Comments Locked

36 Comments

Back to Article

  • gevorg - Tuesday, December 20, 2011 - link

    What an amazing naming convention! Good job AMD! The A8-3870K must be better than upcoming i7-3770K from Intel.
  • Ammaross - Tuesday, December 20, 2011 - link

    Troll much?

    At $130, it will in no way compete against an i7-3xxx chip in price bracket nor performance. I'm just glad they're making sensible number bumps and using a "common" naming convention (the "K" for unlocked). Although BE wasn't too confusing either.
  • euler007 - Tuesday, December 20, 2011 - link

    That's not a troll.

    Back in the Athlon days AMD started naming their CPUs using number to indicate that these chips were comparable (faster actually) with a pentium 4 of that frequency. The Athlon XP 2000+ was faster than a Pentium 4 2.0 MHz but only ran at 1.67 GHz.

    By matching the numbering scheme again (and using the K suffix for unlocked cpus) they are matching their previous strategy, except this time their product is of lower performance then the intel counterpart and they increment the number up.

    Say what you will, but it is an attempt at deception.
  • BSMonitor - Tuesday, December 20, 2011 - link

    On AMD's selected/preferred benchmarks, the XP####+ naming scheme worked. When SSE and newer Apps were involved, the XP showed it's age.
  • backy51 - Tuesday, December 20, 2011 - link

    When AMD released the Athlon64 chips they completely pwnd the P4's which forced Intel out of its complacency and led them to design the Core2Duo which was based on the PIII/Pentium M line.
  • JonnyDough - Tuesday, December 20, 2011 - link

    Yep. First to have on-die memory controller. People forget.
  • Qefxx - Wednesday, December 21, 2011 - link

    Actually IIRC Alpha and HP (PA) had them 1st. AMD was first to the x86 party only.
  • damianrobertjones - Tuesday, December 20, 2011 - link

    I though they released the Pentium D... then the Core duo... THEN core 2 duo?
  • jkostans - Tuesday, December 20, 2011 - link

    Nope. Pentium 4 and D were the same architecture (Netburst), Core 2 Solo/Duo/Quad/Extreme was the next step. There was no Core duo architecture....
  • Klinky1984 - Tuesday, December 20, 2011 - link

    There was a Core Duo, it was a mobile part & never released on desktops.

    Technically the Core lineage started with the Pentium M/Centrino.
  • backy51 - Wednesday, December 21, 2011 - link

    And the Pentium M came from PIII - didn't get anything from the P4 line that I'm aware of.
  • Operandi - Tuesday, December 20, 2011 - link

    From the model number itself to the adoption of the “K” designation for unlocked chips (what exactly was wrong with “Black Edition”?), to the bringing back of the FX name only to have it trashed with a subpar chip. AMD market is absolutely horrible at least in the CPU division, I thought they fired all these idiots?, seriously WTF is going on here?
  • JonnyDough - Tuesday, December 20, 2011 - link

    That's definitely trolling. It isn't saying anything constructive at all. It's just being an anti-AMD Intel fanboy. Truth be told, an all in one CPU and GPU for $135 that is capable of playing any game over five years old is amazing.
  • CeriseCogburn - Saturday, May 26, 2012 - link

    Way to go, total hypocrisy from an amd fan.
    I suspect you posted in anger, disgust, and disbelief over nVidia rebranding for like 5 years straight.

    Doesn't matter if you did or not, own the evil, deceptive, PR fool the public amd lying trickster naming convention. Own it - feel it - let the evil flow through you...
  • StevoLincolnite - Tuesday, December 20, 2011 - link

    No. The PR rating that AMD re-introduced with the Athlon XP wasn't a performance metric that could be compared with the Pentium 4.

    It was actually a benchmark to compare it to the Thunderbird Athlons, the older Athlons without the PR denominator.
    It just so happened that Intel's performance with the Pentium 4 was simply bad, it wasn't until the 3000+ Athlon XP's that the naming scheme wasn't comparable to the Pentium 4 as the IPC/clock speed improvements just wasn't cutting it when compared to Netburst.

    Then... We got the Athlon 64 which left feces in the competitions face.

    The number rating that was used with the K6-2 was a different kettle of fish to as it wasn't a metric used to compare the newer chips with the older ones.
  • silverblue - Wednesday, December 21, 2011 - link

    I thought the K6-2 had actual MHz ratings?
  • chizow - Tuesday, December 20, 2011 - link

    Yeah, funny there's 25 other letters in the alphabet and they settle on K. If AMD wants to stop playing "follow the leader" they could start by not mimicking Intel's naming schema verbatim.
  • JarredWalton - Tuesday, December 20, 2011 - link

    I don't mind the "K" for unlocked being used, and given we had the A8-3850, A8-3870K makes complete sense. Remember that the A8-3850 already existed prior to the announcement of the i7-3770K. Anyway, if someone is so uninformed as to buy an AMD A8-3870K and think it's better than an i7-3770K, they get exactly what they deserve. "Hmm... why is this PC one fourth the price of this other PC when the CPU appears to be a higher model number?"
  • SlyNine - Tuesday, December 20, 2011 - link

    Yea, I always wondered who that wasn't in the "know" actaully cared about the model number, or even bothered to look enough to know the other one had the same model number.

    Unless it was a computer at best buy sitting right next to it. But then the price should give them a clue. But honestly How many of those people would even benifit from spending more.
  • N4g4rok - Tuesday, December 20, 2011 - link

    Doesn't really seem like a bad idea to have a understood standard for unlocked processors either.
  • chizow - Wednesday, December 21, 2011 - link

    Yeah I doubt McDowell's had a problem using the "Golden Arcs" either. AMD just comes off looking like a cheap knock-off every time they try copying Intel's naming scheme.....this classic bit comes to mind:

    http://www.youtube.com/watch?v=CGRiTpCji0I
  • Belard - Tuesday, December 20, 2011 - link

    The DUMBEST thing about the AMD A-Series K chips is that... ONLY computer geeks/techies know the difference between a 2500 and a 2500k (and so on)...

    Those SAME people *KNOW* the difference between an AMD-A and FX and how it compares to the intel i5.

    Yeah, none of the As are in comparison to the i5... and from the looks of things, an overclocked A chip may exceed the performance of an FX.
  • CeriseCogburn - Saturday, May 26, 2012 - link

    LOL - a8 3870 beats i7 3770 !

    Once again amd relies on deceipt and the hope that it's customers are dumb as a rock.

    It must work for them, right, because we were told for years that nVidia renamed and rebranded to fool the public, and the amd fans couldn't repeat it enough, angrily telling us that's why nVidia had such high sales. I suspect amd won't achieve the high sales part.

    What say you amd fans, is amd evil ?
    Since "you're trolling" and silent crickets is the response, we know none of you have any integrity and were not about fairness or defense of the public.

    Pure hypocrisy.
  • fic2 - Tuesday, December 20, 2011 - link

    Ok, so for $135 I can either get an locked part running at 2.9GHz or unlocked part running at 3.0GHz? Why not just take the 3650 and 3850 off the market. Seems rather stupid to have both.
  • Pinin - Tuesday, December 20, 2011 - link

    Not for me. Only will consider it when they fix the OpenGL bug that makes it use the worse GPU's in the system.
  • BSMonitor - Tuesday, December 20, 2011 - link

    This is exactly what these should have been from the start. AMD A8-3870K version that is.

    Now, a system builder can lower the maximum CPU clock and increase the GPU clock. Or vice versa depending on the use.

    Why on earth AMD increased CPU and GPU power as the models increased blows me away. Anyone who wants more CPU power is not going to be satisfied with anything the iGPU can provide. So increasing it's capability is really wasted. But increasing both is likely going to hit instability faster.

    Here's an idea, lower the CPU perfomance, as light gaming rarely needs 4 CPUs clocked insanely high, and use that extra thermal room on the GPU clock!!! Wow. Brilliant!

    Now if only they'd let you shut off 1-2 CPU cores or 200-300 GPU cores as needed as well.

    AMD, if you want to stand out, make a single SKU that can ramped up or down based on the consumers need! Make the ultimate OEM friendly APU.
  • Arnulf - Tuesday, December 20, 2011 - link

    "Here's an idea, lower the CPU perfomance, as light gaming rarely needs 4 CPUs clocked insanely high, and use that extra thermal room on the GPU clock!!! Wow. Brilliant!

    Now if only they'd let you shut off 1-2 CPU cores or 200-300 GPU cores as needed as well."

    Back when Llano came out I would have preferred it to come in a 2 or 3 core + faster GPU flavor. If this could be managed in software on the fly, even better. There are plenty of games that don't even run two threads.
  • JonnyDough - Tuesday, December 20, 2011 - link

    The solution we need is automatically adjusting the frequencies based on the application load. AMD knows this I am sure. Intel does too, and is headed in that direction with their turbo as well. Automatic removal of the bottleneck is an awesome feature to have. Every application would ideally run at the fastest speed the system can provide.
  • tech6 - Tuesday, December 20, 2011 - link

    Given all the bulldozer fiasco and the lukewarm reception to the Fusion parts maybe they should just unlock everything.
  • Taft12 - Tuesday, December 20, 2011 - link

    Of note (to me at least) is that the A6-3600 and A8-3800 seem to be in stockat Amazon. It's the first time I've seen the 65W versions of the Llano desktop quad-cores available anywhere besides an OEM HP or Acer system

    Also somewhat amusing is Amazon's A8-3870's price:
    List Price: $149.99
    Your Price: $160.02
  • fashionbook - Tuesday, December 20, 2011 - link

    fdsafdsf
  • fashionbook - Tuesday, December 20, 2011 - link



    Our Website: ===== www fashion-long-4biz com ====
    Our main product list is as follows:
  • Soulkeeper - Wednesday, December 21, 2011 - link

    "partially unlocked" <-- pisses me off
  • Jambe - Saturday, December 24, 2011 - link

    You'll let us know how they OC, right?

    =)
  • geok1ng - Monday, December 26, 2011 - link

    AMd is almost there now all we need is a dual core CPu with 400 cores on the iGPU, that can reach 2.5Ghz+ on both cores , consumes no mnore than 40w at load and idle below 10w. Did i mention that the 400 cores must reach 500Mhz at load?

    Llano is all about great IGP with fast enough CPU.

    We dont need more CPU cores , we need faster cores.

    We need the optimal balance of frequency and core count on the IGP side also. Llano has as much cores as a 6670, but at less than half frequency, performance suffers a lot.

    AMD should focus on optimazing core count/ frequency on these parts. Why would anyone use 4 cores at 1.6 Ghz? Give us 2 cores at 3Ghz. Heck, even 2 cores at 2.5Ghz would be enough, if the IGP could reach 400cores@600Mhz.

    And lets not forget that Fusion and Llanno used pretty high voltages on the first batchs, that could operate normally at 25% less power using programs to hack the voltage controller.

    These are parts begging to be unlocked, undervolted and overclocked.
  • Tibbz - Tuesday, December 27, 2011 - link

    Dual cores can get hindered by a lot of new games and software's, but there is still little difference between a X4 and X3. If AMD would make a 2.8 ish (give or take) tri core CPU with a 6770 or on the new TBA 7770, it would be a killer entry level system.

Log in

Don't have an account? Sign up now