Scan's TekSpek

Our Aim
To provide you with an overview on New And existing technologies, hopefully helping you understand the changes in the technology. Together with the overviews we hope to bring topical issues to light from a series of independent reviewers saving you the time And hassle of fact finding over the web.

We will over time provide you with quality content which you can browse and subscribe to at your leisure.

TekSpek GPU - Graphics
AMD R9 300 Series

AMD R9 300 Series


Date issued:

AMD's new graphics card products for 2015 are arriving in two different forms - a new R7 and R9 300-series family and a trio of graphics cards with an unconventional naming strategy called R9 Fury, Fury X and Nano that make use of a new graphics processing chip, codename Fiji. The new R7 and R9 300-series graphics cards will cover the majority of price points and performance segments for AMD. The range includes the R7 360, R7 370, R9 380, R9 390 and R9 390X which are all graphics cards that share many similarities with currently offered graphics cards in the R7 and R9 200-series family.

The new AMD Radeon R9 range

 

R9 390X

R9 290X

R9 390

R9 290

R9 380

R9 285

Launch date
Jun 2015
Oct 2013
Jun 2015
Nov 2013
Jun 2015
Sep 2014
GCN version
1.1
1.1
1.1
1.1
1.2
1.2
DX support
12
12
12
12
12
12
Process (nm)
28
28
28
28
28
28
Transistors (mn)
6,200
6,200
6,200
6,200
5,000
5,000
Approx Die Size (mm²)
438
438
438
438
359
359
Full implementation of die
Yes
Yes
No
No
No
No
Processors
2,816
2,816
2,560
2,560
1,792
1,792
Texture Units
176
176
160
160
112
112
ROP Units
64
64
64
64
32
32
Peak GPU Clock/Boost (MHz)
1,050
1,000
1,000
947
970
918
Peak GFLOPS (SP)
5,914
5,632
5,120
4,849
3,476
3,290
Peak GFLOPS (DP)
739
704
640
606
435
411
Memory Clock (MHz)
6,000
5,000
6,000
5,000
5,700
5,500
Memory Bus (bits)
512
512
512
512
256
256
Max bandwidth (GB/s)
384
320
384
320
182.4
176
Default memory size (MB)
8,192
4,096
8,192
4,096
4,096
2,048
Power Connectors
8+6-pin
8+6-pin
8+6-pin
8+6-pin
6+6-pin
6+6-pin
TDP (watts)
275
290
275
275
190
190
GFLOPS per watt
21.50
19.42
18.61
17.63
18.30
17.32
Current price (Newegg)
$429
$329
$349
$280
$249
$199
Frame-rate control*
Yes
No
Yes
No
Yes
No

AMD Radeon R9 390X

Starting off at the higher-end and it will become quite obvious, from referring to the specification table, that the new R9 390X is based on a slightly modified variation of the R9 290X. AMD has decided to retain the fundamental structure of the predecessor, a "rebrand" of sorts, since the R9 290X is still a capable 1440p and 4K gaming graphics card.

Yet the R9 390X isn't a carbon copy of the R9 290X since it ships, by default, with twice as much video memory - 8GB of GDDR5 compared to 4GB. Further still, AMD has elevated clock speeds on the graphics engine and on the graphics memory, with the increased memory frequencies facilitated by the widespread availability of faster SK hynix GDDR5 memory.

The increased frequencies equate to a five per cent bump on the graphics engine and a 20 per cent increase on the graphics memory yet AMD claims a reduced TDP of 275-watts compared to 290-watts on its predecessor. Such a reduction has been enabled by some low-level microcode enhancements and general maturity in the manufacturing procedures that produce the GPUs used in the R9 390X.

AMD directly targets the Nvidia GTX 980 with its R9 390X and with a price point of $429, the R9 390X does hold an advantage against Nvidia's GTX 980 which has a $499 MSRP. The R9 390X has the credentials to tempt PC users with ageing graphics into an upgrade, though R9 290X or R9 290 owners are unlikely to benefit from what is an incremental upgrade.

Radeon R9 390

AMD has followed a very similar strategy with the R9 390 by doubling up the video memory capacity and increasing frequencies across the GPU engine, by 5.6 per cent, and memory by 20 per cent. The R9 390 doesn't benefit from a TDP reduction but maintains the same 275-watt envelope while being able to offer that extra speed - a characteristic that still indicates increased efficiency of the underlying silicon used.

AMD positions the more affordable R9 390 against the Nvidia GTX 970, rather than the GTX 980, which makes sense given that the $349 MSRP of the R9 390 aligns closely to the $329 MSRP of the GTX 970. Given the incremental speed improvements on the R9 390 one might assume that it will benchmark closely with the previous generation R9 290X.

Radeon R9 380

AMD continues the rebranding exercise with the R9 380 which is based on the previous generation 'Tonga' R9 285. Such a move signals the end for the Tahiti silicon that was used in the HD 7950, HD 7970, HD 7990, R9 280 and R9 280X. In a similar fashion to the R9 390(X) AMD has doubled the video memory on offer, from 2GB to 4GB, and increased frequencies on the graphics engine, by 5.7 per cent, and on the graphics memory by 3.6 per cent.

Unlike with the R9 390(X) AMD is granting graphics-card vendors some flexibility in selling both older 2GB and newer 4GB variants of the R9 380. It should be noted that the 2GB models can ship with a lower memory frequency of 5,500MHz. With an MSRP of $249 the R9 380 is a little expensive to compete with Nvidia's $199 GTX 960 so instead sits somewhere between the GTX 960 and GTX 970, despite having performance more similar to the former.

The Radeon R7 range

 

R7 370

R7 265

R7 360

R7 260

Launch date
Jun 2015
Mar 2014
Jun 2015
Oct 2013
GCN version
1.0
1.0
1.1
1.1
DX support
12
12
12
12
Process (nm)
28
28
28
28
Transistors (mn)
2,800
2,800
2,080
2,080
Approx Die Size (mm²)
212
212
160
160
Full implementation of die
No
No
No
No
Processors
1,024
1,024
768
768
Texture Units
64
64
48
48
ROP Units
32
32
16
16
Peak GPU Clock/Boost (MHz)
975
925
1,050
1,000
Peak GFLOPS (SP)
1,997
1,894
1,613
1,536
Memory Clock (MHz)
5,600
5,600
6,500
6,000
Memory Bus (bits)
256
256
128
128
Max bandwidth (GB/s)
179.2
179.2
104
96
Default memory size (MB)
2,048/4,096
2,048
2,048
1,024
Power Connectors
6-pin
6-pin
6-pin
6-pin
TDP (watts)
110
150
100
95
GFLOPS per watt
18.15
12.63
16.13
16.17
Current price (Newegg)
?
$149
?
$109
Frame-rate control*
Yes
No
Yes
No

Radeon R7 370

Interestingly with the R7 370 AMD up-rates a previous generation graphics card, the R7 265, to a higher position in the new generation. The R7 370 is based on the earliest GCN 1.0 revision that formed the HD 7850 which in itself isn't a full implementation of the Curacao GPU that can sport 1,280 shaders. Due to the age of GCN revision the R7 370 is the only R7 or R9 300-series graphics card to lack support for AMD FreeSync technology.

In keeping with the themes of AMD's 300-series rebranding the R7 370 gets a small GPU engine frequency increase, 925MHz to 970MHz, though the memory is left at default speeds. Like with the R9 380 AMD's graphics card vendors will be permitted to sell the R7 370 in either 2GB or 4GB variations to meet different price points.

Radeon R7 360

The R7 360, which is based on the previous generation R7 260, goes through a similar process of frequency increases and memory doubling. AMD doubles the video memory from 1GB to 2GB while boosting its speed from 6,000MHz to 6,500MHz and the graphics engine from 1,000MHz to 1,050MHz. The R7 360 is set to be a touch slower than the previous generation R7 260X which had a full implementation of the 896-shader Bonaire die but should still deliver decent 1080p gaming performance.

Software Tweaks

AMD is fully aware of the fact that launching an entire new series based on rebranded parts is a potentially risky strategy given its typical target audience of enthusiasts and gamers looking to get the most value for money and performance. In an effort to add more value to the R7 and R9 300-series products AMD will be delivering a series-specific software enhancement.

The newest feature, dubbed frame-rate targeting control, allows the driver to enable and maintain a specific frame-rate target. The main benefit of using such a strategy is that the graphics card can be prevented from generating unnecessary frames that result in excess heat, power consumption and noise. The roll-out of frame-rate targeting control is limited to the 300-series despite the fact there is no genuine limitation as to why 200-series cannot access the same feature. Presumably, AMD wants to create the impression that the 300-series brings something genuinely new to the table for existing 200-series owners.

AMD has also rolled out its downsampling technology, called Virtual Super Resolution (VSR), to the new 300-series graphics cards. VSR is a competing equivalent to Nvidia's Dynamic Super Resolution (DSR) which performs an equivalent downsampling operation. Users with GPU horsepower to spare will be able to render the game in a higher resolution, downsample that image to fit their screen and achieve additional visual clarity.

Rebrand and tweak

In summary the new AMD 300-series graphics cards are best on existing technology that has been tweaked and enhanced to offer consumers more performance. AMD implements this through a number of strategies - more video memory, higher frequencies and microcode enhancements. As the transition between the 200- and 300-series begins many prospective buyers may find stock being cleared from the 200-series makes for a better deal.

Enthusiasts may be slightly underwhelmed by AMD's 300-series and the slowing pace of GPU development but, importantly, AMD is set to bring out a trio of new graphics cards shortly. The R9 Fury, R9 Fury X and R9 Nano are all in the immediate pipeline for consumers who find the R9 390X simply doesn't offer enough performance for their needs. AMD's 300-series graphics cards cater to a wide range of price points, performance segments, and graphics cards vendors have been busy releasing many new custom designs with cutting-edge cooling solutions and PCB implementations.