AMD A10-6700 And A10-6800K Review: Richland Hits The Desktop

Richland, code name for AMD's highest-end APUs, finds its way into our lab in the form of A10-6700 and A10-6800K. Based on the Piledriver architecture and VLIW4 graphics, these chips are slight improvements to Trinity. Can they outshine Core i3, though?

AMD's Kabini: Jaguar And GCN Come Together In A 15 W APU showed us what the company's Jaguar and GCN architectures could accomplish between 4 and 25 W TDPs. But, on the desktop, AMD isn't quite ready to make the leap to a next-gen design. It just introduced its desktop-oriented Richland APUs, which aren't really new at all. Rather, you can think of them as power-optimized Trinity parts, sporting the same Piledriver-based x86 cores and VLIW4 graphics configuration. Moreover, Richland-based APUs have been available to mobile device makers for months. The only real revelation is that we're getting this update in the desktop and low-voltage mobile spaces now.

Model
Radeon
Package
TDP
CPU Cores
Base/Max CPU Clock
L2 Cache
Radeon Cores
Base GPU Clock
A-Series Low-Voltage and Ultra Low-Voltage APUs
A10-5745MHD 8610G
FP2
25 W
4
2.1/2.9 GHz
4 MB
384
533 MHz
A8-5545MHD 8510G
FP219 W
4
1.7/2.7 GHz
4 MB
384
450 MHz
A6-5345MHD 8410G
FP217 W
2
2.2/2.8 GHz
1 MB
192
450 MHz
A4-5145M
HD 8310G
FP217 W
2
2.0/2.6 GHz
1 MB
128
424 MHz

In the table above, we see the new mobile-oriented options spanning 17 to 25 W TDPs. Richland isn’t much different from Trinity, but it's more efficient thanks to specific Turbo Core optimizations that include a greater number of P-states to facilitate more granular power and performance levels.

Model
Radeon
TDP
CPU Cores
Base/Max CPU ClockTotal Cache
Radeon Cores
GPU Clock
Unlock
Price
A10-6800K
HD 8670D
100 W
4
4.1/4.4 GHz
4 MB
384
844 MHz
Yes
$149
A10-6700
HD 8670D
65 W
4
3.7/4.3 GHz
4 MB
384
844 MHzNo
$149
A8-6600K
HD 8570D
100 W
4
3.9/4.2 GHz
4 MB
256
844 MHzYes$119
A8-6500
HD 8570D
65 W
4
3.5/4.1 GHz
4 MB
256
800 MHz
No
$119
A6-6400K
HD 8470D
65 W
2
3.9/4.1 GHz
1 MB
192
800 MHz
Yes$77

And then we have the desktop-specific Richland parts. The very fastest model enjoys a 300 MHz base clock rate bump compared to the A10-5800K, along with official support for 2133 MT/s DDR3 memory (the other SKUs top out at 1866 MT/s memory). Also, its GPU is 44 MHz faster than the prior-gen version. And yet it fits within the same 100 W TDP.

On the other hand, the A10-6700 looks a lot like the -5800K, aside from a 100 MHz-lower base clock, a 100 MHz-higher Turbo Core ceiling, and a slightly quicker GPU. That one drops to a 65 W thermal limit.

Like the Trinity-based APUs before them, these Richland designs plug into a Socket FM2 interface. A BIOS update should be all that you need for compatibility with existing A55, A75, and A85 platforms.

And what about those shiny new Radeon model numbers? That's marketing being bad. We were already dealing with Radeon HD 7000-series naming on APUs, which was confusing because AMD's desktop 7000-series GPUs are commonly associated with the GCN architecture. Now we have 8000-series nomenclature. And yet, we're still working with the VLIW4 configuration that was around back when AMD was shipping Radeon HD 6900-series cards. Architecturally, the only difference between Richland's Radeon HD 8000 graphics and Trinity's Radeon HD 7000 graphics is the name.

Again, the top-end A10-6800K gets 2133 MT/s memory support, which is particularly meaningful for its on-die graphics engine given a lack of shared L3 cache. The theoretical 34 GB/s of DDR3 bandwidth should go a long way to improve frame rates in the games we'll be testing. With that said, let's move on to why we aren't able to test Richland's Dual Graphics feature today...

This thread is closed for comments
97 comments
    Your comment
  • Novuake
    VLIW4... Are you kidding me??? WHen is that gonna be scrapped...
  • esrever
    Numbering has never been for the architecture inside the chips. They have always been marketing for the current line of products. Normal people buy a laptop, they see the number, they get the performance associated with said number and they don't care about the inside of the chip. Richland is just a trinity refresh with better power management and higher clocks.
  • griptwister
    AMD, Good show! Now, release Kaveri Already! I need a new SteamRoller CPU!
  • tului
    They say it's VLIW4 then say it's GCN. They can't be both can they?
  • songorocosongo
    The 6800k is good, but I expected it to be a little more powerful.
  • Yuka
    No temps with power metrics page? 8(
    I guess Richland is still very hot going by the power figures alone. Still, it's a good step up (and stop gap) for AMD.
    Nice review still. Are you guys planning on a follow up for Dual Graphics? 8)
    Cheers!
  • kfronda26
    This is dumb. Still 6670 max card for dual graphics. Disappointing.
  • Wes Young
    1302017 said:
    This is dumb. Still 6670 max card for dual graphics. Disappointing.


    No, since driver 13.1 even the 5800k was able to run dual graphics with a HD 7750. I am typing on a system with that exact setup right now. I am not sure if the 6800k will allow anything above the 7750 though. When I tried a 7770 with the 5800k I wasn't given the option to enable dual graphics.
  • unknown9122
    AMD swings and misses once again.
  • unknown9122
    AMD swings and misses once again.
  • slomo4sho
    Couple days late but thanks for the review.

    Richland does appear to be just a refinement of Trinity. This review does explain why Kaveri desktop APUs are due to release at the end of the year.
    Quote:
    The A10-6700 offers similar performance as a stock A10-5800K, but offers greater efficiency. Unfortunately, it's also multiplier-locked. I have to believe that if you're willing to spend $150 on a 65 W A10 that can't be overclocked easily, then you're probably better off with a 55 W Core i3 that's also stuck in place for $10 less.

    I don't see why this argument is even made since both the 6700 and 6800K have the same MSRP considering that the only real difference in power consumption observed between the two chips was in gaming...

    Quote:
    The Intel chip's performance in single-threaded apps is exceptional. It holds its own in more parallelized workloads.

    Your final graph suggests that the overall performance of the i3 is within margin of error of the A8-6800K(for which you didn't even bother to provide overclock benchmarks)

    Lastly, can you confirm the MSRP? Your values seem to differ, other sources suggest $142 instead of your stated $149 (in fact, the pricing on all of the models seems to be off)
  • opmopadop
    Can someone review this chip with the APU turned off and CPU cranked with liquid cooling (or crazy air). That higher RAM timing has to do something positive... I hope.
  • ingtar33
    so... you tested a chip that supports ddr3 2400 ram with ddr3 2133 ram... and no overclocking data? the A10-5800k already supported 2133 (maybe not officially, but it worked), the a10-6800k supports and works with ddr3 2400 ram, as other reviewers have noted elsewhere.

    Isn't this a tech enthusiast site? The few reviews I've seen out there claim Richland overclocks better, cooler and higher then Trinity. Furthermore, they claim the overclocked igpu performs at the level of a 6670... which is a huge jump in performance... as the 5800k, even overclocked and with fast ram, was only about 70% a HD 6670.

    where is the beef? Seriously i expected more from this site.
  • Nintendo Maniac 64
    Two things. First, I really think you should have thrown in a lower-end quad Intel CPU as well. From reading older reviews it seems that an i5-2300 is actually a second SLOWER than the A10-6800k in the TotalCode Studio test.
    Secondly, the comment that Kaveri will require a new socket is largely unknown at this point - all that's been revealed is that it uses an "FM2+" socket - who knows what that'll mean in terms of mobo socket compatibility.
  • envy14tpe
    I still feel like the A10-x800 line is designed for only adding RAM, not a GPU. If you add a GPU then performance parallels (or falls behind in games) a i3.
  • aibenq
    why you not tested with HD7750... (/゚Д゚)/
  • ohyouknow
    So this power consumption chart. Does that include the i3 with a 6670? or is that straight cpu vs apu? The gaming power consumption chart feels a bit wrong if it is running the i3 HD in metro w/o the gpu as that is misleading. Clarification?
  • rmpumper
    Why not include G2020+HD6670 benchmarks? It would be ~the same cost as 6800K and I bet that everyone would want to know which setup is better.
  • abitoms
    @ingtar33, does Richland chips officially support official DDR3-2400 RAM? If so, Toms should have tested with that RAM. However I think Richland supports 'just' 2133 MHz so that's what Toms has tested with.
    @Don, I have 3 queries for you.
    Does Richland crossfire with 7750 (officially or unofficially)? If it does, that is sure an interesting thing to explore.
    Also Don, yes even I've seen the figure of $142 floating around in other tech sites. Why do you quote it otherwise?
    Lastly, you 'could' have added the Core i3 with Radeon 6670 in the power consumption charts just to show the increased performance in games and other applications comes with the added power consumption of the 6670 if it is present.
  • slomo4sho
    235831 said:
    Why not include G2020+HD6670 benchmarks? It would be ~the same cost as 6800K and I bet that everyone would want to know which setup is better.


    Is this necessary? The gaming benchmarks already show the 6800K paired with 2133 ram to perform equal to the i3-3220 + HD 6670.

    365749 said:
    does Richland chips officially support official DDR3-2400 RAM? If so, Toms should have tested with that RAM. However I think Richland supports 'just' 2133 MHz so that's what Toms has tested with.

    Richland officially only supports 2133 but 2400 seems to be compatible.
  • bustapr
    actually performs quite decently against the i3 on the cpu side. has pretty good gaming performace also as a standalone APU(for what it is). Its been said since richland was announced, richland is an improvement to trinity. kaveri is the real successor. just as steamroller has been hyped since before piledriver, kaveri has had hype since trinity.
  • sarinaide
    My first opinion is one of being happy, the x86 performance and OpenCL (HSA) has improved substantially along with impressive power gating which sports a higher clock rate at similar to 5800K power consumption. The other impressive feature and most APU owners like myself would attest to was that the Trinity's were poor overclockers, to get 4.2ghz stable needed a almost exponential boost in vCore voltage and while DDR3 2133 was operable to get DDR3 2400 you were never really stable and what Richland reviews have shown is that 5Ghz is very doable on conventional cooling and in our tests managed to sport DDR3 2600 albeit at a high NB voltage which is neither here or there for acceptable 24-7 operation.
    I was just surprised Don didn't at least do the Dual Graphics test, and use games like BF3 that actually scales well to crossfire and SLI, having run a 6670 in DG in BF3 there is notable gains and doing a review on a APU without reference to that feature is essentially skimping on details.
    Overall all the review shows is on x86 the APU is catching a i3 while showing its real value in entry level gaming perameters essentially showing tremendous price to performance being that it can out pace a i3 + discrete entry level card in demanding titles is testiment to just how impressive these are for low cost systems. It also follows Anandtech's review done on what systems represent value for money where the 5600K showed up tops in low cost discrete gaming systems. AMD is not as far behind as people are led to believe and in Desktop I think most will agree that the i3's power leverage is basically non issue then there is the factors for that;
    1) The i3 is lower clocked
    2) The i3 is a dual core
    3) The i3 has a mediocre iGPU
    Those factors alone attribute to the higher system load which in all honesty can run a full HTPC/Gaming system on 300w or less power supplies, the next factor as to why the A-series is more appealing is the FM2 platform offers high end features at a lowend cost without skimping on a B75/H61 iffy platform with almost no love for the end user. The final factor lies in the fact that the i3 needs a discrete card or its decimated at every level and I will say the HD4600 i3's will also be well beaten. In most books the APU is the undisputed king of budget and this is only enhanced by the fact that Richland still accross the board improved on Trinity with no microarchitecture changes.
  • ingtar33
    365749 said:
    @ingtar33, does Richland chips officially support official DDR3-2400 RAM? If so, Toms should have tested with that RAM. However I think Richland supports 'just' 2133 MHz so that's what Toms has tested with.


    Richland has been tested with and runs well with ddr3 2400 - 2600, Tom's has been testing the A10-5800k with ddr3 2133 ram, even though officially it's not supported. Because it will run with it, just fine. Richland not only runs with ddr3 2400, but it overclocks up to 4.8-5.1 ghz in every review i've seen that bothered to run an overclock test.

    the igpu will overclock up to 1.2ghz as well (which is a tremendous overclock over it's base 850mhz); the net result is the igpu will perform just like a HD 6670, when paired with ddr3 2400 ram. Which is nearly a 40% improvement over the a10-5800k.

    yet we don't get any such testing from tom's... i like to come here because generally i trust their methodology over pretty much everyone elses... yet nothing. I'm not disappointed in the review because it's poorly written (it isn't) but because it's beyond vanilla.

    they took a chip which had a massive efficiency bump in a refresh lineup (it is a refresh) and rather then investigating what the improved efficiency meant (higher clocks/lower voltage/lower heat/better overclocks), they simply confirmed it was a refresh by running it at stock and measuring the turbo performance vs the stock 5800k. That's flat out silly. They could have accomplished the same by locking them both at the same ghz and benching the cpu/gpu and turn up... surprise! the same number (which is exactly what would have happened)

    You don't need a 10 page review to prove it's simply a refresh.

    how about digging a little deeper into what it means? others have done it, and the results is this chip actually overclocks fairly well, and the igpu overclock, matched with the faster ram, seems to have resulted in a significant improvement in performance.
  • sarinaide
    1027081 said:
    365749 said:
    @ingtar33, does Richland chips officially support official DDR3-2400 RAM? If so, Toms should have tested with that RAM. However I think Richland supports 'just' 2133 MHz so that's what Toms has tested with.
    Richland has been tested with and runs well with ddr3 2400 - 2600, Tom's has been testing the A10-5800k with ddr3 2133 ram, even though officially it's not supported. Because it will run with it, just fine. Richland not only runs with ddr3 2400, but it overclocks up to 4.8-5.1 ghz in every review i've seen that bothered to run an overclock test. the igpu will overclock up to 1.2ghz as well (which is a tremendous overclock over it's base 850mhz); the net result is the igpu will perform just like a HD 6670, when paired with ddr3 2400 ram. Which is nearly a 40% improvement over the a10-5800k. yet we don't get any such testing from tom's... i like to come here because generally i trust their methodology over pretty much everyone elses... yet nothing. I'm not disappointed in the review because it's poorly written (it isn't) but because it's beyond vanilla. they took a chip which had a massive efficiency bump in a refresh lineup (it is a refresh) and rather then investigating what the improved efficiency meant (higher clocks/lower voltage/lower heat/better overclocks), they simply confirmed it was a refresh by running it at stock and measuring the turbo performance vs the stock 5800k. That's flat out silly. They could have accomplished the same by locking them both at the same ghz and benching the cpu/gpu and turn up... surprise! the same number (which is exactly what would have happened) You don't need a 10 page review to prove it's simply a refresh. how about digging a little deeper into what it means? others have done it, and the results is this chip actually overclocks fairly well, and the igpu overclock, matched with the faster ram, seems to have resulted in a significant improvement in performance.


    Ditto, the only thing toms can do is talk about efficiency knowing that it means didly squat on desktop, ie: nobody buys a i3 over a i5 unless money is a factor because the i5 gives markedly more gains ove the i3 and thats what matters despite the extra power. Power efficiency seems to be the fall back to defacto position in favor of Intel when in hindsight a desktop owner doesn't really care.

    Ditto on the overclocks, it will be nice when pro's push limits and bench so as to bring clarity on what is achieveable. Basically Don half sold this, either it was to tedious or for some reason its not interesting enough. Also why just not show Dual Graphics in a spate of games so as to see which games benefit and which don't.