AMD: Pirate Islands (R* 3** series) Speculation/Rumor Thread
- Thread starter iMacmatician
- Start date
-
- Tags
- amd
With the early sales, has there been any comparison of the 390x cooler to the 290x one?
It's a non sense comparison, put an AIO on a 980ti and probably you'll have the same temps and decibels
In my opinion, it would be impressive to have those results with a single fan air cooler on a 275w tdp card
Last edited:
No its not, that's the least impressive part of this card, is the AIO, they actually need an AIO to cool this card down? That is a big draw back in engineering if that's the case. AMD won't put an expensive AIO cooler (relatively speaking with current fan driven cooling solutions) if they didn't have to. Everything comes down to $ and sense (yes sense, like what is logical)
Just hypothesizing here but with 275 watts average on this card, they didn't need the AIO they need it for people that overclock it...
The 980ti on average only uses ~210 watts, there is a big difference in terms of efficiency between maxwell v2 and Fury if the performance remains similar on the two cards.
It's unknown if the AIO cooler is really needed or just something they did for, uhm, coolness factor.
Not sure where you are getting your figures from but they are wrong. 980Ti and TX are right up there with 290X as far as power consumption goes.
http://www.anandtech.com/show/9306/the-nvidia-geforce-gtx-980-ti-review/16
As for needing AIO. Who said they needed AIO? Fury X has a 275W TDP. Less than Hawaii. What they did is overbuilt
cooling system so that its silent and cool and gives you large OC headroom. They could have easily slapped on a reference cooler with dual axial fans and it would still be quieter than 980Ti reference radial cooler.
AMD can't catch a break. People complain that their reference coolers are too noisy then when they put a much quieter one on, then AMD still sucks. The amount of FUD and fearmongering in this thread is amazing.
-FUDie
I don't care what cooler they use, if its fast, (relatively) cheap, cool and quiet then its a great combination as far as I'm concerned. This paired with one of the new desktop broadwells would make a small, very fast and very, very quiet PC. Costly, but oh so nice!
Something very fishy going on with Tomb Raider:
Not sure what settings TPU is using @1080p, but judging by FPS probably higher? Definitely fishy TX results.
http://www.techpowerup.com/reviews/NVIDIA/GeForce_GTX_980_Ti/25.html
http://www.anandtech.com/show/9306/the-nvidia-geforce-gtx-980-ti-review/16
Please look at more then one review to make your assertions of different cards, its hardly that hard to do.
Anandtech is only one of them and only one game.
TPU gives it 213
Extreme tech total system usage 330 at the wall
Tom's gives it 233
Hot hardware is 335 at the wall
legit review: 354 at the wall
Hard OCP : 372 at the wall
many more
All of these are lower then the r290x by a considerable amount 70 plus watts.
Fury X has a 275 average power consumption, not TDP, TDP is over 500 watts. And what you just stated is what I just stated.
its not that impressive, if its needed, it actually will have a negative connotation right off the bat, its impressive for what it can do for overclocking, yes it is. I think its going to be an overclocking beast, but its not like someone cannot put an AIO on other cards too.
Fury X has a 275 typical board power. Some early Hawaii reviews gave it a higher typical board power, but AMD's figures have it at 250W.
Typical board power is whatever AMD says is close enough for government work.
TDP is the line between the card running or failing to dissipate enough heat to avoid shutdown or damage.
Well..
http://www.legitreviews.com/nzxt-kr...oler-review-on-an-amd-radeon-r9-290x_130344/4
Now, Fury is significantly larger than Hawaii, so the number of transistors contributing to static leakage is higher, roughly 40% higher.
If static leakage was unimproved, a similar air-cooled to water-cooled Fiji comparison would have around 36W of savings.
That's over 10% of the power budget, stock.
How much is Fury X winning some of the benchmarks again?
I checked the TPU average numbers. 211W for 980Ti 258W for 290X. So 47W difference. But yes that's still some 20W higher than what AT got in FarCry3.
I would guess 290X power consumption is somewhat inflated by very high temps its running at with reference cooler.
Which reminds me, where are 390X power consumption numbers?!
Last edited:
gamervivek
Regular
Tom's have 290X uber at 242W compared to 233W of 980Ti. What exactly you mean to convey by that statement then?
Looking at one review is like looking at none, it doesn't' give a good picture, I just garbed the first few that google pulled up, I can probably get even more, then 70 watts, more like 90 watts if I wanted to cherry pick.
Its just that the r290 is lets say 50 watts higher, it seems fury will be 25 watts more then the r290, its just higher that's it there is no argument to change that. There is a reason when AMD isn't showing benchmarks and power consumption for the Fury cards on their slides, when was the last time we didn't see an IHV not show performance figures vs other cards with marketing slides, err........
I think it was the FX series from nV, not saying this is going to be an FX series, and for AMD or at the time ATI it was the r600 series, not saying its an r600 either, doesn't look to be.
Edit, can't seem to find them anywhere on then net, if someone can put that up that would be great, cause I don't remember them that well either.
In any case its simple when they start swaying away from performance and power usage vs their competition, they are hiding something or something they don't want to give up because its going to hurt them some way.
Sun Tzu's art of war,
“Know the enemy and know yourself; in a hundred battles you will never be in peril” Sun Tzu
AMD knows where they are and where their enemy is, and they can't afford to show any weakness compared to their enemy. Eventually it will come out, but if they can get enough good media before that, it won't hurt them as much.
Its all going to come down to power consumption and possibly memory amounts (this was something they didn't get to much into as well)
Last edited:
