Jump to content

PowerBaller

Member
  • Posts

    139
  • Joined

  • Last visited

Awards

This user doesn't have any awards

2 Followers

System

  • CPU
    9700K 5.0Ghz , 1.0 AVX 2 offset ratio, Vcore1.33V, VCCIO 1.262,(Prime 95 Small LTT Passed)
  • Motherboard
    Maximus XI Hero WIFI
  • RAM
    4 x 8 GB G.skill Trident Z 3200Mhz CL14, daily operate at 4000Mhz, 16-16-16-36. 1.44V Dram voltage,1.275V System Agent.
  • GPU
    EVGA 2080Ti FTW3 Ultra
  • Case
    Coolermaster H500P
  • Storage
    970 Pro 512GB + 860 EVO 500GB + 860 EVO 2TB + PMR WD Blue 4TB + 2 x Toshiba N300 8TB (Synology 918+)
  • PSU
    evga 850W G2
  • Display(s)
    Acer XV273K
  • Cooling
    EWKB velocity + EKWB Vector + 2 * Hardware Lab Nemesis GTS 360, parallel loop config.
  • Keyboard
    corsair strafe gen 1
  • Mouse
    logitech G703 hero
  • Sound
    Creative AE-7 + Klipsch ProMedia 2.1
  • Operating System
    Windows 10 Pro
  • Laptop
    MacBook Air 2019

Recent Profile Visitors

707 profile views
  1. I have been helping my friend planning his new build. My own rig includes a water-cooled EVGA 2080 Ti FTW3U. The overclock outcome is satisfactory to me : +100 Mhz on GPU, + 1100 Mhz on Vram, +50mv on GPU Voltage, and 118% power, has not crash on any games and programs, and 3Dmark stress test at 99.8%, while stay under 55 degree with two 360 rads (Hardware labs Nemesis GTS). There are rumors from multiple uncertified sources about the Overclock outcome of Ampere cards: only a few is able to stay at +40Mhz, and atop of that their OC benchmark scores has only minimum improvement against the stock score (like a less than 50 improvement at Time Spy Non Extreme), though it could be the incompletion of their Driver . Any of your experience with flagship 3090?
  2. Like i state, "inconsistency". You are lucky to get a decent block from them. I used to buy a block from them on JD.com. The contact was not even and after no less than 7 attempts of remount and reapplication of TIM, the stress test temp at stock freq still peaks at 90+, (8700K). Indeed 8700K are pasted beneath the lid, however even a good tower air cooler has lower temp than that block. Report to the customer service, they did not gave a damn, since "the product is already not in the restockable condition", plus they insist it is the defection of other components in the loop. The JD official has limited involvement since their orders are not fulfilled by JD.
  3. Bykski are notable for being a budget product in Chinese communities, however also notorious for its inconsistent workmanship and almost non-existing custom support. Furthermore, the size jet fins are larger than those from euro, lead to subpar performance. Should you look for an entire budget build, it is a decent bet on Bykski or Barrow. But if you prefer "Pudget", consider advanced blocks + advanced pump + budget fittings, tubing and reservoirs.
  4. Have been thinking mod a PS4 pro. But cant find any designated water block, nor enough source confirms any other cpu blocks that are fully compatible. OriginPC sells a premium solution Big O, but they dont appear to sell the block separately. Is fabricating one my own the only option?
  5. As long as your OS is a properly updated Windows 10, you concern is not necessary. Microsoft has coded a basic graphics driver: if the driver isn't compatible with your new hardware(graphic card), or even the driver is lacking at all, either of your old or new card can still output video signal as "Microsoft Basic Display Adapter". The most hassle free solution for you is to install Geforce Experience and has it automictically detect the properly driver to be installed
  6. Windows 10: Pasted the following in your file explorer address bar: Control Panel\All Control Panel Items\Storage Spaces you will then have the option to create an storage pool. Or you can simply ask Cortana: "create a storage pool"
  7. A 16GB stick has 2 ram rank. Usually, overclocking single rank ram (8GB) is much easily than dual rank rams(16GB). And daisy chain mobo works well with 1 dimm (2 x 8GB), while T topology excels at 2 dimm (4x8GB). Msi mobo are typically daisy chain routing; Asus boards are T Topology
  8. If your platform is X570, expect the Samsung 980 Pro.
  9. The stock setting is coded in the bios already. When GPU-Z scan your card, i reads both the stock speed and the current speed. most overclocking software tune your GPU by override that setting, the bios itself is read only. In other words, rebooting with set it to default unless you flash a modified bios.
  10. EKWB's clear concentrate; but half the concentration it recommended in order to avoid sediments.
  11. FYI, i get 42 after playing 90 min of CODMW, with only 30% duty load (fixed regardless of temp. Hardware is OC 9700K at all 5.2, and OC 2080 TI at 2150mhz. This is exactly why this loop is parallel. Some Cold liquid enter the GPU block and leave with the heat, the rest head into the CPU block without any thermal exchange with the GPU. I can assigned dedicated radiators to cool CPU coolant and GPU coolant almost independently , before they converge atop of the reservoir. Doing a loop like it negates the shortcoming of D5 pump somehow. D5 pump moves a lot of fluid at low pressure. Splitting them draw out more potential of the pump. after all. Its just imperial talk. Insufficient data for mine statement to be analytical.
  12. What is the front panel of your case? Open, glass, or mesh? My setup is: one 360 x 30 rad at front for cpu, one 360 x 30 at top for gpu. The GPU and CPU loop are parallel, taking advantage of the D5 pump. One indepedent 140mm fan as rear exhaust. Fan choice? Regular coolermaster fan, not capable of generating high static pressure (only 2.32 mm), but enough for thin rads. Even setting fixing rate at around 850 rpm, after 90 min of CODMW game play, the coolant temp peaked at 42 cecius at 45 min; no future raise. My case is a cooler master H500P. Not a liquid cooled friendly case (but however have top internal ventilation), the front panel is mesh: some dust resistance, but massive merit over glass in terms of intake. By theory, we model how heat dissipate from radiator into a force convective heat transfer: the dissipation power are codetermined by a heat transfer coefficient, temperature difference ( exponent may apply at great gradient) and Surface area. Most research shows that high RH(humidity) level result in lower coefficient; increasing fluid flow rate will also increase coefficient; the return diminishes. If the radiator is too thick for the air flow to move pass, the coefficient greatly reduced; nature convection dominates the force convection. Higher static pressure fan or push pull config will ensure the force convection. In addition higher intake air flow also result in cooler air being feed into the case. Despite more heat is brought in the case from the front rad, it is less significant than the increment of air flux. In practical, my thought includes: to remove front / top panels, to leave a wider gap between the reservoir and the front rad, to reconfig a parallel loop.
  13. Full nickel yield the best performance by theory, but a plexi cover let you see through how much sediments builds up and warn you a major maintenance. If you are using clear coolant or even just distilled water, i will get full nickel .
  14. Keep an Eye on EVGA's official site; their CLC280 is a top performer and are frequently listed as weekly sale.
  15. EK's Magnitude is overpriced, unless you treat it as luxury and lean on its look. It does NOT perform better than velocity. If you are willing to spend a lot on water block, there is a silver cpu block from aquacomputer. Using DDC pump on a loop with those radiator are a little overkill, in terms of their hydrallic pressure. If the fluid resistance is too little for your pump, it will result in higher risk of leaking. D5 pump, on the contrary, are know for high flow rate + decent hydraullic pressure. And finally, if you have source to purchase radiator from Hardware Lab, they will surely amaze you.
×