|Introduction and Related Information|
A few months back, AMD unveiled a handful of mobile Elite A-Series APUs, formerly codenamed Richland. Those products built upon the company’s existing Trinity-based products but offered additional power and frequency optimizations designed to enhance overall performance and increase battery life. In addition to these optimizations, AMD also began offering a host of specialized software with their Richland APUs that’s leveraged their inherent strengths--namely, AMD Face Login, Gesture Control, Screen Mirror, and various video enhancement features like AMD Steady Video, Quick Stream, and Picture Perfect. The ultimate goal was to make AMD hardware and software more appealing to OEMs and consumers alike.
AMD’s A-Series APUs debuted back in 2011 with Llano and were updated in 2012 with more powerful Trinity-based products. Though Richland supplants Trinity in AMD’s APU line-up, the die shots (and specifications) above reveal that Trinity and Richland are actually based on the same silicon.
AMD has put in some additional engineering effort and done some new things with chips, which we’ll discuss a little later, but since Trinity and Richland are fundamentally similar, a refresher on the technology and platform as a whole couldn’t hurt to help lay some foundation for what we’ll be showing you on the pages ahead.
In our AMD Trinity A10-4600M coverage, we go into detail on the Trinity architecture and the Piledriver microarchitecture on which its CPU cores are based. And in the remaining AMD A10- and A8-related articles, we cover last year’s Trinity-based desktop APUs and detail the Virgo platform as a whole, which encompasses the APUs and associated chipsets and motherboards.
|Differentiating Richland 2013 APUs|
As we’ve mentioned, AMD’s new Richland-based Elite A-Series APUs are an update to an existing design (Trinity) that’s meant to improve the company’s competitive positioning against Intel--for the next few months at least. AMD’s upcoming Kaveri APUs, which are slated to arrive before the end of the year, will feature brand-new CPU cores based on the Steamroller microarchitecture and an updated GPU block that leverages AMD’s GCN (Graphics Core Next) microarchitecture, just like current-generation discrete Radeon products.
Richland, however, still features Piledriver-based CPU cores and Northern Island-class, VLIW4 graphics, despite the fact that the GPUs have been branded with Radeon HD 8000-series model numbers. The good news is that Richland is a drop-in upgrade for existing platforms. These new Elite A-Series APUs work with current A55, A75, and A85X chipsets and use the same socket FM2. The bad news is that they don’t offer any true next-gen features, and their performance uplift isn’t huge.
What AMD has done with these new Elite A-Series APUs is leverage data gathered by on-die temperature sensors to offer significantly better power management and dynamic frequency boosting; AMD is calling this Hybrid Boost. The thermal sensors were always present on Trinity, but the engineering effort to use the data provided by them wasn’t completed in time for Trinity’s original launch. With Richland, actual temperature data is used to supplement AMD’s existing power-driven calculations to more aggressively boost frequencies and wring more performance from the parts at similar or lower power envelopes. Tuning of the fabrication process over time has also allowed AMD to improve voltage and frequency margins. What all that means is that Richland, while fundamentally similar to Trinity, should be able to hit higher frequencies at lower voltages and power envelopes and more intelligently use Turbo to boost performance further.
In fact, all of the new Richland APUs launching today can boost well above 4GHz, with the same 100w or 65w TDPs of their predecessors. We’ve got the two highest-end Elite A-Series APUs on tap for you here, the A10-6800K and A10-6700. Like previous-gen APUs (and Intel’s current processors), the “K” in the product name denotes an unlocked processor with CPU, GPU, and Memory multipliers that can be freely manipulated by end users.
The new lineup consists of five APUs, two of which are unlocked. The various CPU, GPU, and cache configurations of the chips are outlined above. As you’ll see, the highest-end parts feature quad-CPU cores with 384 Radeon cores and 4MB of total cache. The top end APUs have GPU cores clocked at 844MHz (a 44MHz increase over Trinity) with boost clocks that top out at lofty 4.4GHz. We should also note that the top-end part, the A10-6800K, has been validated for use with DDR3-2133MHz memory. The rest of the APUs max out at 1866MHz. That additional frequency headroom on the A10-6800K’s memory controller results in a marginal increase in GPU performance, as you’ll see a little later.
To further differentiate their Elite A-Series APUs for 2013, AMD is also offering different software bundles. The top of the line A10 parts will include some games from AMD’s Never Settle Bundle along with AMD’s entire suite of customized apps. The rest of the parts will include some, but not all of the applications. The complete breakdown is listed in the slide above, and we’ll simply quote AMD for descriptions of the applications...
|Test Setup and PCMark 7|
Test System Configuration Notes: When configuring our test systems for this article, we first entered their respective system BIOSes or UEFIs and set each board to its "Optimized" or "High performance Defaults". We then saved the settings, re-entered the BIOS/UEFI, and set the memory speed to each platform's maximum, officially supported speed--DDR3-1866 in the case of Virgo. The solid state drives were then formatted, and Windows 7 Ultimate x64 was installed. When the Windows installation was complete, we fully updated the OS, and installed the drivers necessary for our components. Auto-Updating and Windows Defender were then disabled, and we installed all of our benchmarking software, performed a disk clean-up, cleared any prefetch and temp data, and ran the tests.
Futuremark's PCMark 7 is the latest version of the PCMark whole-system benchmarking suite. It has updated application performance measurements targeted for a Windows 7 environment and uses newer metrics to gauge relative performance.
Below is what Futuremark says is incorporated into the base PCMark suite and the Entertainment, Creativity, and Productivity suites--the four modules we have benchmark scores for you here.
The new AMD A-Series Elite APUs for 2013 shows some nice performance improvements in PCMark 7 due to their higher frequencies and much higher video encoding performance, which seems to leverage the GPU properly now. Even so, Intel's Core processors, however, have no trouble outperforming AMD here.
|LAME MT and SunSpider|
In our custom LAME MT MP3 encoding test, we convert a large WAV file to the MP3 format, which is a popular scenario that many end users work with on a day-to-day basis to provide portability and storage of their digital audio content. LAME is an open-source MP3 audio encoder that is used widely in a multitude of third party applications.
In this test, we created our own 223MB WAV file (a hallucinogenically-induced Grateful Dead jam) and converted it to the MP3 format using the multi-thread capable LAME MT application, in both single and multi-thread modes. Processing times are recorded below, listed in seconds. Shorter times equate to better performance.
Audio encoding with LAME MT is definitely not one of Piledriver's strong suits. The A10-6800K and A8-6700 outpace all of the other APUs we tested, but Intel's parts simply dominate in this test due to their strong IPC advantages over AMD at this time.
All of the systems were tested using the latest version of Internet Explorer 9, with default browser settings, on a clean install of Windows 7 Ultimate x64.
AMD's new APUs showed some nice improvements over last year's 5000 series products in the SunSpider benchmark, but despite the nice speed boosts, Intel's processors finished well ahead.
|Cinebench R11.5 and POV-Ray|
Cinebench R11.5 is a 3D rendering performance test based on Cinema 4D from Maxon. Cinema 4D is a 3D rendering and animation suite used by animation houses and producers like Sony Animation and many others. It's very demanding of processor resources and is an excellent gauge of pure computational throughput.
This is a multi-threaded, multi-processor aware benchmark that renders a photorealistic 3D scene (from the viral "No Keyframes" animation by AixSponza). This scene makes use of various algorithms to stress all available processor cores. The rate at which each test system was able to render the entire scene is represented in the graph below.
The A10-6800K and A10-6700 performed relatively well in the Cinebench benchmark. Single-thread performance continues to be an Achilles heal for AMD, but the new Elite series APUs were able to outpace the similarly priced Core i3 processors in the multi-threaded test.
POV-Ray, or the Persistence of Vision Ray-Tracer, is an open source tool for creating realistically lit 3D graphics artwork. We tested with POV-Ray's standard 'one-CPU' and 'all-CPU' benchmarking tools on all of our test machines, and recorded the scores reported for each. Results are measured in pixels-per-second throughput; higher scores equate to better performance.
The performance breakdown in the POV-Ray benchmark looks much like Cinebench, with AMD's latest APUs besting the Core i3s in the multi-threaded test but trailing the AMD FX and quad-core Intel processors.
|Media Encoding and Image Editing|
|Cyberlink's MediaEspresso is a video conversion tool that imports various video file types and converts them to other formats for publication, portability, and/or streaming. In this test, we take a 277MB high definition 1080p AVCHD video clip and convert it to a an H.264 encoded MP4 compatible file designed for use with an iPhone / iPad (or other portable media playback device).
Here we're going to look directly at AMD's HD Media Accelerator in the new A10 series APUs and compare it to Intel's Ivy Bridge and Haswell Quick Sync engines and competitive CPU-based solutions, as well.
AMD's latest APUs put up some better numbers here, but they still can't come close to matching the performance of Intel's Quick Sync engine.
Musemage is a fully-functional GPU-powered photo editing software suite. Thanks to the GPU-accelerated filters built into the app, Musemage can leverage the additional performance afforded by a more powerful GPU and offer real-time (or near real-time) visual feedback. The application also offers a built-in benchmark, which outputs performance as a final "score". The score put up by each test system is represented in the graph below.
AMD's A10 series APUs all put up similar scores in the MuseMage benchmark, which was significantly higher than anything from the Intel camp.
|Low-Res Gaming: Crysis and ETQW|
For our next set of tests, we moved on to some low-res benchmarking with Crysis (DirectX) and Enemy Territory: Quake Wars (OpenGL). In these tests, we drop the resolution to 1024x768, and reduce all of the in-game graphical options to their minimum values to minimize the load being placed on the GPUs and push frame rates as high as possible. However, the in-game effects, which control the level of detail for the games' physics engines and particle systems, are left at their maximum values, to enhance image quality somewhat.
The use of faster DDR3-2133 memory on the A10-6800K, in addition to its higher base and Turbo clocks, give it a nice boost in performance over last year's APUs. AMD's A-Series' APUs top the charts here as well, due to their much higher-performing GPUs versus Intel's offerings. With a discrete graphics card installed, the results of low-res tests like these would likely change significantly, but when leveraging on-die GPU resources, AMD clearly has an edge.
|3DMark Vantage and Cinebench OpenGL|
The individual GPU tests that partially make up the 3DMark Vantage suite show AMD's latest GPUs with a clear lead across the board, versus last year's APUs and Intel's similarly-priced offerings.
3DMark's overall score reflects the lead AMD's APUs held onto in the individual tests. Once again, we should note, the A10-6800K's official support for DDR3-2133 RAM give it a nice boost in performance versus the 6700.
Cinebench R11.5’s GPU benchmark uses a complex 3D scene depicting a car chase, which measures the performance of a graphics card in OpenGL mode. The graphics card has to display a huge amount of geometry (nearly 1 million polygons) and textures, as well as a variety of effects, such as environments, bump maps, transparency, lighting and more to evaluate the performance across different disciplines and give a good average overview of the capabilities of the graphics hardware. Results are reported in frames per second.
The new A10-6700 and last year's A10-5800K put up very similar numbers in Cinebench's OpenGL benchmark. The higher-clocked A10-6800K, when paired to DDR3-2133 RAM, simply improves the situation. AMD's APUs clearly outpaced Intel's offerings in this test, as well.
|Hi-Res Gaming: ET:QW, L4D2, Metro 2033, JC2|
For this next set of tests, we pit the integrated processor graphics incorporated into AMD's Richland-based APUs against Intel's HD 2500, HD 4000, and HD 4600 series engines and a few of the least expensive discrete GPUs from NVIDIA and AMD: the GeForce GT 430, Radeon HD 6450, and Radeon HD 5550. We tested the games at high-quality settings, at a resolution of 19200x1200 with anti-aliasing and anisotropic filtering enabled to put a significant strain on the various GPUs.
In both cases, the Radeon HD 8670D graphics cores built into the new AMD A10-6800K and A10-6700 APUs offered nearly double (or more) the performance of Intel's integrated graphics solutions. The HD 4600 engine in the Haswell-based Core i5-4670K put up much better numbers than the HD 4000, but it still AMD by a wide margin here.
These Metro 2033 and Just Cause 2 tests are somewhat more demanding than the others posted above, and as such, we had to crank down the image quality in both titles. These tests were run at a lower-resolution (1680x1050) with in-game graphics options set to their medium values, with only 2XAA (JC2) and 4X anisotropic filtering enabled. Still, the trend is the same, with AMD's latest APUs clearly outpacing Intel's various HD graphics options.
Throughout all of our benchmarking and testing, we also monitored how much power our test systems consumed using a power meter. Our goal was to give you all an idea as to how much power each configuration used while idling and while under a heavy workload. Please keep in mind that we were testing total system power consumption at the outlet here, not just the power being drawn by the processors alone.
Color us surprised. Although AMD had touted the increased power efficiency afforded by Richland's new boost modes and more sophisticated power management schemes, we thought the chips' higher frequencies would result in higher power consumption across the board. That was not the case, however; the A10-6700 and A10-6800K both used slightly less power than last year's A10-5800K in our tests, though Intel's processors still sipped power by comparison.
|Our Summary and Conclusion|
Performance Summary: AMD’s latest Elite A-Series APUs offer modest performance improvements over their predecessors. When paired with DDR3-1866 memory, the A10-6700 performs very much like the A10-5800K. The A10-6700 puts up somewhat better numbers due to its slightly higher-clocked CPU and GPU cores (3.7GHz / 4.3GHz / 844MHz vs. 3.8GHz / 4.2GHz / 800MHz), but the differences weren’t dramatic. The A10-6800K fares a bit better. Since AMD has qualified the A10-6800K for use with DDR3-2133MHz memory, and the APU is clocked at 4.1GHz (base) / 4.4GHz (Turbo) / 844MHz (GPU), its performance is better than any previous APU across the board. The higher CPU clocks improve processor performance, and the faster GPU and memory improve graphics performance considerably. Power consumption is also lower than last year's models. Versus Intel’s offerings, Richland doesn’t do much to minimize Intel’s lead in CPU performance, but AMD’s lead in integrated / on-die graphics performance has been maintained. Even the Intel HD 4600-series graphics in the Haswell-based Core i5-4670K doesn’t come close to the performance of the Radeon HD 8000-series GPU in Richland.
The AMD Elite A-Series APUs we’ve shown you here should be available at retail and from AMD’s system partners immediately. Pricing for these new APUs ranges from $69 for the entry-level A6-6400K dual-core to $142 to the top-of-line A10-6800K.
At those prices, AMD’s latest Richland Elite A-Series APUs are somewhat more expensive than their Trinity-based counterparts, which currently top out at about $129, but they’re priced very competitively with Intel’s offerings. The Core i3-3225, for example, is currently selling for $139; the couple of extra bucks for the A10-6800K will get you a better-performing CPU (in multi-threaded workloads, at least) with a much higher-performing GPU. However, Intel still has the edge in single-thread performance and power efficiency. These new APUs are also drop-in upgrades for existing platforms. Looking back at the numbers, we doubt anyone with a high-end Trinty-based APU will feel the need to upgrade, but if you’ve currently got a low-end socket FM2 APU and want to breathe some new life into the system, a Richland Elite A-Series APU is the easiest way to do it and it won't break the bank.