Jump to content

Mr Technician

Member
  • Posts

    761
  • Joined

  • Last visited

Everything posted by Mr Technician

  1. I'm debating if it's worth the $40 just to be safe given that it's already $250 for the 1000 watt one...
  2. That board has an M.2 key and will work fine with the Samsung SSD you listed.
  3. I will edit the description to include the specs.
  4. I am building a high end workstation and am debating between 1000 and 1200 watts for the PSU. 1000 ought to be enough but I like the idea of having a large margin just in case. I will likely get the Seasonic Vertex GX-1000/1200 as the PX-1000 is unavailable thus far. Would 1200 watts be the wisest safe bet? Relevant specs: CPU: 13900k GPU: RTX 4090 RAM: 96GB DDR5 (Corsair 2 x 48GB kit) Storage: 3x M.2 Cooling: Triple radiator custom loop
  5. It's interesting that you mentioned this because I have been looking at the 96 GB kit from Corsair and a good alternative. I'm also likely going to go with the 13900k but that remains to be seen for the next week as I nail down my water cooling parts.
  6. Nice, good to know. I guess it makes sense that it would fix the port forwarding issue with NAT.
  7. I don't believe that fixes the issue with NAT. He would need to use bridge mode as others have suggested.
  8. Unfortunately I exchanged my PNY 4090 for a Gigabyte one specifically so I could get one compatible with EK blocks. Other brands don't have blocks that are as attractive as the EK ones but the prices are steep... I also like the idea of watercooling everything while I am going to the effort of cooling the CPU. This is partly for aesthetics, and partly because I know I can keep the card cooler IF I have enough radiator space, but I understand that the stock coolers are still great. Partly, yes, but I am going with ZMT over hardline. I wasn't planning on the KS, but I might delid and use a direct die block even with the non-s chip. Yes, but I am also getting back into heavy video editing (6k raw, possibly) and blender, so the extra CPU performance matters.
  9. I am in the process of planning an RTX 4090 and 13900k build. I have an NZXT H9 Flow for my case and could run a max of 3 360mm radiators. In terms of thickness: Top: 30mm ideally, 45 mm would be pushing clearance for motherboard Side: 30mm, 45mm would be cramped Bottom: If I don't mount my GPU vertically, 45mm. With vertical GPU, 30mm. I will probably just get 3x 30mm HW labs radiators, but this would give me no options for a bottom drain. I also like the look of flat reservoirs, which I could put in place of the side rad. It seems like triple radiators is the wisest choice for temps and noise considering the hardware, but I am open to other opinions.
  10. Looks like it does have Linux drivers as it is being touted for open source: https://www.phoronix.com/review/intel-arc-a380-linux It has a power connector which isn't ideal (just from a power consumption standpoint) but it sounds like this is because its power limit of 75W is right on the edge of what PCIe can supply.
  11. I am looking to purchase a GPU for use in an HTPC VM attached directly to my 4k TV. I currently have a GT 1030 but it will only run at 4k30 over HDMI 2.0. My requirements: DP 1.4 output (will be converted to HDMI 2.1) Modern decoding support (VP9 for sure, AV1 would be nice but is likely expensive) I don't care about encoding Has Linux drivers Doesn't suffer from the reset bug as this will be used with KVM Should I just buy another GT 1030 that has DP 1.4? I think that would check all but the AV1 box.
  12. What do you mean by "different network"? Is this someone else's home wifi network?
  13. True, but I think the trade off for speed wouldn't be worth it. I will need to check benchmarks.
  14. I'm starting to lean more strongly toward the 13900k with 64GB of memory to start. I don't like being locked to LGA 1700 (no upgrade path) but I doubt I will be getting a new CPU in this build at any point anyway. My other consideration is how well Hyper-V will handle big-little. From what I can tell the scheduler is smart enough to make use of both core types when running a VM.
  15. Is this because of the 3D cache? Otherwise the X3D chip is slower than the non-3D for most multi-threaded workloads.
  16. Maybe, but I don't think my needs are substantial enough to justify the much higher cost of a HEDT chip if I can feasibly upgrade to 128GB in the future as larger DIMMs become available or memory support improves. I also care about single thread performance for gaming so a Threadripper or similar would hurt me in that area.
  17. At this point it's not a strict requirement. I want to run 1-2 Windows VMs for testing software and simultaneously be able to edit 4k ProRes in DaVinci Resolve or Adobe products. 32GB on my current machine is nowhere near enough to do that smoothly. Even software like Jetbrains Rider can eat up 10GB. 64GB is probably enough, but I wanted to go all in- however given the issues with that many DIMMs, I might start with 64GB and upgrade it later.
  18. Are there any 64 GB DIMMs on the market, yet? Doesn't the 13900k only run 128GB max? Given that my choice of 128GB of memory is more for future proofing and because this is still a gaming machine, I think the Xeon line isn't appropriate for me. I could certainly start with 2 x 32GB and upgrade later.
  19. I'm building a top of the line gaming and workstation PC and plan to install 128GB of DDR5 memory in it. At present, this requires me to install 4 32GB (presumably dual rank?) DIMMs, which I understand is problematic at speeds above ~4400 Mhz. However, I prefer having all four DIMM slots full for aesthetic reasons anyway, so waiting for 64GB DIMMs might not help me. I am leaning toward the 13900k because it seems to handle 4 DIMMs better than the 7950X and X3D. Some have had success with 6000 Mhz as in this video: https://youtu.be/_Tnnc9Th13Q. If I want to run this much memory, should I stick with Intel? Also, assuming I get at least an Asus Strix class board, will a better board help me with memory stability? The video I linked above seems to suggest that the Pro Art board works better than other options, but the memory controller is on the CPU.
  20. Why bother with RAID 0 when involving < 10 Gbps networking? It's much less reliable (even if you do have raid 1 to back up to) and you will never actually read or write at that speed.
  21. You could try a different VPN but they might be flagging it as suspicious, either because they know it's a VPN IP or because they know your account previously connected from elsewhere. Aside from that, I think you are out of luck. Although I would suggest not spending too much money on WG anyway, after the release of the BZ 176.
  22. Ok, I have the Asus passively cooled one. The only other thing i can think of is the system only has PCIE 2.0- but that shouldn't matter for displaying Windows.
  23. What model of 1030 specificially?
  24. Positive. I currently use a different system with an AMD GPU in it with a DP to HDMI 2.1 adapter and it works great. The only issue is the card doesn't support VP9 decoding and quite often needs to fall back on the CPU. Is this an issue with all 1030s or just this one? Edit: it seems unlikely it would be arbitrarily limited to 30 fps.
  25. I have an Asus low profile GT 1030 connected to my 4k TV and I am stuck at 30 hz. I have tried: Multiple cables Checked that the card has HDMI 2.0 Used both Windows 10 and Linux No matter what I try, I am stuck at 4k30 even with cables "rated" for HDMI 2.0. Is there anything else I can try, or do I need to buy a GT 1030 with a DP connector so I can use an HDMI 2.1 adapter?
×