In 2011, AMD released their first APU architecture codenamed “Llano” which was eventually followed by its successor “Trinity” in 2012. The Trinity APUs were met with a positive response by consumers due to the value they offered.
Additionally and over the Llano APU, Trinity offered the latest 32nm Piledriver cores which brought x86 architecture and an enhanced Radeon graphics core. The Piledriver cores on the Trinity APU offered while not the best but adequate performance to power applications but the graphics power shined against Intel’s HD counterparts. The only downside remained the high power consumption that needed to be sorted out in a revision later on. The new revision launched three days after Intel released their Haswell processors which brought a new 22nm architecture enhancements but the major improvement on Intel’s front was not the compute performance but rather the graphics performance and overall power efficiency.
While Trinity’s graphic side was strong, the new revision had to tackle Intel on both the GPU front and power efficiency. Its time to see how the third generation APU revision of Trinity codenamed “Richland” fairs in our review.
AMD APUs Get Boosted With The Richland Platform
The third generation APU platform from AMD is codenamed “Richland”. While Richland is considered the successor to the Trinity APU, it still is technically the same architecture with the exact same design. AMD invested their time updating the Trinity architecture, tuning its graphics and power management features.
Essentially, the Richland APU is built on the same 32nm SOI process featuring 1.303 billion transistors on its die which measures 246mm2. The Richland APU makes use of upto four x86 Piledriver cores and the GPU side gets powered by the VLIW4 architecture. One would ask, if its the same thing as was Trinity, why did AMD bother to release Richland as a new platform?
The reason is the time been invested by AMD in tuning the APU architecture, enhancing its power efficiency and boosting the clock speeds far more than ever possible on and APU. These small yet abundant reasons made AMD opt for a new APU platform in 2013. Just like the Trinity APU, bringing the high-performance x86 architecture “Piledriver” wasn’t feasible for Richland APUs which had to accommodate a enhanced graphics core that would result in higher cost and inefficiency so a few features had to be cut down. The major thing that was cut out from the APU model was L3 cache since it was most demanding in power consumption and the least performance awarding. AMD’s APUs were limited to L2 cache since AMD chose efficiency over performance for their new Fusion platform. Like Trinity, Richland supports the latest instruction set from AMD – FMA4/3, AVX, AES, XOP.
A little naming and frequency boost update was received to the graphics side rather than a different architecture. AMD’s Trinity used the Radeon HD 7000 series integrated graphic chip, Richland APUs arrived with the new Radeon HD 8000 series IGP chips. In addition to the new naming, AMD also boosted the clock frequency on their Richland chips which resulted in faster gaming performance which we will see in the performance section of this review.
The most significant update still remained the power management features on the Richland APUs. Previous generations adjusted power between the CPU and GPU through balancing loads between both whether in functional, idle or turbo modes. This management is achieved by deploying several integrated micro-controllers on the APU die. Richland also makes use of the same controllers which sense and adapt to the nature of the cooling solution and ambient conditions and also power usage. In this case, if a CPU is performing with limited potential while the GPU is being bottlenecked, the CPU would be automatically boosted. The other case involves the GPU not being used and the CPU consuming too much power in which case the CPU power would be reduced. Similarly, if an application makes use of only two cores on the APU, the power management features would boost the clock speeds of those two independent cores to boost the performance. Temperature plays a major role here too with boosting the performance of both GPU and CPU when the temps are under limit but the APU throttle down to save power when they reach the TJ max limit.
These “Hybrid Boost” features help improve the power management on Richland APUs, so under the same TDP packages consumers get faster performance.
2013 AMD A-Series Platform Features:
- Turbo Charged x86 architecture featuring “Piledriver” cores
- Supports up to 4 cores and support for the latest ISA instructions including FMA4/3, AVX, AES, XOP
- Up to 2MB L2 cache per dual-core module (up to 4MB total)
- Max Turbo Frequencies up to 4.4 GHz
- Configurable via AMD OverDrive
- AMD Radeon HD 8000 Series GPU Cores
- Featuring VLIW 4 architecture
- Up to 384 shaders
- Up to 844MHz
- Up to 8xAA and 16AF support
- Controllable via AMD OverDrive
- DirectX11 Support
- New DDR3-2133 support on A10 APUs
- Enhanced AMD Turbo Core
- More Frequency/Voltage levels for CPU and x86 cores
- Temperature Smart Turbo Core
- New bottleneck detect algorithms
- Controllable via AMD OverDrive
- UVD and VCE
- Video Encode and Decode Hardware to offload CPU
- AMD Picture Perfect support with HD Post Processing technologies
- Support for latest display technologies
- AMD Eyefinity2 technology for 3+1 monitor support
- Display Port 1.2 support
AMD RichlandTech Specs:
- Die size: 246mm2
- 1.303B Transistors
- Process: 32nm SOI
- Power reduction during HD media playback
- Unified Northbridge (UNB)
- Quad Core and Dual core configurations
- Updated AMD Radeon DirectX11 GPU
- Northern Islands GPU With Upto 384 Cores 2.0
- 3 dedicated display outputs
- 4 independent display controllers
- DisplayPort 1.2 with symbol rates of 1.62, 2.7 and 5.4 Gbit/s
- UVD and AMD Accelerated Video Converter
- IOMMU v2
AMD FM2 Socket and A75 / A85X Chipset
The AMD FM2 Socket and A85X Chipset based motherboards support the latest Richland APUs. With this move, AMD offers an easy upgrade path to consumers currently using the Trinity processors. Users with Llano APUs can’t upgrade to FM2 since the new socket isn’t pin-compatible with the older processors.
However, AMD is still using the A55 and A75 chipsets on their new FM2 motherboards along with the latest A85X chipset codenamed “Hudson D4″ that unleashes complete features available on the Richland APU platform. The A85X chipset is going to be shipped with the high-end FM2 motherboards allowing upto 8 SATA 6GBps ports, four USB 3.0, ten USB 2.0 ports, High Definition audio and AMD CrossfireX support. AMD has arranged their new chipsets in three tiers; A55 Chipset for Entry FM2 solutions, A75 chipset for media users and gamers while the A85X chipset is targeted towards performance users who want to feast upon all the features available on the APU platform.
AMD Richland APU Lineup
AMD has a total of five SKUs under its Richland platform of APUs, six of these APUs fall under the A-Series branding. Each processor under the A10/A8/A6 brand features a fully unlocked ‘K’ design which allows overclocking, should be noted that Intel does not offer fully unlocked K series chips in Core i3 and Core i5 processors except the top tier Core i5 processors so this allows users to gain more performance through overclocking.
The AMD Richland APU lineup is listed below:
|CPU Clock (Base/Turbo)||4.1/4.4 GHz||3.7/4.3 GHz||3.9/4.2 GHz||3.5/4.1 GHz||3.9/4.1 GHz|
|L2 Cache||4 MB||4 MB||4 MB||4 MB||2 MB|
|Integrated Graphics “Radeon”||HD 8670D||HD 8670D||HD 8570D||HD 8570D||HD 8470D|
|GPU Clock||844 MHZ||844 MHz||844 MHz||800 MHz||800 MHz|
|GPU Cores||384 Cores||384 Cores||256 Cores||256 Cores||192 Cores|
|DDR3 Frequency||2133 MHz||2133 MHz||2133 MHz||2133 MHz||2133 MHz|
AMD A10-6800K and A10-6700 – The Top Tier Richland Chips
Today we will be testing not one but two Richland APUs, the A10-6800K and A10-6700. Both Richland chips succeed their successors – A10-5800K and A10-5700. The A10-6800K is powered with four x86 piledriver cores and a VLIW4 architecture based Radeon GPU with 384 cores.
The A10-6800K comes with a clock speed of 4.1 GHz with Turbo core frequency boosting it upto 4.4 GHz which is quite fast for an accelerated processing unit. The HD 8670D graphics core operates at 844 MHz and can be configured to run at idle mode when not under operation or direct load from a demanding application. The graphics core can support “Dual graphics” allowing select motherboards that allow the features to pair up the graphics core with a discrete graphics solution allowing an operation similar to CrossfireX in which power can be utilized from both cores.
The A10-6800K comes with a TDP of 100W which is high compared to its counterparts from Intel but its due to the 32nm process which Intel’s Piledriver architecture is built around. Heat is the least of an issue for the new processors.The A10-6800K currently retails at an MSRP of $142 which is quiet a reasonable price for a chip that packs the CPU and GPU with fast clock speeds.
On the other hand we have the A10-6700 which as the name denotes is a locked chip which cannot be overclocked but comes at a price $20 cheaper than the unlocked A10-6800K at $122 US. Though it does makes some compromises on the CPU clock speed with 3.7 GHz base and 4.1 GHz boost but it is backed by a 65W TDP that could enable for some good budget level HTPC builds.
The GPU core remains the same as the A10-6800K with the HD 8670D die boasting 384 cores at a clock speed of 844 MHz. Fortunately the GPU core can be overclocked but with it the performance is 100% alike to the A10-6800K which is a good thing.
|Power Supply:||Xigmatek NRP-MC1002 1000 Watt|
|Hard Disk:||Intel SSD 520 Series 256 GB (OS)
Seagate Barracuda 500GB 7200.12
|Memory:||4 x 4 GB Kingston HyperX 2133 MHz
10th Anniversary Edition Memory Kit
|Case:||Cooler Master HAF 932|
|Video Cards:||AMD HD 8670D
AMD HD 7660D
Intel HD 4600
Intel HD 4000
Intel HD 2500
|Cooling Solutions:||Corsair H60 Hydro Cooler|
|OS:||Windows 8 Ultimate 64-bit|
We used the high-end Core i7 processors just to compare the fastest available graphics chip from Intel against the fastest graphics chip available on AMD APUs. The compute test was compared against an Intel Core i3-3220 which falls within the same price range as the A10-5800K.
Overclocking the A10-6800K and HD 8670D
Overclocking the A10-6800K was a fairly easy job, we just had to raise the multiplier from 48x without a minor voltage adjustment of 1.476V and the processor was overclocked to 4.8 GHz. This is a good improvement over the base speed of 4.1 GHz. The A10-6800K can easily reach around 4.8/5 GHz overclocked frequencies with better coolers. Although high-performance coolers don’t make a sense with a budget APU like the A10-6800K but those who still want extra performance can go for a nice cooler around the $49-$59 range.
We didn’t face any thermal throttling issues since our setup included the Corsair H60 which is adequate enough to handle overclocked loads. We also bumped the clock speed of the Radeon HD 8670D IGP to 1050 MHz from its stock 844 MHz limit. The overclock was stable for both CPU and GPU, the respective GPUz and CPUz screenshots can be seen below. We have included the overclock scores in the charts which are provided in the performance section.
AMD A10-6800K 4.4 GHz / Radeon HD8670D 1050 MHz Overclock:
X264 HD Encode Benchmark
This benchmark measures the encoding performance of the processor. It offers a standardized benchmark as the clip as well as the encoder used is uniform.
Cinebench is based on Maxon’s Cinema 4D. It is used to compare graphics as well as processor performance. We are using the CPU performance numbers for our comparison.
CineBench R11.5 OpenGL
Winrar Compression Test
Some might argue against using 7-zip’s compression and decompression benchmark as a ‘real world’ test. But if you try and think about it for a minute, the benchmark does show how fast the program will either compress or decompress, while negating the impact of disk transfers.
Super PI is used by many overclockers to test the performance and stability of their computers. In the overclocking community, the standard program provides a benchmark for enthusiasts to compare “world record” pi calculation times and demonstrate their overclocking abilities. The program can also be used to test the stability of a certain overclock speed.
If there’s one really good thing about AMD’s APUs, its that they can run most of current generation games at a decent frame rate. That’s something which has not been possible on IGPs but with the recent introduction of more powerful solutions from both Intel and AMD, users can now get to play their latest games with a decent visual configuration and playable frame rates.
While 3DMark 11 was a success, 3DMark from Futuremark further pushes the boundaries of benchmarking utilies going all out with cross platform support which include Windows, Windows Phone, iOS, MAC and even Android. The utility comes with three benchmark tests configured for different tiers of high-performance PCs, Mid-range PCs/Tablets and smartphone devices.
Futuremark released 3DMark 11 in 2011 bringing support for the latest DirectX 11 GPUs at that time. Since then, the benchmark tool is used widely for evaluating performance of high-end PCs.
3DMark Vantage is still used to this date as a complete benchmark suite for GPU and CPU performance.
Resident Evil 6
Capcom brings the horror back to the screens with their blockbuster Resident Evil 6 title which was well received among the community. The game features three playable campaigns which include Leon, Chris, Jake and downloadable content for Ada Wong.
The Tomb Raider franchise was rebooted this year with the latest title in the long running franchise. The players start off their journey with a younger and under-trained version of Lara who goes off on her first survival action journey.
Grid 2 is the sequel to the highly successful racing game – GRID. While the game runs great on graphic cards, the developers have also optimized their coding for the latest Haswell processors with fourth generation HD graphics core which features AVX2/AVX and DirectX 11.1 support.
Metro Last Light
Metro Last Light once again puts us in the foots of Artyom, a survivor of the nuclear holocaust that shattered Russia. Metro: Last Light is considered as the best looking game to be released to date making use of intensive DirectX11 Tessellation, High-Res Textures, Global illumination lightning and more.
The Elders Scroll: Skyrim was released by Bethesda in fall 2011. The game featured one of the most largest worlds ever created in an Elders Scroll game taking the RPG genre to the next level.
When it comes to power consumption, it should be noted that while AMD has been focusing on increasing the IPC performance of their CPU core and boosting the GPU performance, Intel has opted to improve the power efficiency of their already powerful core processors.
Only a few days before AMD launched their Richland APUs, Intel unveiled the Haswell processors which are based on the most power efficiency 22nm architecture ever built by Intel. Reducing the power is one thing but Intel also boosted some good performance with their new HD Graphics chip so AMD had to tackle them in both fronts with their older 32nm process which could only had been tweaked to gain more efficiency.
Below, you can see that AMD has higher wattage compared to Intel processors due to its 32nm design which has got older, Intel on the other hand aims for 22nm with their Haswell and Ivy Bridge processors.
AMD’s Richland APU improves the power efficiency on AMD’s end but still lacks against Intel’s offerings. Kaveri APU which is planned for Q4 2013 would be based on the 28nm architecture and is supposed to improve performance and efficiency on all ends. But let’s just focus on the Trinity A10-6800K part and see how much watts it consumes in idle and load compared to its older Trinity siblings:
With this review completed, i want to say that APUs are the way forward for AMD and computing. AMD may not be the leader in x86 performance but they can definitely become one in the APU design and tech department. Its known that APUs currently power the next generation consoles from Sony and Microsoft, they power ultra low power tablets and netbooks and lastly the offer amazing value to desktop consumers with the A-series platform.
AMD Richland improves on the design of last year’s Trinity APU. It comes with more value than ever before and provides decent amount of performance for a refresh. The clock frequencies have gone up, the performance has gone up and the overall power consumption has been reduced by a fair amount. For the first time with Richland, AMD now offers select APU with SimCity from EA as a part of their gaming evolved program. The motherboards for the FM2 platform are easily within reach and if you already have an FM2 socket motherboard that you were using with Trinity, you can just take out the older APU and pop in the new one proving as an easy upgrade path.
Gaming performance is strong with AMD’s Richland A-Series APUs. I have been testing out gaming for weeks with the Richland A-Series APU. I tried Skyrim, Borderlands 2, Batman: Arkham City and SimCity which to my surprise ran perfectly well with decent amount of visuals. If someone is preparing a budget built with the aim to play older or current titles on smaller resolution monitors than A-Series should be their definitive choice. Some current gen games can even run at 1080P resolutions provided you have toned down AA and texture details but overall i was impressed.
There’s nothing bad at all about Richland, but when it comes to comparison with Intel AMD gets stuck on an older 32nm design which is obsolete by now. Their power consumption is still much higher compared to Ivy Bridge and Haswell processors. And the Haswell HD graphics chips have come on par with AMD’s offerings while the mobile Iris and Iris Pro graphics processors turn out to be a faster solution. Since Richland was a refresh, it was expected that it won’t offer incredible amounts of performance but it has provided decent value. What AMD has in plans for its next APU architecture is worth to be noted. Richland and Trinity are confirmed to be supported by the new FM2+ socket. The same socket would feature support for Kaveri, AMD’s next generation APU that comes with the latest x86 Steamroller cores based on the 28nm process and a brand new GCN based Volcanic Islands IGP that would almost match some discrete entry level solutions from the Radeon HD 7000 series. All eyes are on AMD’s Kaveri APU since other than its architectural changes, it also comes with HUMA and HSA enhancements which allow a easier memory sharing path along the CPU and GPU. So while Richland is the best APU now, Kaveri is going to be better and moving onward to 2015 we will see Carrizo, the next in line to replace Kaveri.
So all in all, AMD has a strong APU lineup for the upcoming years and about Richland itself, the new A-series APUs offer amazing value to consumers with impressive clock speeds and faster gaming performance in one complete package.