AMD A10-6800K & A10-6700 Richland APUs

A follow up to Trinity, AMD’s Richland APU architecture utilizes an intelligent power management system to hit higher clock speeds without increasing power consumption.

July 1, 2013 by Lawrence Lee

Product	AMD A10-6800K FM2 Processor	AMD A10-6700 FM2 Processor
Manufacturer	AMD
Street Price	US$150

The A10-6800K and A10-6700 represent the best of what AMD currently has to offer in desktop APUs (accelerated processing units). They’re based on a new architecture, codename Richland, which is their third generation APU, a follow up to Trinity using the same FM2 socket. This is a bittersweet update as FM2 will begin to be fazed out late this year to make way for the next generation, Kaveri, which will require new socket FM2+ motherboards. Upgradeability has been a selling point for AMD in the past but those days are long gone, at least as far as their budget APU platform is concerned. FM1 lasted less than a year before it was succeeded and FM2 will not live much longer than that.

The A10-6800K.

With the clock ticking on FM2, you would think they would need to make some drastic improvements to tempt users into investing in an almost redundant platform. Actually, Richland is very similar to Trinity on paper, more so than most processor architecture upgrades. Underneath the hood are the same Piledriver CPU packages and the “new” Radeon HD 8000 series graphics chips are simply rebadged HD 7000’s using the same Northern Islands core. What’s really changed is how things work inside.

Hybrid Boost, AMD’s new power management system, is the key feature that makes Richland an upgrade. The APU is outfitted with multiple sensors to monitor the thermal conditions of every component and region throughout the chip, giving Hybrid Boost plenty of information to make intelligent decisions regarding the adjustment of clock speeds and voltages. Richland also has more P-states and sub-states, that is more defined frequency/voltage levels, to switch between. This gives the system fine-grain control to eke out as much performance as possible when the need arises while staying within the defined power envelope. Hybrid Boost can also detect whether the CPU or GPU is bottlenecking work and adjusts the states of either chip accordingly.

So what does this mean to the end user? The crux of it is Richland can be run at faster clock speeds than Trinity without increasing the power draw, even though the main building blocks, the CPU and GPU cores, are identical. It’s essentially a free performance boost.

Richland Desktop APU Comparison
Model	A10-6800K	A10-6700	A8-6600K	A8-6500	A6-6400K
CPU Cores	4	4	4	4	2
CPU Clock (Turbo/Base)	4.4 / 4.1 GHz	4.3 / 3.7 GHz	4.2 / 3.9 GHz	4.1 / 3.6 GHz	4.1 / 3.9 GHz
Total L2 Cache	4MB	4MB	4MB	4MB	1MB
Max DDR3 Frequency	2133 MHz	1866 MHz	1866 MHz	1866 MHz	1866 MHz
GPU Name	HD 8670D	HD 8670D	HD 8570D	HD 8570D	HD 8470D
GPU Clock	844 MHz	844 MHz	844 MHz	800 MHz	800 MHz
Radeon Cores	384	384	256	256	192
Unlocked	Yes	No	Yes	No	Yes
TDP	100W	65W	100W	65W	65W
Street Price (USD)	$150	$150	$120	$120	$80

The pair of A10’s we’re testing today are the headliners, the A10-6800K and A10-6700. They have very high CPU clock speeds and the best graphics chip available, the HD 8670D, which features 384 Radeon Cores and a GPU clock speed of 844 MHz (a 44~84 MHz increase over Trinity’s flagship HD 7660D). The A10-6800K supports faster DDR3 memory with speeds up to 2133 MHz which is a nice boon for those running on integrated graphics (memory frequency is critical to IGP performance) and an unlocked multiplier for easier overclocking. The A10-6700 lacks these amenities but sports a lower TDP (65W vs. 100W) thanks primarily to a lower base CPU frequency. Lower-clocked quad core A8’s and a dual core A6 is also available.

When considering the cost of a system, the CPU is only part of the equation
as the price of motherboards varies greatly from platform to platform. In the
chart above, we added the current street price of the chips compared today to
those of an average compatible motherboard from Newegg.
The following criteria were used for the motherboards: retail versions, Asus/Intel/Gigabyte/MSI
branded, microATX/ATX form factor, SATA 6 Gbps and USB 3.0 controllers, and outrageously
priced models were omitted. The average motherboard price turned out to be
US$119 for LGA1155, US$105 for AM3+, and US$92 for FM2.

At US$150, the A10-6800K and A10-6700 are strategically priced between Intel’s quad core and dual core processors but AMD’s lower motherboard prices are what really give them an edge.

TEST METHODOLOGY

Common CPU Test Configuration:

• Kingston HyperX LoVo 4GB memory – 2x2GB, DDR3-1600 @ 1333 MHz, 9-9-9-24
• Asus EN9400GT Silent Edition
  graphics card – 512MB
• Western Digital VelociRaptor
  hard drive – 300GB, 10,000RPM, 16MB cache
• Seasonic
  SS-400ET ATX power supply
• Scythe Kabuto
  CPU cooler – stock fan at 800RPM
• Microsoft
  Windows 7 operating system – Ultimate, 64-bit

Common IGP Test Configuration:

• Kingston HyperX LoVo 4GB memory – 2x2GB, DDR3-1600 @ 1600 MHz, 9-9-9-24
• Western
  Digital Scorpio Blue notebook hard drive – 500GB, 5400RPM,
  8MB cache
• Asus
  BC-1205PT Blu-ray drive
• Seasonic
  SS-400ET ATX power supply
• Scythe Kabuto
  CPU cooler – stock fan at 800RPM
• Microsoft
  Windows 7 operating system – Ultimate, 64-bit

AMD AM3+ Platform:

• AMD FX-8150
  processor – 3.6 GHz, 32nm, 125W
• AMD FX-8350
  processor – 4.0 GHz, 32nm, 125W
• Asus Crosshair V Formula motherboard
  – 990FX chipset
• Asus Sabertooth 990FX R2.0 motherboard
  – 990FX chipset

AMD FM1 Platform:

• AMD A8-3850
  processor – 2.9 GHz, 32nm, 100W, integrated Radeon HD 6550D graphics
• Gigabyte A75M-UD2H motherboard
  – A75 chipset

AMD FM2 Platform:

• AMD
  A10-6800K processor – 4.1 GHz, 32nm, 100W, integrated Radeon
  HD 8670D graphics
• AMD A10-6700
  processor – 3.7 GHz, 32nm, 65W, integrated Radeon HD 8670D graphics
• AMD
  A10-5800K processor – 3.8 GHz, 32nm, 100W, integrated Radeon
  HD 7660D graphics
• AMD A10-5700
  processor – 3.4 GHz, 32nm, 65W, integrated Radeon HD 7660D graphics
• AMD
  A8-5600K processor – 3.6 GHz, 32nm, 100W, integrated Radeon
  HD 7560D graphics
• ASUS F2A85-M Pro motherboard
  – A85 chipset

Intel LGA1155 Platform:

• Intel Core i7-3770K processor – 3.5 GHz, 22nm, 77W, integrated HD 4000 graphics
• Intel Core i5-2500K
  processor – 3.3 GHz, 32nm, 95W, integrated HD 3000 graphics
• Intel Core i3-2100
  processor – 3.1 GHz, 32nm, 65W
• ASUS P8P67 motherboard
  – P67 chipset
• Intel DZ77GA-70K motherboard – Z77 chipset

Intel LGA1150 Platform:

• Intel Core i7-4770K processor – 3.5 GHz, 22nm, 84W, integrated HD 4600 graphics
• Intel DZ87KLT-75K motherboard – Z87 chipset

Discrete GPUs Compared: (using our 2012 GPU test system)

• HIS Radeon HD 5550 Silence – 550 MHz, 512MB GDDR5 @ 1000 MHz
• AMD Radeon HD 5570
  – 650 MHz, 1GB GDDR3 @ 900 MHz
• AMD Radeon HD 6570
  – 650 MHz, 512MB GDDR5 @ 1000 MHz
• ASUS GeForce GT 430 – 700 MHz, 1GB GDDR3 @ 8000 MHz
• Zotac GeForce GT 640 ZONE Edition – 902 MHz, 2GB GDDR3 @ 891 MHz

Measurement and Analysis Tools

• CPU-Z
  to monitor CPU frequency and voltage.
• SpeedFan
  to monitor CPU temperatures.
• Real Temp
  to monitor CPU temperatures.
• CPUBurn
  CPU stress software.
• Prime95
  CPU stress software.
• FurMark
  GPU stress software.
• Media Player
  Classic – Home Cinema to play H.264/VC-1 video.
• Cyberlink
  PowerDVD to play H.264/VC-1 video.
• Mozilla
  Firefox with Adobe
  Flash Player to play Flash video.
• Adobe
  Photoshop as an image manipulation benchmark.
• Eset NOD32 as
  an anti-virus benchmark.
• WinRAR as an
  archiving benchmark.
• iTunes
  an audio encoding benchmark.
• TMPGEnc
  Xpress as a video encoding benchmark.
• HandBrake as a
  video encoding benchmark
• Seasonic
  Power Angel AC power meter, used to measure the power consumption
  of the system.
• Custom-built, four-channel variable DC power supply, used to power
  and regulate the CPU fan.

Timed Benchmark Test Details

• Photoshop: Image manipulation using a variety of filters, a derivation
  of Driver Heaven’s Photoshop
  Benchmark V3 (test image resized to 4500×3499).
• NOD32: In-depth virus scan of a folder containing 32 files of varying
  size with many RAR and ZIP archives.
• WinRAR: Archive creation with a folder containing 68 files of varying
  size (less than 50MB).
• iTunes: Conversion of an MP3 file to AAC.
• TMPGEnc: Encoding a XVID AVI file with VC-1.
• HandBrake: Encoding a XVID AVI file with H.264.

3D Performance Benchmarks

• 3DMark11
  DirectX 11 benchmark.
• Unigine Heaven 3.0 DirectX 11 benchmark.
• Lost Planet 2 standalone benchmark, Test “A”.
• Aliens vs. Predator standalone benchmark.
• Crysis demo standalone benchmark.
• Sniper Elite V2 standalone benchmark.

Video Test Suite

Our main test procedure is a series of both CPU (timed tests of real-world applications) and GPU-centric (gaming tests and synthetics) benchmarks. System power consumption is measured during the CPU tests (an average of the first 10~15 seconds) and in various states including idle, H.264 and Flash playback and full CPU and GPU load using Prime95/CPUBurn and FurMark.

Certain services and features like Superfetch and System Restore are disabled
to prevent them from affecting our results. Aero glass is left enabled if supported.
We also make note if energy saving features like Cool’n’Quiet and SpeedStep
do not function properly.

INTEGRATED GRAPHICS TESTING

Our first set of tests focuses on the integrated graphics. Each CPU/APU and motherboard combination was equipped with 4GB of RAM, a 500GB notebook hard drive and a Blu-ray drive.

IGP Energy Efficiency

On lighter loads, the A10-6800K and A10-6700 are slightly more efficient than their Trinity analogs, the A10-5800K and A10-5700. When playing video, Haswell was still king by a healthy margin.

On heavier loads, the A10-6800K used about the same amount of power as the
A10-5800K, so even a small performance boost would make it a more efficient
chip overall. On full synthetic CPU load, the 6800K used only 5W more than the
i7-4770K Haswell chip, but when it came to actual work in the form of video
encoding with TMPGEnc, both the Haswell and Ivy Bridge processors pulled way
ahead.

The A10-6700 exhibited much lower power consumption than its big brother, and in fact, the total draw actually decreased when we ran FurMark on top of Prime95 to stress the GPU and CPU simultaneously. We also observed that the power draw gradually fell off from when we started each test to when it ended by as much as 20W. This was caused by the CPU cores downclocking even though temperature was fine (15°C less than the 6800K). During the Prime95 + FurMark test, the CPU clock speed dipped to just 2.3 GHz. It seems that either the APU or motherboard throttled the clock speeds to artificially keep it within its 65W TDP specification.

There are no motherboard settings pertaining to thermal/power limits as there are on most Intel motherboards but we will test the 6700 on a different board in the near future to see if we can shed some light as to what exactly is responsible. We should also note that despite this issue all the performance numbers we generated was on par with our expectations.

IGP Performance

Note: Discrete GPUs were tested on our GPU testing platform which uses a
Core i3-2100, though CPU scaling shouldn’t be an issue given the relatively low level of GPU performance of the chips compared.

The new Radeon HD 8670D graphics controller is a minor upgrade over Trinity’s HD 7660D, but it’s enough to make it the fastest integrated chip we’ve tested. In our benchmarks, the A10-6800K and A10-6700 caught up to the GeForce GT 640 and Radeon HD 6570 in a couple of instances but generally both discrete cards were superior performers by a comfortable margin. It’s basically the equivalent of a US$50 graphics card, a budget model that’s limited to lower resolutions or less demanding titles if comfortable framerates are to be had.

CPU TESTING

Our CPU testing is conducted with a discrete graphics card (a GeForce 9400 GT) to eliminate integrated graphics as a variable, most notably with regards to power consumption. It’s also necessary for fairly comparing CPUs that do not have an onboard graphics such as Bulldozer and Sandy Bridge Extreme models.

CPU Performance

The A10-6800K and A10-6700’s higher clock speeds give them an obvious advantage over the A10-5800K and A10-5700. The performance gains were modest but what was most impressive was that the power management improvements appeared to be working. During our benchmarks, the difference in power consumption was usually very minor and often in Richland’s favor.

Overall, the 6800K and 6700 were midrange performers, a big step down from Intel’s quad core Sandy/Ivy Bridge parts, but a compelling alternative to AMD’s Bulldozer FX line. Richland’s high clock speeds gave it superior single-threaded performance while the Bulldozer chips only excelled at multithreaded workloads and were power hungry to boot.

CPU Energy Efficiency

If you run on discrete graphics, the A10-6800K and A10-6700 are more power efficient than competing processors near their US$150 price point. On heavy load, they’re roughly on par with the 5800K/5700 and the 6700 comes close to a quad core Sandy Bridge CPU.

For users with balanced workloads, we’ve determined what we call the “average power consumption” which assumes the system is used half the time for light load activities (an average of idle and H.264 playback) and the remaining half for heavy load (an average of the power consumption used running our six benchmarks). We believe this is a very common usage pattern for an average PC — they are often left on for long periods of time, doing little to no work.

In this scenario, Richland is an improvement over Trinity by a couple of watts.

For users with heavy workloads, the total power consumed while running our benchmark suite is of pertinent interest. The total power takes into account the energy efficiency of each CPU while running our benchmark tests as well as how quickly they complete each task. This simulates the power draw of a machine that is purely for doing work and shuts down when its job is finished.

The A10-6800K and A10-6700 finished our benchmarks quicker than their Trinity counterparts and did so using a tad less power. While nothing AMD produces comes close to Intel yet, Richland brings them a little bit closer.

CPU Performance Analysis

We arrived at our overall performance figures by giving each CPU a proportional
score in each real world benchmark with each test having an equal weighting.
The scale has been adjusted so that the A10-6800K is the reference point with
a score of 100.

Overall, all Richland delivered an average improvement of ~11% over Trinity in our CPU performance tests. Despite being short two cores, the Core i3-2100 has been shaming AMD’s budget quad core chips in our benchmark suite for some time thanks to its superb single-threaded performance — the A10-6800K is the first to finally overtake it.

These two new Richland APUs are clearly superior to their predecessors but they are also priced higher, so the overall value they offer is about the same. Of the processors compared today, only the lowly A8-5600K, an older Trinity-based US$100 APU, actually offers significant bang-for-your-buck CPU performance.

To determine performance per watt, we divided the overall performance score
by the average power consumption calculated earlier and again re-scaled with the A10-6800K
as our reference.

The faster, slightly more power efficient Richland APUs offer a sizable edge over Trinity in performance per watt. However, if electricity costs weigh heavy on your mind, Intel clearly prevails. More than two year since its release, Intel’s Sandy Bridge architecture remains a paragon of frugality in this regard.

FINAL THOUGHTS

With some tinkering on the Trinity
architecture, AMD has successfully squeezed out a nice performance boost
while slightly improving power consumption. Richland has higher clock speeds
but doesn’t use any more energy, resulting in a more efficient APU. The graphics
side of their product has also been buffed but to a smaller extent. Overall,
the bump in speed is similar to the move from Llano to Trinity, only this time
it doesn’t require a socket change. That being said, this is the final update
to FM2 platform before AMD launches their new FM2+ socket and APU architecture,
code-named Kaveri, later this year according to most sources.

Everything about Richland is better but it’s a collection of minor upgrades — there is no single impactful improvement or feature that truly makes it stand out against Trinity. In day-to-day operation, you’d be hard-pressed to distinguish between an A10-5800K and an A10-6800K. Furthermore, most of the new Richland line carry a bit of a price premium, so they generally are not better values than their predecessors. The only 6000 series chip that clearly offers a better bang-for-your-buck is the A8-6600K, which is faster by 100 MHz in base clock speed compared to the A10-5800K and carries a US$5 discount.

Trinity owners have no real incentive to upgrade to Richland, and for those
looking into a budget system, the decision between getting an old or new generation
APU may depend ultimately on how much money is budgeted for the processor. Complicating
matters somewhat is upgradeability or rather the lack thereof. After a year,
FM2 is on its way out, just like FM1 before it. (Editor’s Note: Keep
in mind that a CPU upgrade is much less attractive than it used to be —
compared to, say, a decade ago — even for PC enthusiasts. With current
hardware, there usually isn’t enough bang-for-the-buck in a CPU-only upgrade.)
On the Intel side, their dual Core i3 chips are still surprisingly competitive,
boasting better single-threaded performance and energy efficiency. Richland
and Trinity have the edge in graphics and the quad core models have a sizable
advantage in multi-threaded workloads. Which choice is better for any single
user depends on usage.

Our thanks to AMD for the A10-6800K and A10-6700 samples used in this review.

* * *

Articles of Related Interest
Intel Core i7-4770K Haswell Processor
AMD FX-8350 CPU: Piledriver Arrive
AMD A10-5700 APU: Trinity at 65W
AMD Trinity: A10-5800K & A8-5600K 2nd Gen APUs
Intel Core i7-3770 Ivy Bridge CPU
Intel Sandy Bridge Extreme: Core i7-3960X LGA2011 Processor

* * *

Discuss
this article in the SPCR forums.