EVGA

1080ti FTW3 not performing as expected

Author
Paris ._.
New Member
  • Total Posts : 6
  • Reward points : 0
  • Joined: 2017/10/24 12:03:51
  • Status: offline
  • Ribbons : 0
2017/11/18 04:57:35 (permalink)
Hi all I have a 1080ti ftw 3 that is not performing as I would like it. I bought it for high refresh rate (240hz) 1080p gaming (yes I know 1080ti is a waste at 1080p but I wanted the best if I want 1440p of 4k in the future). Anyways the card isn't getting very good frame rates at all, something like 60-90frames on ultra in pubg and 120-150fps ultra in rainbow six siege (Both 1080p). If I lower the settings to medium or low the fps doesn't seem to increase. It seems gpu usage is very low at 30%-70% but my cpu usage is also low at 30% in pubg g and 60% in rainbow six siege. My cpu is a ryzen 1700 oc'd to 3.9ghz so I don't think it should bottleneck that badly. I EXPECT it to bottleneck at 1080p sure, but my fps is far too low. I've tried so far Uninstalling and reinstalling all drivers a few times, have reinstalled os, updated bios to the latest version, tweaked NVidia control panel settings, have ensured the pci-e lane being used is the correct lane, troubleshooted pci-e connectors, etc etc doesn't seem to help. My heaven benchmark scores seem ok but running furmark I was disappointed to see a 10000 ish score which is more like 1060/gtx1070 scores. In fact the card was only hitting 1600mhz during the testing. The gpu usage did spike to 98% but the core clock is far too low. The mem. clock was normal however. I'm starting to think I got a bad card. The only thing wrong with my rig so far is maybe the memory is too slow (2400mhz ram 16gb) but I plan to update to 2933mhz(max as allowed by mainboard) later on next week to see if it makes a difference. Does anyone have any suggestions as to why the card either wants to downclock or wants to sit there and be lazy with its low usage. Again I know that at 1080p there is going to be a bottleneck but I should still have plenty of horsepower for a 240hz display. I SHOULD be able to get more than 60 fps playing PUBG on high settings right? Do I have a bad gpu or is there something I am missing?   
#1

27 Replies Related Threads

    demon09
    FTW Member
    • Total Posts : 1334
    • Reward points : 0
    • Joined: 2016/09/16 21:18:42
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/18 10:48:25 (permalink)
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k
    #2
    schulmaster
    SSC Member
    • Total Posts : 541
    • Reward points : 0
    • Joined: 2011/06/27 19:48:02
    • Status: offline
    • Ribbons : 7
    Re: 1080ti FTW3 not performing as expected 2017/11/18 12:27:20 (permalink)
    demon09
    With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu.


    This. You need to look at the utilization of the 2-4 logical CPU threads that are being pinned by the 3D application.

    Case: --InWin LE S-Frame #111/500--
    Board: ASUS x299 SAGE WS 
    CPU: i9 10980XE (18C/36T) @ [4.0GHz - 4.5GHz]
    CPU Cooling: Noctua NHD15S 
    RAM:
    64GB Vengeance PRO Quad Channel @ 3.2GHz 
    GPU: ASUS RTX 3090 STRIX OC Edition
    Display:
    27" 2560x1440 Asus Swift 165Hz IPS GSYNC
    Storage:
    2x 1TB Micron 1100s SSDs RAID0 (OS)
    2TB Samsung 980 Pro NVMe (Steam)
    4TB WD Enterprise 
    1TB 10000RPM Velociraptor
    SPU: Creative AE7 
    PSU: Corsair AX1600i  
    #3
    CriticalHit_NL
    iCX Member
    • Total Posts : 404
    • Reward points : 0
    • Joined: 2009/09/08 10:23:06
    • Status: offline
    • Ribbons : 17
    Re: 1080ti FTW3 not performing as expected 2017/11/18 12:58:15 (permalink)
    The 1080Ti FTW3 performs great here on an older i7 3930K, and I think the Ryzen 1700 is faster single-threaded, but there is no such thing as buying an high-end GPU and expecting 240fps in every game.
    Some games are simply too demanding on the CPU and especially in DirectX 11 you need double the CPU capacity to get double the framerate due to drawcalls load scaling lineair.
     
    It's pretty simple, if your GPU load is low and only increases when raising graphical settings without losing more fps it's simply a CPU bottleneck.
    Not having 100% total CPU usage or on any core doesn't mean there is no CPU bottleneck, sometimes games even may perform slightly worse with HTT enabled for more logical threads.
     
    1080P and 240Hz is pretty much a peace of cake for a 1080Ti FTW3, but not for todays CPU's, they have been lacking in the performance increase department compared to GPU's the recent years.
    The Ryzen 1700 probably compares somewhat to the i7 6700K or i7 7700K in games, except maybe a few that scale well on many threads.
     
    Oh and a suggestion, try to make your post a little more readable, people would be more easily understanding your question and you may get more answers this way.

    i7 3930K 4.4Ghz - EVGA 1080Ti FTW3 - 32GB Corsair Dominator DDR3-1866 - Asus Sabertooth X79 - Samsung 840 Pro 256GB - Samsung 860 Pro 2TB - 3x HGST 7K4000 Deskstar 4TB - 3x HGST Ultrastar He10 8TB - Corsair AX1200i - Asus PG279Q + BenQ BL2411 - Razer Viper Mini/SBF98 - Logitech Z906 & G35 - Windows 10 Professional

    #4
    rjohnson11
    EVGA Forum Moderator
    • Total Posts : 102215
    • Reward points : 0
    • Joined: 2004/10/05 12:44:35
    • Location: Netherlands
    • Status: offline
    • Ribbons : 84
    Re: 1080ti FTW3 not performing as expected 2017/11/18 12:59:40 (permalink)
    I'll add a +1 to demon09's post about furmark. It's a horrible benchmark and you shouldn't use it.

    AMD Ryzen 9 7950X,  Corsair Mp700 Pro M.2, 64GB Corsair Dominator Titanium DDR5  X670E Steel Legend, MSI RTX 4090 Associate Code: H5U80QBH6BH0AXF. I am NOT an employee of EVGA

    #5
    demon09
    FTW Member
    • Total Posts : 1334
    • Reward points : 0
    • Joined: 2016/09/16 21:18:42
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/18 13:05:44 (permalink)
    CriticalHit_NL
    The 1080Ti FTW3 performs great here on an older i7 3930K, and I think the Ryzen 1700 is faster single-threaded, but there is no such thing as buying an high-end GPU and expecting 240fps in every game.
    Some games are simply too demanding on the CPU and especially in DirectX 11 you need double the CPU capacity to get double the framerate due to drawcalls load scaling lineair.
     
    It's pretty simple, if your GPU load is low and only increases when raising graphical settings without losing more fps it's simply a CPU bottleneck.
    Not having 100% total CPU usage or on any core doesn't mean there is no CPU bottleneck, sometimes games even may perform slightly worse with HTT enabled for more logical threads.
     
    1080P and 240Hz is pretty much a peace of cake for a 1080Ti FTW3, but not for todays CPU's, they have been lacking in the performance increase department compared to GPU's the recent years.
    The Ryzen 1700 probably compares somewhat to the i7 6700K or i7 7700K in games, except maybe a few that scale well on many threads.
     
    Oh and a suggestion, try to make your post a little more readable, people would be more easily understanding your question and you may get more answers this way.
    for games that prefer clock speed the 1700 is below the 6700k and 7700k. It has a single core score of about 160ish where as a 7700k at 4.9ghz is about 200ish and and a 6700k at 4.5 is probably around 190. Hopefully we see games move to better use of more cores with Intel finally jumping on board it's more likely. It's single core score is about a stock 4770k which isn't bad but high refresh rate gaming on some games is very cpu bound
    post edited by demon09 - 2017/11/18 13:08:06
    #6
    arestavo
    CLASSIFIED ULTRA Member
    • Total Posts : 6916
    • Reward points : 0
    • Joined: 2008/02/06 06:58:57
    • Location: Through the Scary Door
    • Status: offline
    • Ribbons : 76
    Re: 1080ti FTW3 not performing as expected 2017/11/18 15:17:21 (permalink)
    "If I lower the settings to medium or low the fps doesn't seem to increase."
     
    And that's the definition of a CPU bottleneck right there.
    #7
    Paris ._.
    New Member
    • Total Posts : 6
    • Reward points : 0
    • Joined: 2017/10/24 12:03:51
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/18 15:43:26 (permalink)
    schulmaster
    demon09
    With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu.


    This. You need to look at the utilization of the 2-4 logical CPU threads that are being pinned by the 3D application.

    I'm not sure about pubg, but I know in Rainbow Six ALL of my cpu cores are at around 50% usage
    while the card is around the same usage 50-60%
    rainbow six siege is utilizing all my cores and none of them are at 100% like I would excpect
    The cpu and gpu only hit 100% when loading in the game and only for a spilt second
     
    #8
    Paris ._.
    New Member
    • Total Posts : 6
    • Reward points : 0
    • Joined: 2017/10/24 12:03:51
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/18 16:00:09 (permalink)
    demon09
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k

    Good advice!
    Superposition gave me a score of 9053 at 4k optimized
    card did hit 99% utilization during that...
     
    Must mean ryzen 7 is crap
     
    Can anyone recommend a better cpu that might be worth buying to play at high refresh rated for 1080p
    don't want to waste money if it not going to help all that much but I will think about it
    I'm not afraid to overclock at all btw
     
    #9
    CriticalHit_NL
    iCX Member
    • Total Posts : 404
    • Reward points : 0
    • Joined: 2009/09/08 10:23:06
    • Status: offline
    • Ribbons : 17
    Re: 1080ti FTW3 not performing as expected 2017/11/18 16:20:54 (permalink)
    demon09
    for games that prefer clock speed the 1700 is below the 6700k and 7700k. It has a single core score of about 160ish where as a 7700k at 4.9ghz is about 200ish and and a 6700k at 4.5 is probably around 190. Hopefully we see games move to better use of more cores with Intel finally jumping on board it's more likely. It's single core score is about a stock 4770k which isn't bad but high refresh rate gaming on some games is very cpu bound

    Ah my bad I thought it was a little bit better than that, but yes I am aware the current AMD Ryzens are still performing below the Intels when it comes to gaming.
    And yeah you're right, and with some games it's simply mission-impossible.
     
    @Paris ._. I can't help with overclocking since I don't do any of that, but a 7700K or 8700K would be a reasonable increase for gaming, but still don't expect maxing out that Hz range, you're only getting closer.
     
    Here are some comparisons of a 7700K vs a 8700K combined with a 1080Ti reference:
    Assassins Creed Origins: https://www.youtube.com/watch?v=BjwqTXkPrew
    The Witcher 3: https://www.youtube.com/watch?v=Tz7seeIarY8
    PUBG: https://www.youtube.com/watch?v=zn88B1X-Ejk

    i7 3930K 4.4Ghz - EVGA 1080Ti FTW3 - 32GB Corsair Dominator DDR3-1866 - Asus Sabertooth X79 - Samsung 840 Pro 256GB - Samsung 860 Pro 2TB - 3x HGST 7K4000 Deskstar 4TB - 3x HGST Ultrastar He10 8TB - Corsair AX1200i - Asus PG279Q + BenQ BL2411 - Razer Viper Mini/SBF98 - Logitech Z906 & G35 - Windows 10 Professional

    #10
    arestavo
    CLASSIFIED ULTRA Member
    • Total Posts : 6916
    • Reward points : 0
    • Joined: 2008/02/06 06:58:57
    • Location: Through the Scary Door
    • Status: offline
    • Ribbons : 76
    Re: 1080ti FTW3 not performing as expected 2017/11/18 22:45:30 (permalink)
    Paris ._.
    demon09
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k

    Good advice!
    Superposition gave me a score of 9053 at 4k optimized
    card did hit 99% utilization during that...
     
    Must mean ryzen 7 is crap
     
    Can anyone recommend a better cpu that might be worth buying to play at high refresh rated for 1080p
    don't want to waste money if it not going to help all that much but I will think about it
    I'm not afraid to overclock at all btw
     




    The highest single threaded performance that you can currently get is going to be a 7700K or 8700K (or i5 equivalent, but it will hurt you in games that can use more cores).
    #11
    demon09
    FTW Member
    • Total Posts : 1334
    • Reward points : 0
    • Joined: 2016/09/16 21:18:42
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/18 22:59:58 (permalink)
    arestavo
    Paris ._.
    demon09
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k

    Good advice!
    Superposition gave me a score of 9053 at 4k optimized
    card did hit 99% utilization during that...
     
    Must mean ryzen 7 is crap
     
    Can anyone recommend a better cpu that might be worth buying to play at high refresh rated for 1080p
    don't want to waste money if it not going to help all that much but I will think about it
    I'm not afraid to overclock at all btw
     




    The highest single threaded performance that you can currently get is going to be a 7700K or 8700K (or i5 equivalent, but it will hurt you in games that can use more cores).


    Paris ._.
    demon09
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k

    Good advice!
    Superposition gave me a score of 9053 at 4k optimized
    card did hit 99% utilization during that...
     
    Must mean ryzen 7 is crap
     
    Can anyone recommend a better cpu that might be worth buying to play at high refresh rated for 1080p
    don't want to waste money if it not going to help all that much but I will think about it
    I'm not afraid to overclock at all btw
     


    I will do some testing in pubg for you with a 7700k if you want when i get some free time. just send me a pm with your settings you would be running at 1080p. If it was me and my mission was best fps id probably buy the 8700k oh lol noticed you have only three posts just list your settings here for pubg that you currently use. 
    #12
    Paris ._.
    New Member
    • Total Posts : 6
    • Reward points : 0
    • Joined: 2017/10/24 12:03:51
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/19 01:15:04 (permalink)
    demon09
    arestavo
    Paris ._.
    demon09
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k

    Good advice!
    Superposition gave me a score of 9053 at 4k optimized
    card did hit 99% utilization during that...
     
    Must mean ryzen 7 is crap
     
    Can anyone recommend a better cpu that might be worth buying to play at high refresh rated for 1080p
    don't want to waste money if it not going to help all that much but I will think about it
    I'm not afraid to overclock at all btw
     




    The highest single threaded performance that you can currently get is going to be a 7700K or 8700K (or i5 equivalent, but it will hurt you in games that can use more cores).


    Paris ._.
    demon09
    I believe furmark gets gimped via drivers to keep the card from using to much power as Nvidia implmented. If you run superposition benchmark at 4k preset what score do you get? Both pubg and rainbow are games that don't easily run at high fps. With your card being at such a low utilization it would point to some bottle neck. Since ryzen has 16 threads and games often don't utilze cores well you can get bottlenecks no where near 100% usage on your cpu. But run super postion at 4k too see if it's in the range it should be it should max your card out at 4k and remove CPU bottle necks. If your card doesn't hit 99% in that then it may be something else then a bottle neck k

    Good advice!
    Superposition gave me a score of 9053 at 4k optimized
    card did hit 99% utilization during that...
     
    Must mean ryzen 7 is crap
     
    Can anyone recommend a better cpu that might be worth buying to play at high refresh rated for 1080p
    don't want to waste money if it not going to help all that much but I will think about it
    I'm not afraid to overclock at all btw
     


    I will do some testing in pubg for you with a 7700k if you want when i get some free time. just send me a pm with your settings you would be running at 1080p. If it was me and my mission was best fps id probably buy the 8700k oh lol noticed you have only three posts just list your settings here for pubg that you currently use. 


    I play everything on ultra screen scale at 100
    #13
    Paris ._.
    New Member
    • Total Posts : 6
    • Reward points : 0
    • Joined: 2017/10/24 12:03:51
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/19 03:23:29 (permalink)
    thanks guys for your help probably going to save up for a better processor and use the ryzen for a non gaming workstation or a future gaming cpu when games utilize more cores more efficiently
     
    one more question is the 8700k worth it over the 7700k?
     
    on youtube ive seen the 7700k get up to over 5ghz overclocked...
     
    that's fairly impressive
     
    it's also a lot cheaper than the 8700k. Is anyone running a overclocked 7700k that can chime in, and tell me about what it takes to cool one of those?
     
    I can go to a different forum if needed lol
    #14
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/19 06:07:31 (permalink)
    rjohnson11
    I'll add a +1 to demon09's post about furmark. It's a horrible benchmark and you shouldn't use it.




    Make that a +2. This was implemented approx 5 years or so ago? Before that it could really damage GPUs if stability testing. They also had an EXTREME setting with a warning. Insane. Those conditions could never exist in any game or app.

    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #15
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/19 06:11:27 (permalink)
    Paris ._.
    thanks guys for your help probably going to save up for a better processor and use the ryzen for a non gaming workstation or a future gaming cpu when games utilize more cores more efficiently
     
    one more question is the 8700k worth it over the 7700k?
     
    on youtube ive seen the 7700k get up to over 5ghz overclocked...
     
    that's fairly impressive
     
    it's also a lot cheaper than the 8700k. Is anyone running a overclocked 7700k that can chime in, and tell me about what it takes to cool one of those?
     
    I can go to a different forum if needed lol




    Absolutely 100%. For a start 8700k is 6+6 cores. Said to OC better, and favourable reviews.
    The 7700k may be a bit cheaper? But forget it. Doesn't even compare. For a start 2 real cores less (4+4). Lower per thread figures. OCs Ok but not like the Coffee lake. Seriously, buying a kaby or skylake now would be a real mistake.
     
    I this info. from reading quite a few review sites - good ones.
     
    Another thing I read was that even with a poor "silicon lottery result," I't pretty much a given that the 8700k can be permanently set to 5GHz. When I get on next year, I expect that. My goal will be 5.1GHz everyday, and if lucky, if heat is under control, if totally stable I'll put together a permanent 5.2GHz setup. I don't expect that though. Just over 5GHz, absolutely.
     
    I'll say again, you will be doing your self a big diservice if you go Kaby now that coffee lake is available. 12 threads (6+6) nice. In a few years games will get better at fully using multiple cores too. Partly because AMD chips tend to have a lot of cores.
    post edited by Gawg36 - 2017/11/19 06:17:11

    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #16
    Paris ._.
    New Member
    • Total Posts : 6
    • Reward points : 0
    • Joined: 2017/10/24 12:03:51
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/19 06:27:27 (permalink)
    Gawg36
    Paris ._.
    thanks guys for your help probably going to save up for a better processor and use the ryzen for a non gaming workstation or a future gaming cpu when games utilize more cores more efficiently
     
    one more question is the 8700k worth it over the 7700k?
     
    on youtube ive seen the 7700k get up to over 5ghz overclocked...
     
    that's fairly impressive
     
    it's also a lot cheaper than the 8700k. Is anyone running a overclocked 7700k that can chime in, and tell me about what it takes to cool one of those?
     
    I can go to a different forum if needed lol




    Absolutely 100%. For a start 8700k is 6+6 cores. Said to OC better, and favourable reviews.
    The 7700k may be a bit cheaper? But forget it. Doesn't even compare. For a start 2 real cores less (4+4). Lower per thread figures. OCs Ok but not like the Coffee lake. Seriously, buying a kaby or skylake now would be a real mistake.
     
     
     
     on
    I this info. from reading quite a few review sites - good ones.
     
    Another thing I read was that even with a poor "silicon lottery result," I't pretty much a given that the 8700k can be permanently set to 5GHz. When I get on next year, I expect that. My goal will be 5.1GHz everyday, and if lucky, if heat is under control, if totally stable I'll put together a permanent 5.2GHz setup. I don't expect that though. Just over 5GHz, absolutely.
     
    I'll say again, you will be doing your self a big diservice if you go Kaby now that coffee lake is available. 12 threads (6+6) nice. In a few years games will get better at fully using multiple cores too. Partly because AMD chips tend to have a lot of cores.




     
    well that makes sense. I don't plan to buy anything right away but within the next month or so probably. Thanks for your insight more than likely going to get a 8700k fairly soon
    #17
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/19 07:19:27 (permalink)
    Paris ._.
    Gawg36
    Paris ._.
    thanks guys for your help probably going to save up for a better processor and use the ryzen for a non gaming workstation or a future gaming cpu when games utilize more cores more efficiently
     
    one more question is the 8700k worth it over the 7700k?
     
    on youtube ive seen the 7700k get up to over 5ghz overclocked...
     
    that's fairly impressive
     
    it's also a lot cheaper than the 8700k. Is anyone running a overclocked 7700k that can chime in, and tell me about what it takes to cool one of those?
     
    I can go to a different forum if needed lol




    Absolutely 100%. For a start 8700k is 6+6 cores. Said to OC better, and favourable reviews.
    The 7700k may be a bit cheaper? But forget it. Doesn't even compare. For a start 2 real cores less (4+4). Lower per thread figures. OCs Ok but not like the Coffee lake. Seriously, buying a kaby or skylake now would be a real mistake.
     
     
     
     on
    I this info. from reading quite a few review sites - good ones.
     
    Another thing I read was that even with a poor "silicon lottery result," I't pretty much a given that the 8700k can be permanently set to 5GHz. When I get on next year, I expect that. My goal will be 5.1GHz everyday, and if lucky, if heat is under control, if totally stable I'll put together a permanent 5.2GHz setup. I don't expect that though. Just over 5GHz, absolutely.
     
    I'll say again, you will be doing your self a big diservice if you go Kaby now that coffee lake is available. 12 threads (6+6) nice. In a few years games will get better at fully using multiple cores too. Partly because AMD chips tend to have a lot of cores.


    Thanks for the reply. From what I've read (a lot) you will be impressed. As stated I am going to purchase one early next year for my new 'Mega build!"
     
    well that makes sense. I don't plan to buy anything right away but within the next month or so probably. Thanks for your insight more than likely going to get a 8700k fairly soon





    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #18
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/19 07:21:45 (permalink)
    @PARIS, sorry, something weird about my last reply format. Anyway, just in case I wanted to thank you for the reply, and say that I'm sure you will be happy with it. I am getting one early next year!
     

    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #19
    demon09
    FTW Member
    • Total Posts : 1334
    • Reward points : 0
    • Joined: 2016/09/16 21:18:42
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/19 08:05:27 (permalink)
    If you want to oc the 8700k be ready to need a delid from what I have seen it follows suit on the 7700ks bad cooling .
    #20
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/19 09:25:46 (permalink)
    demon09
    If you want to oc the 8700k be ready to need a delid from what I have seen it follows suit on the 7700ks bad cooling .

    Yeah, Intel dam cheap skates. All CPUs since SandyBridge i7 2700k etc have used thermal paste rather than soldering. (I mean Sandy was soldered, next Ivy to now, not soldered. True.)
    But many people including myself can still OC their CPUs. Granted, if soldered, or delided and soldered, the OC would be more. But for moderate OCing, it's fine. My Ivy is i5 3570k is over 5 years old.
    It has been running at 4.8GHz, now 4.75GHz for 5 years. It never overheats, and has never caused any trouble. Of course  that's why we have liquid coolers. I use H100i water rad. with my i5.
     
    Deliding WILL give better OCs, but it's only for those with very steady hands, and those who trying to achieve something outstanding. I've never done it, and am too scared to perform this operation.
     
    So, what you said is true for large OCs. But a Coffee lake at 5GHz with liquid cooling and not delided will be no problem at all. (I've read several reviews, Kit guru etc, and they never delid as it only applies to a tiny proportion of overclockers.) Dangerous stuff!! Easy to ruin the CPU.
    It's a mystery why Intel try to save a few cents by using paste instead of a proper solder.
    The vast majority of folks who OC their CPUs do not delid them. That's for the real enthusiasts or professional overclockers.

    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #21
    demon09
    FTW Member
    • Total Posts : 1334
    • Reward points : 0
    • Joined: 2016/09/16 21:18:42
    • Status: offline
    • Ribbons : 0
    Re: 1080ti FTW3 not performing as expected 2017/11/19 10:01:46 (permalink)
    Gawg36
    demon09
    If you want to oc the 8700k be ready to need a delid from what I have seen it follows suit on the 7700ks bad cooling .

    Yeah, Intel dam cheap skates. All CPUs since SandyBridge i7 2700k etc have used thermal paste rather than soldering. (I mean Sandy was soldered, next Ivy to now, not soldered. True.)
    But many people including myself can still OC their CPUs. Granted, if soldered, or delided and soldered, the OC would be more. But for moderate OCing, it's fine. My Ivy is i5 3570k is over 5 years old.
    It has been running at 4.8GHz, now 4.75GHz for 5 years. It never overheats, and has never caused any trouble. Of course  that's why we have liquid coolers. I use H100i water rad. with my i5.
     
    Deliding WILL give better OCs, but it's only for those with very steady hands, and those who trying to achieve something outstanding. I've never done it, and am too scared to perform this operation.
     
    So, what you said is true for large OCs. But a Coffee lake at 5GHz with liquid cooling and not delided will be no problem at all. (I've read several reviews, Kit guru etc, and they never delid as it only applies to a tiny proportion of overclockers.) Dangerous stuff!! Easy to ruin the CPU.
    It's a mystery why Intel try to save a few cents by using paste instead of a proper solder.
    The vast majority of folks who OC their CPUs do not delid them. That's for the real enthusiasts or professional overclockers.
    a guy on here showed 92c on a run of cinnbench at 5ghz on his 8700k under a liquid cooler.but if someone is ok with those temps then no need to delid I gusse. It was a similar story for my 7700k. I5's ussaly dont need the delid to overclock as much as there i7 counter parts
    #22
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/20 04:20:56 (permalink)
    demon09
    Gawg36
    demon09
    If you want to oc the 8700k be ready to need a delid from what I have seen it follows suit on the 7700ks bad cooling .

    Yeah, Intel dam cheap skates. All CPUs since SandyBridge i7 2700k etc have used thermal paste rather than soldering. (I mean Sandy was soldered, next Ivy to now, not soldered. True.)
    But many people including myself can still OC their CPUs. Granted, if soldered, or delided and soldered, the OC would be more. But for moderate OCing, it's fine. My Ivy is i5 3570k is over 5 years old.
    It has been running at 4.8GHz, now 4.75GHz for 5 years. It never overheats, and has never caused any trouble. Of course  that's why we have liquid coolers. I use H100i water rad. with my i5.
     
    Deliding WILL give better OCs, but it's only for those with very steady hands, and those who trying to achieve something outstanding. I've never done it, and am too scared to perform this operation.
     
    So, what you said is true for large OCs. But a Coffee lake at 5GHz with liquid cooling and not delided will be no problem at all. (I've read several reviews, Kit guru etc, and they never delid as it only applies to a tiny proportion of overclockers.) Dangerous stuff!! Easy to ruin the CPU.
    It's a mystery why Intel try to save a few cents by using paste instead of a proper solder.
    The vast majority of folks who OC their CPUs do not delid them. That's for the real enthusiasts or professional overclockers.
    a guy on here showed 92c on a run of cinnbench at 5ghz on his 8700k under a liquid cooler.but if someone is ok with those temps then no need to delid I gusse. It was a similar story for my 7700k. I5's ussaly dont need the delid to overclock as much as there i7 counter parts

    The case you mention is unacceptable of course. Many possible reasons, too much voltage for a start, but many more. If the temp was 92c at 5.2GHz then yes, deliding may well be the only solution.
    But for 5GHz stable reasonable heat OCs (with liquid double fan large rad cooling of course) Should be attainable with an "average," piece of silicon. Above 5GHz would be specualtion. I don't own this CPU, simply repeating what I have read from pretty much all the large sites. You may be correct? Not worth arguing about really.
    Kaby lake, haven't read about that one for a while so I'll agree with you re: Kaby.

    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #23
    Gawg36
    iCX Member
    • Total Posts : 376
    • Reward points : 0
    • Joined: 2016/07/28 15:58:52
    • Location: Japan (Work) U.K (home)
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/21 05:51:05 (permalink)
    @Demon09,
     
    I've spoken with a friend and researched a bit more. It's only fair for me to admit I was too gung ho. You were right, sorry.  

    1. Taichi z370. i7 8086k @ 5,2GHz Stable.
    16GB G.Skill Trident Z @ 3,000MHz.
      EVGA GTX 1080ti  FTW3. Acer XB270HU IPS 1440p 144Hz Refresh with G-Sync. PSU Corsair AX850. Cloud two cans, and Creative T20 stereo. Realtek HD on board sound.
     
    #24
    w318ti
    New Member
    • Total Posts : 84
    • Reward points : 0
    • Joined: 2017/04/24 01:16:45
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/21 08:25:40 (permalink)
    Couple of notes:
     
    1. PUBG is not very well optimized. The frame rates you are seeing sound about right.
    2. At 1080p resolution, you will run into CPU bottlenecking regardless of using AMD or Intel.
    3. Before you run out and get a new processor and motherboard, test the games you're playing at higher resolution and see if the GPU utilization and performance increases. Try using DSR (dynamic super resolution) to mimic a higher resolution if you only have a 1080p display. You should see performance increases.
    #25
    CriticalHit_NL
    iCX Member
    • Total Posts : 404
    • Reward points : 0
    • Joined: 2009/09/08 10:23:06
    • Status: offline
    • Ribbons : 17
    Re: 1080ti FTW3 not performing as expected 2017/11/21 09:12:49 (permalink)
    w318ti
    Couple of notes:
     
    1. PUBG is not very well optimized. The frame rates you are seeing sound about right.
    2. At 1080p resolution, you will run into CPU bottlenecking regardless of using AMD or Intel.
    3. Before you run out and get a new processor and motherboard, test the games you're playing at higher resolution and see if the GPU utilization and performance increases. Try using DSR (dynamic super resolution) to mimic a higher resolution if you only have a 1080p display. You should see performance increases.

    You'll see a higher GPU utilization and possibly less high framerate, it will not increase his framerate any further because this is a CPU bottleneck.
    If he wishes to get more fps to get more out of his 240Hz monitor currently the only way is to get a better processor.

    i7 3930K 4.4Ghz - EVGA 1080Ti FTW3 - 32GB Corsair Dominator DDR3-1866 - Asus Sabertooth X79 - Samsung 840 Pro 256GB - Samsung 860 Pro 2TB - 3x HGST 7K4000 Deskstar 4TB - 3x HGST Ultrastar He10 8TB - Corsair AX1200i - Asus PG279Q + BenQ BL2411 - Razer Viper Mini/SBF98 - Logitech Z906 & G35 - Windows 10 Professional

    #26
    w318ti
    New Member
    • Total Posts : 84
    • Reward points : 0
    • Joined: 2017/04/24 01:16:45
    • Status: offline
    • Ribbons : 1
    Re: 1080ti FTW3 not performing as expected 2017/11/21 09:58:19 (permalink)
    CriticalHit_NL
    w318ti
    Couple of notes:
     
    1. PUBG is not very well optimized. The frame rates you are seeing sound about right.
    2. At 1080p resolution, you will run into CPU bottlenecking regardless of using AMD or Intel.
    3. Before you run out and get a new processor and motherboard, test the games you're playing at higher resolution and see if the GPU utilization and performance increases. Try using DSR (dynamic super resolution) to mimic a higher resolution if you only have a 1080p display. You should see performance increases.

    You'll see a higher GPU utilization and possibly less high framerate, it will not increase his framerate any further because this is a CPU bottleneck.
    If he wishes to get more fps to get more out of his 240Hz monitor currently the only way is to get a better processor.


    Please explain to me what is better about the processors recommended here vs. what the OP is currently using. Higher clock speed?
     
    Neither the GPU or CPU is being fully utilized in the OP's case. So why not run some tests where the GPU is closer to 100% utilization and see how the performance changes. 
     
    Using DSR or running benchmarks at higher resolutions is free. Upgrading the CPU and mobo not so much.
     
    BTW I don't agree with you that framerate will not increase. For example in Unigine Superposition you almost always see lower scores and framerates with a 1080Ti on the 1080 resolution tests vs the 4k test. Same reason... the GPU is not allowed to be 100% utilized.
     
     
    #27
    CriticalHit_NL
    iCX Member
    • Total Posts : 404
    • Reward points : 0
    • Joined: 2009/09/08 10:23:06
    • Status: offline
    • Ribbons : 17
    Re: 1080ti FTW3 not performing as expected 2017/11/21 13:38:03 (permalink)
    w318ti
     
    Please explain to me what is better about the processors recommended here vs. what the OP is currently using. Higher clock speed?
     
    Neither the GPU or CPU is being fully utilized in the OP's case. So why not run some tests where the GPU is closer to 100% utilization and see how the performance changes. 
     
    Using DSR or running benchmarks at higher resolutions is free. Upgrading the CPU and mobo not so much.
     
    BTW I don't agree with you that framerate will not increase. For example in Unigine Superposition you almost always see lower scores and framerates with a 1080Ti on the 1080 resolution tests vs the 4k test. Same reason... the GPU is not allowed to be 100% utilized.

    I'm not going to be arguing here but Unigine Superposition is not a game but a synthetic test, which is by a long shot not a relevant test for practical gaming circumstances.
    Besides the 4K test is with optimized settings and thus less heavy than the extreme setting 1080P test, if you're talking about the Extreme test on 1080P that is.
     
    In both situations my 1080Ti is fully loaded in those tests, GPU temperature confirms this.
    1080P extreme:
    Min: 35.88
    Avg: 44.16
    Max: 52.19
     
    4K Optmized:
    Min: 57.64
    Avg: 72.96
    Max: 94.98
     
    Turning on DSR (supersampling) is simply giving the GPU a higher workload because of a higher resolution, this does not increase the framerate at the same settings, but only stay roughly the same given the GPU load is below 99-100% or decrease.
    If you're going to increase the resolution the GPU utilization will ofcourse increase, but if you get a lower framerate as a result the CPU utilization will also be lower because there's less drawcalls per second.
     
    And no it's not simply for the clockspeeds, the Intel processors simply have a better IPC (Instructions Per Cycle), which means more performance per clockcycle.
    That means Intel could do with less high clockspeeds to reach the same performance level, and you guess what happens if they increased the clockspeed even further.
     
    The problem with games is that they are often single-thread bottlenecked and most games are DirectX 11 where this is the biggest issue, and they only utilize limited CPU threads in most cases.
    It means that even though you have many CPU cores, you can still have a bottleneck yet not reaching 100% CPU.
     
    Though that will likely happen on regular quad core i5 processors because games often use 4 CPU cores these days, but since the main thread is usually having the highest load on most games you usually do not reach 100% on all cores, the main thread is bottlenecking the rest.
     
    But in the quad core i7 case that means it's performing on it's peak at 50%~ in a game, whatever you do you won't see it reach 100% ever because the extra logical threads end up unused.
    Even if the CPU load is balanced over all the logical threads, it still performs like the performance of 4 physical CPU threads.
     
    Unless you have something like a dual core i3 with HTT, which has two physical cores and two extra threads, making a total of 4, then games will take advantage from the extra logical HTT threads and thus a higher load is possible. (because the games support 4 threads)
     
    Now the same is true for processors that usually use more than 4 physical cores, games end up mostly using only 4 of them and the rest is either idling or used for other background tasks. (less game interference because more CPU buffer, yay for that)
    Some games however are pretty well optimized and can take advantage of more than 4 threads, so seeing 50%~ usage on a 6-core i7 is then also possible.
     
    You surely don't want to go near 100% of total CPU load either way.
    Because games have a very unstable workload, it rapidly changes from low to high usage and whatever you see inside taskmanager is not the whole picture.
    Taskmanager only refreshes at a high interval, while the CPU is doing much more in the background per second, so you can't really tell it never hits the ceiling except when noticing very short hiccups in the games. (that is pretty much always a CPU load spike)
     
    And the higher the framerate the more performance your CPU needs to deliver to feed the GPU enough juice to keep using it's full potential.
    The required CPU performance scales lineair with the amount of fps you want to render in DirectX 11's case, so every bit of power is very welcome.
     
    And DirectX 11 is still very relevant today, looking at the amount of games being made and old games being available versus Vulkan / DirectX 12 games it's night and day.
    I would rather have something that performs on the bulk of the games rather than a few titles.
     
    Edit: here are some benchmark comparisons:
     
    Battlefield 1 (DX12):
    https://nl.hardware.info/reviews/7588/16/intel-core-i7-8700k--i5-8600k--i5-8400-coffee-lake-review-betaalbaar-zes-cores-gaming-benchmarks-gtx-1080-ti-battlefield-1-dx12
    Doom 2016 (Vulkan):
    https://nl.hardware.info/reviews/7588/17/intel-core-i7-8700k--i5-8600k--i5-8400-coffee-lake-review-betaalbaar-zes-cores-gaming-benchmarks-gtx-1080-ti-doomnvulkan
    GTA V (DX11):
    https://nl.hardware.info/reviews/7588/18/intel-core-i7-8700k--i5-8600k--i5-8400-coffee-lake-review-betaalbaar-zes-cores-gaming-benchmarks-gtx-1080-ti-gta-v
    Prey (DX11):
    https://nl.hardware.info/reviews/7588/19/intel-core-i7-8700k--i5-8600k--i5-8400-coffee-lake-review-betaalbaar-zes-cores-gaming-benchmarks-gtx-1080-ti-prey
    Rise of the Tombraider (DX12):
    [link=https://nl.hardware.info/reviews/7588/20/intel-core-i7-8700k--i5-8600k--i5-8400-coffee-lake-review-betaalbaar-zes-cores-gaming-benchmarks-gtx-1080-ti-rise-of-the-tomb-raider-dx12]https://nl.hardware.info/...f-the-tomb-raider-dx12[/link]
    post edited by CriticalHit_NL - 2017/11/21 13:51:51

    i7 3930K 4.4Ghz - EVGA 1080Ti FTW3 - 32GB Corsair Dominator DDR3-1866 - Asus Sabertooth X79 - Samsung 840 Pro 256GB - Samsung 860 Pro 2TB - 3x HGST 7K4000 Deskstar 4TB - 3x HGST Ultrastar He10 8TB - Corsair AX1200i - Asus PG279Q + BenQ BL2411 - Razer Viper Mini/SBF98 - Logitech Z906 & G35 - Windows 10 Professional

    #28
    Jump to:
  • Back to Mobile