• Content count

  • Joined

  • Last visited

Community Reputation

465 Excellent

About w6kd

  • Rank
    Global Forum Moderator

Flight Sim Profile

  • Commercial Member
  • Online Flight Organization Membership
  • Virtual Airlines

Profile Information

  • Gender
  • Location
    Woodland Park, CO

Recent Profile Visitors

10,927 profile views
  1. I'm saying that buying and trying may well give you a usable P3D platform for just the price of the GPU, and if it doesn't you're not out anything because it could then be used as part of a new system build. It shouldn't require *too* much adventure, but you have to be the judge. The primary fit issue would be the length of the card...they're both PCIe devices so the card slot is no different. You'd want to measure that out to ensure the card isn't too long for the slot it'd be in. If your case is really packed in tight, card width might be a factor, but that's rare except in mass-produced OEM machines. The card length should be in the manufacturer's specs. The 1070Ti also needs a single 8-pin VGA supplementary power pigtail from the PSU, where the 660 had (I think) a 6-pin, so you'd need to make sure that your PSU has that cable/connector. I still have an i7-2600K machine (not my sim machine for a long time now) and I upgraded it three years ago from a 680 to a 970 and that was a pretty good performance bump (with FSX, not P3D). P3Dv4 will make even better use of the 1070Ti...the question is whether a 4.6GHz Ivy Bridge CPU (low range of acceptable now, IMHO) can keep it gainfully employed at a high enough frame rate to do what you want, and that's gonna depend on how you run the sim--monitor size, autogen, complex acft, vector scenery etc. It's probably doable, but not unconstrained with the whole kitchen sink thrown in. P3D makes much more effective use of the GPU...I would imagine that your current 2GB 660, which was a third-tier performance card six years and three GPU generations ago, is pretty underwhelming with current software. Cheers
  2. Well, what are you trying to do with P3D? Are you using v4? That 660 GPU is the most limiting component in your system for P3D. You could decide what GPU you'll want in the next full system build (whether that be now or later) and try it in the current machine to see if it'll give you what you want together with the current Ivy Bridge CPU...if not, then time to move on to a new system build with the new GPU already in-hand. An 8GB 1070Ti would probably be the best bang for the buck for P3D use now with VRAM usage trends being what they are, but it'd depend on what you want to do with the sim and your machine later. Regards
  3. I don't see that as ruling timezone issues out...it might rule out bad data around the problem locations defined in the timezone definition bgl, but it doesn't necessarily address process/algorithm issues. I do still wonder if crossing a TZ boundary that moves the date back (in either/both of UTC or local time) might cause some sort of logic fault, as a programmer might have assumed that date always moves forward. At any rate, a corrected TZ definition bgl would not necessarily alter that sort of behavior (if it existed). Regards
  4. IFPro? (a Google search yielded pages and pages of nonsense)
  5. I think the data that initially needs logged includes: position (lat/long), heading (true and mag), time (UTC and local), date (UTC and local), altitude, COM1/COM2 freq, wind (dir/speed), and outside air temp.
  6. The only potential issue with using FSUIPC's logger is that it is a DLL, so it dies with the sim in a crash, so the most useful recent data might be lost. Another thought comes to mind as I ponder this (I've never actuially seen this problem here since I generally use QuantumLeap to beam myself ahead to somewhere close to the TOD): it seems these incidents are happening at high lattitudes--another potential issue might be something unhappy in the magnetic <-> true conversion math in locations close to the magnetic pole(s). Regards
  7. w6kd

    LOWS DD no dynamic lights

    Make sure you have the dynamic lighting tickbox checked in the P3D Options menu. DL active in a scenery still won't work if it's not enabled in P3D. It's a common boo-boo. Regards
  8. I wonder if there's some sort of flight logger that could be used to track a flight and log some metrics so that place/time at least would be known. Something that runs outside P3D, using FSUIPC to gather data for example, would be ideal, as it wouldn't crash with the sim. Regards
  9. w6kd

    Top end hardware - poor results

    Also, 2x16GB DIMMs is a suboptimal memory config--X299 is a quad-channel architecture, so 4x8GB would use all four memory channels, where your 2x16GB configuration uses only two. And fast memory would be worth having--2666MHz at CAS16 ain't zackly tearing it up. The hybrid HDD looks a little anemic for a performance system build. SSDs are pretty inexpensive these days...you don't need a high-end one to really speed up your load times from that hybrid drive. Good luck
  10. I'm wondering if you've taken note of both the local and UTC time when these crashes occur. Is it near midnight either local or UTC in the local timezone or in the timezone you're crossing into? IOW, would crossing into the next timezone result in changing the date in local or UTC time (forward or backward)? Regards
  11. w6kd

    TFDI 717 Version 1.1 Released.

    Thanks for the update.
  12. w6kd

    TFDI 717 Version 1.1 Released.

    Argh...thanks for the heads-up.
  13. w6kd

    TFDI 717 Version 1.1 Released.

    Does anyone know if the problems with TFDi's Trueglass/RealLight in a dual-GPU SLI config ever got resolved? I'd be interested to hear from anyone running this updated version in an SLI config, as I have stayed away from this one due to those problems. Regards
  14. w6kd

    Top end hardware - poor results

    The "heartbeat" of the sim is the main thread that runs on Core 0...if it runs slow, so does everything else. 3.3 GHz is...slow. You want to get that up near 5 GHz to make the most of the rest of your hardware. Your friend's conclusion about loading time being an indicator of poorly written code is wrong, IMHO. The load time is driven by processing hundreds of thousands of scenery files, AI aircraft, etc to enable the simulation to run anywhere in the world. I run a quad-core 7700K at 5 GHz and a pair of 1080Ti GPUs in SLI (properly bridged with a high-bandwidth SLI bridge, as Michael noted above)...I use 4xSSAA in daytime (4xMSAA at night with DL on) and I use a 30 Hz 4K TV for a primary display, and I can keep it locked at 30 fps most of the time. I suspect that if you can get that i9 up to the high 4 GHz range you'll see a major bump in performance. Regards