When we first heard about the new Radeon R9 285, our first impression was pessimistic. The new Tonga GPU sports specifications that are nearly identical compared to the Radeon R9 280, but with a slightly lower GPU clock and a sizable memory bandwidth deficit thanks to its thinner 256-bit memory interface. While the Radeon R9 285's onboard RAM runs at a higher 1375 MHz clock, the net result is 176 GB/s of memory bandwidth - significantly less than the 240 GB/s memory bandwidth of the R9 280. In addition, the R9 285 has 2 GB of RAM, whereas the R9 280 has 3 GB.
| Products |
|
|
|
| Pricing |
|
|
|
| GPU (Fab. Process) | Tahiti (28 nm) | Tonga (28 nm) | Tahiti (28 nm) |
| Shader Cores | 1792 | 1792 | 2048 |
| Texture Units | 112 | 112 | 128 |
| ROPs | 32 | 32 | 32 |
| Core Clock | Up to 933 MHz | Up to 918 MHz | Up to 1000 MHz |
| Memory Clock | 1250 MHz | 1375 MHz | 1500 MHz |
| Memory Bus | 384-bit | 256-bit | 384-bit |
| Memory Bandwidth | 240 GB/s | 176 GB/s | 288 GB/s |
| Onboard Graphics RAM | 3 GB | 2 GB | 3 GB |
| Max TDP | 250 W | 190 W | 250 W |
| Power Connectors | 2x 6-pin | 2x 6-pin | 2x 6-pin, 1x 8-pin |
The tale of the tape is not kind to the Radeon R9 285. Yes, the new card's power usage has gone down significantly compared to its predecessor, but from a gaming perspective, that would be small consolation in exchange for lower performance. If raw specifications were all that mattered, and if the Tonga graphics processor was merely a re-spin of the Tahiti GPU in the Radeon R9 280, we wouldn't have a lot of nice things to say about the new Radeon R9 285.
After testing it, though, we were pleasantly surprised to find that the Radeon R9 285 has a few tricks up its sleeve thanks to some re-engineering of the Graphics Core Next (GCN) architecture.
Despite what its specifications may suggest, Tonga is not a spin on the Tahiti GPU in the Radeon R9 280 and 280X. Rather, it is a new and condensed version of the Hawaii GPU in the Radeon R9 290 and 290X. Among other things this means it has four times the number of asynchronous compute engines, that's eight instead of the Radeon R9 280/280X's two. According to AMD this can improve tessellation performance from two to four times, and facilitates effects that rely on GPU compute. In addition, the Radeon R9 285 inherits the 290 series' quad-shader layout, allowing four primitives to be rendered per clock cycle instead of two. Also note the CrossFire XDMA block, which provides the possibility of multi-card operation without a bridge connector.

Tonga features four shader engines, each carrying seven compute units (CUs). Just like previous GCN-based GPUs, every CU is host to 64 shaders and four texture units, adding up to a total of 1792 shaders and 112 texture units in the Radeon R9 285. These numbers are equal to the cut-down Tahiti chip in the Radeon R9 280, but the arrangement of resources is different.
Speaking of cut-down, we're told that Tonga is slightly handicapped for use in the 285, and that the uncut GPU has the potential to utilize eight compute units per shader engine for a total of 2048 shaders and 128 Texture units. If this sounds familiar, it's because this is the same number of resources available in the Radeon R9 280X. Perhaps AMD has some bigger plans for the Tonga GPU in the future.
Regardless, AMD did scale back some parts of the new GPU. Each of the four shader engines carries two render back-ends - instead of four as per the Radeon R9 290 series. Each of these is capable of rendering four full-color pixels per clock, for a total of 32 pixels per clock cycle in total. This is half of what Hawaii can process but equals the Tahiti GPU in the Radeon 280 and 280X.
Improvements are always welcome but with the memory interface cut by a third compared to the Radeon R9 280, AMD needed to accomplish some magic to compensate for that 27% drop in available bandwidth. The company's solution was to enable the GPU to read and write frame buffer color data in a lossless compressed format, a technique that it claims can deliver 40% higher memory bandwidth efficiency. We're somewhat skeptical that this will completely compensate for the Radeon R9 285's lower available bandwidth compared to the Radeon R9 280, but we'll see what the benchmarks have to say about it.
That's not all that's been improved over the Hawaii GPU in the Radeon 290 series, though. Tonga boasts other new features such as the ability to process instructions in parallel between SIMD lanes, improved compute task scheduling algorithms, and even new 16-bit floating point and integer instructions for compute and media processing tasks.
Of course, Tonga has inherited Hawaii's fixed-function hardware, too, such as TrueAudio (AMD's new audio processor) and project FreeSync (the open-source answer to Nvidia's G-Sync) support. In addition, AMD claims to have revamped the unified video decoder (UVD) and video coding engine (VCE) with improvements specific to the Tonga GPU. The Radeon R9 285's UVD now supports H.264 playback at high frame rates on 4K displays, and the VCE has been improved with faster performance in addition to 4K resolution support. AMD claims a 31%-47% transcoding advantage over the GeForce GTX 760, but we'd prefer to run our own tests on that so expect some in-depth coverage from us in the future.
For the first time, AMD has added the ability to specify the maximum fan speed in the Overdrive overclocking utility. This means that users who prioritize noise can set the maximum fan RPMs, and the driver will automatically adjust clock rates to fit power and heat into the user-specified envelope. This feature will be available in the 14.8 release driver, but only to models with second- and third-iterations of GCN processors: the Radeon R7 260, 260X, R9 285, R9 290, and R9 290X. This functionality didn't seem to work in our pre-release beta driver, unfortunately.
Those are the updated details we have to share from an architectural perspective, now let's get down to the nitty gritty: the test hardware, specifically the Radeon R9 285 card.
AMD has not provided a reference design for the Radeon R9 285, leaving the manufacturers to put their own spin on the card. Our test sample is from Asus, marketed under their performance-oriented Strix line, and we'll be using it to provide game benchmark results.
| Asus Strix Radeon R9 285 | |
|---|---|
| Dimensions | 276 mm (L) x 135 mm (H) x 40 mm (D) |
| Weight | 859 g |
| Form Factor | Dual Slot |
| PCIe | 2x 6-Pin |
| Connectors | 1x DVI-I (Dual Link, + analog) 1x DVI-D (Dual Link) 1x HDMI 1x DisplayPort |
Asus' Radeon R9 285 sports the company's trademark black highlighted with a dark red tribal art design. The PCB is 9" by 4.6", but the cooler shroud extends to 10.5", giving the impression of a larger, more serious piece of performance hardware. At 1 lb 14 oz, it doesn't feel as light as we'd expect it to from the size of the PCB, probably thanks to the metal backplate. Despite the plastic cooler shroud, the card feels very sturdy and well put together.
Asus' card has a 954 MHz target GPU clock rate, 36 MHz faster than AMD's reference specification. The 2 GB of onboard graphics RAM running at a 1375 MHz GDDR5 memory clock is in line with AMD's guideline, though. We may see 4 GB models in the future, but these will likely be a rare exception rather than the rule.

The aluminum heatsink features three large 10mm heatpipes and is cooled by two 95 mm low-profile fans. The card's 190 Watt TDP is satisfied by two six-pin PCIe power connectors. ASUS cuts a relief out of the PCB so that the power connector clips face the opposite direction of the heatsink, which makes for much easier removal.
Both connectors feature an LED that glows red when a power cord isn't attached, and blue when it is connected properly. These details might not sound important but it really leaves a positive impression. Note the lack of Crossfire connector on the card, as the Radeon R9 285 does not need a bridge to operate in tandem.
Asus' new card comes with a DVI-I, DVI-D, full-sized DisplayPort, and full sized HDMI output.
|
|
|
|
|
Another board partner graphics card will serve as the basis for our power consumption and temperature measurements going forward. The Gigabyte R9 285 WindForce OC comes overclocked to 973 MHz from the factory.
| Gigabyte R9 285 Windforce OC | |
|---|---|
| Dimensions | 250 mm (L) x 120 mm (H) x 38 mm (D) |
| Weight | 609 g |
| Form Factor | Dual Slot |
| PCIe | 2x 6-Pin |
| Connectors | 1x DVI-I (Dual Link, + analog) 1x DVI-D (Dual Link) 1x HDMI 1x DisplayPort |
The relatively light graphics card uses the new WindForce cooler with only two fans, which have a larger diameter of 96 mm and fan openings with a diameter of 103 mm.
The cooler and board also represent an interesting new design. The voltage converters are now located near the back of the card, as opposed to the area close to the PCIe power connectors. The aluminum cooler has three parts and uses two 8 mm heat pipes that run through its center to transport waste heat to the two side sections.
We see that the power converters have their own separate aluminum cooler and aren’t connected to the bottom of the main cooler. The airflow is supposed to go through the fins of the main cooler and reach all the way down to the other cooler. We’ll see later if, and how well, this might work.
The top of the Gigabyte R9 285 WindForce OC is dominated by the two 6-pin PCIe power connectors, which have been turned to face toward the board. This perspective also provides a good overview of the cooler’s design.
The end of the graphics card affords a good view of one of the two large heat pipes made of composite material. The orientation of the cooler’s fins shows the direction of the airflow from the board to the top of the card.
The connectors are standard fare. Details can be found in the table above.
|
|
|
|
|
Power Consumption Measurement Methodology
Tom’s Hardware Germany’s test system for the exact measurement of graphics cards’, CPUs’, and other components’ power consumption was developed in cooperation with HAMEG (Rohde & Schwarz). It was designed for particularly precise measurements with very small time intervals and has a temporal resolution of up to 1 ms.
Only sophisticated technology is able to handle the challenges presented by AMD’s PowerTune and Nvidia’s Boost technologies. These generate changes in core voltage in time frames of under 10 ms, which results in very large and quick voltage fluctuations. Let’s take a look at what can happen in the space of only one single millisecond, using measurement intervals of 10 μs.
This is why we’re evaluating all measured currents and voltages with a 500 MHz four-channel oscilloscope with a data logger, the HAMEG HMO3054, and extremely fast current probes. This setup also allows for unified data storage and remote control.
The measurements provided by the three high-resolution DC current probes, all HAMEG HZO50s, are taken via a riser card for the 3.3 V and 12 V rails, which we constructed specifically for this purpose. It supports PCIe 3.0 and uses short signal paths. The remaining probes are connected to the PCIe power cable that we modified for this.
We measure the voltages directly at their respective rails. We’re now working with a temporal resolution of 1 ms, since this allows us to record and evaluate the fluctuations caused by AMD’s PowerTune and Nvidia’s Boost technologies with confidence.
We’ve limited the measurement duration to 1 minute due to the very high volume of data when we measure all channels. We only shorten the measurement intervals all the way to the minimum that the physical capabilities of our setup allows for our more detailed measurements.
| Test Methodology | No Contact Current Measurement at All Rails Direct Voltage Measurement IR Real-Time Monitoring |
|---|---|
| Test Equipment | 1 x HAMEG HMO3054, 500 MHz Four-Channel Oscilloscope with Data Logger 4 x HAMEG HZO50 Current Probe 4 x HAMEG HZ355 (10:1 Probe, 500 MHz) 1 x HAMEG HMC8012 DSO with Data Logger 1 x Optris PI450 80 Hz Infrared Camera + PI Connect |
| Test System | Intel Core i7-5960X MSI X99 Gaming 7 16GB G.Skill Ripjaws DDR4 2666 (4 x 4GB) Samsung 850 EVO 512 GB Raijintek Water Cooling be quiet! Dark Power Pro 1200W Microcool Banchetto 101 |
Benchmarking Hardware And Software
The Radeon R9 285 was tested with the 14.8 beta launch driver, but all other AMD cards were outfitted with the Catalyst 14.7 RC1 for testing. The GeForce cards used the newest option, which at the time of testing was the 340.52 WHQL driver.
We selected a variety of newer game titles with high detail settings at a resolution of 1920x1080 in order to give the Radeon R9 285 and its competitors a solid, real-world workload that this class of card should be able to handle.
The Asus Radeon R9 285's core clock was dropped to the 918 MHz reference specification in order to show what a typical Radeon R9 285 should be able to accomplish. Keep in mind that there is no reference cooler for this card, so all Radeon R9 285 will be unique in this respect.
Some readers will note that the Radeon R9 280 results are those we collected from our Sapphire Dual-X Radeon R9 280 review. Sapphire's card comes with a 940 MHz core clock, a mere 7 MHz over the reference specification. We took a few benchmarks with the reference clock but quickly realized that the results are within the margin of error, so we're using the Sapphire numbers to represent reference Radeon R9 280 results. Keep in mind that there is no reference Radeon R9 280 cooler, either.
Two of the games we're testing have an option to use a Mantle code path, so we're running those benchmarks (Thief and Battlefield 4) with Mantle enabled and disabled to measure the API's impact.
Graphics cards like the Radeon R9 280 require a substantial amount of power, so XFX sent us its PRO850W 80 PLUS Bronze-certified power supply. This modular PSU employs a single +12 V rail rated for 70 A. XFX claims continuous (not peak) output of up to 850 W at 50 degrees Celsius.

We've almost exclusively eliminated mechanical disks in the lab, preferring solid-state storage for alleviating I/O-related bottlenecks. Samsung sent all of our labs 256 GB 840 Pros, so we standardize on these exceptional SSDs.
| Test System | |||||
|---|---|---|---|---|---|
| CPU | Intel Core i7-3960X (Sandy Bridge-E), 3.3 GHz, Six Cores, LGA 2011, 15 MB Shared L3 Cache, Hyper-Threading enabled. | ||||
| Motherboard | ASRock X79 Extreme9 (LGA 2011) Chipset: Intel X79 Express | ||||
| Networking | On-Board Gigabit LAN controller | ||||
| Memory | Corsair Vengeance LP PC3-16000, 4 x 4 GB, 1600 MT/s, CL 8-8-8-24-2T | ||||
| Graphics | Asus Strix Radeon R9 285 954 MHz GPU, 2 GB GDDR5 at 1375 MHz (5500 MT/s) (underclocked GPU to reference 918 MHz specification for benchmarks) AMD Radeon R9 280X 850/100 MHz GPU, 3 GB GDDR5 at 1500 MHz (6000 MT/s) Sapphire Dual-X R9 280 OC 850/940 MHz GPU, 3 GB GDDR5 at 1250 MHz (5000 MT/s) AMD Radeon R9 270X 1050 MHz GPU, 2 GB GDDR5 at 1400 MHz (5600 MT/s) Nvidia GeForce GTX 660 980/1033 MHz GPU, 2 GB GDDR5 at 1502 MHz (5008 MT/s) Nvidia GeForce GTX 760 980/1033 MHz GPU, 2 GB GDDR5 at 1502 MHz (5008 MT/s) Nvidia GeForce GTX 770 1046/1085 MHz GPU, 2 GB GDDR5 at 1752 MHz (7008 MT/s) | ||||
| SSD | Samsung 840 Pro, 256 GB SSD, SATA 6Gb/s | ||||
| Power | XFX PRO850W, ATX12V, EPS12V | ||||
| Software and Drivers | |||||
| Operating System | Microsoft Windows 8 Pro x64 | ||||
| DirectX | DirectX 11 | ||||
| Graphics Drivers | Radeon R9 285: AMD Catalyst 14.8 beta All other Radeon cards: AMD Catalyst14.7 RC 1 All GeForce Cards: Nvidia 340.52 WHQL | ||||
| Benchmarks | |||||
|---|---|---|---|---|---|
| Watch Dogs | Version 1.04.497, Custom THG Benchmark, 90-sec FRAPS, Driving | ||||
| Arma 3 | V. 1.26.126.789, 30-sec. Fraps "Infantry Showcase" | ||||
| Battlefield 4 | Version 1.3.2.3825, Custom THG Benchmark, 90-Sec | ||||
| Assassin's Creed IV: Black Flag | Custom THG Benchmark, 40-Sec | ||||
| Thief | Version 1.6.0.0, Built-in Benchmark | ||||
| Titanfall | Version 1.0.5.7, Demeter Map, Custom THG Benchmark | ||||
| Grid Autosport | Version 1.0.101.4672, Built-In benchmark | ||||
| Far Cry 3 | Version 1.05, Custom THG Benchmark, 55-sec FRAPS | ||||
|
|
|
|
|
We'll begin our analysis with a couple of synthetic benchmarks, starting with 3DMark Firestrike:

We don't believe that the newest iteration of 3DMark is an ideal measure for comparing GPU architectures, but it does give us a useful glimpse into the Radeon product stack. Despite the slightly lower core clock and memory bandwidth compared to the Radeon R9 280, we see the Radeon R9 285 perform faster than its predecessor and land right between the Radeon R9 280 and 280X. Indeed, the improvements that AMD has made to the Tonga GPU have a substantial impact here.
Now let's see if Tonga's strengths will manifest in a GPGPU processing scenario. We took Sandra 2014 SP3's GPGPU cryptography benchmark for a spin:

The Radeon R9 285's lossless color compression can't help it here, and the new card's performance reflects its slightly lower core clock rate when it comes to GPGPU processing. Having said that, performance per watt have increased as we'll see in the power test results later on. Now let's move on to the game performance benchmarks.
|
|
|
|
|
We'll begin our game benchmarks with Titanfall. We aren't able to test this game with FRAPS, as the only way to get consistent results is to use a private match that requires EA Origin's overlay, which is incompatible with this frame-rate measuring tool. Nevertheless, we found a particularly demanding part of the Demeter map and use it to provide a rudimentary performance measure:

Titanfall's Source engine appears to work well with the Tonga GPU's special sauce, allowing it to improve slightly over Radeon R9 280 performance and approach the GeForce GTX 770.
Next up, Battlefield 4. We're able to include detailed benchmarks for this game using both DirectX and AMD's Mantle graphics API. We know that Mantle only plays nice with 3GB of onboard memory or more, so let's see how the 2GB Radeon R9 285 handles it:




When using the DirectX code path the Radeon R9 285 is slightly bested by the Radeon R9 280, but with Mantle enabled the 2GB cards suffer a performance penalty while the 3GB options show a slight improvement. Indeed, when it comes to Battlefield 4, Mantle is a memory-hungry beast.
Let's now consider frame time variance. Battlefield 4 is a well-coded engine that doesn't show any weaknesses in this respect. Having said that, there are other sources of lag. While these results don't reveal it, the Mantle code-path definitely demonstrates notable skips in monitor output, especially on the Radeon R9 270X and 285.
|
|
|
|
|
Thief doesn't suffer from Mantle's high memory requirement in Battlefield 4, showcasing gains with AMD's API across all Radeon cards with a single exception: the Radeon R9 285. AMD's new card actually shows a performance deficit with Mantle enabled. It looks like the new Tonga driver needs some work to accommodate the new product.




As for frame time variance, there is definitely some latency with this engine. It is mostly within acceptable limits with the exception of the GeForce GTX 660, a card that demonstrates significant spikes above 15 milliseconds.




Next up is ARMA 3, a game that slightly favors the Radeon R9 285's architectural improvements and color compression over the Radeon R9 280's brute force approach. It is interesting to see the frame time latency results, with the GeForce cards suffering consistently higher average and 75th percentile numbers, while the Radeon R9 285 and 270X have the highest 95th percentile spikes.
|
|
|
|
|




Grid Autosport is a new addition to our benchmark suite, and the Radeon R9 285 virtually ties the Radeon R9 280 in this game engine. There are small frame time latency spikes of 5 milliseconds consistently throughout the benchmark, but it doesn't manifest in any skippy performance we could notice in this well-coded title.




Assassin's Creed IV delivers the big surprise in our benchmark suite, with the Radeon R9 285 surpassing the high-end Radeon R9 280X. We took this benchmark multiple times to make sure what we we're seeing isn't a glitch. Assassin's Creed IV is known to make heavy use of tessellation, and we can conclude that the extra geometry units in the new Tonga GPU are earning their pay when it comes to this visually demanding title. As for frame time variance, there's very little to report with exemplary low lag across all the cards we tested.
|
|
|
|
|




Now for Watch Dogs, and another sterling example of the Radeon R9 285 keeping up with the Radeon R9 280. There's not much else to say here except there are some significant frame time variance spikes across all cards, and this game isn't particularly smooth when travelling the highway at high speeds into new areas that probably cause a lot of map loading on-the-fly.




Finally, let's consider Far Cry 3. We suspect this game engine doesn't give the Radeon R9 285 a lot of room to make use of lossless color compression, as the Radeon R9 280 bests it here. That might make sense considering the detailed gradients on the lush foliage. Regardless, the Radeon R9 285 doesn't do badly here, either, tying with the GeForce GTX 760. As for frame time variance, this game displays spikes at various points in the game regardless of graphics hardware.
|
|
|
|
|
Power Consumption: Idle (2D Desktop)
Unfortunately, AMD doesn’t fare very well here. We’ve repeated these measurements, reviewed running processes, and even switched systems altogether. We just couldn’t get the Gigabyte Windforce Radeon R9 285 to draw less than 15 W at idle.

Looking at just one minute’s worth of power draw curve after smoothing it over a bit, it’s plain to see that there are significant fluctuations even at almost 0 percent load.
The situation at the motherboard slot looks perfect. We never exceeded or even reached 75 Watts.
Let’s take a look at how that 15 W idle draw is distributed:
| Minimum | Maximum | Average | |
|---|---|---|---|
| PCIe 12V | 2 W | 21 W | 8 W |
| Motherboard 3.3V | 1 W | 3 W | 2 W |
| Motherboard 12V | 0 W | 31 W | 5 W |
| Graphics Card Total | 2 W | 45 W | 15 W |
List of All Individual Values per Supply Line
For those readers who like lots of detail, we’ve put together all of the idle power consumption values for each supply line in the gallery below:
Voltages
The voltage values are very important because they’re used to calculate power consumption by multiplying them with the amount of current. Looking at the 12 V rail, we see very clearly that the voltage fluctuates extremely if measured in small increments of time. The average voltage is 12 V, but the switched-mode power supply architecture and the different phases of capacitor load leave a very distinct mark. We’d like to tease our reference article about current graphics cards in relation to common power supply units here, which will cover how the two interact in detail. This is a much more interesting topic than it might appear to be at first glance.
We don’t know yet if the high idle power draw represents a problem in our specific board partner’s card or if it’s a problem inherent in the Tonga architecture. Comparing reviews proves difficult once again in light of the scarcity of actual homogenous reference cards. This is too bad, since we’d have liked the Radeon R9 285 to provide a stronger showing in this arena.
|
|
|
|
|
Power Consumption: Gaming Loop
For the gaming lop measurements we firstwarm up the graphics card for 20 minutes until a stable GPU temperature of 64 degrees Celsius has been reached. At this point we proceed to measure the card’s power consumption again. Our gaming loop is relatively challenging for the GPU, so we’re confident that the average 176 W we measured is a good representation of today’s demanding game engines. less demanding titles should draw between 165 to 170 Watts depending on the particular title, of course.
Once again, lets focus on just one minute of the smoothed curve. This shows us that the way the load is distributed between the PCIe and the motherboard power connectors conforms to the applicable norms.
The control exercised at the motherboard slot is reassuring. There’s never a sustained load that exceeds the slot’s maximum 75 Watt rating.
Let’s take a look at how the 176 W of power consumption while gaming are split in the table.
| Minimum | Maximum | Average | |
|---|---|---|---|
| PCIe 12V | 33 W | 209 W | 124 W |
| Motherboard 3.3V | 3 W | 7 W | 5 W |
| Motherboard 12V | 9 W | 93 W | 47 W |
| Graphics Card Total | 47 W | 290 W | 176 W |
List of All Individual Values per Supply Line
Again, we’ve put together all power consumption values for each supply line in a handy gallery.
Voltages
The average voltage is exactly 12 V, just like it was at idle. However, the fluctuations are present and accounted for as well.
The Gigabyte R9 285 WindForce OC and its 176 W for gaming comes in almost 40 W lower than a moderately overclocked AMD Radeon R9 280 reference graphics card, which also provides slightly lower performance on average. When we referenced the Gigabyte R9 285 WindForce OC’s efficiency on the graphics cards in our 2014 VGA charts, it looks like it’s comparable to Nvidia’s older Kepler-based GeForce GTX 760. If winning the efficiency war was the goal, then it most certainly hasn’t been reached, but at least AMD was able to catch up. This probably won’t be enough to compete with Maxwell, though.
|
|
|
|
|
Power Consumption: GPGPU Stress Test
We'll take one step further and push the Gigabyte R9 285 WindForce OC as far as it will go. The resulting 191 W (without changing the power target) are right around where AMD’s TDP said the card would max out. This also shows that the manufacturer’s specifications can generally be trusted, especially since increasing the clock frequency doesn’t change the power draw as long as the other settings are kept the same.
And now lets focus on a perfect minute’s worth of gently smoothed curve.

The measurements at the motherboard slot confirm, once again, that all of the power values we saw fell well within the acceptable range.
Let’s take a look at the individual measurements again in our table.
| Minimum | Maximum | Average | |
|---|---|---|---|
| PCIe 12V | 57 W | 175 W | 140 W |
| Motherboard 3.3V | 3 W | 6 W | 5 W |
| Motherboard 12V | 19 W | 83 W | 47 W |
| Graphics Card Total | 87 W | 250 W | 191 W |
List of All Individual Values per Supply Line
Once again, the following gallery shows all power consumption values for each supply line.
Voltages
The average voltage is 11.9 V. Those fluctuations we saw are making an appearance yet again.
The 191 W torture test result is almost the exact TDP as stated by the manufacturer. In light of the higher computing power requirements of OpenCL and DirectCompute, the Gigabyte R9 285 WindForce OC’s performance yield is higher than that of a comparable Kepler-based GeForce graphics card in places. Still, we’re talking about catching up to the competition from an efficiency perspective, not surpassing it. Tonga will probably have trouble going up against Maxwell.
|
|
|
|
|
Temperatures during Continuous Operation
Lower power consumption should result in lower GPU temperatures under load, which, in turn, should lead to a simpler and quieter cooling solution. It sounds good in theory, but let’s take a look at how the Gigabyte R9 285 WindForce OC actually handles heat.
Infrared Measurements with the Optris PI450
Interestingly, we’ve identified a method to confirm what our sensors tell us in the form of the PI450 by Optris. This piece of equipment is an infrared camera that was developed specifically for process monitoring. It supplies real-time thermal images at a rate of 80 Hz. The pictures are sent via USB to a separate system, where they can be recorded as video. The PI450’s thermal sensitivity is 40 mK, making it ideal for assessing small gradients.
Warmup Process with Pictures, Video, and Numbers
The ambient temperature was 22 degrees Celsius during all measurements. On the open bench table, the Gigabyte R9 285 WindForce OC comes in at 29 degrees Celsius at idle, 64 degrees Celsius during gaming, and 65-66 degrees Celsius during the stress test.
It’s plain to see that, even under a constant load, the only area that really heats up is around the voltage converters. The cooler’s surface comes in at 72 degrees Celsius, which is definitely acceptable.
Things don’t look as pretty when turning the graphics card around. Low fan RPM are certainly nicer to listen to, but they’re not so great when it comes to cooling the voltage converters. It’s a pretty safe bet that the back of the board gets just as hot as the components on the front after the warmup phase. Consequently, the almost 90 degrees Celsius measured at the area on the back of the board that has the voltage converters on the front are a realistic result.
This doesn’t cause any damage, of course, since the board and all active components involved are designed for high temperatures like these. Then again, we also see how the temperatures spread across the board to areas where the connectors of sensitive components have been permanently soldered to the board. We’ve illustrated the whole process with our 2-minute (originally 20-minute) time-lapse video.
Incidentally, this finding isn’t unique to the Gigabyte R9 285 WindForce OC. All current high-end graphics cards’ boards get excessively hot. A backplate doesn’t help either and might well be counterproductive. The table below provides a summary of all temperature results.
| Idle | 3D Workload - Open Benchtable | 3D Workload - Closed Case | |
|---|---|---|---|
| Gigabyte R9 285 Windforce OC | 29 °C | 64 °C | 66-67 °C |
| Gigabyte R9 280X Windforce | 30 °C | 73 °C | 75 °C |
Noise Measurement
Graphics cards’ noise measurements are performed using a calibrated high-end studio microphone (supercardioid) positioned perpendicular to the middle of the graphics card in question at a distance of 50 cm. This distance, in conjunction with the very strong cardioid directionality of the microphone, represents a compromise between avoiding noise due to fan turbulence and avoiding ambient noise, which can never be completely eliminated. We performed all of the noise measurements at night for this reason.
We’ve also used a closed case to simulate everyday-life conditions. After the warmup phase, we don’t just measure the temperatures, but also the fans’ RPM. We then manually set the fans’ RPM to these values after shifting to the open setup. Doing it this way allows for real comparison measurements under identical acoustic conditions.
So how does the Gigabyte R9 285 WindForce OC fare when it comes to noise? Pretty well, actually.
This is what the cold hard numbers look like:
| Idle | 3D Workload - Open Benchtable | 3D Workload - Closed Case | |
|---|---|---|---|
| Gigabyte R9 285 Windforce OC | 30.9 dB(A) | 35.5 dB(A) | 36.8 dB(A) |
| Gigabyte R9 280X Windforce | 31.4 dB(A) | 42.7 dB(A) | 44.9 dB(A) |
We also captured a few more load noise comparison data points in our Canadian labs. These were performed in an open case environment, from 3" behind the output bezel of the graphics cards, after 10 minutes of Battlefield 4:

While the Radeon HD 7870 reference card is the loudest, all of these cards deliver acceptable acoustic performance. Having said that, Asus' Strix Radeon R9 285 stands out as extra quiet in this particular comparison.
|
|
|
|
|
When we average out the benchmark performance across all of the tests we've taken for this article, we are left with the following aggregate results. These are normalized to the Radeon R9 270X, which represents the 100% baseline. Note that these numbers do not include Mantle performance. We also decided to leave Assassin's Creed IV out of this average result as we're a bit skeptical that it represents a typical scenario.
This is not a situation that we expected to see. Based on the specifications alone, we thought that the Radeon R9 285 would fall behind the Radeon R9 280 when it comes to average game performance. In actual benchmarks it comes very close, sometimes beating and sometimes losing to the Radeon R9 280 by a small amount, and slightly besting its predecessor on average. This indicates that AMD's new lossless color compression scheme is effective enough to compensate for the raw memory bandwidth deficit that the 256-bit Radeon R9 285 suffers compared to the 384-bit Radeon R9 280. That alone is an impressive technical accomplishment.
From the perspective of a gamer, the Radeon R9 285 isn't quite as impressive when compared to the Radeon R9 280 it will replace. Sure, performance-per-watt has improved significantly, and it's nice to have access to new features such as TrueAudio, a revamped 4K-compatible UVD/VCE, and bridgeless CrossFire. But the most important metric to a gamer is FPS-per-dollar, and this doesn't change much vs. the Radeon R9 280.
Having said that, the Radeon R9 285 is a wonderful option at $250 and probably represents the best product a gamer can buy for that price, just as the Radeon R9 280 did. I certainly wouldn't recommend that current Radeon R9 280 owners upgrade to the Radeon R9 285, as there wouldn't be a noticeable performance improvement. But for folks coming from entry-level graphics cards, the Radeon R9 285 is an excellent upgrade choice and delivers true high-detail 1080p gaming.
We'd also like to address the GeForce GTX 760's position on our average performance chart. Seeing the 760 sit just below the Radeon R9 270X wasn't something we expected, and this situation caused us to re-run the majority of benchmarks and cross-reference them with other tests we've taken. The results were consistent, though, and we believe this situation is the result of a combination of factors including some newer game titles such as Thief (that favor the GCN architecture), mixed with some high-detail settings that don't sit well with the GeForce GTX 760's 192 GB/s of memory bandwidth (bandwidth that doesn't benefit from lossless color compression). We wouldn't count out the GeForce GTX 760, a card that uses even less power than the Radeon R9 285, but we're beginning to wonder if it's time to re-assess its position over a wider range of benchmarks.
As for Nvidia's other offerings, the GeForce GTX 770 performed admirably, beating out the Radeon R9 280X. It's also no secret that this company is rumored to have its next-generation GPU architecture poised on the horizon. If so, this could have a major impact on the $250 market, depending on what kind of competing products are released over the next quarter. Of course we won't have anything concrete to conclude about that until Nvidia tips its hand.
Before tying this up, we should also mention some updates to AMD's Never Settle game bundle promotion. Dubbed the Never Settle: Space Edition, new titles have been added to the mix including the upcoming Alien: Isolation (a gold-tier option), and Star Citizen (which includes an AMD-themed in-game racing spaceship, a gold- or silver-tier option). While Nvidia is offering Borderlands: The Pre-Sequel with its GeForce GTX 770/780/Titan cards, we do think that AMD's bundle offers more choice.
In the final analysis, we have no reservations that would prevent us from recommending the Radeon R9 285 at the $250 price point. Having said that, we're just as enthusiastic about the end-of-lifed Radeon R9 280, especially if we see prices drop as it is cleared out. While the Radeon R9 285 does offer compelling new features, lower power usage, and a good value proposition right now, the best part may come to consumers after some time has passed: a cheaper-to-manufacture 256-bit memory interface may provide AMD with more room to lower prices in the future.
Pros:
Performance on par with Radeon R9 280. New features including TrueAudio and bridgeless CrossFire. Lower power usage and improved performance-per-watt.
Cons:
Lower raw memory bandwidth than Radeon R9 280 that may result in slightly lower performance in some cases. 2 GB of onboard RAM instead of 3 GB. Does not notably increase performance over Radeon R9 280 on average.
Verdict :
A good $250 replacement for the Radeon R9 280 that doesn't up the ante when it comes to game performance, but adds new features and lower power usage to the mix.









































