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


  • Content Count

  • Joined

  • Last visited



About ShadowChaser

  • Title


  • CPU
    Intel i7-8700K
  • Motherboard
    Asus ROG Strix Z390i
  • RAM
    Corsair Vengence LPX 2x16GB 2666
  • GPU
    MSI Aero 1070Ti + EKWB Titan XP Block
  • Case
    Fractal Node 202
  • Storage
    Samsung 970 Evo 500GB + HP EX920 1TB
  • PSU
    Corsair SF450 Gold
  • Cooling
    Corsair H75
  • Keyboard
    Razer Blackwidow 2016
  • Mouse
    HyperX Pulsefire Core
  • Laptop
    Gigabyte Aero 15x 2018

Recent Profile Visitors

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

  1. I will have to in order to use 3 gpus (technically 4 total) on this board. I will have 2x Tesla M40s for the heavy machine learning, the P106 handles my 3d printer spaghetti detection, and I have a GT 730 idling 99% of the time in a modded 1x slot in case I need to pull up a gui. If all 3 big gpus are running full tilt (which may very well happen frequently) they'll be pulling 225W from the pcie slots, plus another 5 or 10 for the GT730, which I am not comfortable with. Considering that I will also be powering dual channel memory, a pcie ssd, and potentially more, I'm not feeling v
  2. M40s have a 250W max power draw and a 150W operational power draw, how would they pull that from the motherboard when the pcie slot power spec is rated for 75W? Am I missing something?
  3. As title states - I'm throwing 2 custom cooled Tesla M40s into my machine learning rig alongside an ex-mined P106-90. I'm worried about blowing the pcie slots on my motherboard (Asus Z170-A) as if all 3 gpus take the 75W from the mobo, I am quite worried about the motherboard just going poof. I'm hoping that someone who's worked with these gpus know whether they get all their power from the power cable or if they draw power from both the 8 pin and the pcie slot, or at least can point me towards documentation on the matter.
  4. I'm trying to get an extreme undervolt on my rx580 to get it running cooler and quieter (performance is not a concern) since it's effectively running passive. In both Afterburner and Wattman I can lower the voltage frequency curve to any point I'd like, but any value below 930mv won't apply to the core once I enter it in (I'm trying to hit 1000mhz@0.85v, which shouldn't be a problem). At 90W of power draw reported in software it's still running a little too warm for my taste, so is there anything I can do at this point? Card is question is a Sapphire Nitro+ if that matters at all
  5. I'm struggling to find anything about the MSI B550M-A Pro. It's not on the list but I'd assume it's similar to the B550 Pro-VDH/Pro-DASH just without the heatsinks on the VRMs. Am I correct in thinking this? Or is this actually a $50 board priced at $90 because it's B550?
  6. I'd love to see if you can figure out what platform it's on, such as an Andyson or something. Or at least if the components seem fine if it's group regulated.
  7. I would do that if power spikes weren't a thing. I'm currently pulling roughly 430-450W under a high mixed load, 500ish under full synthetic, and I'd rather not redline my psu since I will have this doing cuda and essentially be maxed out as many hours per day as possible. I'm also toying with the possibility of putting a second gpu in here (Tesla M40, which is basically a 980Ti afaik) and I'm definitely going to need 750W. I guess at the end of the day it's not a big deal since a decent Montech or a Segotep is only $10-15 more, but I'm a cheapass so I try and save :)
  8. This is the one I'm currently interested in: www.amazon.com/dp/B08JM12SQ5/ Just a generic 750W bronze unit that doesn't look like a bomb. Not sure what the internals are, hoping it can do 4-500W of power without breaking anything. Anyone able to point me towards a review or anything? Can't find anything myself.
  9. You could look at the mounting holes and pcb to see if another of msi's coolers will just work with it (similar to how evga hydro coppers are just FTWs with a waterblock attached) or you could grab a universal gpu cooler like an arctic accelero or something along those lines. edit: from my very limited research it appears that this has a founders pcb and thus an universal cooler should work
  10. I think I posted some models of the cross section before, I think those would show very well how much space I'm working with. On top of the gpu anything >10mm won't work so if I were to add a fan where the vent is it would have to be a 60mm fan, which noctua does not make in a 10mm thickness (or at all, iirc. They only do 40mm and 80+mm)
  11. The cpu cooler needed a shroud when I was just channeling air over it, but I eventually realized that there was enough clearance for the fan to fit and not ruin the airflow. The spacing between the motherboard and gpu means that the bottom half of the gpu heatsink does most of the work, though I think for this to be most optimal I'd either have to place a fan flowing air down on the heatsink or flip the exhausts. Either way, it would make the system even louder and I'd rather just take the thermal penalty and underclock both parts some more.
  12. oh yeah it's on 2004, but prior to this morning it was just fine and it's been on 2004 for a while now. There must've been another update for 2004 after I updated initially that rolled out to me last night >_>
  13. for reference, this is what it looks like on my 1903 desktop so much better >_>
  14. that's about all that I could find. I'm rather disappointed that this isn't something that can be easily adjusted. I previously had the same problems with my icons at the right side of the taskbar being made wider than they were before
  15. they are the same image with different button widths highlighted, the difference is night and day, as you so helpfully demonstrated with your side by side comparison but that's beside the point, I have no idea how to change the width x_x Edit: see how the first three occupy a mostly square space while the last three are clearly rectangular