ASUS Radeon RX 470 STRIX OC 4 GB Review 126

ASUS Radeon RX 470 STRIX OC 4 GB Review

(126 Comments) »

Introduction

ASUS Logo


Even as NVIDIA has a somewhat top-down approach to the market with this generation by launching its top-performing "Pascal" GPUs in quick succession, including the "big chip" with the recent TITAN X Pascal, AMD's approach appears to be the opposite - bottom-up. The company debuted its "Polaris" architecture with GPUs that target the most voluminous market segment, $100-300, which AMD says to constitute the vast majority of discrete GPU sales.

Following up on the June launch of its $199-239 Radeon RX 480, AMD is launching the second, cheaper GPU based on the "Polaris" architecture, the Radeon RX 470. This SKU's price is expected to be below $200, although some custom-design cards, such as the ASUS Radeon RX 470 STRIX 4 GB we're reviewing today, could go slightly over it. This could confuse buyers because the 4 GB variant of the faster RX 480 is priced at $199 and gives you more performance.

The Radeon RX 470 is based on the same Polaris 10 "Ellesmere" silicon as the RX 480, featuring fewer shaders and lower clock speeds. The chip features 32 out of the 36 Graphics CoreNext (GCN) compute units (CUs) physically present on the chip, which makes for 2,048 stream processors. The TMU count is proportionately lowered to 128. The memory bus is untouched at 256-bit GDDR5, and the ROP count stays at 32. Reference clock speeds are set at 928 MHz core and 1206 MHz boost, and the memory runs at 6.6 Gbps, which works out to a bandwidth of 211 GB/s. ASUS overclocked its Radeon RX 470 STRIX graphics card to feature a boost frequency of 1270 MHz and left the memory untouched.



In this review, we have the ASUS Radeon RX 470 STRIX with us, which features the coveted Republic of Gamers STRIX branding and the company's all new DirectCU II cooling solution with an aluminum fin-stack heatsink that relies on two copper heat pipes that make direct contact with the GPU to draw heat away and to the fin stack. This heatsink is ventilated by a pair of 100 mm spinners, which don't spin at all when the GPU is idling. You also get an ROG emblem that glows in any of the 16.7 million colors across the RGB palette and a unique GPU-synced 4-pin PWM case-fan header feature ASUS is pioneering with this generation.

Radeon RX 470 Market Segment Analysis
 GeForce
GTX 960
Radeon
R9 380
Radeon
RX 470
ASUS RX
470 STRIX OC
Radeon
R9 380X
Radeon
R9 390
GeForce
GTX 970
Radeon
RX 480
Radeon
R9 390X
GeForce
GTX 980
GeForce
GTX 1060
Shader Units10241792204820482048256016642304281620481280
ROPs3232323232645632646448
Graphics ProcessorGM206TongaEllesmereEllesmereTongaHawaiiGM204EllesmereHawaiiGM204GP106
Transistors2940Munknown5700M5700Munknown6200M5200M5700M6200M5200M4400M
Memory Size2 GB2 GB4 GB4 GB4 GB8 GB4 GB4 GB / 8 GB8 GB4 GB6 GB
Memory Bus Width128 bit256 bit256 bit256 bit256 bit512 bit256 bit256 bit512 bit256 bit192 bit
Core Clock1127 MHz+970 MHz1206 MHz1270 MHz970 MHz1000 MHz1051 MHz+1266 MHz1050 MHz1126 MHz+1506 MHz+
Memory Clock1753 MHz1375 MHz1650 MHz1650 MHz1425 MHz1500 MHz1750 MHz2000 MHz1500 MHz1750 MHz2002 MHz
Price$170$165$180$210$210$260$265$200 / $240$310$360$250 / $300

Architecture

At the heart of the Radeon RX 470 is the new "Ellesmere" (Polaris 10) GPU built on the 14 nanometer silicon fab process by Samsung and GlobalFoundries. The wafers are made in Upstate New York, USA, and are then bumped and packaged at a facility in Taiwan to be sent to the various graphics card manufacturers located there and across the straits.



This GPU is based on AMD's fourth generation Graphics CoreNext architecture codenamed "Polaris." According to AMD, Compute Units (CUs) based on Polaris are 15% more efficient at number crunching than CUs based on the preceding Graphics CoreNext 1.2 architecture (R9 Fury, R9 380X). Pay attention to the numbers here. While the number-crunching machinery is 15% more efficient, the chip is claimed to have a 2.5x leap in overall energy-efficiency over the previous generation. This is because AMD is cashing in on the immediate gains a new silicon fab process brings to the table, the 14 nm FinFET process in this case, to increase transistor counts and clock speeds.



The component hierarchy in the Polaris 10 "Ellesmere" silicon is similar to older-generation chips, although each of the components received major updates. We begin with the chip featuring two hardware schedulers and the introduction of dedicated real-time asynchronous compute with spatial and temporal scheduling. The chip also features four async compute engines (ACEs). AMD optimized the async compute engine with new quick-response queue tech.



There's a design focus on stepping up geometry processing performance and blunting the brute-tessellation advantage NVIDIA traditionally enjoyed over AMD. For a chip of this segment, Polaris 10 features four independent geometry processors. Their functionality is upgraded over the previous generation, featuring a primitive discard accelerator which culls (discards) triangles in the pipeline with zero area or no inclusive sample points. The geometry engine now features a tiny cache called the Index Cache, which cushions small instanced geometry and reduces data movement to improve primitive throughput during instancing.

The Polaris 10 silicon features 36 Compute Units (CUs). 32 of these are enabled on the RX 470, spread across four shader engines, each with a dedicated geometry processor, a raster engine, and two render backends. The four shader engines are supported by a large 2 MB L2 cache, which acts as the town-square for the GPU's various key components.

Most of the architecture-specific innovations are centered on the CU, which now features a hardware instruction prefetcher, a larger instruction buffer, and native half-precision (FP16/Int16) support, which should reliably crunch numbers for gaming applications with significantly reduced memory and register footprints while lowering power execution. Altogether, the "Polaris" CU is claimed to have up to 15% higher performance than CUs based on the GCN 1.1 architecture (R9 390X). Each CU features 64 stream processors, which has the 32 CUs amount to 2,048 stream processors. In summary, the Polaris 10 chip features 2,048 stream processors, 128 TMUs, and 32 ROPs.



The Radeon RX 470 features a 256-bit wide GDDR5 memory interface, holding 4 GB or 8 GB of memory, clocked at 6.6 Gbps. The actual memory bandwidth of this interface at its given clock speeds is rated at up to 211 GB/s, although its effective bandwidth could be higher thanks to an updated lossless delta color compression (DCC) tech with full 2/4/8:1 compression ratios. AMD claims that its new gen DCC tech can provide an effective bandwidth uplift of a staggering 30 percent. The ASUS Radeon RX 470 STRIX we're reviewing today features 4 GB of memory.



The multimedia accelerators receive a major update, now supporting H.265 Main10 decode hardware acceleration and 4K60 HEVC encode hardware acceleration. The other components with big updates are the display controllers, which now support DisplayPort 1.4 (DP 1.3 HBR3 and DP 1.4 HDR) and HDMI 2.0b. FreeSync is supported over both DP and HDMI. Resolutions as high as 5K60, 10-bit 4K96 HDR, and 4K120 SDR are supported.

Packaging

Package Front
Package Back




You will receive:
  • Graphics card
  • Driver CD + documentation
  • ASUS zip ties
  • ASUS stickers
  • World of Warships coupon

The Card

Graphics Card Front
Graphics Card Back

ASUS has engineered a completely new cooler for their RX 470. It is dominated by plastic and has no colored highlights. Some stickers are included in the package, so you can put them on if you prefer some color instead of the all-black theme. Dimensions of the card are 24.5 cm x 13.0 cm.

Graphics Card Height

Installation requires two slots in your system.

Monitor Outputs, Display Connectors

Display connectivity options include two DVI ports, one HDMI port, and one DisplayPort.

The HDMI port is now version 2.0b, and DisplayPort has been updated to 1.3 HBR3/1.4 HDR ready, which enables support for 4K @ 120 Hz and 5K @ 60 Hz, or 8K @ 60 Hz with two cables. GPU accelerated encoding is now supported for H.264 at up to 4K30, and HEVC is supported at up to 4K60. Accelerated decoding is supported for HEVC at up to 4K60 Main-10, VP9 is supported at up to 4K, and H.264 works at up to 4K120.

Multi-GPU Area

AMD CrossFire has been running over the PCI-Express bus for a few generations now. The Polaris Series is no different.

Graphics Card Teardown PCB Front
Graphics Card Teardown PCB Back

Pictured above are the front and back, showing the disassembled board. High-res versions are also available (front, back).
Our Patreon Silver Supporters can read articles in single-page format.
Discuss(126 Comments)
Mar 28th, 2025 01:53 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts