Why Benchmarking CPUs with the GTX 1080 Ti is NOT STUPID!

Scroll this

Check prices now:
Intel Core i5-8400:
Ryzen 5 2600:
Ryzen 5 2600X:
X470 Motherboards:
ASRock X470 Taichi Ultimate:
Aorus X470 Gaming 7 WiFi:
MSI X470 Gaming M7 AC:

Read the written version of this editorial on TechSpot:

Support us on Patreon

Why Benchmarking CPUs with the GTX 1080 Ti is NOT STUPID!

Twitter –
Facebook –
Google Plus –
Instagram –

Music By:


  1. m running my core i3 4170 with a 1080ti
    m saving for my 8700k tho

  2. Hey Steve, its good to see you are doing well! Excellent content as always, keep it up mane!

  3. My last system went through 3 graphics cards, (maybe even 4), and the last most powerful card was transplanted my current system.
    It makes sense because the VGA card is what has the greatest effect on FPS.

  4. This question and its subsequent video response is just wasting Steve's time. He should be engaged in other benchmarking matters.

  5. Great explanation! Could you make a video about what CPU pairs best with what Gpu without creating a bottleneck? For example, if I use my i5 8400 with a gtx 1080ti, what will be the limiting factor ? I plan on using a 1080p high refresh panel.

  6. while i agree with the testing done with the 1080ti, the logical future-proof choice is still the 2600x because it self overclocks and has double the threads. @3:40 you can see the difference between an i5 and i7 starting at gtx 1060 performance

  7. Save up for Titian v that should get rid of the gpu bottle neck at 1444p

  8. Of course you benchmark with the best GPU available…and guess what AMD fanboys…next gen GPU's will start to reveal how weak Ryzen is in games…and the generation after that, ryzen will look like faildozer compaired to intel's flagship.

  9. we should benchmark using bum vodoo 3d card at 1080p this will make bum Athlon and Pentium 4 as fast as ryzen 2700

  10. i will be honest i did not comment but i did struggle to comprehend why you use 700 to 900 quid gpu on a low end cpu test. i think the reason why is i could not find the pairing relate-able, but i completely understand now nice explanation πŸ™‚

  11. Removing bottleneck is important to see the actuall difference between the raw CPU power , but comparing CPU with high price difference should take into account that the money you do not spend on your CPU you spend on GPU if you are actually buying one.
    Maybe you should test with COMBINING CPU and GPU prices. For example are you going to get more performance from 100$ cpu and 500$ gpu OR 300$ cpu and 300$ gpu.

  12. no testing with a 1080 is not stupid..and testing in 720p is not stupid ether..even testing a 1080ti in 720p makes sense to see what impact it has on cpu(even tho i doubt anyone will game in 720p with a 1080ti) personally i like to just enjoy my game not really caring about the settings if there on low or med if i get playable frames im good. even low as 30-40fps is still playable for me 50-80fps is more than playable. so good call on that steve πŸ™‚

  13. I’m not a fanboy on either side, mainly because I have money to afford both. But it’s been proven impossible to have absolutely no GPU bottleneck your always going to have a little. So therefore it’s an incredibly stupid topic to even touch especially with a bias of any sort.

  14. I found a used RX570 4GB expedition at 200€, should I go for it?

  15. I know this is just another request within the sea, but I would appreciate lower settings tested at higher resolutions, like 1440p for example, between different cpus. It's hard to tell when the CPU bottleneck enters the picture at 1440p. It's usually very subtle at ultra settings for mid range clue and above.

  16. It's baffling how many people in the computer and gaming community don't understand what a bottleneck is and how to benchmark. Benchmarking a CPU with a mid range GPU or at a high resolution tells you absolutely nothing about the CPU. It would be similar to benchmarking a GTX 1080 ti and a 1060 against eachother while using a dual core CPU on a modern CPU intensive AAA game.

  17. Hey guys iam building my first build ever so if anyone can help me with my first build i would appreciate it iam new to this stuff and iam very confused and my budget is around 1300 usd . Thank you πŸ˜€
    EDIT = if possible i will increase my budget to 1600 usd
    Here are the prices in my country :

    core i5 8600k = 300 usd
    core i5 8400 = 250 usd
    core i7 8700k = 463 usd

    ryzen5 1600x = 235usd
    ryzen5 2600x = 250 usd
    ryzen7 1700x (no fan ) = 330 usd
    ryzen 7 2700x = 370 usd

    as for gpu i was thinking either
    zotac gtx 1060 3GB = 357 usd
    zotac gtx 1060 6GB = 463 usd
    gtx 1070ti mini = 622 usd
    1070 ti Amp = 660 usd
    my local store doesn't sell gtx 1070 πŸ™


    AM4 :
    x370 SLI ATX = 190 usd
    b350 =145 usd
    ASUS ROG STRIX x470 =262 usd
    ASUS PRIME X470 pro = 251 usd
    MSI b350 Tomahwak = 137 usd

    my local store doesn't have alot of AM4 motherboards

    LGA 1151:
    msi Z370 =206 usd
    ASUS ROG STRIX Z370-E Gaming =251 usd
    ASUS prime Z370 quad crossfire and SLI = 235 usd
    ASUS TUF Z370 PRO Gaming ATX =209usd
    ASUS TUF b360 PRO Gaming =182 usd
    ASUS TUF b360M-E Gaming = 129 usd
    GIGABYTE Z370 MicroAtx =172 usd

    and 600w gold psu = 92 usd

    8GB one stick of ram = 103 usd
    8GB ram dual Channel = 132 usd -2666mhz
    8GB ram Dual channel = 137 usd -3000mhz
    16GB ram dual channel = 264 usd -3200mhz

    as for the case they only come in 2 prices for all cases

    1-120usd : Lame and simple few with plastic glass
    2- 250 usd = comes with build in water coolers and flow meters and tempered glass
    but the case i like is the corsair 470x = 209 usd its on discount

    1TB WD = 79usd
    1TB WD black =132usd
    250GB M.2 960EVO= 164 usd
    500GB M.2 960EVO= 304usd
    500GB samsung SSD 860=198
    120GB Mushikin SSD =52 usd
    250GB Mushikin SSD = 111usd
    500Gb Mushikin SSD = 182 usd

    Thank you for your help it means alot πŸ˜€

  18. It is not stupid but it is misleading to newbies. You should add a couple mid to low end gpu when doing this..
    And the assumption that you doing this for the intention to show which cpu will serve us for the long run is BS, nobody build an i5 now thinking in buying a 1000 dollars card 2 years from now. Maybe 4 years from now it will have a mild impact if you buy a new mid-range card, but then again 4 years from now you might want to upgrade the cpu too.

  19. what percentage of the cpu is left though if its tapped out then it has nothing left to give … dont know the answer but would like to know ..

  20. We Live in fucking 2018

    1080p is a bit 2015 right !

    so testing 1080Ti is the best choice as you explained

  21. Why not do a few select games with a midrange GPU so as to convey where you are spending more than you need to "today"

  22. Mr hardware unboxed . Pls help me .

    So what should i choose better ?
    i3 8100+1050ti or g5400+1060 3gb?

    Im stuck at this choices -_- help

  23. I think it is more important to test with the highest possible GPU for all the reasons stated, but I believe there is value to showing how much more frames each CPU will give at a specific resolutions with a suitably paired cards.

  24. Mr I'm Your Host Steve is my favorite personality on this particular YouTube channel! Love this guy to death!

  25. Thanks for the explanation, as I was one of those that asked myself. Great content !!!

  26. Out of curiosity, has this reasoning ever been validated in a retrospective way? I.e. taking a benchmark that was done a few years ago, with a 980 ti, and replicating the performance difference between lower and higher end CPUs using mid- or low-range 1000 series cards in today's games?

    Not saying you're wrong at all, just interested in seeing if it's been proven.

  27. Could you find a bottleneck by low clock speed or higher clock speed with less or more core

Submit a comment