AMD's Ryzen CPU's (Ryzen/TR/Epyc) & Vega/Polaris GPU's

Discussion in 'Hardware Components and Aftermarket Upgrades' started by Rage Set, Dec 14, 2016.

  1. ajc9988

    ajc9988 Death by a thousand paper cuts

    Reputations:
    1,156
    Messages:
    4,739
    Likes Received:
    6,619
    Trophy Points:
    581
    Except, Sandy and Ivy bridge still have a huge part of the market. How many years is that? Not everyone upgrades on the cadence, only enthusiasts. With the power in the modern chips on both sides, and Ryzen and Intel's 8700K being the main increase making upgrading worth it if you had a sandy bridge CPU, I'd argue that it is equally future proof, except you would have more upgrade options down the road on the same platform on the AMD side.
     
  2. Vasudev

    Vasudev Notebook Prophet

    Reputations:
    3,290
    Messages:
    6,323
    Likes Received:
    4,128
    Trophy Points:
    431
    Still use my Ivybridge pentium laptop and have an old Radeon 5750.
     
    Robbo99999, ajc9988 and hmscott like this.
  3. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    4,650
    Messages:
    16,175
    Likes Received:
    19,884
    Trophy Points:
    931
    Thoughtless newbies throwing money out the window aren't really considering much more than straight line performance, FPS, when making their buying decision.

    Cost and price limit's hone the buying choices and help force the buyer to consider other criteria than straight-line performance when making the purchase decision.

    When you have $M's to spend on computer hardware, you might think cost isn't an object of consideration, but considering price / performance plays a vital role in focusing the spend on value.

    The considerations that come from the price / cost of choice help make smarter and better solution choices than only using straight-line performance at any cost.

    In fact you'd quickly lose your purchasing ability making decisions based only on the numbers from straight-line performance metrics.

    Consumer purchases are the same, you need to be price sensitive and also look at cost, don't overspend on the wrong metrics, missing out on the wider picture.

    If you can get more cores for the same cost providing better multitasking performance along with future multi-core software performance improvements, and only give up a non-noticeable percentage of straight-line performance, then an AMD CPU with more cores is a better buying choice than a higher clocking Intel CPU.

    If you are going to get fooled chasing metrics meaningless in everyday use, you'll quickly go broke on upgrades constantly trying to chase the numbers.

    In real life use, those numbers get exceeded by new products on a regular basis, that top performance metric machine is going to quickly lose it's value to you, and you'll never be happy with what you have.

    Figuring out how to be happy with a little less performance for less cost is a good way to start being happy over the long term with your purchase. :)
     
    Last edited: Jul 11, 2018 at 12:51 PM
    ajc9988 likes this.
  4. TheReciever

    TheReciever D! For Dragon!

    Reputations:
    238
    Messages:
    1,434
    Likes Received:
    762
    Trophy Points:
    131
    Yeah I used to be one of those guys chasing the benchmarks.

    I should've kept the 1055T system I had back in the day instead of trying to get the first gen i7's. That 1055T was the easiest 4Ghz overclock I ever had. I sold it to make a point to my brother though. Told him he could keep the PC if he gets a job even if its just some part time gig. 1055T and 4870x2 was a good combo back then. He refused to get a job so I sold it.

    Its why I am ok with my Ranger, I already knew Haswell was good enough for my needs and can upgrade the GPU through MXM.

    Next system will likely be be the Acer Helios 500 provided the firmware is sound. If the 7950 can last as long as it has, I think Vega 56 and 2700 will be fine for a while yet. Here's hopin'.
     
    James D and hmscott like this.
  5. hmscott

    hmscott Notebook Nobel Laureate

    Reputations:
    4,650
    Messages:
    16,175
    Likes Received:
    19,884
    Trophy Points:
    931
    Vasudev likes this.
  6. Deks

    Deks Notebook Prophet

    Reputations:
    984
    Messages:
    4,054
    Likes Received:
    1,256
    Trophy Points:
    231
    Some news regarding AMD Vega on 7nm:
    http://www.itp.net/617514-amd-demos-7nm-gpu

    "AMD claims the new 7nm process is twice as dense as its 14nm process, and the 7nm Vega die appears to be roughly 40% smaller than its predecessor. The new process also affords a 2x increase in power efficiency and AMD also claims it provides a 1.35x increase in performance."

    Note that they don't seem to mention that the process will result in 2x increase in efficiency OR provide 1.35x increase in performance... but rather they continue to describe both 2x increase in efficiency and 35% increase in performance as something that goes hand in hand on 7nm TSMC.

    So, if Vega gains 35% increase in performance and is double the efficiency at the same time... I wonder how this new process might affect the yields (and subsequently voltages) on 7nm vega (and Navi) and if it might result in Vega and Navi being clocked much higher and gain more performance given that Vega on 14nm suffered from high voltage problems and mauf. process designed for low clocks which forced AMD to use low clocks.

    Thinking about Pascal, it was clocked about 30% higher than Vega was, mainly because the process was suitable for high clocks and PAscal had auto voltage tuning, and less compute units... (though this produced roughly similar or lower performance to lower clocked Vega 56 and 64 - except of course the 1080ti)

    Turing architecture from what I understand is more AI oriented... so, no real benefits for games.
    And Nvidia also said it won't be releasing a new GPU for a long time (though things are open to change)... at best, Nvidia could refresh Pascal on 7nm.
    It would be an interesting 'treat' if Vega suddenly on 7nm ends up equating or beating 7nm refreshed Pascal and with Navi just around the corner - but I guess only time will tell.

    Right now we're getting figures from the manuf.process improvements... no word on whether AMD improved/optimized Vega's internals, etc.

    EDIT: Still bearing in mind that the 2x increase in efficiency and 35% performance increase could be an error on the article writer (or algorithm) and could easily still end up being one OR the other.
    It's just that until now, articles seem to say that both will occur (which of course is not consistent with history - but then again, that 35% increase in performance could be due to the node eliminating the bottlenecks of 14nm LPP).
     
    Last edited: Jul 12, 2018 at 5:00 PM
    hmscott and Vasudev like this.
  7. ajc9988

    ajc9988 Death by a thousand paper cuts

    Reputations:
    1,156
    Messages:
    4,739
    Likes Received:
    6,619
    Trophy Points:
    581
    I explained this **** to you already. How you achieve both on a curve is due to other factors at work, such as faster memory and other refinements.

    Meanwhile, what I didn't explain was that AMD confirmed they are doing multi-die, just not for consumers because ISVs will not program for NUMA on GPUs for games. Software on the server side doesn't care and can use as many cards as you drop in. Nvidia's multi-die paper shows them also doing NUMA for multi-die GPUs. That means that until software vendors designing games add support, you have to go monolithic on consumer graphics. Also, 35% improvement on Vega puts it at or above 1080 Ti levels. Navi is rumored to be around 1080 to Ti levels also, but with less power to be used for embedded solutions.
     
    Vasudev and hmscott like this.
  8. ajc9988

    ajc9988 Death by a thousand paper cuts

    Reputations:
    1,156
    Messages:
    4,739
    Likes Received:
    6,619
    Trophy Points:
    581
    [​IMG]
    This is the curve. You cannot exit the curve. That means that if you exceed the values on the curve, it comes from other changes, such as memory, architecture tweaks, etc. I also showed that GF is using the same pitch as TSMC, and by some measures, GF is slightly denser on 7nm, but density is hard to calc between fabs.
     
    jaybee83, Vasudev and hmscott like this.
  9. ajc9988

    ajc9988 Death by a thousand paper cuts

    Reputations:
    1,156
    Messages:
    4,739
    Likes Received:
    6,619
    Trophy Points:
    581
    On other news, in the past week, Asrock has released BIOS updates for the B350, the X470, the X370, and the Z370. So, the X399 should be coming soon.
     
    Vasudev and hmscott like this.
  10. Deks

    Deks Notebook Prophet

    Reputations:
    984
    Messages:
    4,054
    Likes Received:
    1,256
    Trophy Points:
    231

    As I said, the way the articles up until now were wording it all is weird and probably inaccurate.
    However, bear in mind you also discarded my pricing projections of Threadrippers before, even though they turned out to be accurate.
    Not saying that this will be a repeat of that (given that there are obvious limits of the manuf. process curve that can't be exceeded), I'm just indicating that 14nmLPP was a proverbial bottleneck that contributed towards Vega's higher power consumption and frequencies considering the kind of process it was vs 7nm and that with those bottlenecks gone, we could (at least hypothetically) see further improvements (but admittedly, none of that is guaranteed and remains hypothetical).

    I'm taking everything you posted into account and keeping it at the forefront, but I'm also waiting for final release of Vega 7nm from AMD and see what the end product does as all this talk doesn't always manifest itself accurately in the final product (the numbers from the process improvements alone are probably best case scenario - but, we know that final products sometimes exceed expectations due to other optimizations (which we have 0 clue if AMD decided to do for Vega).

    On another end, here's what TwakTown stated on the subject - though they just copied most info from WCCFTECH:
    https://www.tweaktown.com/news/62491/amd-vega-20-7nm-32gb-hbm2-up-tflops-400w-tdp/index.html

    "The new details on Vega 20 see it coming in at just 360mm2 compared to Vega 10 at 510mm2, a huge 70% reduction in total die size thanks to the fresh 7nm node. This is where a fork in the road happens: AMD can choose power savings with the die size savings, or they can ramp clock speeds up. Vega 20 could benefit from around 55% power savings over Vega 10, or GPU clocks can be cranked by up to 40%, which would also see the now 4-stack HBM2 drawing more power. This would drive the Vega 20 card with 32GB of HBM2 to 400W, or more. AMD could use a moderate 20% gain in GPU clock speeds and reach 300-350W with a performance leap of 65% all while saving 30-40% on power consumption. This would allow AMD's new Vega 20 GPU to beat NVIDIA's current Tesla V100 in compute at 300W with around 20.9 TFLOPs compared to the V100 with 15.7 TFLOPs.

    Read more: https://www.tweaktown.com/news/62491/amd-vega-20-7nm-32gb-hbm2-up-tflops-400w-tdp/index.html"

    TweakTown and WCCFTECH - or for that matter neither one of us) don't really know what kind of power targets and performance is AMD chasing with Vega on 7nm.
    All we have are estimates and projections from non industry sources based on what 7nm shrink in its own right provides.

    I'm a bit discouraged if Vega's power consumption doesn't go down more than what 7nm provides... but then again... balanced power profiles tend to drop power consumption quite a lot (so do power saving modes for a limited/small reduction in performance) as does obvious undervolting.

    On another end, with 7nm, you'd expect Vega to be able to clock much higher without suffering same power consumption increases due to the new process being optimized for high frequencies and efficiency.
     
Loading...

Share This Page