Latest CPU & GPU Discussion (Intel/AMD/Nvidia)

Thread Rating: 4 votes, 5.00 average.
(4 votes)
Results 10,551 to 10,575 of 10583

  1. Post
    nzbleach wrote:
    Massive win for AMD. Minor win for Intel. More AMD radeons = more optimization and application support for radeon features, which hopefully means we see perf gains.

    In terms of competition, I still think Raven Ridge will be a viable alternative.
    This except I would guess the performance on the Igpu on the Intel would be better because of HBM, I wonder if AMD will release an APU with higher TDP part and HBM. AMDs HBCC with vega and say 1gb of vram would be interesting.

  2. Post

  3. Post
    This is a more full explanation:

  4. Post
    The head of AMD's Radeon Technologies Group has left to join Intel and will head a new division of the company developing high-end discrete GPUs.
    https://newsroom.intel.com/news-rele...i-joins-intel/

  5. Post
    They probably quadrupled his salary.

  6. Post
    mhm Intel apu inside the PS5?

  7. Post
    SirGrim wrote:
    mhm Intel apu inside the PS5?
    Doubtful IMHO
    First of all this is not an "APU"

    Second intel would have to pay AMD to develop the GPU/CPU combo for the consoles to tender for the console. The console space is really low margin space AMD would be able to under cut intel if they are developing in house unless Intel would want to make a loss.

    Third: There is no real motivation to compete in this space for Intel, Nvidia would be more motivated to do this but they don't have X86 IP they could partner with intel but I don't see that happening ever, intel and AMD got together for this CPU combo, intel and Nvidia are not exactly friendly much less so than AMD and intel or even AMD and Nvidia.

  8. Post
    Intel/AMD Radeon chip
    Name:  DOLwFP4X4AUuLSd.jpg
Views: 334
Size:  94.8 KB

  9. Post
    Its Polaris, not Vega. Probably a win on the efficiency front. Interesting to see it paired with HBM though (i guess that was the whole point). Far out HBM has a small physical footprint.

  10. Post
    nzbleach wrote:
    Its Polaris, not Vega. Probably a win on the efficiency front. Interesting to see it paired with HBM though (i guess that was the whole point). Far out HBM has a small physical footprint.
    At least one revision is, may not be final silicon. Vega scales well at lower clocks too apparently

  11. Post
    I wonder if this amd/intel chip is for the intel nuc range as I assume that would classify it as a desktop chip.

  12. Post
    A big part or it is targeted and thin and light gaming laptops, will prob end up in the high end NUC too though.

    They make a big deal about it being thinner, whilst getting to take a shot at AMDs tech than an interposer.

  13. Post
    reubi-one wrote:
    I wonder if this amd/intel chip is for the intel nuc range as I assume that would classify it as a desktop chip.
    I read somewhere that this was happening

  14. Post
    Nek minnit, GPUs are now on die, you have to choose a tiered socketable processor(want a 1080 Ti equiv? Well it only comes with the X series processor equiv) and have to buy an entire CPU/GPU combo if you want to upgrade (to much expense) otherwise everything else is non-socketable and ATX is all but dead.

  15. Post
    it wouldnt matter for me, i only used to buy a new gpu every 3 years at which point it required a new cpu/mobo anyway to make the most of it

  16. Post
    Is Cinebench alone enough to get an accurate measurement of a CPU's overclock performance increase or would I need to run a whole bunch of different benchmarks?

  17. Post
    WhangaDan wrote:
    Is Cinebench alone enough to get an accurate measurement of a CPU's overclock performance increase or would I need to run a whole bunch of different benchmarks?
    Cinebench is as good as any other synthetic test.

    I assume you mean is it a good way to verify that your performance is scaling with your clock speed increases?

    If that's all then yes. If you wanted to test more for thermals then AIDA64 can be better at stressing/heating up cpus - if your setup can handle those two without any glitches that tends to be a safe bet your OC and cooling setup is solid.

  18. Post
    WhangaDan wrote:
    Is Cinebench alone enough to get an accurate measurement of a CPU's overclock performance increase or would I need to run a whole bunch of different benchmarks?
    It depends on what you want to say about your over clock.

    It gives you an accurate measurement of what you test, and a potential measurement of others. If you say "Cinebench increased by X amount after a Y increase in clocks" then it's fine, but it doesn't guarantee you to say, "Shadow of War will be X faster".

    Just test what you care about. If a game is CPU limited, it could gain the whole frequency increase in FPS. Or it could not change at all.

    If you run fluid simulations, then test fluid simulations and skip Cinebench.

  19. Post
    Ok thanks. My 3770k is at 4.6ghz from 3.5 and I get a 23% increase in performance in Cinebench. Guess I wanted to find out if that's a decent gain for the 1.1ghz bump.

    Some games it's almost doubled my average FPS (Arma 3).

  20. Post
    Some consumers have got some of the new Ryzen mobile chips (Ryzen 5 2500U on the HP 360) looking forward to some gaming benchmarks but here are some initial results

    https://www.3dmark.com/compare/spy/2727277/spy/14997

    45w TDP vs 15W TDP and not even the top of the line Ryzen 7 model



    Not to bad at all for a 15w chip (the intel I5 8250u sits on average at 50 points less)

    Not a "you where right" from Suntoucher yet but in synthetics this is smoking the iris 640 so far

  21. Post
    how are ryzen chips looking against 7700hq in games?

    ok so in cinebench r15 the 7700hq score is 750, 25% over the r5 ryzen which is not that much. also there is the 7820hq cpi which scores 850 but is only found on gtx1080 laptops

    really keen for gaming benches if it can out match the 7700hq then it may pay to wait for laptops with these
    Last edited by SirGrim; 14th November 2017 at 9:34 pm.

  22. Post
    SirGrim wrote:
    how are ryzen chips looking against 7700hq in games?

    ok so in cinebench r15 the 7700hq score is 750, 25% over the r5 ryzen which is not that much. also there is the 7820hq cpi which scores 850 but is only found on gtx1080 laptops

    really keen for gaming benches if it can out match the 7700hq then it may pay to wait for laptops with these
    I have a 7700hq in my gaming laptop its a big thick thing because its a 45w part the ryzen 5 is a 15w part its more ultrabook low end gaming not max details.
    Some manufacturers where down clocking desktop Ryzens and putting them in laptops I don't know if thats still a thing.

  23. Post
    Look at this thing, can't wait to see benchmarks (as in how long it takes to blow up): http://www.xoticpc.com/np9877-clevo-p750tm1-g.html

    Yes that is the 95w desktop 8700k inside that laptop with a GTX1080 (they will even let you do 1080 SLI if you want)

  24. Post
    SirGrim wrote:
    Look at this thing, can't wait to see benchmarks (as in how long it takes to blow up): http://www.xoticpc.com/np9877-clevo-p750tm1-g.html

    Yes that is the 95w desktop 8700k inside that laptop with a GTX1080 (they will even let you do 1080 SLI if you want)
    Lol look at the thickness I have a 17 inch gaming laptop and it looks sleak by comparison.
    It will probably be fine I would rather a 1080ti than 2x 1080s.

  25. Post
    SL1CKSTA wrote:
    Not a "you where right" from Suntoucher yet but in synthetics this is smoking the iris 640 so far
    *makes claim*
    *doesn't provide any evidence*

    I am shocked, SHOCKED!

    Don't forget that AMD lives for the hype train, "accidentally" leak some convenient stats that don't give a proper representation. Don't let any one review it between soft and hard announcement.

    Sell units via people googling and just get positive results for the first few months. Eventually people learn its complete trash. The only outlier so far is Zen on the desktop but even that isn't the massive price/performance difference it was hyped to be.

    It's competitive, but was hyped to be much more than that.

    Will wait for the device release, Notebook Check has a good representation of how much performance can differ when shoved into a chassis on the 640.
    Last edited by suntoucher; 15th November 2017 at 7:49 am.