blkspade

On my Strix X470 board PBO is raising base and boost clocks, with scalar set to 10x and 200mhz offset.

Do you have PBO turned on? My 3900x get up to 4540, but I use PBO. There is also evidence of clock variance between motherboards that isn't the CPU itself being the source of the problem. A 200mhz deficit is a lot, and leads me to think its an Asus issue with the x370 boards. https://www.guru3d.com/news-story/ryzen-7-3800x-show-different-boost-clocks-with-different-motherboards.html

Well it's not technically a bus clock problem, but a system timer issue that is effected by the bus clock. It's only in that Microsoft changed the timer source in Windows 8 and after where the issue arose. It only effect benchmarks where completion time is the chosen metric a opposed to a throughput measurement.

I have a 3900x that hits a max multiplier of 45.5 on 3 cores, 45 on another and 44.3-44.8 on the rest. The bus clock is consistently shy of 100 at 99.8, which is the only variable that really seem to change between motherboards. I have to wonder if the boosting logic will allow the multiplier to overshoot 46 to make up for boards that have underclocked bus speeds. Currently overclocking my board's bus clock to 101 breaks boosting. Its really in the spirit tinkering and testing for me as I'm not sweating it that much.

I would much rather have the ECM actually work against the F-14 and F-18.

Well conveniently enough the latest Ryzen chips are near to equal the single core performance of the 5.0Ghz Intel chips out of the box.

Well DCS is currently seems to be capable of using more cores for loading resources into memory/vram. You'll see lots of threads spike when rotating the external views suddenly. While that isn't proving us much on the surface, I suspect that such ability to even address extra threads could potentially translate into better utilization when they switch to Vulkan. There is a thread on the forums about there actually being some degree of performance scaling in DCS with more threads. The ability for render operations to take place independently (Asynchronously) of the main thread is a major feature for both Vulkan and DX12. Xplane is also transitioning to Vulkan, and the new MS Flight that'll be coming out next year will very likely be DX12. Star Citizen devs announced plans to use Vulkan, and its base engine (Cryengine/Lumberyard) is already MT aware. Nothing is available right now in the sim space that is using the full potential, but they don't have the resources of AAA game devs.

1st Gen Ryzen was positioned against the 6th Gen Core series, if that tells you anything.

Various games and game engines have been pushing beyond 4 cores for a while now. In those scenarios, the low points are the worse resulting in noticeable stutters.

The principle reason your statement is not accurate, is that DCS is very multithreaded in IO operations. So while the general simulation and audio are not as much, the process of loading assets into memory loads way more than 2 cores. Its very noticeable when panning or rotating the view as you'll see many cores suddenly get loaded up. People have demonstrated real performance differences as a result going above 2 cores.

https://www.asus.com/us/Motherboards/ROG-CROSSHAIR-VI-HERO/HelpDesk_CPU/ Umm, kinda your fault there, and partially Asus. Even though there are bioses available newer than whats listed on the compatibility list, they haven't given details on which CPU are supported. The chart for my X470 board at least has some of the new chips up to the 3800X from the previous bios, but no details yet on what the latest BIOS that's available added. Edit: Well they just added the 3900X to the list for my board as part of the previous bios.

I'm probably gonna move from my 2700 to a 3900X. A potential 33% more performance in DCS should lead to more stable frames in VR. As far as the 3600 vs the 3700X, the 3700X will offer 200Mhz more in the boost clock, which will likely provide some improvement. Given my experience with a 2700 non-X, I'd say go for an X model if its in the budget. Also OC3D's review claimed that the single NVME port directly attached to the CPU gets full PCIe4.0 performance on X470. Not that it'll net much real performance in game loading DCS.

N Sent from my Pixel using Tapatalk

I can forsee this causing problems, if a manually set contact is mislabeled foe and acts as a 2nd donor for everyone else's SA page. Sent from my Pixel using Tapatalk

I'm pretty sure if they have an intended target for a release, somebody is already "harassing" them regardless of the amount of noise the community is making. Someone is a manager already demanding status updates. Not saying that I agree with either POV outright, but it seems the statement of non-communication is perhaps about a lack of preemptive information initial news, as opposed to expecting regularly updated throughout the day. That would be totally unreasonable. Someone makes a post when its on time though.

I have the 1080ti because it was the only thing that could handle VR with a significant pixel density bump at the time. Which would also translate into decent performance for higher res headsets. While the Radeon VII is likely a marginal improvement over it, the shift to Vulkan will likely make it the better choice in the long run for the same money as the non-Ti 2080 or less. Its a cut down workstation card that actually costs more to produce with 2x the frame buffer of the 2080. Yeah I have problem giving my money to companies that are regularly anti-consumer and do arbitrary things to manipulate their consumer base. Maybe I'm weird for not wanting to give rewards for bad behavior. The 2080ti is not priced on its own merits, but to inflate the perceived value of older inventory. Nvidia learned they could charge $1200 for a non-gaming card sold to gamers, which they explicitly justified themselves as not marketed to gamers (yet not calling it a Quadro). So they could definitely charge even more for their entire "for gamers" product stack. Its all margin. They were already making stupid high margin at $700, and I would have considered $800 "fair". Its on the same node as the previous gen which gets cheaper to produce on over time. It's genuinely not worth $1200, no matter how much value an individual ascribes to it being the best available option. Well 2nd best since the $2400 RTX Titan is a thing.

Your POV on value is subjective, while I'm saying rather objectively you are solely being charged for the companies manufacturing mistake. They are literally about to repeat what they did with the original Titan GPUs, where it was the best gaming (even though it wasn't supposed to be for gamers) until a "normal" priced Ti card for practically the same performance was released. They just shifted the entire product stack up a tier to save face (with investors) for the 10 series over production. They are going to release RTX Super cards in roughly a months time where in all likelihood the $700-800 card supplants the $1200 ti card. The 2080ti offers the same percentage performance increase over the 1080ti, that the 1080ti did over the 980ti. Only 1080Ti was just $50 more, where as the 2080ti is nearly twice the cost of it predecessor.

AMD is not releasing Ray Tracing hardware until it can be offered across the entire product stack without halving performance. That's basically next year with Navi 2 and the new consoles coming out. Nvdia isn't doing non-raytracing equivalents as not to allow consumers a path to actually quantify the monetary value of the tech. The 2080ti isn't worth its asking price, and only costs what it does (the entire 20 series) so Nvidia could clear 10 series inventory they over produced.

Current info indicates that the new AMD chips IPC improvement basically has their 4.4Ghz part equivalent to Intel chip at 4.7Ghz in single threaded performance. At 4.6Ghz it would definitely be encroaching upon Intel's 5.0 Ghz performance. If there is any overclocking headroom at all, then it blows the 9900k away. Simple explanation of PCI-E 4.0 is that its doubles the bandwidth over 3.0. So more potential performance for various components.

Another thing is that you should try to use an "Every Frame" option Cockpit displays, because the others also induce stutter.

People fly F-5s on servers with no weapon restrictions. So it's basically gonna be like that, except in a jet with datalink. If you want to fly on servers with unrealistic limitations, then fly fly your F-16 like its a flanker.

Ideally you shouldn't even bother entering a 2v1 engagement without TWS. While TWS would be ideal, the missing ability to direct the azimuth when the scan width is reduced also hampers the hornets effectiveness. Azimuth control is huge factor in trying to engage 2 fighters solo. What you should be able to do even without TWS and even using AIM-7s is fire early and F(A)-pole, to force the first one defensive. The first shot should be at a range that still threatens, but you need to drop that lock if/when the bandit beams or turns cold (even before pitbull). The missing azimuth control is important for the next step, because you want to point your radar back at where is wingman should be before pointing your nose. The decreased scan width would give you a much faster update to acquire and fire on the likely closing wingman and force him defensive. At that point you should be within critical range of the first bandit that you might be able to engage during his turn to recommit. If you catch him near the start of his turn fire and immediately break across him, so he has to turn further to get his shot off.

I think the point you are missing is that in DCS everyone is the same pilot in every jet, so the only variables are literally what the plane offers.

Given the way the missile is designed, this behavior is likely closer to reality, excluding the fact that it always lofts now. The Mk58 rocket motor is a 2 stage where the first boost stage is only 4.5 secs long and a sustain stage of 11secs. So it gets up to speed in 4 secs levels out and sustains said speed for 11 secs. Its a way more efficient profile than what it used to be, and gets to target with way more energy. ATM it out ranges non lofting MRMs that are otherwise faster. This does mean setting up the BVR engagement to where you should try to take a shot before an opponent armed with ERs or 530s, from further out. Its the DCS AIM-7F that has "interesting" behavior. Given that it doesn't loft and just drops before the rocket ignites it ends up climbing up to the target in the terminal phase when launched on co-alt non-maneuvering target. This hampers low altitude launches if fired without pitching up, but is something noted as an improvement over the F by the M IRL. The M/H in game will always pitch up at least 5 degrees at launch. The one other odd thing about the F is that it always rolls out 20 degrees from side its fired. If this is actually intentional then I've already worked out the logic behind it.

Imagine how mad you'd be getting killed by a manpad flying over open water.