Gearbox

Yes, but what I think he was saying is without eye tracking you could still render the middle of the FOV in high quality and leave the edges at lower quality and still get an FPS boost, but with more head movement needed to look around.

I guess it could work, your neck would just get an extra workout. It would require learning some unnatural movements of your eyes/neck combo, but so does using TrackIR and most people figure that out pretty quickly.

Am I correct that the MST hub makes the computer see the monitors as separate devices? An old doublehead2go or triplehead2go might work, it treats the monitors on it as one large one.

What's the price difference between the two and how long would it take you to save up the extra bucks?

Lame. From what I gathered from this thread though, VR in DCS doesn't seem to be constrained at the GPU level if you have a high end GPU: https://forums.eagle.ru/showthread.php?t=157374 That's why I was wondering about the idea of running two separate engine threads both sending their own outputs to the GPU to separate the computing tasks for the left and right eyes. (disclaimer: I last wrote a computer program in the year 2000)

If you've got a new enough motherboard you can give Intel Optane a try, it uses a specialized small but super fast M2 drive for caching your most recently used files. It's a lot cheaper than a second entire drive, and if it dies the system should keep chugging instead of losing everything the way a RAID 0 would. You should still keep backups of course, there are very cheap cloud backup services available now. Weird, I figured that the main bottleneck with syncing a full copy of a blockchain would be your internet bandwidth, not your hard drive bandwidth.

It has to know where you're looking so it can render that area in high fidelity.

I've seen that kind of thing caused by bad GPU memory. Do you get graphic glitches anywhere else?

Weird, the whole point of g sync and freesync is to eliminate tear at super high frame rates.

I appears that Spectre is worse than initially thought and affects AMD, and even affects iphones and androids. edit: forgot link https://www.pcworld.com/article/3245790/mobile/spectre-cpu-faq-phones-tablets-ios-android.html

Try it in a different PCIe slot on the mobo if possible.

If you opt to leave windows on the old drive, you should consider moving the windows virtual memory swap space to the SSD to get better VM performance.

I remember back in my Falcon 4.0:AF days going from a stock Cougar to uber2-modded one made airborne refueling go from impossible to pretty fun.

Ergh, my new in-progress build uses an ASUS mobo and I was planning to OC the crap out of it. How essential to overclocking is the AI suite?

You may find this A-10 pilot's coloring book educational: https://imgur.com/gallery/fd4sK

Why not use displayport to HDMI for the rift then? https://forums.oculusvr.com/community/discussion/29713/just-to-confirm-displayport-and-dvi-adapters-do-work

Even without the ability of the GPU to generate true 4k, it could be worth it for reduction of screen door effect and the wider FOV until GPUs catch up.

Will this thing have eye tracking and foveal rendering? Otherwise even a Titan V would be overloaded wouldn't it?

You can see the actual impact bursts, you'll even see some further downrange from the target from rounds that skipped off the ground and impacted later. At least this happened back in the beta days. I've been away but reinstalling this weekend.

FWIW I found this, but these games are likely not CPU limited. https://www.phoronix.com/scan.php?page=news_item&px=x86-PTI-Initial-Gaming-Tests Other benchmarks are a mixed bag. The synthetic I/O one gets scary but the actual software tests don't seem too bad. https://www.phoronix.com/scan.php?page=article&item=linux-415-x86pti&num=2

Microsoft hasn't released the patch to the general public yet.

I work in IT, and the news on this is freaking everyone out because of the performance implications: https://www.theregister.co.uk/2018/01/02/intel_cpu_design_flaw/ tl,dr: patching a bug in the way virtual memory is handled is expected to cause 5% to 30% hits to performance on some things. Anyone have any thoughts? I JUST bought the bits to build a new i7-8700K system. I had considered a Ryzen (which turns out isn't affected) but I heard that they aren't as good for DCS. Everything is unopened though so I could send back the CPU and mobo if necessary.

I wonder if it would be feasible to run an entire second copy of the engine on a second CPU core and send the output to a second GPU for one GPU per eye. RAM usage would increase but in theory it shouldn't double if you can get both threads to share the RAM that is storing the models and textures. Are the devs reading this thread?

A-10s are slooooow, that would be really long and boring. The book about A-10s in the gulf war has a section about the ferry flight, it sounded pretty terrible.

In one of those a lucky round strikes the side of the barrel an penetrates, and you can see flames shoot out the muzzle!