A Primer on AMD's Radeon R9 290 Series

After much waiting, the true new heir of AMD's next generation graphics cards is finally unveiled as the Radeon R9 290X and is now officially available. We give you a run-down of what you need to know of the new card and its "Hawaii" GPU core.

AMD's Radeon R9 290X is Finally Out

Following the Radeon R9 280/270 series that were made available since 8th October, today, the Radeon R9 290X is now officially available with true next-generation graphics core and the focus of this article.

Unfortunately, due to the delayed arrival of our test sample card from AMD, we will have to reserve performance comparisons for a future article. This article will squarely discuss on what one can expect from the new GPU core. Also, without the final test card in our hands, we’ve recycled the shots we’ve obtained from our previous preview article:-

Here she is in all its glory, the heir to the "Tahiti" based Radeon HD 7970, This graphics card will boast the "Hawaii" based R9 290X GPU and is just as large as any other top-tier discrete graphics card.

Here she is in all its glory, the heir to the "Tahiti" based Radeon HD 7970, This graphics card will boast the "Hawaii" based R9 290X GPU and is just as large as any other top-tier discrete graphics card.

This is another shot showing its top view. The card seems to be fully covered in a new cooler shroud that completely envelops the top, front and most of the bottom of the graphics card. Also noticeable is that the card is no thicker than two expansion card slots and there isn’t the usual CrossFire connectors. In fact, it doesn’t need physical connectors to enable CrossFire operation – just plug and play!

This is another shot showing its top view. The card seems to be fully covered in a new cooler shroud that completely envelops the top, front and most of the bottom of the graphics card. Also noticeable is that the card is no thicker than two expansion card slots and there isn’t the usual CrossFire connectors. In fact, it doesn’t need physical connectors to enable CrossFire operation – just plug and play!

Over at the rear side of the card, this reference unit is left bare and you can notice solder ball joints of the numerous memory chips that are mounted on the forward side of the PCB. We counted a total of 16 memory IC spots, which should total up to 4GB of graphics memory (GDDR5) screaming at 5000MHz on a wide 512-bit memory bus.

Over at the rear side of the card, this reference unit is left bare and you can notice solder ball joints of the numerous memory chips that are mounted on the forward side of the PCB. We counted a total of 16 memory IC spots, which should total up to 4GB of graphics memory (GDDR5) screaming at 5000MHz on a wide 512-bit memory bus.

Power input is handled by a pair of graphics power connectors that require a 6-pin and an 8-pin connection. This is no different from the current Radeon HD 7970.

Power input is handled by a pair of graphics power connectors that require a 6-pin and an 8-pin connection. This is no different from the current Radeon HD 7970.

Display output options are available in the form of dual dual-link DVI outputs, an HDMI and a full-sized DisplayPort.

Display output options are available in the form of dual dual-link DVI outputs, an HDMI and a full-sized DisplayPort.


The Build-up To "Hawaii"

Last month, AMD held their GPU14 Tech Day event to stress on their new gaming-oriented initiatives and the team’s motto of “AMD = Gaming”. The star of the show is of course AMD’s new Radeon R9 and R7 series of products, but alas, we figured out that most of the ‘new’ products are just rebadged Radeon HD 7000 parts.

When we asked AMD of the reasons behind the new naming scheme, they pin it squarely on their new marketing and initiatives for the GPU division as a whole and thus a new brand name to reflect this change. Primarily, AMD is working more with game developers and making waves in the gaming industry with their new Mantle API initiative. While that bodes well with AMD's directions, this will undoubtedly cause some confusion with end-users who expect a radical change when they purchase newly branded products and not variations of existing products.

While many of those R9 and R7 products are steadily populating the retail channels and booting up AMD’s visibility since 8th October, such as the string of Radeon R9 280X cards we’ve reviewed, the star of the line-up is none other than the focus of today’s article - the Radeon R9 290X (and the slightly slower R9 290 counterpart). The R9 290 series are the real torchbearers for the new GPU line-up as it sports a refreshed GPU core codenamed "Hawaii".

In the below table comparison, here’s how the new GPU stacks up against the competition and you’ll notice that the Radeon R9 290X is also available in a slightly tamer R9 290 version . The latter model will be available at a later date, but the R9 290X is now officially available.

Model
AMD Radeon R9 290X / 290
AMD Radeon R9 280X
AMD Radeon HD 7970 GHz Edition
NVIDIA GeForce GTX 780
NVIDIA GeForce GTX Titan
Core Code
Hawaii
Tahiti XT variant
Tahiti XT
GK110
GK110
Transistor Count
6.2 billion
4.3 billion
4.3 billion
7.1 billion
7.1 billion
Manufacturing Process
28nm
28nm
28nm
28nm
28nm
Core Clock
Up to 1000MHz / 947MHz
Up to 1000MHz
1050MHz
863MHz
836MHz
Compute Performance
5.6 / 4.9 TFLOPS
4.1 TFLOPS
3.79 TFLOPS
N.A.
N.A.
Stream Processors
2816 / 2560
2048
2048
2304
2688
Texture Mapping Units (TMUs)
176 / 160
128
128
192
224
Raster Operator units (ROP)
64 / 64
32
32
48
48
Onboard Memory
4GB GDDR5
3GB GDDR5
3GB GDDR5
3GB GDDR5
3GB GDDR5
Memory Clock
5000MHz
6000MHz
6000MHz
6008MHz
6008MHz
DDR Memory Bus
512-bit
384-bit
384-bit
384-bit
384-bit
Memory Bandwidth
320GB/s
288GB/s
288GB/s
288GB/s
288GB/s
PCI Express Interface
PCIe ver 3.0 x16
PCIe ver 3.0 x16
PCIe ver 3.0 x16
PCIe ver 3.0 x16
PCIe ver 3.0 x16
Molex Power Connectors
1 x 6-pin, 1 x 8-pin
1 x 6-pin, 1 x 6-pin
1 x 6-pin, 1 x 8-pin
1 x 6-pin, 1 x 8-pin
1 x 6-pin, 1 x 8-pin
Typical Board Power
~300W
250W
250W
250W
250W
Multi GPU Technology
XDMA AMD CrossFire
CrossFireX
CrossFireX
SLI
SLI
Display Connectivity
  • 2 x Dual-Link DVI
  • 1 x HDMI
  • 1 x DisplayPort
  • 2 x Dual-Link DVI
  • 1 x HDMI
  • 1 x DisplayPort
  • 1 x Dual-Link DVI
  • 1 x HDMI
  • 2 x mini-DisplayPort (mDP)
  • 2 x Dual-Link DVI
  • 1 x HDMI
  • 1 x DisplayPort (mDP)
  • 2 x Dual-Link DVI
  • 1 x HDMI
  • 1 x DisplayPort (mDP)
AMD TrueAudio Technology
Yes
No
No
N.A.
N.A.
API Support
DirectX 11.2, OpenGL 4.3, AMD Mantle
DirectX 11.1, OpenGL 4.3, PhysX, CUDA
Launch Price
US$549 OR US$579 (with BF4 bundle) / US$399
US$299
US$549 (Current street price: ~US$299)
US$649
US$999

We've a lot more about the Radeon R9 290X covered over the following few pages, so read on for the details:-

At the Core of Things - An Updated GCN Architecture

Last generation’s Southern Island GPUs have made significant changes to the graphics processing architecture that debuted as the Graphics Core Next (GCN) architecture. The new Hawaii-core based GPU that will be featured on the Radeon R9 290 series take it up a notch with further refinements and an updated core layout to account for even more firepower. The core layout, while revamped, is not a radical departure from the Radeon HD 7970 and it's first generation GCN architecture. This time round, you can notice that there are 4 main functional processing blocks called shader engine blocks that make up the Radeon R9 290X GPU. All other supporting functional units are encapsulate the shader engine blocks:-

This is the full functional block diagram of the Hawaii GPU found on the Radeon R9 290X GPU.

This is the full functional block diagram of the Hawaii GPU found on the Radeon R9 290X GPU.

Depending on future GPU models, these large shader engine blocks can be cut back accordingly to make up a smaller GPU die. Within each, shader engine block, you'll find one geometry engine. With four shader engine blocks, the R9 290X has four geometry processors - that's double what the previous generation Tahiti core can offer. The core processing unit in any of AMD’s modern GPU is a GCN compute units (CU). Within each shader engine block, it can support 1 to 11 compute units. In the case of the fully decked Radeon R9 290X, there are 11 CUs in each shader engine. With four such shader engines, the Radeon R9 290X offers a total number of 44 compute units, which is more than the 32 offered on the previous generation.

A closer look at each shader engine block where you can see it architecture supports up to 11 CUs per block. With the R9 290X, you've four such shader engines for a total of 44 CUs.

A closer look at each shader engine block where you can see it architecture supports up to 11 CUs per block. With the R9 290X, you've four such shader engines for a total of 44 CUs.

Each of these GCN CU units have largely remained identical to that of the GCN CU of the Southern Island GPUs (Radeon HD 7000 series), but has a few updates such as support for a Flat Addressing support that now allows the hardware to determine direct addressing, improved media processing instruction support – especially to the Maskable Quad Sum of Absolute Difference (MQSAD) that was introduced in the previous generation whose function is to allow background pixels to be ignored while helping isolate moving objects. So yes, on the whole each GCN CU still has quad 16-processing element vector units, which gives you 64 stream processors per GCU block (or per GCN CU).

The Radeon R9 and R7’s GPU compute unit (CU) largely remains similar the previous generation, but what differs is the number of compute units available per GPU. This also means the number of other supporting processing engines and blocks that co-work with the basic GCN compute unit also differs in each GPU configuration.

The Radeon R9 and R7’s GPU compute unit (CU) largely remains similar the previous generation, but what differs is the number of compute units available per GPU. This also means the number of other supporting processing engines and blocks that co-work with the basic GCN compute unit also differs in each GPU configuration.

Supporting the graphics processing blocks is other functions like the rasterizer, render back-ends, geometry processors, L2 cache (total of 1MB, up from 768KB) and the memory interface – all of which have been incrementally updated but the biggest change is the allocation of the number of units per shader engine block. We’ll detail this when we’ve obtained clearance, but two aspects that have been publicly acknowledged is the doubling of the render back-ends on the top tier R9 290 series (to cater to 4K resolution gaming) compared to the Radeon HD 7970 and the much higher density 512-bit memory interface used on the new R9 290X that consumes much less die space in a bandwidth per mm2 dies size used.

The new R9 and R7 series of GPUs are still manufactured based on the 28nm processing node, but given all the enhancements and increased number of processing units/blocks and other other aspects, the top-end R9 290 series carries over 6 billion transistors and naturally a larger die size compared to its predecessor. Despite that, AMD assured us that the R9 290 series uses 25% less die size compared to NVIDIA’s Titan and is more efficient per mm2 die size (though there’s no mention of how the performance will stack up). In place, we've some stats from AMD when comparing the Radeon R9 290X against the Radeon HD 7970 GHz Edition:-

 

Power to the People - A New PowerTune for 2013/2014

AMD’s PowerTune technology is the company’s version of the more popular GPU Boost used on NVIDIA’s products – even though AMD debuted this technology earlier. Essentially, AMD PowerTune that is featured on the previous generation Southern Island GPUs (Radeon HD 7000 series) analyzes the ‘active power signature’ of the card to utilize the unused thermal headroom. This is because most use-case scenarios hardly approach the graphics card’s TDP and technologies like AMD PowerTune help utilize balance power budget to push the core clock speeds and provide enthusiasts with increase performance.

The design goals of the new AMD PowerTune on AMD Radeon R9 and R7 graphics cards.

The design goals of the new AMD PowerTune on AMD Radeon R9 and R7 graphics cards.

After more than 1.5 years since the Radeon HD 7970 first debuted, we’re glad to know that the Radeon R9 290 parts will also feature a much more comprehensive PowerTune technology, more so because NVIDIA’s GeForce Boost 2.0 in its current generation of offerings has been available since earlier this year. As painted in AMD’s PowerTune manifesto, it aims to be the most advanced controller to-date as it will now not only factor active power consumption, but also factor in other attributes such as real-time temperature monitoring, voltage draw and even fan speed.

As shown in the block diagram above, a Digital Power Management (DPM) arbitrator checks on the card’s temperature, power consumption and voltage draw to determine how best to increase another attribute to maximize the potential of the hardware. Factoring temperature targets (default threshold set at 95 degrees Celsius) and fan noise are the newer aspects of the new PowerTune architecture to provide more control and optimization. For the longest time, we’ve been complaining about the AMD reference coolers being rather noisy when we’re in the thick of gaming. Fortunately, the ability to set fan speed to your preference helps one control the optimal acoustics of the card and also ensures there’s no drastic changes in noise levels as the card enters various stages of usage.

At the end of the day, overall performance of the Radeon R9 and R7 graphics cards are determined by the overall balance power budget available and due to the dynamic nature and the various parameters that are in control by PowerTune (and further user inputs), AMD is officially acknowledging that the R9 and R7 cards will no longer have a single advertised clock speed but they will be advertised as “Up to xxxxMHz”. This is again noticeable by the large specs table we’ve tabulated above that reflects this change of marketing.

The 4 pillars of the new AMD PowerTune that’s featured in the Radeon R9 290X.

The 4 pillars of the new AMD PowerTune that’s featured in the Radeon R9 290X.

Alas, all of this is playing catch-up as NVIDIA’s GPU Boost 2.0 and partner utilities have supported all of these in the GeForce 700 series of graphics cards from earlier this year. Nevertheless, it’s good to know AMD recognizes what needs to be done to appeal to the modern gamer. Performance numbers isn’t everything as the overall user experience is important too. Having said that, we notice that AMD hasn’t talked about any improvements to the cooler/shroud/fan used and we suspect that without the updated PowerTune, perhaps the reference cards will once again rear their ugly side of the previous generation Radeon cards. We’ll find out when and if we get hold of a reference-based Radeon R9 graphics card.

Designed for High Resolution Gaming (Ultra HD / 4K)

While AMD was first to support 4K resolution gaming, setting up such monitors wasn't a straightforward affair and might need Eyefinity setup to manually configure such monitors. With the new graphics card series bringing in more firepower, an improved Catalyst driver suite promises to support popular 4K resolution monitors out of the box without configuration. On that note, AMD has also proposed to the Video Electronics Standard Association (VESA) to update its standard to support for displays larger than 4K resolution, tiled display technologies and stereo 3D formats among others, which VESA accepted and updated their DisplayID standard to version 1.3.

In that sense, the new Radeon graphics cards are designed to be 4K-ready, but only the Radeon R9 290X has enough processing throughput to really deliver high performance gaming at high quality settings at such resolution. For standard non-gaming display output needs, both the old Radeon HD 7000 series and the newer Radeon R9/R7 models support 4K resolution output via HDMI at 30Hz and via DisplayPort at 60Hz (with or without the MST hub).

The new Radeon graphics cards are forward looking enough to be 4K-resolution ready.

The new Radeon graphics cards are forward looking enough to be 4K-resolution ready.

The new VESA Display ID v1.3 standard to address tiled displays in handling large screen resolutions and how the updated AMD Eyefinity configuration can fast track setup and optimal use of such displays.

The new VESA Display ID v1.3 standard to address tiled displays in handling large screen resolutions and how the updated AMD Eyefinity configuration can fast track setup and optimal use of such displays.

Most existing Ultra HD displays aren’t yet updated to the latest display connectivity standards (VESA Display ID v1.3 came out only a month ago) and support either Ultra HD resolutions at less than 30Hz or need to be configured as dual tiled displays of 2K x 2K resolution at 60Hz. While the former is fine for viewing movies, it’s not ideal for fast paced activity like gaming and thus requires the increased refresh rate. There will soon be newer displays that can deliver Ultra HD resolutions at 60Hz in a single stream and when they are made available, AMD mentions that is Radeon R9 290 series will be able to support it and drive high pixel rates of up to 600MHz which is required for Ultra HD resolutions with higher refresh rates.

In terms of display connectivity, you might notice from the table of specs that the new graphics card models adopt a single normal sized DisplayPort connector in favor of the previous generation’s dual mini-DisplayPorts. While it is one port less, most people don’t really require that many outputs and might in fact be more content with standard sized connectors that won’t need a converter before being able to hook up to a monitor. For those who demand more display support, the DisplayPort still supports multiple monitors via an MST (Multi-Stream Transport) Hub which the DisplayPort supports. From a GPU perspective, the new Radeon series still has 6 display controllers integrated on the die and thus a single card can handle 6 displays with the help of the MST Hub.

Not factoring DisplayPort options, the new range for Radeon R9 and R7 products are able to run 3x DVI/HMDI connections simultaneously. Previously on the Radeon HD 7000 series you could only run any two combinations of DVI/HDMI connections. This enablement is however not a GPU enhancement, but more of a board-level enhancement because the GPUs have had adequate display controllers since the earlier generations.

What about 3D gaming? Like its predecessor, 120Hz 3D gaming monitors are supported via graphics card's DisplayPort connection.

 

AMD TrueAudio Technology

AMD was the first to combine audio stream output via HDMI many years back, but this time around, they've taken a radical step forward to incorporate a fully programmable audio engine as part of the GPU die. The idea is to move away from simplistic audio reproduction with minimal effects and have real-time positional audio rendered, just like how games have tremendously improved visually over time when moving from fixed function programming to programmable shaders. Among the new R9 and R7 series of graphics cards, the AMD R9 290X and the R7 260X are the only GPUs to have this feature.

We've got a full article dedicated to AMD TrueAudio, our brief experience and the interviews from AMD engineers, so check it out!

 

Project Mantle

We've got a detailed news piece on this development, but to summarize it here, AMD is releasing a new API (codenamed Mantle) to directly tap into the hardware capabilities of AMD's GCN-based GPUs and unleash the full performance potential of the GPU in an efficient manner. While it may seem to alienate Intel and NVIDIA, the core reason for this initiative came about to facilitate game developers to concentrate on improving what matters most to gamers instead of time spent on porting and optimizing games on different platforms. This is all the more feasible because AMD now produces the processing brains behind all of the next generation consoles.

This initiative is supported on all GCN-architecture based graphics cards, so the brand new Radeon R9 290X and the older Radeon HD 7000 series will benefit from Mantle-powered games. We'll have to assess its true merits once games like Battefield come out with a new update to support the Mantle API.

AMD's slide neatly sums up the benefits of the Mantle API initiative. In the words of AMD, this has the potential to deliver performance of 2018, today thanks to the vastly increased draw calls per second that the light API enables.

AMD's slide neatly sums up the benefits of the Mantle API initiative. In the words of AMD, this has the potential to deliver performance of 2018, today thanks to the vastly increased draw calls per second that the light API enables.

Our articles may contain affiliate links. If you buy through these links, we may earn a small commission.

Share this article