Fury Nano Discussion Thread

[cen1]

Yes, pretty much every review! Pcper, guru3d, hwcanucks..

 

[SimianR]

It's definitely a very interesting card. If AMD ramps up production of the Fiji based cards - it would be a very compelling product at $499-549. Maybe someday..

 

[Techhog]

I guess I just don't care enough about small cards to be impressed. :\

It comes more into play in the future. Just think about it. This is a GCN card which outperforms the 290X significantly while using 100W less power and is much smaller. That says A LOT about where HBM can take this industry. Imagine if Maxwell had HBM. Faster than Titan X, while using less than 200W, while still on 28nm. What does this mean for 16nm?

Still not worth the price, though.

 

[RussianSensation]

It's definitely a very interesting card. If AMD ramps up production of the Fiji based cards - it would be a very compelling product at $499-549. Maybe someday..

I think AMD needs positive PR / word of mouth advertisement from excitement of having 1 solid product that punches above its weight class. During HD4800 both the 4850/4870 cards accomplished this. During HD5800 series, again both the 5850 and 5870 cards did. During HD6000 series, HD6950 was the star of AMD's line-up offering performance far above its asking price once unlocked and overclocked. During HD7000 series, HD7850's overclocking and HD7950's overclocking made them stellar values. It was possible to beat HD7970Ghz/GTX680 with an overclocked 7950. With R9 200 series they had R9 290. With this generation, they do not have such a product.

I think AMD should have priced Fury X at $549-579 because GTX980Ti is a better product overall. After that, they should have decided which card would be that HD4870/5850/6950/7950/R9 290 successor and priced it at $449. This could have been the Fury or the Nano. If they had done so, it would have made a lot of people think twice about buying a 980 or spending the extra $ on a 980Ti. As it stands though, both the Nano and the Fury X are going to be niche cards while the Fury is priced too closely to the 980Ti and too far above the 980/390/970 cards to make sense.

I guess AMD cannot afford to sell those cards at lower prices and while supplies are low, they can keep their prices high but I don't know how long this can continue as their market share will keep dropping more and more.

That says A LOT about where HBM can take this industry. Imagine if Maxwell had HBM. Faster than Titan X, while using less than 200W, while still on 28nm. What does this mean for 16nm?

Still not worth the price, though.

Exactly. The Nano is 78% faster than the R9 280X at Computerbase.de at 1440P HQ. And people said HBM did 0 over GDDR5. Since AMD didn't have a completely new architecture, it is because of using HBM that they were able to increase the die size and reduce power usage/die space over GDDR5. With 2016-2018 GPUs we will start to see the true benefits of HBM with 1-1.5TB/sec memory bandwidth GPUs, new GPU architectures, etc. If Maxwell had HBM instead of GDDR5, it would have been insanely impressive and that's why many expect amazing things from Pascal. I am honestly expecting 80-100% increase in performance over GTX980Ti in the fastest Big Daddy Pascal chip because NV will have HBM2 + 16nm node + new architecture. The last time NV had a full node shrink and new architecture, they doubled the performance moving from GTX580->780Ti.

I presume Fiji/Fiji XT will be shrunk & improved to the mid-range segment next generation and replace existing R9 390/390X level of cards so that perf/watt and size of cards like the Nano would travel down to $300-400 segments.

I think the Nano is priced too high but its performance is miles ahead of the fastest miniITX GTX970 level card. Because it's a fully unlocked Fiji XT GPU, with increased powertune and overclocking, near Fury X performance is possible in a 6 inch videocard.

 

[Hitman928]

Yes, pretty much every review! Pcper, guru3d, hwcanucks..

Ok, see it in Guru3d now, they tacked it on in a one liner at the very bottom of the page.

Here's Guru3d's compiled results for those that like this sort of thing. I'll try and do for others later if I have time,

 

[RussianSensation]

Ok, see it in Guru3d now, they tacked it on in a one liner at the very bottom of the page.

Here's Guru3d's compiled results for those that like this sort of thing. I'll try and do for others later if I have time,

Nice summary Hitman! Looks like the >30% faster than GTX970 was dead on.

 

[Hitman928]

Anandtech

 

[tential]

Because RS, most people looked at the marketing slides and didn't actually rationally think about it. This is why I stressed so much that the marketing was hurting AMD. This is a fast card. It's a slightly slower Fury X. Why it would be in a conversation with the GTX 970 is beyond me when the card is up to 40% faster than a 970 at times.

So yes, I think a LOT of people were wrong. I expected more performance due to Silverforce's numbers on the R9 290x/290, and due to Tomshardware Fury X powerlimit numbers.

I think this is a good start.

I think Arctic Islands will build on the Fury X, Nano, and Fury and if Nvidia doesn't improve it's product lineup (from a build standpoint not a performance standpoint), I think AMD will have a good advantage from a build quality standpoint on the high end.

they just need to ensure there is no stupid mistake like faulty pumps, coil whine, stating it's an OC dream when it's not, etc. and I think Arctic Islands will be a great improvement on the product stack as a whole.

 

[flopper]

they just need to ensure there is no stupid mistake like faulty pumps, coil whine, stating it's an OC dream when it's not, etc. and I think Arctic Islands will be a great improvement on the product stack as a whole.

Not counting on that yet, no evidence that Lisa Su been able to correct the system itself at amd. The recent comment of the OC dream is just plain showcase that the guys at amd engineers builders have no clue of the real world they sell their cards to.

 

[Hitman928]

Here's nordichardware, it's not done, there's 3 games I missed but the site seems to have crashed, I'll update when it's back online.

 

[AnandThenMan]

Thanks for compiling those graphs very very useful.

 

[Hitman928]

I should also say that all of them are taken from the 1440p results, I'll add that to the graphs.

 

[jpiniero]

It's about what I expected. Seems like it is in stock, which I didn't expect given that the Fury X isn't available. Maybe AMD has given up on the water cooler?

 

[crashtech]

So, can we slap a water cooler on a Nano and OC it?

 

[crisium]

This card definitely exceeded my performance expectations. I took AMD's 30% to mean 15% in reality. But it often delivers.

It's a very niche product though. It is only for itx cases that will not accommodate larger cards, and that's an utterly small part of the market. It loses in raw performance and/or performance per dollar to every other card. Sapphire Tri X Fury is likely slightly faster overall, and certainly quieter and cheaper if you can fit it in your case. 980 is barely slower and way cheaper. Fury X and 980 Ti beating it for the same price... etc. Terrible PPD, incredible PPW and Performance Per "Size".

Niche to the point of obscurity perhaps.

 

[jpiniero]

The problem is if you cut the price of the Nano, it kind of makes the regular Fury redundant and they'd have to cut the price of that too. If it doesn't sell and AMD manages to get enough volume they will have to do something but it wouldn't be cheap. You'll notice that nVidia hasn't released any other chips based on GM200 other than the Titan X and 980 Ti.

 

[3DVagabond]

Did anyone else comment on the coil whine besides HWC? I read through a couple of others and didn't see any mention of coil whine.

This card definitely exceeded my performance expectations. I took AMD's 30% to mean 15% in reality. But it often delivers.

It's a very niche product though. It is only for itx cases that will not accommodate larger cards, and that's an utterly small part of the market. It loses in raw performance and/or performance per dollar to every other card. Sapphire Tri X Fury is likely slightly faster overall, and certainly quieter and cheaper if you can fit it in your case. 980 is barely slower and way cheaper. Fury X and 980 Ti beating it for the same price... etc. Terrible PPD, incredible PPW and Performance Per "Size".

Niche to the point of obscurity perhaps.

Nice to see AMD make a perf/W champion so that metric will no longer be used as trumping everything else.

 

[AtenRa]

Nice to see AMD make a perf/W champion so that metric will no longer be used as trumping everything else.

Perf/watt will be forgotten now, perf/$ will be the new mojo from now on

 

[Qwertilot]

Prob only in rather small numbers anyway, so niche/objectively overpriced is OK.

They can't really cut the price - die picked fully capable cards? Automatically a very expensive way to behave. NV does similar of course, but they can push them into laptops where people seem happy to pay the consequent premiums.

All of these Fury cards are about them trying to keep people thinking of them as technically relevant so their next gen stuff gets a fair hearing. Seems reasonably effective in those terms.
(Just think about how it'd look without any of the Fury cards for why they're needed.).

Tiny cases are definitely good Would be nice if this form factor catches on properly. Not perhaps so much for cards this big/fast - this would be much quieter with a biggish cooler and 2 fans - but there's plenty of lower powered GPUs where people seem to love sticking huge coolers on them. Little reason for a 960 to need 2 fans, let alone the 3 they stuck on some models!

No big reason to have a >35w cpu nowadays with where Intel are optimising for, so you can get away with a tiny cooler there.

Cases rare mind. Not sure that Q21 has the air flow for it? The Q33 likely does - and has always looked rather nice to me. Not terribly small though, if quite a small desktop footprint.

Wonder how much length you could take out of something like a Node 304 if not needing full sized graphics? Maybe not so much with the motherboard + PSU to fit in to one level.

 

[Enigmoid]

Perf/watt will be forgotten now, perf/$ will be the new mojo from now on

Obviously.
If they are the same there is no competitive advantage there anymore.

If I'm looking for a fuel efficient car and the best two cars have very similar fuel efficiency obviously I move onto other factors as neither offers an advantage in that particular criteria anymore.

 

[Gikaseixas]

Obviously.
If they are the same there is no competitive advantage there anymore.

If I'm looking for a fuel efficient car and the best two cars have very similar fuel efficiency obviously I move onto other factors as neither offers an advantage in that particular criteria anymore.

Very well said as long as the other metrics still valid. Problem is people tend to find excuses to hate on certain hardware no matter what. It applies to several brands, not just AMD.

 

[3DVagabond]

Obviously.
If they are the same there is no competitive advantage there anymore.

If I'm looking for a fuel efficient car and the best two cars have very similar fuel efficiency obviously I move onto other factors as neither offers an advantage in that particular criteria anymore.

But when AMD was 45% better in terms of perf/w the standard retort was efficiency of Maxwell.

Remember the perf/$ mantra when DX12 comes out. Right now a lot of GCN's hardware isn't even being used efficiently.

 

[Elixer]

So, can we slap a water cooler on a Nano and OC it?

Well, you could, since you have 1 8 ping power plug + 75W from PCIe, then the max the card can get is 225W, and not the 175W it is limited to now.
This would require a BIOS mod I bet.

 

[RampantAndroid]

Perf/watt will be forgotten now, perf/$ will be the new mojo from now on

Yeah, because I totally wasn't reading reviews for *years* that evaluated perf/$.

Right.

Well, you could, since you have 1 8 ping power plug + 75W from PCIe, then the max the card can get is 225W, and not the 175W it is limited to now.
This would require a BIOS mod I bet.

Is the 75W from the mainboard a solid number? I seem to remember someone implying that those 75W weren't reliable?

 

[Shehriazad]

Seems this card is also the easiest to "overclock" out of the Fiji bunch since all you'll have to do is ramp up the power target/voltage a bit so it more consistently stays at its' 1000 Mhz. Most tests (at least the German ones') I've seen so far all seem to be stuck with a max OC of +50 Mhz.

But cranking up the power as far as possible gave them up to +15% in benchmarks or some games thanks to the fact that it will not downclock as much anymore.

Then again, the card will remain limited to the 8pin unless AMD opens that up to other Card Vendors...which I almost doubt. xD Either way...bios mods will probably still try to do the full 225W no matter the reliability here. 220W should put it at a somewhat stable Fury X scenario...but I doubt that the Nano cooler could still keep up with that without going crazy (or can keep up with it at all).


Either way...despite the initial price this card still looks like a very good success. Once I can get one later during a sale or something for around $500...I will certainly get one. For now I'm just worried about audible coil whine...but only 1 out of 4 tests I've read complained about that. Haven't checked out English reviews yet.