Comments Locked

52 Comments

Back to Article

  • Chaitanya - Friday, January 11, 2019 - link

    Maybe like Threadripper on next gen AMD will add 2nd CPU chiplet. But before that I would really like to see AMD bring out 8Core APU based on Matisse.
  • DigitalFreak - Friday, January 11, 2019 - link

    If you're wanting an APU, you're not likely to need the extra speed of Zen2 over Zen+ anyway.
  • darkswordsman17 - Friday, January 11, 2019 - link

    Disagree, especially since the idea for this type of APU assumed a larger GPU (I think many assumed it would also include some HBM memory to help with bandwidth bottlenecks). At minimum you'd like the transistor and power characteristics of 7nm Zen 2. And if the I/O module would help with memory coherency (between the GPU and CPU) and latency.
  • neblogai - Friday, January 11, 2019 - link

    That is not true. High performance mobile needs it, and AMD needs it for gaming (and used with dGPU) if they want to compete vs Intel in gaming laptop market, and embedded would be very happy with a crazy-efficient, 8-core Zen2 APU.
  • teldar - Friday, January 11, 2019 - link

    Completely disagree. I would love to have a reasonably high end 8 core processor with a very reasonable (read not crappy) video unit built in so I can change the home computer to a SFF instead of a tower. I have a SFX case and PSU I bought for HTPC use and recently switched to Roku instead. I would love to make use of my SFX case. I'm not a computer gamer but want some video capacity. More than what the current APUs have. And I want more CPU as well.
    I would be very willing to pay the $300-400 for a good CPU and a good VPU in one chip. Particularly if it had some HBM on board.
  • Boxie - Friday, January 11, 2019 - link

    I personally would love a powerful SFF workstation
  • GreenReaper - Tuesday, April 16, 2019 - link

    You're likely to want the power savings from 7nm on an APU-based laptop - the most practical kind.
  • Teckk - Friday, January 11, 2019 - link

    If they do launch APUs as 3000 series processors it'll add to the confusion 🤔🤦🏻‍♂️
  • gavbon - Friday, January 11, 2019 - link

    They probably will!
  • edzieba - Friday, January 11, 2019 - link

    3000 series APUs will be Zen+, like the Ryzen 3000 series mobile APUs announced earlier (and same as the 2000 series APUs being one gen behind). APUs using Zen 2 would be 4000 series.
  • Cooe - Friday, January 11, 2019 - link

    Already have. Just like all 14nm Zen Raven Ridge SKUs were under the 2000 series umbrella, all 12nm Zen+ Picasso parts will have 3000 series model numbers (i.e. expect a 3200G & 3400G on desktop). See the just launched 2nd Gen Ryzen Mobile SKUs (using Picasso) for an example.
  • Teckk - Saturday, January 12, 2019 - link

    Cool, thanks for clarifying. Not consistent but that's how they're doing it every time then. 👍
  • nemi2 - Friday, January 11, 2019 - link

    Please let there be an option for the 2nd chiplet to be a massive amount of level 3/4 cache...
  • PeachNCream - Friday, January 11, 2019 - link

    Cache size increases experience diminished returns so an absurdly large SRAM cache, while admittedly cool on paper, may not justify the cost increase with greater performance. Also the off-die nature of a chiplet being used as a cache pool would likely further reduce cache benefits by adding distance and communication related latency. The nerdy side of me is cheering for HBM, but that Mattise IC package doesn't look like an interposer which, as far as I know, is mandatory for current HBM implementations. That leaves DRAM or some eDRAM/Crystalwell-like animal. DRAM might not like sharing the same IHS and HSF as the rest of the CPU package though it would be a potential competitor for that real estate given the locality of the IO die with its memory controllers. The temperatures DRAM would experience could be detrimental to performance and eDRAM was an Intel thing that offered about 50GB/s which is what DDR4 in dual-channel is now delivering albeit at higher latency.
  • Santoval - Friday, January 11, 2019 - link

    "The nerdy side of me is cheering for HBM, but that Mattise IC package doesn't look like an interposer which, as far as I know, is mandatory for current HBM implementations."
    An interposer is not "mandatory" for HBM. If it was mandatory Kaby Lake-G could not have had its HBM stack connected to the GPU via EMIB. Now, EMIB is quite faster than IF so it would have a bandwidth and latency advantage, at the trade-off of being very close range. Yet I don't see why IF could not be used for HBM.
  • Spunjji - Monday, January 14, 2019 - link

    EMIB is an Intel technology, for the time being at least. I very much doubt we'll see it on an AMD CPU any time soon.
  • CajunArson - Friday, January 11, 2019 - link

    While a 16 core RyZen 2 that for whatever reason AMD couldn't get running for CES will make the fanboys happy (even though they already have Thread rippers... Right?) it's a flat out dumb idea to leave a giant swath of the market that uses integrated graphics for basically everything that's not a AAA game unaddressed.

    You'll also not that AMD used to love to mention Navi at it's talks. Funny how it disappeared at CES. Connection?
  • ArcadeEngineer - Friday, January 11, 2019 - link

    Given the Radeon VII announcement lack of Navi mention doesn't require additional explanation. They aren't going go 'here's our fancy new GPU, and here's the one that will make it obsolete in six months'.
  • StevoLincolnite - Friday, January 11, 2019 - link

    Navi is likely a Polaris replacement, not a Vega replacement.
  • Dragonstongue - Saturday, January 12, 2019 - link

    Navi is NOT a desktop card replacement, this has been known about for a few months, the name itself was unknown, Radeon VII (seven, or five too when spoken) was not known before CES and the replacement for Vega was known to becoming, so the V in the name, second gen of VG, makes sense to me.

    If AMD CPU teams took the ideas to heart of what they did going from Athlon 64 coming forward to Bulldozer leading to Ryzen truly to heart, to keep the ball rolling, they will be doing exceptional in a couple of years time...

    If I had $10-150k to plunk down on them today (could afford it) I would, but I cannot :( T_T

    On the GU side of the equation, they went from the Radeon 1900 series to the 2k forward to 7k and you can tell that their funding went quite down hill as although there were some "beast" cards they lost that sweet spot middle that they established a few years prior in the 4k series (4870 was the top...5k 5970 was top level enthusiast but 5870 had very decent performance, 6870 was "close" could overclock a touch to match the 5870 for less power, 6970 was a "new" VLIW approach, n using 4 medium fat process everything shaders, instead of 1 fat process all and 4 process specific tasks in a more threaded fashion toe the meaning hundreds of dispatchers..

    Anyways, ahead we come to polaris, which effectively uses a modernized combination or fusion of VLIW-4 and VLIW-5 at GCN7 (I believe for Vega II, for GCN version as well) Polaris is level 6 (Vega is as well)

    Polaris was the "top" for gaming grade but the enthusiast branding "morphed" a few times over their years offering highest performance for the workstation offerings and so it did with the desktop grade as well not being known "simply" by the highest model number but also like Nano/Fury, and it seems like now the Vega will be the "top" and become its own "chain" opr at least should IMHO such as best cutting edge VR grunt etc and the more "main stream" like the 7870 many years ago will be the new NAVI when it releases.

    To establish two lines for their long term ideas would not kill them, it might actually help them focus their "assembly line" for ideas.

    I get mainstream, I budget, I get mobile, I get high end, I get enthusiast, workstation, enterprise, server,

    Titan is a "good example" of workstation grade, IMO this is where AMD has done will over the years with dual-gpu on one card (often times better then Nv did) then they also did with the 90 versions or the X numbers just being a high shader/clock card type thing, Vega in its own right is a "titan" for desktop use, and the x70 and x80 naming they now seem to use will become the "Navi" using numbers in its name.

    I do not see them doing much cut down versions of Radeon VII otherwise it would just be a Vega 64, and if "true" they only have another 4 CU to go before it reaches GCN design limits of 4096 or was 4560 or something like that shader count?
  • eldakka - Sunday, January 13, 2019 - link

    "Radeon VII (seven, or five too when spoken) "

    Huh? VII is the Roman numeral for 7, just like X (O/S X) is the Roman numeral for 10.

    VII is not pronounced "five too" it is pronounced "seven", IX is pronounced "nine", XII is pronounced "twelve", etc.
  • cheshirster - Friday, January 11, 2019 - link

    Not enought 7nm capacity, so APU and midrange NAVI has to be dropped/postponed?
  • dan_ger - Friday, January 11, 2019 - link

    there is huge 7nm capacity. if anything it is currently undersold.
  • FullmetalTitan - Saturday, January 12, 2019 - link

    There really isn't though. The foundries were constrained in initial 7nm capacity by having to choose between SAQP/SAHP or EUV, and EUV route was constrained by ASML production of scanners, while SAQP adds a LOT of cycle time.
    Going forward the constraints will ease, especially as TSMC moves to their 2nd gen 7nm with EUV inserted, but there is also the massive 7nm commitment to mobile chips. TSMC is building the new snapdragon part, and that eats a huge amount of 7nm capacity.
    So far Navi is still exactly on track with roadmaps AMD has provided for the last 2 years. I expect we will hear a lot more at computex, and see Navi parts in the wild around the end of the summer.
  • Dragonstongue - Saturday, January 12, 2019 - link

    It is, I read about this a few times, many (such as apple, Nvidia etc) would not be stepping in till 2Q-4Q 2019 into 2020-2021 if they see it being worthwhile.

    TSMC is able to give huge volume to AMD of total capacity and not hurt themselves $$$$ overall, which means, if Ryzen 1-2 sell/continue to sell well, TR same, EPYC which is selling very well and soon to be Ryzen 3xxx also do well, sky is the limit ^.^
  • Alexvrb - Friday, January 11, 2019 - link

    We knew Navi wasn't going to hit for a while yet. The only surprise was they released a consumer Vega 20 as a stopgap. As for a Zen 2 APU, they prioritized their GPU-less desktop variant, so the APUs will come later. It's a smart move - strike while the iron is hot, before Intel can get their 10nm chips cranked out en masse.

    Meanwhile on mobile they just released Zen+ designs including 35W models. That will help them end up in more laptops across the board, now including gaming models. Of course this is made possible in part by Intel's struggle to produce enough chips. I predict Zen 2 mobile APUs will eventually span an even larger range, possible down to sub 10W models. Surface Pro?
  • shadowx360 - Friday, January 11, 2019 - link

    Read the Q&A with the CEO, AMD chose to showcase an 8 core to show core to core parity with Intel, instead of relying on more cores to achieve the same performance.
  • PixyMisa - Saturday, January 12, 2019 - link

    They just announced Ryzen 3000 APUs. Just that they're 12nm, not 7nm.

    So there won't be a second range of Ryzen 3000 APUs. They'll be called something else.
  • Jmce - Saturday, January 12, 2019 - link

    Or... As AMD has said in their notes for the event on their website that they specifically chose an 8c/16t 3000 series because it had the same core/thread count and thus allowed for a core for core performance comparison. The idea being "we've always had more cores, this time were going keep it a as level a playing field as possible in order to show our improvements in both efficiency and performance".

    Personally I think it screams, "our mid tier is as fast as your flagship. The best is yet to come".
  • MonkeyPaw - Friday, January 11, 2019 - link

    No APU version now makes sense, as they announced 3000 series APUs based on 12nm Zen+ just a few days before.
  • neblogai - Friday, January 11, 2019 - link

    No- 8 core Zen2 chiplet APU, made with a different, energy efficiency oriented IO die that has a tiny iGPU makes sense (but not sure if AMD will put effort to make it). At this point- AMD's Raven Ridge/Picasso APUs serve only ultrathin laptop, and entry level desktop markets. And a faster 8 core CPU packaged with iGPU could offer a whole new lineup, as it is needed if AMD want to seriously compete in gaming laptop market and high performance laptop market; and a crazy efficient 8-core Zen2 with iGPU would be very welcome for embedded/enterprise.
  • StevoLincolnite - Friday, January 11, 2019 - link

    I would rather AMD stuck with Quads for the laptop market... The more CPU cores you throw at it... The less TDP the APU has in driving up GPU and CPU Clock rates.

    Raven Ridge will sacrifice GPU/CPU clocks when running a game after it has hit it's TDP wall... Sometimes you can get more gaming performance if you limit your CPU clocks as the GPU can clock up higher/more often.

    Now idle power consumption is where the real issue is with AMD's notebook efforts right now, that needs a good looking at.
  • neblogai - Friday, January 11, 2019 - link

    Thing is- AMD's quad APUs slow down the dGPU, compared to Intel quads (even U-series) and 6-cores. Not sure why- maybe 4MB L3, and high latency from all the IF. So, AMD needs better mobile CPUs to run dGPUs properly- that is, if AMD want to compete in laptop gaming market with their CPUs.
    As just RR APUs- I agree, those by themselves are fine for media consumption laptops, and some gaming. And regarding idle power consumption- I believe it is more of a laptop manufacturer problem, because Huawei managed to make a cheap, fast, and great value MatebookD with Ryzen- that somehow offers great battery life. Other manufacturers should learn from them, how to achieve such energy efficiency. Plus, just released Raven Ridge 12nm refresh will help efficiency too.
  • nandnandnand - Saturday, January 12, 2019 - link

    Simply put, if desktop Ryzen goes to 16 cores, mobile Ryzen should go to 8 cores. And Intel is already putting out 6-8 core mobile chips.

    Ryzen 3000-series mobile APUs are a big disappointment. Wait a year, and they will go to 7nm, probably doubling the core counts, lowering power consumption, increasing IPC, including Navi graphics instead of Vega, etc.

    Also, too many people are confusing mobile APUs with desktop APUs in these comments.
  • Dragonstongue - Saturday, January 12, 2019 - link

    I agree with you, the throttling stuff, and being high clocked low battery size sucks $%^ use 4 cores, cooling is easier to deal with and power use is "easy" (4 core 8 thread of course) but the makers of laptop etc need to stop being douches with "has to be as thin as possible" instead use a good to great size battery with good connectivity options. do not gimp paired memory or speed amounts, easy migration to other drives included "in the box"

    Anyways, 4 core, 8 thread, 7 nm, means they have a hunking big chunk available to "plop" a graphics core next to it *Drum roll **************** Navi...........
  • nicolaim - Friday, January 11, 2019 - link

    Typo "amenable to using a graphics."
  • Ryan Smith - Friday, January 11, 2019 - link

    Thanks!
  • costeakai - Friday, January 11, 2019 - link

    apus with navi.gpu and zen2.cpu (must) have to be incoming sooner rather than later.
  • neblogai - Friday, January 11, 2019 - link

    Actually, next 'confirmed' AMDs APU, as per leaks, will be a 7nm, 4-core Zen2 + Vega, not Navi, and considering Raven Ridge refresh just launched as 12nm- it should be coming no sooner than late 2019.
  • GreenReaper - Friday, January 11, 2019 - link

    It better have a tweaked video block with AV1 support or it'll be outdated almost as soon as it arrives, especially for power-conscious mobile devices.
  • yeeeeman - Friday, January 11, 2019 - link

    It actually makes sense since current APUs have 4/8 core/threads configurations in a single die. They can probably up that to 8/16 core/threads + some Vega/Navi (hopefully) cores and call it a day. You don't need/want more than 8 cores in an APU be it in laptop or in desktop config.
    Another thing to consider is the fact that chiplet design is good for yields and scalability, but it is a source of power waste and/or performance loss compared to a monolithic die.
  • pfdman - Friday, January 11, 2019 - link

    They also need at least 6 cores in their laptop offerings in order to compete with Intel's 6 core laptop chips. So the next round of APUs after the 3000 series will have at least 8 cores. But since the 3000 series laptop APUs were just announced, we won't see 8 core chips in their laptops till next year.
  • neblogai - Friday, January 11, 2019 - link

    Next monolithic APU is Renoir/Grey Hawk: 7nm 4-core with Vega graphics, 10-35W TDP.
  • Cooe - Friday, January 11, 2019 - link

    The 4-cores & Vega graphics parts of that statement are nowhere even remotely close to being "confirmed facts". At this point, we know next time nothing about Renoir aside from the fact it'll be fabbed on 7nm & be a Picasso successor thanks to it's placement on leaked AMD product development time-tables. Nothing more beyond that information exists at this point; ESPECIALLY fine details like core count & GPU arch.
  • neblogai - Friday, January 11, 2019 - link

    https://www.reddit.com/r/hardware/comments/aemjgn/...
  • nandnandnand - Saturday, January 12, 2019 - link

    If they don't put 6-8 cores in a mobile APU, I'm not buying it. End of story. I'd wait another year.

    However, if they are using the 8-core chiplets across all product lines, I don't see why they would not have 8 cores for laptops on 7nm. Especially if they are going to double the max Ryzen desktop core count to 16.
  • darkswordsman17 - Friday, January 11, 2019 - link

    I'm not sure this changes much as far as the prospect of them making APUs that use separate modules instead of monolithic die. The reason being that pairing a CPU and a decent sized GPU would be less than optimal for the AM4 platform. While it could still be decent and well built boards could support it. I would think this type of APU would skip a socket for now (or require something like the TR4 socket). This way there's less limitations on power and memory setup, and they'd likely be custom built for the customers (for instance console companies). That would make more sense for laptops as well (although I don't really expect such a product, it'd be nice to see). And All-in-One systems would probably be better off with dGPU on a card, and SFF - like Mac Mini type - would be better with an eGPU box.
  • BigMamaInHouse - Saturday, January 12, 2019 - link

    Can you Imaging 8 chiplet like this on TR3? I think over 15K on CB-R15 going to be easy (on stock or P.B.O), It's Crazy time Saver for Professionals!.
  • dromoxen - Tuesday, January 15, 2019 - link

    Corsair has already produced a module to be fitted to the space on the CPU.. it consists of a small capillex screen which can display any cpu or other system information, all in a breathing pattern.

Log in

Don't have an account? Sign up now