What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

The AMD R9 Nano Preview


HardwareCanuck Review Editor
Staff member
Feb 26, 2007
During AMD’s E3 press conference there was a whole lineup of new products shown. While some like the Fury and Fury X were completely new and others were essentially updates of existing architectures, the R9 Nano was arguably the most talked-about product there. It promised to offer an insane performance per watt ratio alongside an enviably small form factor, providing the perfect demonstration of AMD’s HBM initiative and where it could lead in the future. Unfortunately, details about the Nano were kept behind closed doors…..until now.

AMD promised the R9 Nano in August and that’s exactly what we are getting but maybe not in the way some of you may have been thinking. Today is considered the official launch date but actual press reviews and availability have been pushed back a bit to September 10th. To a certain extent, we have to respect AMD’s stance on this since they are trying (in their words) to build “critical mass” for a broad scale launch a few weeks from now. In plain English that means cards will only be available for purchase in a few weeks.


There has been a lot of speculation about how what AMD would sacrifice to get their Fiji core into a SKU with a 175W TDP and now we have an answer: surprisingly, not all that much. The R9 Nano (this card doesn’t carry the Fury designation) boasts a full-enabled Fiji XT core with 4096 Stream Processors, 64 ROPs and 256 texture units. It also comes equipped with 4GB of HBM that runs at 500MHz with a total bandwidth of 512 GB/s.

The only real difference between this card and the Fury X are the core frequencies which are rated to “up to 1000MHz”. At first glance this may give some hope that the R9 Nano will reach the performance levels of AMD’s flagship but that isn’t quite the case. We are told that the core will likely hit 900MHz to 950MHz in actual gaming scenarios and there are some situations that may push clock speeds even lower. Now this shouldn’t be too much of a problem considering the substantial core count on tap but the lower clock speeds may push real-world framerates down into R9 Fury territory or below depending on the situation. We do however have to emphasize: for a card with a TDP of just 175W this isn’t bad at all.

One of the largest points of consternation in this particular launch will be the R9 Nano’s price. At $649 it aligns perfectly with the LCS-equipped R9 Fury X without offering equal performance. Some may roll their eyes at this but I happen to think it is right in-line with expectations since this card offers space-saving convenience alongside performance that should handily beat an NVIDIA GTX 980. It is however almost double what the next most powerful mini-ITX GPU -the $339 ASUS GTX 970 Mini- goes for these days.


AMD’s focus with the R9 Nano is to offer high end 4K performance in a compact form factor. While most mini ITX chassis do support longer graphics cards, this GPU doesn’t require builders to sacrifice GPU throughput for size. Whether or not that will resonate with today’s gamers remains to be seen but I personally think AMD’s marketing of this card will hit the right chords among their intended market. Many buyers are looking to downsize their systems’ footprints and the Nano should allow them to do exactly that, but without throwing away performance.


So how has AMD reduced TDP to this extent without sacrificing cores or clock speeds? We’re not all that sure but there’s a high likelihood that some serious core binning is taking place here. What this means for long-term availability is anyone’s guess but we’re willing to bet the Nano will be in short supply for the foreseeable future. To actually find Fiji XT and 4GB HBM combinations that require under 175W of power can’t be easy. In addition, we hear that overclocking will be strictly limited which also points towards stringent TDP constraints.


While power input is handled by a single 8-pin connector, the real head scratcher for many reading this introduction will be how AMD managed to include a fully-enabled Fiji XT core without an integrated water cooler. As you may remember, the Fury X required a high performance closed loop liquid cooler to bring its temperatures to reasonable levels and yet here is a card that is supposed to offer comparable specs in an air cooled sub-7” form factor. Confused yet?

According to AMD, much of the Nano’s relative strengths lie within its heatsink design. While the main fin array looks like many other cards we have reviewed in the past, there’s some substantial technological feats going on below the surface.


Alongside high static pressure but low RPM 90mm fan, the Nano incorporates a pure copper core contact plate which interfaces directly with a dual vapor chamber. This vapor chamber makes direct contact with the aforementioned aluminum fin array to insure quick transfer of heat and efficient cooling.

While we have seen this approach in the past, AMD’s temperature claims are impressive to say the least. Supposedly, this card will have a nominal operating temperature of just 75°C while still operating at just 42 dbA. That may seem like an impossibility but there’s nothing we can see to invalidate this claim.


Past the advanced heatsink design, AMD is also addressing the cooling for their VRM components since there really isn’t all that space to optimize airflow. In this case they have added a secondary copper heatpipe and heatsink which are supposed to remove any worry about components overheating.


Under the heatsink is a relatively anemic setup with a 4-phase GPU PWM along with a single phase for secondary functions. AMD isn’t expecting this card to be an “overclocker’s dream”; that title has always been reserved for the Fury-series parts.

This isn’t to say that the R9 Nano won’t overclock though. While we have yet to test this, supposedly there is some headroom available….just not as much as the R9 Fury X or Fury. That certainly isn’t saying all that much and given the Nano’s abnormally low TDP we highly doubt records will be set. However, for its intended niche, overclcokig shouldn’t be too much of a concern.


The I/O outputs on all of AMD’s newer products have caused a bit of controversy. There is the usual trio of DisplayPort 1.2 outputs and single HDMI 1.4.

Since these cards are based off of a revised and upscaled version of AMD’s Hawaii architecture, they don’t have baseline support for HDMI 2.0. This means their compatibility with 4K UHD TV’s (many of which don’t have DisplayPort inputs) will be limited to sub-4K resolutions.

Luckily, some board partners are coming out with active DisplayPort to HDMI 2.0 adapters but these potentially costly additions still won’t have HDCP 2.2 compatibility so you can forget about playing protected 4K content through this card. Will this hurt sales? Likely not since streaming 4K content from the internet isn’t tied to an antiquated HDCP standard.


Past all of the cool pictures, AMD provided us with some convenient numbers as well. According to them, the R9 Nano doesn’t really have any competition other than the GTX 970 Mini. This is partially true considering mini-ITX class graphics cards are as rare as hen’s teeth but it doesn’t take into account that many of today’s ITX chassis like the Silverstone Raven RVZ-02 and Corsair’s AIR 204 fully support 11” and longer GPUs without sacrificing footprint size. We would assume that the Nano will bring about a whole new case category which is compatible with mini ITX motherboards and sub-7", dual slot GPUs but we’re talking about a niche within a niche now. Indeed, like other cards before it we expect to see this card in some downright tiny systems.

Regardless of our suppositions, it looks like the R9 Nano will achieve performance that’s around 30% higher than the GTX 970 Mini. Now this is in 4K, a resolution the GTX 970 was never meant to compete in, so we have no idea how this card will line up in an apples to apples comparison. However if we extrapolate this upwards and recount that the GTX 980 is roughly 16% faster than the GTX 970, it’s safe to assume AMD’s ultra compact card will overcome the GTX 980 by about 15%. If that figure sounds familiar, it is roughly where the $549 R9 Fury lies right now.

There are a number of questions surrounding this "launch" that will likely be answered once we actually get our hands on the R9 Nano and the weeks following availability when we can gauge the market's reaction. Will such a high cost of entry entice or turn off customers? From where I'm standing right now with limited knowledge of real-world performance, the $649 price point doesn't necessarily represent value on the cost / performance front but it does demonstrate the amount of binning that must go into finding Fiji XT cores that fit the Nano's aggressive 175W TDP profile.

Will the binning negatively affect availability in the long run? That really depends on how many of these cores can actually meet that 175W figure given their continued use of a mature but old 28nm manufacturing process. The most likely possibility is that HBM's continued yield challenges will play more of a factor in AMD's rollouts than anything else. We're already seeing this in the Nano's true launch being pushed back to September 10th.

The R9 Nano certainly looks like an enticing product that will act as a perfect showcase for AMD's new HBM-based dies. While true performance remains to be seen, things on this front are looking very good considering the impressive specs. Naturally, noise will be a concern as well but supposedly AMD has that well in hand too.

This is where we will leave this quick preview since talking about anything else would be nothing but educated assumptions. Join us back here in a few weeks for a full review where we will hopefully be able to put a number of the still-unanswered questions to rest.
Last edited:

Latest posts