Exclusive review: Gigabyte GeForce GTX 980 Ti G1 Gaming SOC Edition

Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Jun 11, 2015.

Thread Status:
Not open for further replies.
  1. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    40,047
    Likes Received:
    8,707
    GPU:
    AMD | NVIDIA
    ShockG - I want this thread back on topic, if you feel the need to complain to Gigabyte over a 250~300 point differential in 3DMark, by all means please do so. I am a bit done with the tone and endless argument though.

    Best of luck to you, guys back on topic please.
     
  2. ShockG

    ShockG Banned

    Messages:
    18
    Likes Received:
    0
    GPU:
    GTX 250 1GB
    He didn't debunk anything.
    Debuking means providing evidence contrary to what is being stated.
    Thus far he has said nothing.

    the P score is an algorithm which combines all the individual scores with different weighting. The GPU score is incorrect at 9,600!
    The 980Ti at the factory clock speeds cannot generate that GPU score, even with LOD tweaks, etc.
    That kind of GPU score can only come about with a much higher clock speed. It is not within margin of error. There isn't a single person who can run identical clocks claimed in the review on a similar system who will get the results the review has shown.

    I've provided the Futuremark links and the screenshots. All my results are in line with the rest at HWBOT. These alone are not ;)
     
  3. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    40,047
    Likes Received:
    8,707
    GPU:
    AMD | NVIDIA
    Yes that is in the pipeline. As you know sometimes up-to 30 cards are listed in a the test, hence I cannot update everything with the latest driver all the time. But the Kepler situation is unique and I will have a look at that later on.
     
  4. moab600

    moab600 Ancient Guru

    Messages:
    6,260
    Likes Received:
    185
    GPU:
    1080TI Gaming editi
    THANK YOU! it is indeed unique and requires a case of it's own, i myself noticed maybe 3 more fps in witcher 3 and that all, game is not smooth as it should be... other games like FC4 and project cars didn't gain anything.

    With your test, we finally know the truth.
     

  5. Denial

    Denial Ancient Guru

    Messages:
    13,226
    Likes Received:
    2,721
    GPU:
    EVGA RTX 3080
    Honestly, I'm sure its a lot of work, but I think it would be cool to see a 680/780/980/7970/290x with launch drivers compared to current drivers. You would only need to do like a synthetic, but I'd be curious to see how much Nvidia/AMD have actually improved performance over the years via drivers.
     
  6. Fender178

    Fender178 Ancient Guru

    Messages:
    4,184
    Likes Received:
    207
    GPU:
    GTX 1070 | GTX 1060
    I am really impressed that the 980ti really out performs the Titan X in Metro Last Light @4k an 11fps difference. While in other games tested it was on par or 1-2 fps difference.

    The guy he banned was a troll and you are showing disrespect and not being on topic and I reported you as well.
     
    Last edited: Jun 12, 2015
  7. Denial

    Denial Ancient Guru

    Messages:
    13,226
    Likes Received:
    2,721
    GPU:
    EVGA RTX 3080
    Metro seems to be weird in general. G1 is a 18% overclock yet in metro it's nearly 27% faster. Although I don't get how that game has such poor performance. It doesn't look that good..

    Which actually makes me curious. There was an article back when the 980's came out about Nvidia reference TDP power limiting. Some of the 980 reference cards were limited to like 230w tdp, while the non reference would blow past it. I wonder if something similar is going on here?
     
  8. Reardan

    Reardan Master Guru

    Messages:
    367
    Likes Received:
    56
    GPU:
    GTX 3080
    I find it kind of upsetting that users can't question results and just be wrong about it.

    When I read this article, I saw a 25% increase in Metro LL and also checked all the numbers myself. I still have questions about that result and a couple others, but on the whole, they're close enough that I just let it slide.

    I don't see anything disrespectful in his posts. So basically what I am to take from this is: Do not question results unless you're right about it.

    Which is a really unhealthy relationship to have with an enthusiast user base who is really just concerned about getting the right numbers.
     
  9. DiceAir

    DiceAir Maha Guru

    Messages:
    1,356
    Likes Received:
    14
    GPU:
    Galax 980 ti HOF
    Yes. I normally look at graphics score cause this has been tested with much better cpu than what i have so physics score will affect overall score. I still see it's much better. i'm just awaiting some more reviews to see performance between brands.
     
  10. Denial

    Denial Ancient Guru

    Messages:
    13,226
    Likes Received:
    2,721
    GPU:
    EVGA RTX 3080
    Yeah idk, I half agree with you. I think that if there is a discrepancy it should definitely be pointed out. I don't think every discrepancy is malicious -- it could lead to learning something about how the architecture works, or why there is a difference in performance (970 memory issue comes to mind). GShock kind of ignored that portion of it though and just automatically assumed it was Gigabyte/Hilbert at fault -- and not like some strange underlining cause. I don't think he should have been banned personally but I don't own the site.

    Yeep, ASUS is the one I'm waiting on, they always do funky cool stuff.
     

  11. zcess81

    zcess81 New Member

    Messages:
    8
    Likes Received:
    0
    GPU:
    EVGA GTX 980 Ti SC
    I'm looking to get GTX 980ti and I've got a question regarding noise levels. I realise that non-reference designs run cooler and overclock better than the reference models, but I'm not interested in those advantages. My primary concern is the noise. Considering that the reference models are much cheaper right now, I was wondering whether reference/non-reference noise levels are more or less the same?
     
  12. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    40,047
    Likes Received:
    8,707
    GPU:
    AMD | NVIDIA
    If anyone feels a need to dispute anything, totally fine. You can have your word but at one point the discussion stops. I asked nicely to get back on topic, then the first post he makes is to thread crap again. He's out, as simple as that.
     
  13. headapohl

    headapohl Member

    Messages:
    40
    Likes Received:
    0
    GPU:
    1080Ti SC2 Hybrid
    Typically non reference cards are more quiet in addition to cooler. The only time they run into issues is with SLI as the top card chokes on the bottom's heat.
     
  14. zcess81

    zcess81 New Member

    Messages:
    8
    Likes Received:
    0
    GPU:
    EVGA GTX 980 Ti SC
    Thanks for the reply.
     
  15. Netherwind

    Netherwind Ancient Guru

    Messages:
    7,622
    Likes Received:
    1,519
    GPU:
    MSI 3080 Gaming X
    So...to change the topic - Hilbert, any chance on future reviews of the ASUS and MSI cards? Will you also include 970SLI or 980SLI in the benchmarks? Or will there be a separate 980Ti SLI test with other SLI setups?
     

  16. headapohl

    headapohl Member

    Messages:
    40
    Likes Received:
    0
    GPU:
    1080Ti SC2 Hybrid
    This card looks like a beast. I am pretty much set with my 295x2 for this generation but will be cool to see the 980ti vs Fury X race. I think at the end of day they will both be comparable.
     
  17. Denial

    Denial Ancient Guru

    Messages:
    13,226
    Likes Received:
    2,721
    GPU:
    EVGA RTX 3080
    Isn't he not comparing the same thing though? The one he linked is a reference that's overclocked. Not the G1. The reference Ti's probably have TDP limiting on the boost, similar to the reference 980's. Which would definitely impact the score, especially in a synthetic benchmark. And we're not talking about a 10-15% gap, the difference between his result and Hilberts is 5%. If you run 3DMark 20 times you get 5% margin of error. This isn't to mention that none of the other gaming really show anything weird or strange for a card that's clocked 18% higher.

    Again, I don't think he should have been banned but he doesn't really have an argument against that result. Hilbert's score seems fine.
     
  18. MADOGRE

    MADOGRE Member

    Messages:
    11
    Likes Received:
    0
    GPU:
    Gigabyte 980ti G1
    Card looks nice, my build is all black and white, so this is probably what I will go with. I dont think MSI will have a white one any time soon.
     
  19. morbias

    morbias Don TazeMeBro Staff Member

    Messages:
    13,444
    Likes Received:
    36
    GPU:
    -
    Maybe I'm missing something but surely anyone can grab a validation link that will support their argument, eg. http://www.3dmark.com/fs/5017586

    There is always a margin with these things that can be affected by RAM speed, other stuff running in the background, even which SSD you have in your system.

    And I'm pretty sure Hilbert was not particularly pleased with the fact he was, along with Gigabyte, being called a liar. Because let's not beat around the bush, that's what was pretty blatantly being suggested here.
     
    Last edited: Jun 12, 2015
  20. Hilbert Hagedoorn

    Hilbert Hagedoorn Don Vito Corleone Staff Member

    Messages:
    40,047
    Likes Received:
    8,707
    GPU:
    AMD | NVIDIA
    You seem to have the notion that I need to answer your questions just because you say so ? Not really. I am done with this topic. Best of luck though.

    Oh and btw, your attitude is what in the end will get you banned, that is pretty much the only part you have correct.
     
Thread Status:
Not open for further replies.

Share This Page