R9 Nano benchmarks surface

[Megahurt]

Compute performance has it at almost the same performance as Fury, maybe priced between Fury X and Fury or the same as Fury X.

[chillxnate]

Makes we want to attempt a watercooled Hadron Air again. 

[Cookybiscuit]

Does anyone even care about Fury anymore?

[nims0c]

Will we see custom PCBs or coolers from AIB partners?

[Misanthrope]

I'm really curious about the price of this thing.  If it costs anywhere close to the Fury X, it's just not going to be worth it.  Most ITX cases are built to accommodate large cards so unless OEMs like Origin or Digital Storm make an ultra-tiny case specifically for the Nano, it makes no sense.

 

Now if it's priced lower between $450-$500, it might prove to be the overclocking card everyone will want to buy.

 

Errm yes and no. Most ITX are build to accommodate physically big gpus, but almost none of them are build to accommodate proper cooling: you're expected to use a loud and ineffective blower card. More over, most cases open up other options like larger PSUs and extra 3.5 slots or radiators if you use shorter cards so it's still an overall better option to get short itx cards since they are designed to work efficiently within all constrains, not just size constrains.

[Virgule]

its not locked, just no one has figured out how to change the voltage. you can already change the GPU/MEM frequencies.

 

just wait for MSI afterburner to get support.

 

 

that's not how it works. there is no driver unlocking of voltage control, the maker of msi afterburner needs to physically have a fury x sample that he can then work with to unlock it. unlike nvidia, AMD has never "locked the voltage" and they've never released an API that allows voltage control.

 

 

I was not referring to current voltage "lock", but to current poor overclocking limitations. I'm not sure if it has to do with the driver or the API but reviewers have typically reported having a 10% if not less overclocking limit on the base clock. I know it is up to MSI afterburner to support voltage modification. However, in the case of memory clocks, it is unavailable to modify in MSI afterburner AND in AMD Catalyst, making me think that THIS could have some thing to do with drivers, therefore making me looking forward to a possible driver update "unlocking overclocking potential".

[The Benjamins]

I was not referring to current voltage "lock", but to current poor overclocking limitations. I'm not sure if it has to do with the driver or the API but reviewers have typically reported having a 10% if not less overclocking limit on the base clock. I know it is up to MSI afterburner to support voltage modification. However, in the case of memory clocks, it is unavailable to modify in MSI afterburner AND in AMD Catalyst, making me think that THIS could have some thing to do with drivers, therefore making me looking forward to a possible driver update "unlocking overclocking potential".

 

again it is not locked, the GPU goes unstable and need more voltage to operate at high frequencies. to go past the 10% you need to add some voltage, that is how overclocking works.

 

people have already found a way to change the memory clocks, as far as I know.

 

every one of these in that list has overclock memory. and I read a article how its in MSI AB now

http://www.3dmark.com/search?_ga=1.35996646.716712013.1433948432#/?mode=basic&url=/proxycon/ajax/search/gpuname/fs/P/AMD%20Radeon%20R9%20Fury%20X&gpuName=AMD Radeon R9 Fury X

[Virgule]

again it is not locked, the GPU goes unstable and need more voltage to operate at high frequencies. to go past the 10% you need to add some voltage, that is how overclocking works.

 

people have already found a way to change the memory clocks, as far as I know.

If it's supported, why does AMD lock it in Catalyst ?

[StareX]

HBM2 is not in production yet

Yes it is..?

[The Benjamins]

If it's supported, why does AMD lock it in Catalyst ?

 

it is new tech and they don't need just anyone able to mess with it. no one knows how far you can push HBM or what happens if you go too far.

[Virgule]

it is new tech and they don't need just anyone able to mess with it. no one knows how far you can push HBM or what happens if you go too far.

...meaning it's not supported yet, hence my statement.

[MEC-777]

you could rape the connector like the 295x2

 

HA! That made me chuckle. So true.

 

I literally lol'd the moment I saw the 295x2 only had two 8-pin connectors. Only takes some simple math to realize they're pulling WAY more than those connectors are rated for.  :lol:

[The Benjamins]

...meaning it's not supported yet, hence my statement.

 

you can still do it, overclocking is never fully supported.

 

also I am noticing alot more Fury X's in 3Dmark having 1100+ GPU and 600+ Mem clocks. they must be upping the voltage.

 

http://www.3dmark.com/fs/5432448

[dalekphalm]

If it's supported, why does AMD lock it in Catalyst ?

AMD has never allowed voltage control via AMD Catalyst - even when a GPU had fully unlocked voltage. You always had to use a 3rd party overclocking tool.

 

Yes it is..?

HBM 2 manufacturing is ramping up, but no current products are using HBM2. Fury Nano will be using HBM1.

 

...meaning it's not supported yet, hence my statement.

See above. Supported does mean they provide the option. AMD has never allowed voltage control via AMD Catalyst. You always had to use MSI Afterburner or some other 3rd party tool for voltage control. Ultimately it's up to the actual board partner whether they wish to lock voltage control via the BIOS or not.

[Virgule]

AMD has never allowed voltage control via AMD Catalyst - even when a GPU had fully unlocked voltage. You always had to use a 3rd party overclocking tool.

 

Yes, but in the case of memory clocks, AMD does not support it. They always provide the option in Catalyst, they don't with Fury lineup.

[dalekphalm]

Yes, but in the case of memory clocks, AMD does not support it. They always provide the option in Catalyst, they don't with Fury lineup.

True - but I think there's some technical reason why Memory control hasn't been enabled yet - they are probably still refining some of the Drivers and will likely release Memory Control at the same time that Voltage Control becomes actually usable.

 

I personally don't think that Memory OCing is "official" yet. People have just managed to unlock it unofficially.

[Virgule]

Memory control hasn't been enabled yet - they are probably still refining some of the Drivers and will likely release Memory Control at the same time that Voltage Control becomes actually usable.

Exactly what I'm looking forward to.

[dalekphalm]

Exactly what I'm looking forward to.

Keep in mind what I'm saying is 100% speculation.

[Sammael]

I'm curious what the pricing will be.  The fury is already 550, fury x 650....  could they make the nano 499?  Even though it's a full fury?  Maybe the part is binned lower that can't hit clocks as high so that makes it easier to discount.  

[Spoonfull]

I'm curious what the pricing will be.  The fury is already 550, fury x 650....  could they make the nano 499?  Even though it's a full fury?  Maybe the part is binned lower that can't hit clocks as high so that makes it easier to discount.  

 

Im thinking the 470 dollar mark. It would be so much easier to predict its pricing if the 390x didnt exist.

[dalekphalm]

I'm curious what the pricing will be.  The fury is already 550, fury x 650....  could they make the nano 499?  Even though it's a full fury?  Maybe the part is binned lower that can't hit clocks as high so that makes it easier to discount.  

You're going to be limited by how much power it can draw anyway - it has a single 8-pin PCIe power connector.

 

I suspect pricing will be around $450-500 personally.

 

It very well might OC like a beast, but we just don't know. Even if it does OC well, we're still likely limited by the power draw - so that means it probably won't beat out an R9 Fury.

[mr moose]

Does anyone even care about Fury anymore?

 

Quite a few, while for many it does not offer compelling reason to buy it right now, it is still only the first release and we have no idea what will happen as they iron out the kinks.

[Dabombinable]

Hang on, the R9 290X doesn't actually consume more than 100W over the GTX 970, and the 290X performs around the same, so AMD is actually leading in efficiency now  :blink: ?

[Curufinwe_wins]

Hang on, the R9 290X doesn't actually consume more than 100W over the GTX 970, and the 290X performs around the same, so AMD is actually leading in efficiency now :blink: ?

You can't do math. 290x tdp is 290W, 970 is 145. Now I know tdp doesn't mean consume BUT that's the number AMD is talking about.

[Curufinwe_wins]

Double post. My bad