Specs and Hardware

What’s 6 inches long and costs $650? The AMD Radeon R9 Nano!

The AMD Radeon Nano graphics card is unlike any product we have ever tested at PC Perspective. As I wrote and described to the best of my ability (without hardware in my hands) late last month, AMD is targeting a totally unique and different classification of hardware with this release. As a result, there is quite a bit of confusion, criticism, and concern about the Nano, and, to be upfront, not all of it is unwarranted.

After spending the past week with an R9 Nano here in the office, I am prepared to say this immediately: for users matching specific criteria, there is no other option that comes close to what AMD is putting on the table today. That specific demographic though is going to be pretty narrow, a fact that won’t necessarily hurt AMD simply due to the obvious production limitations of the Fiji and HBM architectures.

At $650, the R9 Nano comes with a flagship cost but it does so knowing full well that it will not compete in terms of raw performance against the likes of the GTX 980 Ti or AMD’s own Radeon R9 Fury X. However, much like Intel has done with the Ultrabook and ULV platforms, AMD is attempting to carve out a new market that is looking for dense, modest power GPUs in small form factors. Whether or not they have succeeded is what I am looking to determine today. Ride along with me as we journey on the roller coaster of a release that is the AMD Radeon R9 Nano.

AMD Radeon R9 Nano Specifications

I don’t want to spend a whole lot of time on this part of the review if only because we had another story posted just a couple of weeks ago where AMD released most of the technical information about the product. A quick recount of the information is provided here but for the full marketing detail about the card, check out my previous piece.

  R9 Nano R9 Fury R9 Fury X GTX 980 Ti TITAN X GTX 980 R9 290X
GPU Fiji XT Fiji Pro Fiji XT GM200 GM200 GM204 Hawaii XT
GPU Cores 4096 3584 4096 2816 3072 2048 2816
Rated Clock up to 1000 MHz 1000 MHz 1050 MHz 1000 MHz 1000 MHz 1126 MHz 1000 MHz
Texture Units 256 224 256 176 192 128 176
ROP Units 64 64 64 96 96 64 64
Memory 4GB 4GB 4GB 6GB 12GB 4GB 4GB
Memory Clock 500 MHz 500 MHz 500 MHz 7000 MHz 7000 MHz 7000 MHz 5000 MHz
Memory Interface 4096-bit (HBM) 4096-bit (HBM) 4096-bit (HBM) 384-bit 384-bit 256-bit 512-bit
Memory Bandwidth 512 GB/s 512 GB/s 512 GB/s 336 GB/s 336 GB/s 224 GB/s 320 GB/s
TDP 175 watts 275 watts 275 watts 250 watts 250 watts 165 watts 290 watts
Peak Compute 8.19 TFLOPS 7.20 TFLOPS 8.60 TFLOPS 5.63 TFLOPS 6.14 TFLOPS 4.61 TFLOPS 5.63 TFLOPS
Transistor Count 8.9B 8.9B 8.9B 8.0B 8.0B 5.2B 6.2B
Process Tech 28nm 28nm 28nm 28nm 28nm 28nm 28nm
MSRP (current) $649 $549 $649 $649 $999 $499 $329

AMD wasn’t fooling around, the Radeon R9 Nano graphics card does indeed include a full implementation of the Fiji GPU and HBM, including 4096 stream processors, 256 texture units and 64 ROPs. The GPU core clock is rated “up to” 1.0 GHz, nearly the same as the Fury X (1050 MHz), and the only difference that I can see in the specifications on paper is that the Nano is rated at 8.19 TFLOPS of theoretical compute performance while the Fury X is rated at 8.60 TFLOPS.

The memory system is also identical between the R9 Nano and the Fury X: 4096-bit wide high bandwidth memory bus, 4GB of capacity, 500 MHz memory clock rate and up to 512 GB/s of available memory bandwidth. Again, very impressive!

There is one very big difference that we have to point out though: The R9 Nano is rated with a 175 watt TDP, the Fury X at 275 watts. That is a difference in power consumption that we just haven’t seen in any other card to card variance when based on the same GPU.

That difference in power consumption is possible due to AMD “underclocking” the Fiji GPU: bringing the clock speed lower to meet a specific TDP target. And as we’ll see in an upcoming page of this story, the clock speed variance from game to game and even between resolutions is fairly substantial; I saw clocks as high as 1000 MHz and as low as 825 MHz.

It's important to note that even though AMD says the GPU is not thermally constrained, all aspects of the ASIC performance come into play during use. Efficiency is defined as the ability to run at a certain performance level and clock speed within a set TDP (which AMD has already defined at 175 watts). So if a particular game plays with a heavy GPU workload and attempts to draw more power than the 175 watts allowed, AMD's Fiji implementation will downclock until it arrives a voltage that hits ~175 watts draw. That will be different for all different kinds of software so its something we will pay particular attention to in our review.

For users that want to tweak things you will be able to adjust the power limit in the AMD Catalyst Control Center, thus decreasing efficiency, but pushing you closer to that 1.0 GHz frequency regardless of the GPU workload. This obviously takes AMD's new R9 Nano outside the range that it wanted to be able to claim for this product, and what it could guarantee in tight quarters that might be more thermally constrained, but gamers will have that flexibility if they wish, gaining as much as 10% in graphics performance based on AMD's quotes. Once you hit that 1000 MHz mark, however, you revert to the more standard overclocking models that we have already seen which aren't spectacular with Fiji GPUs.

The AMD Radeon R9 Nano – A Sexy Beast

From a design standpoint there is very little that was unknown about the R9 Nano hardware. This is a 6-in PCB design with a vapor chamber cooler that helps keep that sizeable Fiji GPU and HBM memory in the 75-85C range during gaming. Once again, our previous story has all the details on AMD’s position of the R9 Nano and its size, claiming that it can be used in cases that would otherwise not have access to cards in this performance class. Still, this is dead-sexy hardware so let’s take a look around our sample.

It’s hard to really even quantify with a single image, but the R9 Nano is incredibly tiny, measuring just 6-inches long. The fit and finish on the design is top notch, nearly matching the build quality of the Fury X in terms of the finish and rubber-coated plastic body. To be clear – the Nano is entirely plastic, none of that nickel plated gunmetal color to be found here. The single fan in the middle is responsible for moving around the necessary air to keep the Fiji GPU cool – this card is quiet but far from silent.

You won’t find any fancy LED lights on the card, which is a bit of a letdown considering how apt AMD is pushing this card for custom mods and windowed chassis. The Nano does support CrossFire but it’s using XDMA so no CrossFire connectors are needed.

There is no back plate on the R9 Nano which I am sure will aggravate some users immediately. I assume that this is not a cost saving measure and instead is to ensure that the card will fit onto some motherboards with very limited spacing between the primary PCIe slot and the CPU socket.

For power delivery the R9 Nano only requires a single 8-pin PCIe power connection, facing out the back of the card. This should help the Nano fit in more tightly-spaced cases than with the power connection up top, but I am sure there are cases where the opposite is true as well.

Display connectivity on the R9 Nano consists of a set of three DisplayPort connections and a single HDMI port. There are no DVI connections – kind of a bummer. Also, the HDMI port here is still HDMI 1.4a, not HDMI 2.0, which is one of the biggest drawbacks to Fiji over NVIDIA’s Maxwell architecture. This is to an even greater degree in the case of the R9 Nano as it fits perfectly into the stereotype of an HTPC card, with previously unseen graphics performance, though it won’t be able to output to 4K 60 Hz TVs without  an active adapter. An adapter, mind you, that I have yet to touch or see or price or validate.

« PreviousNext »