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


  • Content Count

  • Joined

  • Last visited


This user doesn't have any awards

1 Follower

About DuckDodgers

  • Title
    Junior Member
  • Birthday 1980-02-21

Contact Methods

  • Steam

Profile Information

  • Gender
  • Location
  • Occupation
    Web Developer, Administrator


  • CPU
    Core i7-6800K
  • Motherboard
    ASRock X99 Extreme6/3.1
  • RAM
    32GB G.Skill Trident-Z DDR4-3200
  • GPU
    KFA² GeForce GTX 1080Ti EXOC
  • Case
    Gigabyte MARS
  • Storage
    Samsung 970 Pro 1TB; WD Black 2TB
  • PSU
    Corsair AX860
  • Display(s)
    Dell U2515H, 25" IPS 2560x1440
  • Cooling
    Noctua NH-U12S
  • Keyboard
    Logitech G710+
  • Mouse
    CoolerMaster Inferno
  • Operating System
    Windows 10 x64

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. Go to the Processes tab and see what application or process is pegging the drive. This might be a clue for resolving the issue.
  2. This is the proverbial 99% completion bug, while downloading more RAM. Just leave it to finish and don't worry.
  3. This case is not a new precedent. Many other games were marked with RC in Australia, like L4D2 a decade ago was stuck in the same limbo for long time.
  4. Too low temperature shortens the NAND life-time (number of erase-program cycles). A better solution would be to isolate the Flash chips from the water block and only make thermal contact with the controller, since it's the controller that gets really hot during intensive I/O work.
  5. This thing is an ordinary actively cooled Peltier Element with some small battery attached. I hope Sony doesn't try to break the laws of thermodynamics with this, but most probably the only "cooling" effect is the feel of cold, so your body won't sweat excessively. Another issue is how healthy would be to fool your body's cooling reaction in a hot summer day, by tampering with the natural feedback.
  6. AdoredTV is becoming the YouTube equivalent of SemiAccurate -- fanboi drama and wishful speculations for a nickel. The bite is on the stick already.
  7. Intel limited a single ring up to 10 cores. Dual-ring Xeons were essentially two processors "glued" together on a single chip, since each ring had its own memory partition (Home Agent) and on those CPUs Intel provided an option to treat the memory access as NUMA-like, to reduce the cross-ring latency penalty for NUMA-aware applications at a cost of slightly lower L3 hit-rate.
  8. It is promising to see developers finally getting to master the DX12 API with positive results, but still disappointing that it takes so much time and resources after the game's release. Source #1: https://www.dsogaming.com/news/hitman-2-runs-significantly-faster-in-dx12-15-40fps-increase-in-cpu-ram-limited-scenarios/ Source #2: https://www.reddit.com/r/HiTMAN/comments/b5p79r/some_hitman_2_directx12_benchmarks_massive_fps/
  9. Apple holds the user cipher keys for the iCould accounts. Data stored in the device itself is encrypted locally, so as long as it's not synced with iCloud, the government can't ask Apple for something it doesn't have.
  10. The DXR extension in DX12 doesn't specify how ray-tracing is implemented. The only requirement is that the device driver exposes the minimum functionality to perform RT to the API. It is up to the hardware vendors how to do that -- dedicated logic or software (compute) emulation including any combination of both.
  11. The really short explanation of why this is not practical is that the PCIe link between the GPU and the CPU is not cache-coherent. What you ask for is the GPU to act as a second "CPU" so both memory locations (system RAM and graphics RAM) can be pooled together as a single addressable space -- that's essentially how multiprocessor (NUMA) systems work. Even if there was a way to put the page file in the graphics memory (and program the GPU to do the paging), the CPU would have no information what is being done by the GPU, until the result is being copied back to the system memory for verification and thus nullifying the benefit of the extra memory space and performance.
  12. This is what I have used in my main rig, since my first PC: (certainly not a complete list) 1996 - Trident 9000i ISA 1997 - S3 ViRGE 1998 - T'seng ET6000 1998 - S3 ViRGE/DX 1998 - ATi 3D Rage II 1998 - ATi 3D Rage II+ DVD 1998 - Nvidia RiVA 128 PCI 1998 - Nvidia RiVA 128 AGP 1998 - Matrox Millenium G200 1999 - 3Dfx Voodoo Banshee 1999 - Nvidia RiVA TNT 2000 - S3 Savage 4 Pro 2001 - GeForce 2 MX400 2001 - GeForce 2 Pro 2003 - GeForce 4 MX440-SE 2003 - GeForce 4 MX440 2003 - GeForce 4 Ti4200 2004 - Radeon 9600 XT 64-bit 2004 - Radeon 9600 XT 2005 - GeForce 6600 GT AGP 2005 - GeForce 6600 GT PCI-E 2005 - GeForce 6800 GS 2006 - Radeon X1900 XT 2007 - Radeon HD 2900 XT 2008 - Radeon HD 3870 2x CrossFire 2008 - Radeon HD 4870 2009 - Radeon HD 4870 1GB 2009 - Radeon HD 4890 2009 - Radeon HD 5870 2011 - GeForce GTX 570 2013 - GeForce GTX 580 2015 - GeForce GTX 780 Ti 2015 - GeForce GTX 980 Ti 2018 - GeForce GTX 1080 Ti <<< current
  13. Not really. Turing inherits Volta architecture and adds the RTX hardware. The only missing thing is the HBM2. Regarding the roadmap, Volta was simply moved as an exclusive HPC product to promote Nvidia's Tensor DNN acceleration and to serve as a platform to integrate the expensive HBM2 memory -- none of those were suitable for the consumer market. Then Pascal came to fill the gap between Maxwell and Turing for the consumers, as a test vehicle for the 16nm production, with minimum investment in the architecture to speed up the roll-out. It was essentially Maxwell 3.0 with several tweaks for faster VR, updated video codec, but mostly utilizing the smaller fabrication node to bump up the clock rate, as the main performance benefit.