Looking for Nano or Fury (Vanilla) owner for bench off vs 980

[tviceman]

As title says. I'd prefer Fury Nano at the minimum, but a 3 way bench off would be best. Best OC through normal software welcome (afterburner, precision, etc.), but no GPU or bios mods. 1080p and 1440p (because I don't think these GPUs have enough power for most 2014-2015 AAA games at 4k max settings). Prefer kill-o-watt but not required. We can agree on games and settings through private email.

 

[Azix]

290x 1440p if that's ok and if I can get the games. Guess I can downscale to 1080p

 

[Termie]

You may find a review I published conducting a Fury vs. 980 vs. 980 Ti shootout helpful in setting a baseline for your expectations.

The real issue here is that the Fury has so little overclocking headroom that a bench off really isn't necessary. You can more or less take Fury's stock scores and compare them to yours. The 980 has at least 20% more performance on tap versus reference, making it a better 1080p card, but still leaving it behind or tied at 1440p.

 

[james1701]

You may find a review I published conducting a Fury vs. 980 vs. 980 Ti shootout helpful in setting a baseline for your expectations.

The real issue here is that the Fury has so little overclocking headroom that a bench off really isn't necessary. You can more or less take Fury's stock scores and compare them to yours. The 980 has at least 20% more performance on tap versus reference, making it a better 1080p card, but still leaving it behind or tied at 1440p.

Sapphire has also updated to the Nitro card. They fixed some of the issues in your review. I love mine, but after only a week I have not looked at what the overclock can do. It will never touch the amount of overclock head room of Nvidia's cards, but it's still a nice improvement.

Also OP if you would like some bench numbers message me, I get what I can for you.

 

[AtenRa]

Count me in with a R9 390 for 1080p.

 

[Termie]

Sapphire has also updated to the Nitro card. They fixed some of the issues in your review. I love mine, but after only a week I have not looked at what the overclock can do. It will never touch the amount of overclock head room of Nvidia's cards, but it's still a nice improvement.

Also OP if you would like some bench numbers message me, I get what I can for you.

Does your Nitro have coil whine? I'm thinking of picking one up - it's on sale at Newegg, featuring a factory overclock equivalent to what I could get manually with the Tri-X: http://www.newegg.com/Product/Product.aspx?Item=N82E16814202186

 

[james1701]

Mine is in my 900D at my feet under my table/desk. No coil whine that I have noticed. The fan is the quietest I have ever heard on a GPU. Talking to sapphire ED, the US customer support rep, he will say the card might get some whine in menus, where FPS is really high, but mine had not done that yet.

I came from crossfire Gigabyte 7970GHz cards. The high FPS is a little lower going with a single card, but I'm playing Crysis 3 in eyefinity at medium settings doing 4960x1600 staying around the 100FPS It looks fantastic. Even doing 2560x1600 the minimums are higher than the crossfire setup and much smoother. I really like this card.

There is Valley at out of the box clocks, using a 3770K at stock speeds and no driver manipulations.

There is my Eyefinity Valley bench at the largest resolution that the program will allow with both CPU and GPU at stock speeds.

http://www.3dmark.com/fs/7547568

Firestrike results with 3770K @ 4.5Ghz.

 

[Silverforce11]

The real issue here is that the Fury has so little overclocking headroom that a bench off really isn't necessary.

Afterburner has support for Fiji vcore mod for awhile now, if you still have your Fury, give it a go at proper overclocking.

 

[n0x1ous]

Termie, give the vcore boost a go but my Fury X is basically unstable at anything over +25mhz even with +100mv just as a point of reference. you may get something out of it, but don't be surprised if it makes no difference.

 

[Termie]

Afterburner has support for Fiji vcore mod for awhile now, if you still have your Fury, give it a go at proper overclocking.

Termie, give the vcore boost a go but my Fury X is basically unstable at anything over +25mhz even with +100mv just as a point of reference. you may get something out of it, but don't be surprised if it makes no difference.

That's why I'm curious if the new custom PCB on Sapphire's Fury Nitro allows the extra voltage to produce some real performance gains.

 

[tviceman]

I see there are 390 and a 290x users who wouldn't mind a bench off. I'm ok with that, the problem is though the I wanted to test out the "theories" that OC Fury Nano is faster (and thus a better buy) than OC GTX 980 at 1080p and 1440p.

 

[Flash831]

I have a R9 Nano, and a kill-o-watt.

What benches you want me to run? ()

 

[Azix]

I see there are 390 and a 290x users who wouldn't mind a bench off. I'm ok with that, the problem is though the I wanted to test out the "theories" that OC Fury Nano is faster (and thus a better buy) than OC GTX 980 at 1080p and 1440p.

OC fury nano is fury x. OC 980 is about as fast as Fury. At 1440p it should be no match.

 

[Termie]

OC fury nano is fury x. OC 980 is about as fast as Fury. At 1440p it should be no match.

Except this isn't true. It's unlikely an OC Nano can even match an OC Fury, let alone a Fury X. It doesn't have the power circuitry to sustain Fury X clocks. If it did it would need far better cooling. The Fury is going to be the superior card under almost all circumstances.

The OP has a taker on the benchoff, let's see what it shows.

AMD really should have released an air-cooled Fury X at $550 or $600. I assume it's not allowing board partners to do so, as Sapphire for one certainly has a design capable of cooling it. Instead we get silly things like a liquid-cooled Fury from XFX for $600, which makes no sense.

 

[tviceman]

I have a R9 Nano, and a kill-o-watt.

What benches you want me to run? ()

Sweet I will PM you.

 

[IllogicalGlory]

Except this isn't true. It's unlikely an OC Nano can even match an OC Fury, let alone a Fury X. It doesn't have the power circuitry to sustain Fury X clocks. If it did it would need far better cooling. The Fury is going to be the superior card under almost all circumstances.

The OP has a taker on the benchoff, let's see what it shows.

AMD really should have released an air-cooled Fury X at $550 or $600. I assume it's not allowing board partners to do so, as Sapphire for one certainly has a design capable of cooling it. Instead we get silly things like a liquid-cooled Fury from XFX for $600, which makes no sense.

TPU's overclocking results show OC Nano > Fury X > OC Fury (Asus Strix) in BF3 1440p.

 

[MrTeal]

TPU's overclocking results show OC Nano > Fury X > OC Fury (Asus Strix) in BF3 1440p.

That results is really weird. They got the Nano to 1060MHz core and 565MHz memory, which is 6% on the core and 13% on the memory. That improve performance 12.4% in BF3. That's basically linear scaling with memory speed.

 

[Termie]

TPU's overclocking results show OC Nano > Fury X > OC Fury (Asus Strix) in BF3 1440p.

That results is really weird. They got the Nano to 1060MHz core and 565MHz memory, which is 6% on the core and 13% on the memory. That improve performance 12.4% in BF3. That's basically linear scaling with memory speed.

HardOCP's more thorough overclocking review shows that the Nano cannot reliably emulate a Fury X:

http://www.hardocp.com/article/2015...orm_factor_overclocking_review/5#.VseOp_ErIuU

 

[tential]

HardOCP's more thorough overclocking reviews that the Nano cannot reliably emulate a Fury X:

http://www.hardocp.com/article/2015...orm_factor_overclocking_review/5#.VseOp_ErIuU

You basically proved his point that the OCed nano is basically a Fury X....

Given the Nano's price it's the better deal between the Nano and the Fury. Easily.

$460 for the Nano currently or $500 for the Fury.

Either card is better than the Fury X though right now at $630.

I'd take the Nano though.

Right now I'd take,
R9 390
R9 390x
R9 Nano
GTX 980 Ti

 

[crisium]

As a side note, according to an AMD rep, the HBM in Fiji can only run at 500, 545, 600, or 666.6MHz. It rounds to the closest, and any other reading displayed via software is incorrect.

If this is true, then any review where they overclock the HBM between 525-570 is actually 545. It appears that 545 is virtually guaranteed, and 600 is rare without a good liquid setup.

I'm not really interested in testing myself, unless there is an easy canned benchmark that usually has parity (which I would define at 1440p as a 970 performing in-between a 290 and 290X/390, as this is where it usually lands on aggregate index summaries.) I can push my Fury Core 3840 higher than 1066MHz if I overvolt it, so I'd do any testing at 1100MHz which has proved stable thus far with an overvolt. And for the record I would not be interested in your arbitrary no bios mods, as this almost seems like leftover brand wars salt from 6950 to 6970 unlock > 570 days. It's a variable much like max stable overclocking is.

 

[Termie]

You basically proved his point that the OCed nano is basically a Fury X....
....

I know what the review says, and that's why I don't post just the graphs. Go and read it carefully. A 100% fan speed was required to hit those speeds, and voltage did nothing to stabilize core clocks, meaning only in games where the memory overclock had an effect could the Nano match the Fury X.

In other words, buy a Nano because it's a good deal, not because it will emulate a Fury X. It won't.

 

[tential]

I know what the review says, and that's why I don't post just the graphs. Go and read it carefully. A 100% fan speed was required to hit those speeds, and voltage did nothing to stabilize core clocks, meaning only in games where the memory overclock had an effect could the Nano match the Fury X.

In other words, buy a Nano because it's a good deal, not because it will emulate a Fury X. It won't.

I agree, Nano should be bought because it's a good deal at this point.

It doesn't matter though, good deals don't seem to help AMD. There are few reviewers doing full roundup reviews/GPU recommendations. Just basically all PR for the companies of GPU's they're sent.

 

[tviceman]

I sent Flash831 a PM, waiting to see what he says.

 

[3DVagabond]

That results is really weird. They got the Nano to 1060MHz core and 565MHz memory, which is 6% on the core and 13% on the memory. That improve performance 12.4% in BF3. That's basically linear scaling with memory speed.

It's because increasing the power slider stops throttling which gives you ~15% clock increase on the GPU.

HardOCP's more thorough overclocking review shows that the Nano cannot reliably emulate a Fury X:

http://www.hardocp.com/article/2015...orm_factor_overclocking_review/5#.VseOp_ErIuU

[H] have a real hate for Nano though.

 

[AtenRa]

The NANO is just thermally limited, with enough cooling it can reach Fury X performance easily. But it will cost you more to install a WC Full Cover Block and that will bring you close to Fury X price. It is better to raise the power limit to +20-30% and increase the fan speed up to a limit that the noise is not bothering you.