AMD Trinity A10-5800K APU Preview

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

The AMD Trinity A10-5800K APU is set to hit the market in the near future. To that end, we are able to share a limited amount of details at this point, but we thought a teaser preview might pique your interest. This 2nd generation AMD-A series APU boasts an unlocked quad core which comes in at a lofty max turbo frequency of 4.2 GHz and a base clock of 3.8 GHz.

The new Piledriver x86 core technology will have several enhancements over the previous Bulldozer APUs, and will include the following changes:

  • Supports up to 4 cores and support for the latest ISA instructions including FMA4/3, AVX, AES, XOP
  • Branch Prediction and Cache enhancements over the previous “Bulldozer” cores
  • 2 MB L2 cache per dual-core module (up to a total of 4 MB)
  • Max Turbo Frequencies up to 4.2 GHz
  • Configurable via AMD OverDrive

Much like Bulldozer, a Trinity ‘module’ is two integer cores and one floating point core. So “quad core” means that there are four integer cores, but there are only two floating point cores.

On the integrated graphics front, the A10-5800K will feature the HD 7660D graphics processor. The GPU clock speed is set at 800 MHz and includes 384 Radeon cores (Shaders). AMD A-Series APUs natively support AMD Eyefinity Technology, for up to four monitors.  Additionally, by adding a discrete graphics card to the fold, the dual graphics support can deliver a performance boost of up to 75 percent.

New GPU Core Basic Specifications:

  • Featuring VLIW 4 Architecture
  • Up to 384 Shaders
  • Up to 800 MHz
  • Up to 8xAA and 16AF Support
  • Controllable Via AMD OverDrive
  • DirectX 11 Support


As is customary here at Overclockers, we’ll have a full review ready for launch day, so stay tuned for that! It’s still unclear when the Piledriver CPUs will be released, but when we dive into the Trinity A10 APU we should get our first glimpse of the performance that AMD’s future Piledriver architecture will offer.

-Dino DeCesari (Lvcoyote)


Leave a Reply

Your email address will not be published.


  1. This version is desktop.
    There is an interesting take on today's 'preview' NDA at TechReport. I think it's possibly a little much, but he does make some great points. TBH, we don't have benchmarks today because I was out of town and unable to review this one, so we had to ship the hardware over to Lvcoyote. Now that I've read that, I'm kind-of glad it worked out that way.
    Out of that piece, I'd like to stress one important thing:
    We don't guarantee a positive review; we don't agree to mention certain product features; and we certainly don't offer any power over the words we write or the results we choose to publish. In fact, by policy, these companies only get to see our reviews of their products when you do, not before.

    I'm not sure how many sites do such things, but let me stress that Overclockers has never done any of that. We will show features the manufacturer requests occasionally, by which I mean we'll say "ManufacturerX wants to point out Y", which you've seen in my reviews from time to time. We have never and will never extol features our audience finds useless as great just because the manufacturer says to. Even so, we either list or link to specification lists in pretty much ever review, so that's pretty moot.
    Anyway, I saw that and thought it was interesting.
    Hmm, not sure what the issue is about AMD saying what can be shown in the preview. Sounds reasonable to me. The email TRP shows says they could use any combination of allowed materials not that they were mandatory.
    Am I wrong?
    Yes, that's correct; but the point he is making is that AMD is controlling the launch by only allowing sites to publish GPU results. Obviously Trinity's GPU is strong and should (in theory) beat the pants off Ivy Bridge's GPU (while, in theory, costing less...we can't share that information at this point). AMD should very well be proud of that, they've done an amazing job with on-die GPUs.
    The two-module CPU portion, however, won't exactly set any records. So they're literally putting their best foot forward...and holding back the negative. We all know initial launch articles get the highest number of views. The follow-up articles (when they come out, which we can not say but it's not too far away) won't get near the eyes the ones today will. Thus it sort-of equates to sweeping the less-than-stellar x86 performance under the rug.
    That makes sense when put like that, but any websites that are that concerned about it aren't obligated to do that. Clever of AMD, but not too surprising since they are killing it in GPU's and barely getting by in the x86 side of things. They are a business after all, can't expect them to not try and find a way to look as good as they can.
    Guess it now becomes the review sites job to ensure they produce unbiased reviews. Regardless of what companys try and do.
    whaaaat a 7660 on chip?! i see amd slowly making a comeback if by next gen of these they can have some x750 or x770 on chip, they will be kicking some buttox:clap:
    The 7660 is a vliw4 gpu right? So, a variant of the 6000 gpu's. Is it confirmed that these have a new socket? Fm2?
    Edit: now that I think about it, I think I saw pictures of some of the new boards. Are they backwards compatible with fm1?
    It looks like AMD want the review restricted to GPU benchmarks until it's released, at which point it's ok to review CPU x86 performance.
    That seems perfectly reasonable to me.
    There not trying to stop it's CPU performance from being tested full stop, they just want it to be taken on merit for the market it's aiming at.
    It's not intended for x86 performance, it's an APU. (an IGPU with on DIE GPU Hardware acceleration)
    They don't want it's x86 CPU performance to dominate the headlines when that's not what it is.
    I'm fine with that, but still interested in there x86 performance :)
    The Vishera chips are intended for x86 performance. those are also what i'm most interested in. (for x86 reasons)
    Really looking forward to this review as it'll give some idea of the discrete desktop piledrivers due out soon. My 1090T feels a little long in the tooth after two years in my PC :)