Tuesday, August 25th 2015

AMD Radeon R9 Nano Core Configuration Detailed

AMD's upcoming mini-ITX friendly graphics card, the Radeon R9 Nano, which boasts of a typical board power of just 175W, is not a heavily stripped-down R9 Fury X, as was expected. The card will feature the full complement of GCN compute units physically present on the "Fiji" silicon, and in terms of specifications, is better loaded than even the R9 Fury. Specifications sheet of the R9 Nano leaked to the web, revealing that the card will feature all 4,096 stream processors physically present on the chip, along with 256 TMUs, and 64 ROPs. It will feature 4 GB of memory across the chip's 4096-bit HBM interface.

In terms of clock speeds, the R9 Nano isn't too far behind the R9 Fury X on paper - its core is clocked up to 1000 MHz, with its memory ticking at 500 MHz (512 GB/s). So how does it get down to 175W typical board power, from the 275W of the R9 Fury X? It's theorized that AMD could be using an aggressive power/temperature based clock-speed throttle. The resulting performance is 5-10% higher than the Radeon R9 290X, while never breaching a power target. Korean tech blog DGLee posted pictures of an R9 Nano taken apart. Its PCB is smaller than even that of the R9 Fury X, and makes do with a slimmer 4+2 phase VRM, than the 6+2 phase VRM found on the R9 Fury X.
Sources: VideoCardz, IYD.kr
Add your own comment

101 Comments on AMD Radeon R9 Nano Core Configuration Detailed

#76
Tatty_One
Super Moder@tor
Sony Xperia S
Let's make these things clear:

- R9 Nano will be faster product than R9 290X but not that fast to threaten any Fury;
- R9 Nano won't cost more than 630$, actually we expect price tag in line of around 450$;
- It won't be niche - it should be the new standard or paving the way for new generations of small cards;
- Fury X doesn't need custom boards - you won't achieve anything if you are seeking for guinness record clock heights. It's the best with water and let it please stay with WATER !
So based on your theory, where does the 390X fit in? As for it's pricing, taking into account the price of a 390X at around $420 - $450 how will the 2 fit and who are "we expect a price of 450"?
Posted on Reply
#77
FordGT90Concept
"I go fast!1!11!1!"
Nano should be at or below 390X. The question now is whether or not that is before or after the thermal throttling. We won't get that answered until benchmarks are in.
Posted on Reply
#78
Fluffmeister
Xzibit
VideoCardz.com


I must confess I would expect AMD's top of the range Fiji 596 mm2 HBM monster chip to do a little better against a heavily sliced GM204 398 mm2 chip, but I guess a win is a win.
Posted on Reply
#79
lilhasselhoffer
Sony Xperia S
It is silly from your side to prefer air rather than water. You know that lower temperatures have positive effects - they lower power consumption and increase life time.

But these things are mysteries for you.

Probably you are typing just to argue with someone who has better points than you.
I got used to your points and honestly - I am sick of them and want something better.
Two words: Prove it.

Lower temperatures have positive effect: Theoretically true, but a non-point. Cards without extreme overclocking generally have fans and TIM fail before thermal degredation. Additionally, you could theoretically clock you GPU at 200 MHz and decrease power consumption drastically. You'd have a GPU that might well last you your entire life. Of course, that'd be stupid. After a few years GPU performance is dramatically better. Additionally, under clocking would mean your card can't really run any modern games, but superior users have imagination, no. All they need is a text based RPG and they can imagine the rest of it. Kinda seems like cards are clocked such that performance is as good as possible, with the intended lifetime before most cards die being 3-5 years. Overclockers might burn through that lifetime faster, but the replacement cycle generally reaps cards before physics.


You have "better points" than me: In what strange world do you live? Perhaps it's different, where the physical laws of the universe are somehow fundamentally different. Cooling with water is not better than cooling with air, because it's the same physical process. An area of higher thermal energy has a fluid of lower thermal energy pass over it. Due primarily to conductive transference, a part of the thermal energy is transferred to the fluid, and the fluid is moved from the object. The things that influence how much energy is transferred is the thermal capacity of the fluid, the surface area of contact, and the velocity the fluid is flowing. As such, given enough time and resources I could design an air cooler that beats any water cooler on the market. This is physics in action, and your statement about water is better than air demonstrates you are either ignorant or an idiot. Hopefully the former, but given our past discourse the later.


You want something better: Yeah, so do I. I want real engineers to have a crack at the Fury, and have them show us what it could be. By your logic, once a single car is put out on the market we should be done. No SUVs, no trucks, no motorcycles, we only need a motorized one seat car. That's limiting the playing field, because you're too stubborn or have something up your sleeve that you aren't sharing. In the case of the former, you're killing potential sales with inflexibility. In the case of the later, you are trying to demonstrate how clever you are by lopping of the end of your nose. Stupidity through and through. Or perhaps I'm reading this point wrong.

Maybe your logic is that we only need one video card from each team. Nobody wants the 960, 970, 280, 380, 390, 930, or the plethora of other cards. We only need the 980 and Fury competing. Everyone playing PC games should own one of these two cards, and the day they are released every other card should automatically combust. Heck, we only need one card from each team. Let's all make our computers shoot us, should the Steam survey detect anything but the latest piece of hardware in our system.

No matter how I look at your point, it is backwards. Choices are demonstrably what consumers want, and choices actually make things sell. Market research proved this year ago. If you don't agree, walk down a grocery store pasta aisle, and tell me how many brands have only one formulation. How many have just chunky, just smooth, or even just extra thick. None. There is no perfect pasta sauce, and there should be no perfect Fury. You should choose the one that best fits you. AMD is taking this choice away, and it is stupid.




Let's be frank. You've demonstrated unflinching loyalty to a brand that has not earned it. You've demonstrated a fundamental lack of understanding of physics, so much that your understanding of anything must first be given as a matter of faith. You've dismissed accurate criticism, because in your own head there is one right answer, and you always have it. You've done all this completely unflinchingly, and without ever admitting to the vast gulf between your knowledge and abilities. I can't even begin to ascribe anything to you but an unfounded zealotry that could potentially make a suicide bomber slightly jealous.

Realizing that this is inflammatory, perhaps I should end here. You have demonstrated almost no reasonable opinions, and when asked where you come from you wave your hands in the air and call anyone asking questions an idiot. Please, peddle your brand of crap elsewhere. I can take Nvidiots. I can take die hard members of the Red team. I can even stand members of the Intel superior race. What I can't take is people who won't admit they are wrong, and those people who when confronted with their error fling crap at anyone who pointed it out. You are a member of the group of people I despise.
Posted on Reply
#80
HumanSmoke
Aquinus
There is an "11" footnote marker on the "Compute Units" section of the specs. I would really like to know that footnote actually says.
Here is the relevant footnote slide. Also, AMD's internal benchmarks vs the GTX 970 mITX card were once again done with no anisotrophic filtering for the most part, so numbers probably will differ from actual reviewers and real world usage scenario's.
Posted on Reply
#81
tabascosauz
lilhasselhoffer
--
I'm reliving the Skylake thread all over again. Time and time again, he has demonstrated that the comprehension of how the graphics market works, the knowledge of how graphics cards and CPUs work, and the ability to perceive a certain issue from another user's perspective are all far beyond what he is capable of grasping.

Of all the words I typed into that thread, all of them were wasted. It's akin to arguing with someone behind a soundproof wall. He doesn't understand anything that you say, but the bullshit that flows from him is endless.
Posted on Reply
#82
LightningJR
HumanSmoke
were once again done with no anisotrophic filtering
I am not sure how anisotrophic filtering affects anything other than image quality, the FPS change is <1.
Posted on Reply
#83
EarthDog
LightningJR
I am not sure how anisotrophic filtering affects anything other than image quality, the FPS change is <1.
That depends...
Posted on Reply
#84
newtekie1
Semi-Retired Folder
uuuaaaaaa
Maybe the pasting job was not so good, Hawaii runs hot, but that should not happen on the tri-x cooler...
Nah, it is just the fact that when you jam 2 of them in a closed case, even with 3 120mm high CFM exhaust fans and a 200mm intake on my 650D, the heat from two of the cards just builds up, there is no avoiding it. The top card just runs a good 10°C hotter than the bottom. Swapping the cards didn't change anything, whichever card was on top was running 10°C hotter than the bottom and would hit 95°C and start to throttle.
Posted on Reply
#85
uuuaaaaaa
newtekie1
Nah, it is just the fact that when you jam 2 of them in a closed case, even with 3 120mm high CFM exhaust fans and a 200mm intake on my 650D, the heat from two of the cards just builds up, there is no avoiding it. The top card just runs a good 10°C hotter than the bottom. Swapping the cards didn't change anything, whichever card was on top was running 10°C hotter than the bottom and would hit 95°C and start to throttle.
I guess AMD wisely chose "Hawaii" for a reason x)
Posted on Reply
#86
EarthDog
Perhaps focusing more on getting cooler air TO the cards as well as getting air out of the case would have helped. Those 200mm fans barely move any air (most of them).

Anyway, that has nothing to do with the thread title, so I digress. :)
Posted on Reply
#87
newtekie1
Semi-Retired Folder
EarthDog
Perhaps focusing more on getting cooler air TO the cards as well as getting air out of the case would have helped. Those 200mm fans barely move any air (most of them).

Anyway, that has nothing to do with the thread title, so I digress. :)
If airflow was a problem with my setup, no average user is going to be able to avoid throttling. The rear exhaust in the case is 75CFM, the two top exhausts are pushing 100CFM each minimum and kick up to 150CFM when things get hot.

The front 200mm is rated for 130CFM and runs full blast all the time. Yes, the 200mm fans seem like they don't push a lot of air, because you can't feel it, but they actually do move a lot of air, just at a lower pressure. But even still, the exhaust is pulling air in from the other vents in the computer, so the intake isn't as important. And there was plenty of heat coming out of the top of my case, it was insane.
Posted on Reply
#90
FordGT90Concept
"I go fast!1!11!1!"
Fury X = desktops
Nano = SFF

That's what I was afraid of and it is entirely possible. It puts me in a pickle. :( I might have to go Fury or 290X.
Posted on Reply
#93
HumanSmoke
tabascosauz
Yeah, looks like the GTX 970 then. No one in their right mind is buying that thing for $650 when even the already overpriced GTX 970 DC Mini sells for about $440-460.
I think the assumption is that the $649 is MSRP in $US.
Having said that, Gigabyte's GTX 970 mini is the same length as the Nano/Asus GTX970 DC2 at 170mm/6.7" , and is $300. ( the Asus card is presently $335 at the Egg). It would be interesting to see how many people would justify a ~30% increase in performance for a 216% increase in price - especially when AMD themselves are offering a better equipped card offering better performance at the same (supposed) price point.
Posted on Reply
#95
FordGT90Concept
"I go fast!1!11!1!"
Welp, that $649 price tag made me buy a 290X for $310. I have to repost HumanSmoke's picture because it is so true:

I went half AMD! :D
Posted on Reply
#96
john_
AMD is like a team that scores two own goals in the last five minutes of the match, while winning 1-0, out of fear that it might win.

I was expecting a price above $500. But the same price as 980Ti? YOU F#$%^&*ING MORONS IF EVGA (for example) COMES OUT WITH A 980TI ITX WHERE ARE YOU GOING TO PUT NANO? The fact that a 980Ti ITX would be a couple of inches longer, is not going to save Nano.

Nano will probably go only in ready systems, like Alienware's where you can hide the price, or sell it to Alienware for $550 and not $650. But in retail they will probably sell less cards than the number of posts in here.
Posted on Reply
#97
EarthDog
LightningJR
on?
The game, and resolution. It's true it isn't a huge penalty but to make a blanket statement like that makes it false as there are plenty of games/resolutions that show WAY more loss than 1 fps.
Posted on Reply
#98
LightningJR
EarthDog
The game, and resolution. It's true it isn't a huge penalty but to make a blanket statement like that makes it false as there are plenty of games/resolutions that show WAY more loss than 1 fps.
I have never experienced it, I give the blanket statement because afaik it's correct. If you've played PC games since 1998 and ALWAYS mess with graphics options this would be clear. I have had more FPS increase from closing background apps.

Plenty of games? WAY more loss than 1 fps? Would love to see these results.
Posted on Reply
#99
Aquinus
Resident Wat-man
LightningJR
I have never experienced it, I give the blanket statement because afaik it's correct. If you've played PC games since 1998 and ALWAYS mess with graphics options this would be clear. I have had more FPS increase from closing background apps.

Plenty of games? WAY more loss than 1 fps? Would love to see these results.
It wouldn't take too much to do a couple benchmark runs with and without. If I wasn't limited to just my laptop for the week, it would test it out myself.

Personally, I've seen AF reduce frame-rate by more than 1 FPS but, it's no where as heavy as AA is. From a technical standpoint, AF is highly dependent on memory bandwidth unlike AA which is more dependent on the speed and quantity of the ROPs.
Posted on Reply
#100
EarthDog
LightningJR
I have never experienced it, I give the blanket statement because afaik it's correct. If you've played PC games since 1998 and ALWAYS mess with graphics options this would be clear. I have had more FPS increase from closing background apps.

Plenty of games? WAY more loss than 1 fps? Would love to see these results.
more loss than 1 fps.. but as Aquinus said, it's not much. Sorry to be so pedantic about jt
Posted on Reply
Add your own comment