ASUS Radeon R9-280X DirectCU II TOP review - A new Island and Technology

It's them Islands Again !

So let's deal with a quick overview of the AMD new product range first. Over the past two years you have been hearing about several codenames and that can be a little confusing. It's simple really, in the market we have entry level, mainstream and high-end products. When you notice 'Cape Verde' that's entry level, the 7700 series that was just released. Pitcairn would be the codename the mainstream products (7800) will hide under, and Tahiti is the codename for the GPUs used on AMD's most high-end graphics cards. 

• Entry level = Cape Verde = Radeon HD 7700 series / R5 series
• Mainstream level = Pitcairn = Radeon HD 7800 series / R7 series
• High-end level = Tahiti = Radeon HD 7900 series / R9 series
• Enthusiast level Hawaii = R9-290 and 290X

The entire segment from top to bottom has been released in Q1 2012, but most of the 2013 models are respins, the exception being the R7-260X and the R9-290/290X graphics cards.

R7 240

• Stream Processors 320
• Clock Frequency up-to 780 GHz
• 499 GFLOPS compute performance
• 1 or 2 GB memory at 4.6 Gbps / 128-bit
• 30W TDP
• PCI-E 3.0
• API - DirectX 11.2 / OpenGl 4.3 / Mantle

R7 250

• Stream Processors 384
• Clock Frequency up-to 1.05 GHz
• 806 GFLOPS compute performance
• 1 or 2 GB memory at 4.6 Gbps
• 65W TDP
• PCI-E 3.0
• API - DirectX 11.2 / OpenGl 4.3 / Mantle

R7 260X

• Stream Processors 896
• Clock Frequency up-to 1.1 GHz
• 1.97 TFLOPS compute performance
• 2 GB memory at 6.5 Gbps / 128-bit
• 115W TDP
• PCI-E 3.0
• API - DirectX 11.2 / OpenGl 4.3 / Mantle

R9 270X (previously the R7870 Ghz / Pitcairn)

• Stream Processors 1280
• Clock Frequency up-to 1.05 GHz
• 2.69 TFLOPS compute performance
• 2 or 4 GB memory at 5.6 Gbps
• 180W TDP
• PCI-E 3.0
• API - DirectX 11.2 / OpenGl 4.3 / Mantle

Respin much ?

Yes, much like NVIDIA has done with the 700 series AMD is doing something very similar. The R7 260X is a new GPU, but the R9 270X and R9-280X both are derivatives of the Pitcairn and Tahiti GPUs respectively. And who can remember Oland, that mobile GPU ? Well peek at the R7 250 specs, looks familiar right ? There are a number of changes though to be found in clock and memory frequencies so overall there will be slight performance increment's. And much like NVIDIA does, AMD now fully implemented a 'Turbo' feature much like NVIDIA's boost technology on all of their products.
 

This means that there no longer is a fixed clock on these cards. The GPU clock will get a baseline and maximum frequency, in-between these two values the card will clock up/down based on power draw/limiters, performance and heat. In return the card will adaptive manage fan control, clock frequency and voltages (dynamic). AMS always has been much more reserved with the difference inbewteen baseline and maximum frequency, so the difference in values won't be extremely big.

28nm Technology

The GPU architecture has remained the same though comparable towards the last-generation products, AMD still uses the 28nm process technology, the cards are PCIe gen 3 compatible and there have been significant changes on power consumption. We'll address all the features separately of course. With the launch of the R7 and R9 series you will also see Eyefinity updated towards version 2.0 DDM, audio is now fully supported (you hear audio on the actual monitor it's played off), the 5x1 landscape mode is introduced, and you may now create custom multi-monitor resolutions.
 

Interesting is that AMD now implemented next generation digital multipoint audio with smart channel splitting. So if you setup Eyefinity with three screens the six speakers in your monitors will now be configured as front left/mid/right channels. Very clever.
 

Eyefinity also has received a small update. You can now mix and match any monitor (similar monitors) output and create an Eyefinity setup. So if you have three of the same monitors you could hook up one towards HDMI, once to DVI and once to DP. That is the kind of flexibility we like a lot alright.

Um wait .. so the R280X is a Tahiti / R7970 GHz then ?

Yes, we start off with the GPU overview (we'll discuss architecture later). AMD bakes GPUs on the all new 28nm node, in very simple wording the smaller the die size the more more transistors AMD can put on a smaller area, often at a lower voltage. Honestly, we think the R9-280 cards will appeal the most to you guys so we'll place the most focus on these cards in this article, these are Tahiti XT based. The Tahiti core has a nice 4.3 Billion transistors, 4,312,711,873 to be precise. As a result the GPU is packed with 2048 shader processors harbored in Compute Units segments (32 of them). Memory volume wise we see 3 Gigabyte of DDR5 memory. This is not done for bragging rights, but AMD simply takes Eyefinity and multiple monitor usage seriously, it is there where the extra memory makes sense. The memory bus is 384-bit.

• Stream Processors 2048
• Clock Frequency up-to 1 GHz
• 4.1 TFLOPS compute performance
• 3 GB memory at 5.6 Gbps / 384-bit
• 250W TDP
• PCI-E 3.0
• API - DirectX 11.2 / OpenGl 4.3 / Mantle

The reference R9-280X is clocked with an up-to 1 GHz clock frequency, with room to tweak towards a nice 1100 MHz or even higher. That 384-bit memory is clocked at impressive speeds as well, 5.6 Gbps that's bandwidth. Summed up, this product should be a notch faster then a R7970 (reference).