2004
8,518
8
91
http://cdn3.wccftech.com/wp-content/uploads/2015/06/AMD-Radeon-R9-Fury-X-3DMark-Firestrike.png

http://cdn4.wccftech.com/wp-content/uploads/2015/06/AMD-Radeon-R9-Fury-X-3DMark-11.png

Not solid sourcing but at least something.

Fury X 10% better than 980ti at stock? We know 980ti has headroom for OC, need to see how much Fury X has now.

We need more actual performance benchmarks though. Definitely promising...

3D Mark can be affected by memory clocks and so forth in odd ways. HBM could either be over or under-represented in the real-world benches.

Can't wait to see some real reviews soon. Any benchmarks should be read with caution until then IMO.
 
Last edited:

exar333

Diamond Member
Feb 7, 2004
8,518
8
91
Efficiency does not mean less power draw. It means better performance for a given amount of power usage. If you take a card that draws 100W and gets 50fps, and another card that draws 100W but gets 100fps, the second card has 2x the efficiency.

Yeah, and it gets even more confusing because you could be talking about 'peak efficiency' as well. These are all terms that frame the best-case, generally, so more info to come I am sure...