Sn8ke_iis

CPU cooler? Looks like a good overclocking build to me. I don't think you'd get any kind of significant gains from a new CPU/MOBO/RAM combo. Just maybe better thermals and slightly higher clock speeds. 8 fast cores will be plenty for gaming for the next 2-3 years. One exception I would look at is the GPU. A new 3080 coming out in the next month or so should beat 2080 Ti performance but we'll see once it's released and benchmarked. These cards are also typically hard to get a hold of when first released so you might want to enjoy your build till you have one in hand.

MSFS2020 is on the DirectX11 API and only uses 4 cores. It CPU bottlenecks the top 4-5 GPUs with an AMD 3950X depending on resolution and settings. I haven't seen anyone run a benchmark comparing hyperthreading/SMT on and off. So it's "multithreading" capabilities are unknown AFAIK. I'd test it real quick but don't own it.

I often read this opinion on the forums. There's roughly 200 million consoles from the previous gen still out there. I really don't think Microsoft and Sony will completely abandon that market and only make games that are compatible with and perform properly on the new consoles. Although, I'm sure there will be a few exclusives that may benefit from the extra cores. I think it will be a good 2-3 years before you'll need an 8c/16t CPU to play games, 6c/12t will be fine and the next gen of Intel's CPUs should be compatible with the Z490 boards so you'll have that option down the road. That being said, from what I read online the kids these days like to stream on Twitch, watch a video on YT, keep Discord open, and listen to Spotify all at the same time while gaming so it might help for that. I don't stream and just use an old tablet as an extra screen for checklists, Chuck's guides, and other reference materials.

Was it your goal to buy a fast gaming rig or a top of the line video editing workstation? Digital Storm sold you a video editing workstation. You way overspent on CPU/Mobo/RAM. You could have bought a second 2080 Ti with that budget. The Threadripper 3970X ($1900) is a great CPU for production work that will smoke my CPU on a Cinebench run but it won't run games faster. You would have been better off with the $400 3900X or even better the $250 Intel 10600K and overclocked it. 32GB of RAM is sufficient for DCS and games and will be for a while. 128GB will help with video editing. If you want to spend extra on RAM, buy a high binned b-die RAM kit that can stay stable at 4000MHz+ at the lowest latency possible. Now since you said capping your framerate at 60fps didn't work I'd double check that since that's the cause of most people's issues with stuttering/TrackIR. Firstly, check that high performance is enabled for your CPU in Windows, as well. If a tech remoted in he should have done so but maybe not. Next, make sure to enable the XMP profile for your memory. Even if your system was shipped with a factory tune, something could have easily got reset. Third disable SMT in your BIOS, it provides no benefit in most games. In fact throwing 64 threads at DCS may be your issue. After that I'd look into a free utility called Process Lasso and what "core affinity" is. Assign 4 specific cores for DCS and 1 for TrackIR. DCS will use 3 tops. You can also easily put your rig into a high performance mode that limits the impact of background programs. Windows will by default switch DCS and TrackIR rapidly between cores in an attempt to be more efficient and keep the CPU cool. You want to prevent that, it doesn't help framerate or stuttering. And since your system has a cooling loop it won't be an issue, it's more helpful with a laptop running day to day programs. And as always, delete FXO and Metashaders folders to rule that out.

That's a really good benchmark but there really isn't much in the way of explanation regarding 1% lows. It was more about memory tuning. Unless I missed a paragraph as I skimmed it. The 3950X is a good chip. From the benchmarks I've seen it has the best single core of AMD's current lineup. Haven't seen any XT benchmarks yet. It's a $750 chip in the States though. You can buy a binned 10600K and a custom cooling loop for that. The 3950X is for people who cant afford Threadrippers and want to do production work. I wouldn't recommend it just for gaming. The 1% low is the average frame rate of the lowest 1 percentile. It's what you are more likely to perceive as a stutter. I seriously doubt even pro gamers can tell a 10-20 framerate difference at high framerates without looking at a counter. Frametime graphs are nice as well. You want a nice smooth line for whatever your target frametime is. e.g. 16.7ms for a 60 fps framerate. I'm really interested to see the new gen of AMD chips and how they do in gaming. Even if it's just on an enthusiast level machine and wins by a few frames it will be an impressive accomplishment. I can just see all the online nerd fights that will ensue. Intel will be right on their heels with a new uarch release though. It's a great time as CPU consumer. Now if we can get this kind of competition in GPUs...

[quote=Peter97;4403300} Also, why do I always read DCS takes advantage of single core performance? Doesn't DCS operate with all the cores the CPU has? Many questions... hehehe To answer your second question, No, DCS only uses 3 cores. The core that is handling the graphics rendering thread that sends draw calls to the GPU is the one that will bottleneck and decrease framerate, cause stutters, etc. The other 2 cores are sitting around waiting most of the time. This single core bottleneck is common to most games, not just DCS. Now an application that will actually use and scale with cores and threads is a rendering program like Cinema4D or Blender. If you run a benchmark in Cinebench, you can see the program split up the rendering job into "tiles". If you do a single core/thread test it will render one tile at a time. For a multicore test on my system it will render 16 tiles at a time, 8 when I turn off Hyperthreading corresponding to the 8c/16t of my CPU. For this type of computation, it's relatively easy to split the work up among cores and threads and see a performance benefit. Now for gaming graphic engines rendering in real time it's not so easy to split up work among threads and still keep everything in sequence. We notice if the AI doesn't shoot right or a missile doesn't track properly. Intel consistently wins in gaming benchmarks because of clockspeed, low latency, and overall sinlgle core performance. Again, when comparing these CPU benchmarks make note of the GPU used. It's a 2080 Ti to rule that out as a limiting factor. You won't see the same differentiation if you are using a 2070 S as that will likely be the limiting factor. And I concur with what a previous poster said about 1 and 0.1% low comparison. They are a good objective metric of framerate drops that you'll actually notice. Frametime graphs as well.

This was constantly stated by tech "journalists" before the new Intel chips were even released. It didn't really bear out in the performance data. AMD has a really effective marketing team that uses those articles and social media. Best to look at benchmark data for your use case. Unfortunately, in our case, except for games in general that favor single core performance, no one has done a breakdown of how the different CPUs compare in DCS. With a 10600K a good Noctua air cooler is all you need if you are on a budget. On my system while playing DCS a 9900K draws about 60-70 watts at stock settings. For most people current draw is an overstated issue. I've tested my chip drawing 200w + but only during stress tests, never playing games. Unless you have multiple computers in a small room rendering 24/7 for $ paying European rates for electricity, most people won't notice any difference in power draw. This guy is one of my new favorite content creators. He actually compared power draw for the different CPUs. This was on stock settings. As you can see, the 10600K (6 core) draws the least, the 3900X (12 core) draws the most. The two 8 core CPUs are neck and neck. Depends on the game and even what part in the game. Only a few watts difference. In the US, assuming you are not buying from scalpers, the 10600K is cheaper than the 3700X. I'd put the savings towards a 2080 S GPU or fast, low latency RAM. Then you'd see a performance benefit in games. OP, to answer your question, no one here can definitively say which one will be faster in DCS. If I was building a system just for gaming and DCS I would go with the 10600K. If I was planning on also using it for rendering, encoding, compiling, etc. I'd consider going with the 3700X. Either way I'm confident your experience in DCS would be agnostic to either CPU. If you did a blind comparison, I doubt most people would be able to tell the difference without looking at a frame rate counter.

Personally, I'd go with the 10600K just because I enjoy overclocking. That being said, I'd bet you would see no frame rate difference at 1080p with a 2070S as the GPU would be your limiting factor. You'd need a 2080 Ti or something from the high end of the new 3000 series to actually see a frame rate difference. Just an educated guess from all the benchmarks I've seen. Might make a difference in VR in CPU limited situations.

I don't have any benchmark data but I would bet that there are more consistent frametimes and higher 1% lows especially on larger more complex missions, especially online. Would be an interesting test, curious myself. Just as a general PSA, make sure you are turning on the XMP profile in BIOS. You can get 32 GB binned at 3200-3600 MHz for a decent price, at least before Covid. Just sure how supply chain issues are affecting prices right now. But you won't actually see those frequencies by default. You have to turn on the profile or manually overclock the memory to get those advertised speeds. After an overclocking session it never hurts to double check as you might have had to go back to a default BIOS profile. And that had a noticeable difference on my system versus going from 16 to 32GB. People are reporting good results overclocking the 10600K. A good Noctua air cooler or a 240mm AIO should cool it nicely for OC to 5+ GHz.

All the benchmark data coming back is showing the 10600K as the sweetspot for performance per $ right now. Gamer's Nexus rates it as the best Gaming CPU overall currently. Key is to look at objective benchmarks. All the hype and rumors about the heat and power usage of Intel CPU's turned out to be unfounded clickbait. Under normal gaming loads Intel uses less power. You have to put the CPU through benchmarks/stress tests/high voltage overclocks to get the wattage high enough to cause issues for most cooling solutions. You can build a cheaper system with a 3600 if you are on a budget but the 10600K surpasses AMD's high end in gaming performance. Even more so when properly tuned and overclocked. If you use a cheap AMD motherboard you might not be able to upgrade with that chipset so do your homework. You should get at least one more CPU gen on Z490 motherboards. What differentiates the 10700K and 10900K that affects performance is the quality of the silicon for overclocking/voltage and the amount of cache. Very few games benefit from more than 6c/12t right now, DCS uses 3 last time I checked. It's all about singe core performance and latency for gaming with real time graphic engines. Rendering in Blender or encoding are tasks that benefit from more cores and threads. So far these are the best benchmarks I've found. One other key takeaway, until the new 3000 series graphics cards are out most of these differences are academic unless you are using a 2080 Ti and playing at 1080p. Prices can also vary a lot from country to country. So a 3700x might be the better deal in yours. Any savings in your budget from CPU choice should be put towards the GPU and then cooling. There's some measurable difference at 1440p but at 4K it's almost always a GPU bottleneck. Then specific to DCS and sims you have the issue with TrackIR and locking to 60fps. So for most people the 10600k or the 3600 is the best choice right now unless you are trying to squeeze frames out of a 2080 Ti or better. AMD is supposed to be releasing new CPUs later this year with better IPC and clockspeeds but by that time the new Intel gen will be close to, you can always wait for something new and shiny.

This isn't an accurate statement. Same clockspeeds across 7700K and more recent chips will be similar due to the same micro architecture and IPC inherent in the design of the chip. So all "Skylake" chips running at 5.0 GHz will perform similarly in single core metrics all else equal. Previous gens of chips will have lower IPC even when running at higher clockspeeds. I'd recommend anyone without at least a 7700K or newer that it's time to upgrade. Based on benchmark data the 10600K is the sweetspot for price performance. After that it's your GPU holding you back on most games. Get the 10600K, and use the savings for a better GPU or a good cooler and then OC it and you can closely match 9900K/10900K framerates. And the $2500 Titan will only get you a few frames over a typical 2080 Ti. DCS and other games don't use all that VRAM. It's a developer/AI card. At some point in the future, more games will start to use 8 CPU cores but very few including DCS do right now. Tasks like production rendering and encoding are relatively easy to split up among cores/threads and you will see measurable gains with more cores. Rendering graphics in real time for games doesn't benefit from more cores with most current graphic engines.

There really should be a sticky covering this issue for new people. As Smokey and Vertigo mentioned this is an issue with TrackIR. TrackIR isn't compatible with the Gsync features for variable refresh rate. It probably still helps with input lag but you have to lock the frame rate at 60 or 120 using the Nvidia driver or Riva Tuner. Maintaining 120fps even at 1080p with a 2080 Ti is difficult so locking to 60 fps is what works for most people. If the framerate varies between 60-120 the Gsync chip on the monitor can't match the output of the GPU with TrackIR controlling the viewpoint. It's really easy to test this yourself by pausing TrackIR and using the mouse to pan your view. You will see a noticeable difference in the smoothness of the animation for framerates between 60-120. As previously mentioned, rtings.com has the best reviews of large panel TVs for gaming purposes. They test for input lag and compatibility with Gsync and Freesync for gaming without TrackIR.

32 GB of GDDR7 no less. That and a 2080 Ti only has 14 tflops. PCIe 5.0 should have been a big red flag too. PCIe 4.0 is on the latest high end AMD motherboards and the new Intel chipset coming out soon is probably still 3.0. For the record, I hope AMD can put out a flagship GPU at a good price, competition is a good thing, but this was a hoax. I'm not sure if April Fools is a thing in Germany.

Windows power management profile. It should be set to "high performance". The equivalent setting under the Nvidia Control Panel is covered in this: https://nvidia.custhelp.com/app/answers/detail/a_id/4958/~/max-frame-rate%3A-cap-frame-rates%2C-save-power%2C-and-more Reducing System Latency: Enable Max Frame Rate and set your power management mode to “Prefer maximum performance” to reduce latency. While in this mode, the GPU is kept at higher frequencies to process frames as quickly as possible. To maximize latency reduction in GPU bound scenarios where FPS is consistent, set Max Frame Rate to a framerate slightly below the average FPS and turn Low Latency Mode to Ultra." Not to insult your PC savvy it's just the most common mistake I see. :) They both also tend to reset themselves to default during a Windows update or Nvidia driver update so always good to double check they are still maxed out. By default they're both set to save power for day to day computing, web browsing, etc.

The original source for that article (wccftech) says it was an April Fools joke. Most of those leak articles are just clickbait. Fun to read but you really need to wait till products are actually released to 3rd party testers. https://wccftech.com/amd-radeon-rx-gamma-gpu-big-navi-benchmarks-specs/

Both EBay and Paypal have policies that favor the buyer. If someone sends you a fake CPU or bad power supply it will probably show in the first 30 days and you can file a claim and get your money back. Sounds like a hassle but so is RMAing a new part. Right now used components are in really high demand in the States on Ebay so you might not get as good a deal as under normal circumstances. But there is a big market for used components and people like myself are always selling their old stuff for upgrades or changes in circumstances. EBay does have to charge sales tax now which is a bummer.

1) I keep it off. I saw about 10 fps difference on TF-51 Caucauses freeflight in 1080p High Preset. On higher resolutions (GPU limited) you probably won't notice a framerate difference but could help in other CPU limited scenarios. DCS isn't really coded to take advantage of HT and SMT in our CPUs. It's on the roadmap though for long term development. Probably going to take a while. Gamer's Nexus does benchmarks with HT and SMT off on CPUs and sees higher FPS. That might change in the future but will probably be typical for the next couple of years even on new titles. Another advantage is you get more overclocking headroom with multithreading off, which essentially makes our 9900Ks a high binned 9700K but there aren't any games I'm aware of that use more than 8 physical cores. Older chips with less cores/threads might see benefit though. I also have core affinity limited to 3 cores in Process Lasso. Jumping from core to core is intended to keep any one core from throttling due to heat buildup. If you have adequate cooling it's not necessary for DCS. 2) Never tested that specifically but I wouldn't worry about unless you are trying to free up storage space on a small drive. Would be interested if you did a controlled test though. MSI Afterburner has a benchmark that collects average FPS, 1% lows, etc. Don't forget to max out your power/performance settings for CPU and GPU and turn on the XMP profile for your RAM. Those are the most common mistakes I see that hold back performance in games. DCS specific make sure you are deleting your FXO and metashaders folder in your saved games folder. I've gotten in the habit of doing it everytime I start DCS but is most important when updating to a new version.

AI objects and pathfinding is mostly CPU limited. Not sure graphics drivers will help much but it was worth a try. When I've tested with uncapped framerate I can get high FPS in single player free flight missions but drops considerably when running a mission with wingmen and adversaries. Static ground objects not so much but a moving convoy for instance has a noticeable effect.

Looks like a solid build. Only thing I would suggest changing is the air cooler to a Noctua. The DH-D15 is widely considered the standard by which air coolers are rated. Even better would be a Corsair AIO. You'd get more OC headroom for your CPU. Pushing that 9600K up to 5.0 GHz will closely match a 9700K/9900K in DCS. Things to consider if you can afford it. Get a 1 TB NVMe or SATA SSD and wait on the platter hard drive if you need it for storage. Easy to add later. And a better power supply. It might be better to drop to 650w and get higher quality/efficiency for about the same price. EVGA is a reputable brand so good choice there.

The most common issue I see that new builders overlook are Windows and Nvidia power management settings and XMP memory overclocking profiles. Both Windows and Nvidia Control Panel need to set to be on "High Performance" and "Prefer Maximum Performance" respectively. They are both set to save power by default for day to day computing that doesn't need the full resources of your computer. I'm not exactly sure how AMD's XMP profiles work in BIOS but whatever the equivalent is you need to turn it on otherwise buying 3200 MHz RAM was pointless. "Reducing System Latency: Enable Max Frame Rate and set your power management mode to “Prefer maximum performance” to reduce latency. While in this mode, the GPU is kept at higher frequencies to process frames as quickly as possible. To maximize latency reduction in GPU bound scenarios where FPS is consistent, set Max Frame Rate to a framerate slightly below the average FPS and turn Low Latency Mode to Ultra." https://nvidia.custhelp.com/app/answers/detail/a_id/4958/~/max-frame-rate%3A-cap-frame-rates%2C-save-power%2C-and-more Those 3 things should get you most of the way. After that make sure your BIOS and drivers are all up to date as mentioned and ensure you have adequate cooling and fan curves to keep the processors from throttling. Also, assuming you are using TrackIR make sure your framerate is capped at 60fps in NCP otherwise you are just putting out more heat than you need to and will help stabilize frametimes.

No, that sounds about right if you are looking at total usage for an 8 core single threaded CPU with background apps like TrackIR, etc. You need to look at per core usage to see bottlenecks for DCS. It only uses 3 cores simultaneously. Either in Task Manager or even better with MSI Afterburner set to per core usage in an on screen display it should show one core hitting 90%+. If not, you can turn up visib range, shadows, and anisotropic filtering if you are still able to maintain 60 fps. If you look at core 15 (which corresponds to core 8 on a 9700K) you will see it's at 97%. That's the bottleneck. Note, I usually keep Hyperthreading off when I play DCS as it provides no benefit for DCS and most games on an 8 core CPU. This screenshot was taken while doing some benchmarks. Keeping Hyperthreading off on a 9900K essentially makes it a 9700K with more cache and a very small difference in single core performance which would likely be unnoticeable in normal game play unless taking FPS benchmarks. In general, you want to adjust graphic settings so that you see 90%+ on both CPU (single core) and GPU assuming you have adequate cooling to stay in thermal limits.

https://www.tomshardware.com/news/folding-at-home-worlds-top-supercomputers-coronavirus-covid-19?region-switch=1584847592 "Propelled by average enthusiasts in their shared quest to defeat COVID-19, the Folding@Home network is now pushing out 470 PetaFLOPS of raw compute power. To put that in perspective, that's twice as fast as Summit, the world's fastest supercomputer, making the network faster than any known supercomputer. It's also faster than the top seven supercomputers in the world, combined." "Folding@Home reports that it has seen a 1,200% increase in contributors, with Bitcoin miners also joining the fight, and over 400,000 new volunteers have joined over the last two weeks." And here's a visualization of the spike protein of the virus particle.

That's a bummer man. Bad power supply? Tough call on a tight budget. Be aware that putting an unlocked K processor into that chipset motherboard will limit any overclocking you want to do. You need a "Z" motherboard to fully utilize an unlocked "K" processor. To be honest, I have no idea if it will limit the stock turboboost. The spec page for that motherboard does say it supports Turboboost 2.0. The more expensive chipsets should support 3.0. As I understand it, when Intel fabricates those processors they are the same chip. Might even be on the same silicon wafer. Just the "F" designation means there was a defect in the integrated GPU and it was disabled. Same deal with 9700 vs 9700K. The K means there was a higher quality yield of transistors (MOSFETs) for that chip and it passed the binning process to be overclocked to higher frequencies and still stay within the voltage specification. As far as 16 vs 32 GB RAM. Unless you plan on playing a lot online during the next month, you might not even be able to tell the difference except maybe on complex single player missions. That's something that's really easy to upgrade later as well. I have read anecdotally that the more VRAM on your GPU card you have the less likely you are going to use all your RAM. Never tested that myself though. If you can manage a Z370 or Z390 motherboard with the 9700K I would go with that and upgrade to 32GB RAM later. Consider getting a SATA SSD instead of the M2 drive to save a few bucks. You won't notice the difference in load times. Good luck and stay safe.

Hey guys, Really not trying to get into a whole AMD vs Intel thing. If you are partial to one company, it's your money do what you want. Before we get too off topic the OP was asking about 3600 vs 9600 and specifically said they were the same cost for him in his region. I'm just trying to equip people with sources of objective data so they can make a decision without having to rely on internet rumors and hearsay. Steve from Gamer's Nexus and Der8auer have both mentioned in their videos some of the responses they get from people when they recommend Intel over AMD for gaming. I just don't get it. They're both multi-billion dollar corporations that are beholden to their shareholders. They don't care about us. They just want our money. Hope everybody is doing OK. Stay safe and healthy. And as a former Army medic, wash your hands, cough and sneeze into the corner of your elbow, and drink lots of fluids. If your local store is short on disinfectants you can use household bleach 5-10% solution of a given volume of water (50-100 ml per L) to clean surfaces. It will strip the finish off of wood floors and furniture so be careful what you use it on. Take care.