Jump to content
Search In
  • More options...
Find results that contain...
Find results in...

SkilledRebuilds

Member
  • Content Count

    12,497
  • Joined

  • Last visited

Everything posted by SkilledRebuilds

  1. Redo Front Panel Power Connectors after rechecking cables inside. Sometimes our fiddly fingers and eyes insert them backwards. Take usual precautionary steps with power on standby vs off at the PSU/Wall. You don't want it starting with your hands in there so remove power before going back in. Best of luck, I'm sure more suggestions will come soon.
  2. I'd choose myself... the 8Core+3070 vs 12core+3060 By the time you'd need to replace either the 8core or 12 core, new stuff exists (trade up, sell old, offset new) 3070 is always a nice to have.. depends on games you play though... either way, both CPU's will suffice for your needs (depending how heavy the needs truly are) 3070 and 4K, just be realistic.. High settings, Not Ultra. or a Mix or MediumHighUltra that you barely notice downgrades (optimization videos exist for video games) Whatever you get at 4K, pretty much doubles at 1440p (given no CPU/RamLatency bottleneck)
  3. SkilledRebuilds

    4/15/2021 No patience today... but here is my U…

    Tomorrows Target for Stability testing.
  4. Just a part of how some effects can still induce fps drops, non-solid effects still have to render things behind the non-solid, so much tiny variation/lighting/shadows/depth.. So when you go point blank drilling an enemy with heaps of blood splatter, if FPS chews down HARD, it's likely the transparency effects (Fire/Blood/Glass/Water/Smoke) and such where you can still see through these things, rendering a multitude of depths (each splatter is a new layer) when you go hard at it.
  5. Transparent effects (not solid, see through) like blood particles and water particles can on some games take a huge hit, usually a temporary spike that memory or gpu bandwidth can't overcome.
  6. 4/15/2021

    No patience today... but here is my UEFI.

    Did clock 10x5150Mhz from 10x5107Mhz

    Ring/Cache is 4950Mhz from 4907Mhz

     

    Chasing Core Maximums before I got bored from two hours of stability testing in games and apps.

    DDR4 4446Mhz C181818 37 315 passed Warzone with 3 matches.

    37.5ns huge read and write boosts. 67/69 GB/s

     

    Trying 171818 tomorrow (36.6ns) see if its stable.

    It should be... might have to backoff the 315 trfc.

    20210415_215449.jpg

    20210415_215542.jpg

    20210415_221025.jpg

    1. SkilledRebuilds

      SkilledRebuilds

      Tomorrows Target for Stability testing.

      Screenshot 2021-04-15 225512.jpg

  7. I'd use the TV given it'll do the job, Or use NaTIVE 720p/900p on your 1080p monitor (No Res Scale, Manually go to 720p or 900p) When using non-Native RES on TV (even using it natively for GPU scaling not TV scaling) it's best you activate and use this setting in the Nvidia Control Panel for GPU scaling to happen on the SmarterGPU before sending it out, not the DumbTV scaling that everyone complains about across multitude of instances. Let the GPU do the work first and scale accordingly.
  8. High frequency 4x8GB Dual Channel, Dual Rank

    These are 3200C141414 1.35 at 4500C181818 1.5v

    MSI z490 Gaming Plus

    20210414_192700.jpg

  9. If I remember right, at the extreme levels (4.7-4.8+++) people would maybe drop the HT bus speed a step/bin down and increase NB as best you can (also good if motherboard has a NB-Voltage setting) and then Go for a RAM OC/TUNE on top after CPU/NB are done.
  10. Allcore 5.1 is stable so I'm going back to turbo mode but editing it for 53-54x single, 52x3core, 51x4cores or more.

    My 10850k wont do 5.2allcore really..

    So ill try to maximise DX9/11 and Apps Singlethreaded performance now.

  11. I go for Quiet keyboards due to roomies. As long as it has NKEY rollover (2-3+button mashing doesnt stop inputs going in) and doesn't flex too much I'm good. Extra hotkeys a bonus. Also easy to find/replace for whatever reason. I also standby rock solid mechanicals with good reputations, no flex and goodfeels.... due to friend choices and using them as well.
  12. Open Task Manager (Rightclick Taskbar, Open Task Manager) Scroll down the process list to find START, click START to highlight and END TASK, it should automatically reload, and start menu should now work. I END Task on it when it has issues and it usually clears it up.
  13. https://www.anandtech.com/bench/product/2355?vs=2365 2500K vs 3770K https://www.anandtech.com/bench/product/2365?vs=2362 3770K vs Ryzen 2600 (1600AF) https://www.anandtech.com/bench/product/2365?vs=2594 3770K vs Ryzen 3600 https://www.anandtech.com/bench/product/2692?vs=2709 Ryzen 2600 vs Ryzen 3100 CPU 2021 doesn't have the 3770k in the list so I had to use 2019 which then neglects 10series budget. (But these should suffice) I'd spend $200-300 and replatform over $80 sticking to similar platforms, and enjoy MAX performance for years to come. Not having to (twice
  14. 3200C14 at 4200C17 37.7ns and I upped cache from 48x to 49x, keeping my 51x allcore CPU ratio.
    Latency dropped to 37.3ns (every transaction) just from the +100Mhz cache increase on Intel arch.

    I was already at a much higher Read and Write (3200 at 4200) 47GB/s to 55GB/s so I don't mind sacrificing some frequency to allow for lower latencies..
    Might try to see if I can squeeze 4133Mhz to CL16/Tuned for sub37ns...

  15. 1440p Max

    PreOC Specs  i9 10850k 10Core 20 Thread

    z490 MSI Gaming Plus Motherboard

    G.Skill TridentZ 3200Mhz C14

    MSI RTX 2080TI

    Thermaltake 850w PSU

    Older Cosmos S Case by Coolermaster

  16. My 3200c14 at 4200c17 1.5v (now 1.45 but was testing 1.5v) Never broke 32*c in my room. Idles at sub30*c. 60*c seems a bit much..
  17. Then its Win Win. Usable, still comfy. GJ instead of just binning them.
  18. Pointless. Unless ONLY for learning... You wont gain anything worth noticing vs the extra stress on parts (volts or frequencies) that you wont notice,yet could introduce HW errors or Software bugs from instability... That GPU IS LOWEND, it failed to be something better, so it got cut down many iterations lower just to be sellable. Also a reason you wont notice it is your CPU can and likely is a huge bottleneck, even for a GT710 in some cases. But overclocking for "performance" isn't the way to go here. Whatever it struggles with, will struggle the
  19. CPU is too old. The whole platform is outdated for Gaming for many a year... The CPU (Dualcore and Slower than modern dualcores) Just can't power through the huge instruction sets of modern computing very well.
  20. Sorry bud,just saw this, 1am, not likely. There are plenty of videos on the subject to learn how to vizualise this... and then still using my methods. As the post above states, best to guage stock performance and have values ypu can refer to as comparison.
  21. Varies GPU TO GPU. Nvidia uses 15mhz steps so use that going forward... Load up Game in a window. Open OC program. Set PWR TO MAX Initially... +90Mhz to core (6x15Mhz), run EXTENDED TESTING across multiple game scenarios (heavy, light, both) Passed? Good. If not backoff 15Mhz and retest. Rinse and repeat.. If you pass at +90.. Now use +15 in SINGLE STEPS between testing. You might get +120, it might top out higher at +150 or +180 wont know until you test. VRAM - Remove Core OC. Add +200 each step and test.
  22. Doubt it. If you set particular variables... and keep them between systems you'd tend to think its going to encode faster but look the exact same due to your particulars. Specifically... say using a x264 or H265 Preset.. the preset doesnt change so I doubt the video will. Hope I make sense. You said encode time isn't an issue so I guess if I was you I'd leave it until speed or quality is a concern.. then you make the adjustments you need. Be it hardware for faster encodes ,or using differing quality options for encode quality boosts, which you are already
×