EVGA

3090 FTW3 very bad

Author
andi0139
New Member
  • Total Posts : 89
  • Reward points : 0
  • Joined: 2021/06/17 02:54:39
  • Status: offline
  • Ribbons : 0
2022/03/20 10:25:37 (permalink)
Hey guys, so I got my 3090 Ftw3 from step up and played a bit around with it. When I set my card to 119% Powerlimit = 500 Watts the boosts starts at 1965 mhz and goes down to 1890 mhz fast as soon as the temps rise. I know 1890 ist more than the advertised 1800 mhz but dang, from a 500 watts i expected the card to boost to around 2010 mhz and going down to around 1950 mhz when temps rise. Overclockingwise i can't go past + 120 on core in timespy which results in dissappointing 2055 going down to 2010 mhz when temps rise and the memory doesnt go further than +500 also very disappointing. I already tried undervolting and the card cant even manage 1860 mhz @ 0.9, i have to go down to 1800 mhz @ 0.9v. I have never had such a bad 3090 till yet. Also I have never seen the card using more than 470 watts. In games its most of the time around 430 to 460w (playing in 4k) Has anyone else had such a bad card? Other cards I had could easily hit 2100 mhz or more and +1000 or even more on memory. Is Nvidia keeping the good chips for the 90ti and only puts chips on 3090 or is there something else going on with my card?
post edited by the_Scarlet_one - 2022/03/21 04:27:21
#1

8 Replies Related Threads

    Cool GTX
    EVGA Forum Moderator
    • Total Posts : 31353
    • Reward points : 0
    • Joined: 2010/12/12 14:22:25
    • Location: Folding for the Greater Good
    • Status: offline
    • Ribbons : 123
    Re: 3090 FTW3 very bad 2022/03/20 10:59:15 (permalink)
    Rig details?
    PSU ... brand, size & age
    Nvidia drive & Control P settings
    Room temp
    Case & case fans & fan speeds
    Sensor tab of GPU-Z while benchmarking would go a long way in to helping members have a much better picture of your results. 
     
    You might even want to post some benchmark links
     
    OK, you did not win the silicon lottery ... OC is never guaranteed
     
    Watts drawn is a bad way to compare performance of cards ...it only draws what it needs for the task, until a Perf-Cap is hit or NVIDIA Boost Control overrides to manage the load & temps
     
    What is the Perf-Cap reason shown in GPU-Z ?
     
    NVIDIA Boost technology always has the last word
     
    What about the stability of your OC on the rest of your Rig  - did you retest stability after changing the GPU ?

    Learn your way around the EVGA Forums, Rules & limits on new accounts Ultimate Self-Starter Thread For New Members

    I am a Volunteer Moderator - not an EVGA employee

    Older RIG projects RTX Project  Nibbler


     When someone does not use reason to reach their conclusion in the first place; you can't use reason to convince them otherwise!
    #2
    andi0139
    New Member
    • Total Posts : 89
    • Reward points : 0
    • Joined: 2021/06/17 02:54:39
    • Status: offline
    • Ribbons : 0
    Re: 3090 FTW3 very bad 2022/03/20 11:11:05 (permalink)
    Asus Z690 Formula
    I9 12900k
    Corsair Dominator 6200 mhz Cl36 DDR5
    Seasonic 1300 Watts Platinum
    1 TB Samsung 980 Pro
    500 GB Samsung 980 Pro
    Lian Li O11 Dynamic Evo
    Nzxt Kraken z73 Aio
    10x Lian Li Al120 Casefans

    Asus Rog Swift PG32UQ 4k144hz Gsync Monitor

    Thanks, so my timespy graphics score is 21505, my previous cards had more or around 22k graphics score there. Perfcap reason is PWR even though it never shows 500 Watts. It already shows pwr at around 440 Watts. Pcie Slot never saw more than 66 Watts so it runs within specs. GPU Temp max was 74 degrees in Timespy with fans set to 100 percent . In Games it reaches 78 - 80 degress with 119% Powerlimit and stock fancurve

    In afterburner i set + 105 core + 500 memory (anything above crash) +120 is only possible with no memory oc.

    In games + 90 core is stable, everything above crash

    Nvidia Driver ist the newest available
    Control P settings in Timespy was prefer max performance

    Roomtemp around 23 degrees

    Memory Temp never went above 80 degrees.

    I have my 12900k running stock with only xmp enabled for the ram
    #3
    andi0139
    New Member
    • Total Posts : 89
    • Reward points : 0
    • Joined: 2021/06/17 02:54:39
    • Status: offline
    • Ribbons : 0
    Re: 3090 FTW3 very bad 2022/03/20 11:14:09 (permalink)
    Cool GTX
    Rig details?
    PSU ... brand, size & age
    Nvidia drive & Control P settings
    Room temp
    Case & case fans & fan speeds
    Sensor tab of GPU-Z while benchmarking would go a long way in to helping members have a much better picture of your results. 
     
    You might even want to post some benchmark links
     
    OK, you did not win the silicon lottery ... OC is never guaranteed
     
    Watts drawn is a bad way to compare performance of cards ...it only draws what it needs for the task, until a Perf-Cap is hit or NVIDIA Boost Control overrides to manage the load & temps
     
    What is the Perf-Cap reason shown in GPU-Z ?
     
    NVIDIA Boost technology always has the last word
     
    What about the stability of your OC on the rest of your Rig  - did you retest stability after changing the GPU ?

    Asus Z690 Formula
    I9 12900k
    Corsair Dominator 6200 mhz Cl36 DDR5
    Seasonic 1300 Watts Platinum
    1 TB Samsung 980 Pro
    500 GB Samsung 980 Pro
    Lian Li O11 Dynamic Evo
    Nzxt Kraken z73 Aio
    10x Lian Li Al120 Casefans

    Asus Rog Swift PG32UQ 4k144hz Gsync Monitor

    Thanks, so my timespy graphics score is 21505, my previous cards had more or around 22k graphics score there. Perfcap reason is PWR even though it never shows 500 Watts. It already shows pwr at around 440 Watts. Pcie Slot never saw more than 66 Watts so it runs within specs. GPU Temp max was 74 degrees in Timespy with fans set to 100 percent . In Games it reaches 78 - 80 degress with 119% Powerlimit and stock fancurve

    In afterburner i set + 105 core + 500 memory (anything above crash) +120 is only possible with no memory oc.

    In games + 90 core is stable, everything above crash

    Nvidia Driver ist the newest available
    Control P settings in Timespy was prefer max performance

    Roomtemp around 23 degrees

    Memory Temp never went above 80 degrees.

    I have my 12900k running stock with only xmp enabled for the ram
    #4
    kraade
    CLASSIFIED Member
    • Total Posts : 2433
    • Reward points : 0
    • Joined: 2020/10/18 09:05:52
    • Location: Georgia
    • Status: offline
    • Ribbons : 7
    Re: 3090 FTW3 very bad 2022/03/20 11:26:18 (permalink)
    I can make my kingpin pull more than 470 watts but the card slows down , your card seems to be average and you had good samples before
    #5
    rjbarker
    CLASSIFIED Member
    • Total Posts : 2361
    • Reward points : 0
    • Joined: 2008/03/20 10:07:05
    • Location: Vancouver Isle - Westcoast Canada
    • Status: offline
    • Ribbons : 21
    Re: 3090 FTW3 very bad 2022/03/20 15:10:56 (permalink)
    Is your Dominator 6200 C36 100% stable? I couldnt get mine stable at all at that speed....one day it would pass Memtest and the next day would fail.
    Also when it did pass Memtest I would get random crashes gaming.
     
    The Dominator Kit I have running at 6000 C32 and is rock solid stable.
    Check that it is 100% stable or try setting to default 4800 and then play with GPU OC'ing.....

    I9 12900K EK Velocity2 / ROG Z690 Apex/ 32G Dominator DDR5 6000/ Evga RTX 3080Ti FTW3  EK Vector / 980 Pro 512G / 980 Pro 1TB/ Samsung 860 Pro 500G/ WD 4TB Red / AX 1600i /  Corsair 900D & XSPC 480 * 360 * 240 Rads   XSPC Photon 170 Rez-Vario Pump Combo - Alienware 3440*1440p 120Hz/ W11
     
    #6
    andi0139
    New Member
    • Total Posts : 89
    • Reward points : 0
    • Joined: 2021/06/17 02:54:39
    • Status: offline
    • Ribbons : 0
    Re: 3090 FTW3 very bad 2022/03/20 15:45:27 (permalink)
    rjbarker
    Is your Dominator 6200 C36 100% stable? I couldnt get mine stable at all at that speed....one day it would pass Memtest and the next day would fail.
    Also when it did pass Memtest I would get random crashes gaming.
     
    The Dominator Kit I have running at 6000 C32 and is rock solid stable.
    Check that it is 100% stable or try setting to default 4800 and then play with GPU OC'ing.....


    I dont know if its stable, memtest had no errors for me so i believe yes.

    Can you show me a picture of your 6000 cl32 settings please :)
    #7
    rjbarker
    CLASSIFIED Member
    • Total Posts : 2361
    • Reward points : 0
    • Joined: 2008/03/20 10:07:05
    • Location: Vancouver Isle - Westcoast Canada
    • Status: offline
    • Ribbons : 21
    Re: 3090 FTW3 very bad 2022/03/20 22:21:07 (permalink)
    Dominator 6200 C36 - could not for the life of me get it stable at 6200 C36
    Change Memory: 6000 C32-35-35-76 2T
     
    Set to XMP I then change to 6000 Mhz, then set Timings manual , I applied the following voltages:
    VDD/Q 1.325v VDDQ Transmit 1.325v Memory Controller 1.285v SA 1.22v
     
    BIOS 1302
     

    I9 12900K EK Velocity2 / ROG Z690 Apex/ 32G Dominator DDR5 6000/ Evga RTX 3080Ti FTW3  EK Vector / 980 Pro 512G / 980 Pro 1TB/ Samsung 860 Pro 500G/ WD 4TB Red / AX 1600i /  Corsair 900D & XSPC 480 * 360 * 240 Rads   XSPC Photon 170 Rez-Vario Pump Combo - Alienware 3440*1440p 120Hz/ W11
     
    #8
    03whitegsr
    Superclocked Member
    • Total Posts : 160
    • Reward points : 0
    • Joined: 2021/01/10 20:11:36
    • Status: offline
    • Ribbons : 1
    Re: 3090 FTW3 very bad 2022/03/21 16:30:02 (permalink)
    I wouldn't worry too much about GPU memory OC for games, it does VERY little. Also, I've noticed Afterburner seems to want to crash 3D mark when messing with memory OCs. I never have both programs going at the same time when setting memory speeds and it helps greatly.

    It's likely crashing when load drops and clocks jump up. Limit the maximum frequency. If I were you, if start at the stock max frequency of 1965MHz. Do like +150 on the offset, apply it then go into the F/V curve and flatten the curve out once it hits 1965MHz. If that's stable, push the offset higher but keep it capped at the same max frequency. If it's not stable, lower the offset and try again. There are quick ways to do this, reach point does not have to be set individually.

    Doing this will minimize power use and temperatures, which will keep the boost clocks high. It will outperform stock settings while using a lot less power and run a lot cooler. You can use this same setup too go higher max frequency, but it will be diminishing returns.
    post edited by 03whitegsr - 2022/03/21 16:34:50
    #9
    Jump to:
  • Back to Mobile