kksnowbear

It is a bit off topic, I agree - but I was simply commenting on something that someone else brought up. It's not difficult to describe a way to conclusively demonstrate that you couldn't tell the difference in a very low frame rate and a very high one. You're making the same mistake that (some? most?) people have been making forever: Confusing frame rate with the perception of motion, thus assuming high frame rates will automatically guarantee smooth video. It's just not true. The concept is touched upon by the article linked above. There is nothing, at all, to guarantee that even at 100 FPS you will not have stutters, pauses, hitches (or whatever we're calling it now). And as I mentioned above, it's not worth getting into the argument to convince people who refuse to "think outside the box". (Come to think of it, maybe the problem is really with people who aren't thinking "Inside the box" lol)

Finally...Hopefully this is absorbed and retained by all the decades of idiots I've dealt with who assume and insist that higher frame rates automatically 'fix' stutter. Best part of the entire article is where it's stated plain as day: "This might shock you, but an animation that is capped at 30 fps looks much, much better than the same animation varying between 40 fps and 50 fps." I've maintained for years - because I actually understand how it works - that I'd take a lower, stable frame rate (within reason) over even 100 frames bouncing around all over the place. It's not the lower frame rate that is distracting; it's constant changing between rates in real time. In fact, there's a simple way to show that that no one can really tell the difference in lower and higher frame rates anyway; they simply assume a higher rate will automatically be smoother (not true). Interestingly, there has always been a correlation between those who insist on this, and those having money to throw at top-end hardware, claiming they can see the difference in 45 and 60 FPS (for example, or even between 60 and 100). We all know the post: "I have no stutters because I'm getting 60 FPS rock solid" Someone figures out at least a couple times a year that you can completely eliminate stutters if you throw enough money at hardware Typically, what's taking place is someone wanted 'bragging rights' for the latest, most expensive hardware, but didn't want to admit that's really what they were after. So, the way to justify throwing money at bragging rights is "I can see the difference in 45 and 60 FPS". No, you can't, and it's actually easy enough to prove you can't see the difference between even 1 and 100 FPS. What you notice is variance in the perceived rate of movement, which in reality has nothing at all to do directly with frame rate. (As an aside, the same guy could be expected to be proclaiming a year later - when the next newest GPU came out - that he now has "NO STUTTERS" because he's getting 100 FPS with the new $2000 GPU he just "pulled the trigger on"...lol HINT: It's the same guy who just a year prior said he had no stutters because he got 60FPS. So how do you eliminate stutters that you claim you didn't have a year ago?) Smoother does not necessarily come from higher frame rate, and higher frame rate does not necessarily guarantee smoother. (EDIT: I fully expect to be attacked for having stepped on toes concerning frame rates and expensive, latest-gen hardware. Comes with the territory; been dealing with that for the same aforementioned decades)

I'm sorry if I've frustrated you and my point wasn't clear. I won't say what you asked me not to (but TBH it doesn't matter whether I say it). Regardless, it doesn't appear the solution lies with DCS. I could be wrong and I'm completely prepared to face that outcome, should it wind up being true. Again, I do sincerely wish you the best of luck with your issue

In the very same post you quoted, I already mentioned the exact MicroCenter promo - though it does appear that very recently they've dropped the price another $30. However, as I said above: It's a 32G kit, and the OP already has 64G. Even moving to DDR5/AM5, it makes zero sense to cut your RAM capacity by half. I also said that promo includes a basic, entry-level B650 board, and as I mentioned: I don't want to start a pissing contest here, but my strongest professional recommendation would be that if you're paying to change to an AM5 platform anyway, it makes no sense to go with a PCIe 4.0 board. Stick with X670E or B650E. Just to address some points you have obviously overlooked. I love MicroCenter, we have two near where I live. But that promo is an obvious effort to move hardware that is poorly matched, by bundling it with a 7800X3D (which is undoubtedly the name that they're betting novices will recognize). That board would be better matched with a lower-end AM5 processor, not the top end. You also can't buy anything at any MicroCenter without paying sales tax. Here it's a fairly reasonable 6%; I know people here on this forum that live where it's over 8.5% - which bumps the actual price up to right about $510. The price I cited above ($925) was inclusive of sales tax where I live, plus 64G of the same RAM, but with a board that actually makes sense with a top-end AM5 CPU, supporting not only PCIe 5.0 but also a few other nice features that make a difference in a high-end build. (*NOTE: I'm actually laughing because, on the web page for that B650 board, Gigabyte has linked a SkatterBencher overclocking video that isn't even for that board, and uses a X670E board that has features which the B650 in the MicroCenter promo does not have. The board Gigabyte has linked, with the features that make the SkatterBencher overclocking even possible in the first place is $500 by itself after taxes. And it's not part of any bundle deal at Microcenter either...in other words, just like I'm telling you: People who do this professionally know there's a difference in these parts, and they also know it's not at all the same thing to throw in a cheap-ass low end board with the best gaming CPU on the market right now.) You add a $500 board with the CPU and the 64G RAM, you're now over $1100 (and that's if you only pay 6% sales tax). The $925 price I cited is $175 less and still includes a decent PCIe 5.0 board.)

There are some good points above, but I'd like to throw in a factor that seems to be missing (and is always a part of the considerations): Budget. The point is made above about the 10700K you have is true: It's not too bad even by current standards; this is a product of the last several generations of CPUs (Intel at least) being only so-so in terms of performance increase above the prior gen: Since the 9th gen, not a huge leap forward til maybe the 13th/14th gen. (And yes, I have tested examples through this range first hand). There is also the well-taken point that, considering you're talking about VR performance, while the 2080Ti is adequate (as you've described) , VR is really demanding on a GPU. What all this means taken together, then, (to me) says: 1. You'd need to upgrade the GPU (to improve VR performance). The more you want to improve that performance, the more GPU you will require (and thus, budget is a factor). 2. I strongly prefer to avoid the term "bottleneck", instead I prefer to look at it more as how balanced or matched the CPU and GPU are, because I believe this presents a more accurate technical picture. That said, right now you have a fairly 'matched' system (the 10700k with the 2080Ti). The 2080Ti is very slightly over-matched if you were running 1080P, but certainly not at VR resolution. However, once you change GPUs in order to see better performance in VR as per step 1 above, you're now approaching a situation where the CPU and GPU are not as closely matched (since the GPU was increased). So, you'd need to seriously consider increasing the CPU side of the equation - and, as mentioned above, because your 10700K isn't terrible, and because the last few Intel generations are sort of 'meh' comparatively...then you'd have to take a pretty big step to get an upgrade that's "worth it". And, realistically, whenever we start saying "worth it", what we're doing (like it or not) is we're now folding in another factor: Budget. There is very little doubt that the 7800X3D is the best you could hope for in terms of upgrading your CPU. Of course, as above, upgrading your CPU will require changing motherboards and (for an upgrade that's "worth it) may require new RAM (if you land on a platform that requires DDR5). I follow prices routinely, and I can tell you that if we consider worst case/retail right now, a 7800X3D platform upgrade (CPU, motherboard, and 64G RAM) can quickly approach or even exceed $800 (depends on specific board and RAM). I looked not long ago and came up with $925 after local 6% sales tax for 6000MT RAM and a B650E board. I don't want to start a pissing contest here, but my strongest professional recommendation would be that if you're paying to change to an AM5 platform anyway, it makes no sense to go with a PCIe 4.0 board. Stick with X670E or B650E. Even if it's not apparent today, for the difference in cost it will be apparent soon enough. And this is why I say budget: We're looking somewhere north of $800 or even $900 for the platform upgrade (and this is *after* the GPU). Yes, there are some deals you can find to save a good deal of money on an AM5 platform, but they usually involve getting a basic board and RAM, which themselves aren't really appropriate to the top-end 7800X3D. You can go with a lesser CPU and save money, of course...but then, you're basically 'undoing' the entire point of upgrading the platform in the first place: If you go far enough, it's not going to be enough of a change to make it "worth" the cost. Right now, the local MicroCenter has a 7800X3D bundle that would run $525+ depending on taxes, but that's with a basic B650 board and only 32G RAM. It is NOT AT ALL the same thing as what I'm citing (at $925), which is why it's only $525. Budget: I'm not sure why, but people seem funny about discussing budget...but the reality is, it's typically easiest to set a budget first, then try to find something "worth it" within the budget. The budget can be adjusted if you really want to get more of a performance upgrade. But there is absolutely no point to go over all the possible options if you already know you don't intend to spend more than "X" (whatever that is for you). When we're looking at 700, 800, 900+ in costs plus a GPU, and (strictly an example here) you know you only have 600 absolute max...well, not much point drooling over a 7800X3D If your budget is constrained below the range of AM5 + GPU, then I'd say you should consider a good used GPU and platform upgrade (CPU, motherboard) for a 5800X or X3D which will allow you to re-use your RAM and thus knock a good chunk out of the cost. The 5800X/3D *will * perform better than a lot of the 9th-10th-11th gen Intel stuff, (I own some of all these, and have tested first hand and extensively)...but it is absolutely true to say the increase will not be the "WOW" that some stuff you read leads you to believe. I also believe the impact of CPU upgrades is often overstated on this site and others - this is a big reason I do not care for the 'bottleneck' concept. The CPU will have impact of course, but GPU is always the one doing more of the real work (and that's first hand, from someone who went through 3090 to 4090 and 9900K to 5800X3D to 7800X3D in less than the past year). I've also built many of each of these class machines for others recently and still do it every day. I've recently sold/built/tested two 3090s, a 3090Ti, a 4090 GPUs, on 7800X3D, 5800X3D, 5800X, 11700K, 9900K etc etc etc...lol Best deal you're going to make, IMHO: Find a used GPU and high-end AM4 CPU/board, keep your DDR4 (or trade it - read on). Work with someone who will offer you trade in on used parts if you have anything to trade. It is not always true that buying used means no warranty, nor is it always a mystery/question mark in terms of what you're getting. Yes, this is true if you buy from auction sites or even Facebook marketplace...but deal with reputable people who sell and guarantee used hardware, plus allow returns and warranty, and take trade ins...can't go wrong.

It's a cause-and-effect relationship. Understanding that matters, because the reality is that you cannot separate the two. In my professional opinion, you're trying to solve the symptoms of a problem, not the problem itself. It's a lot like me expecting a doctor to fix my lungs, even as I continue to smoke. (Just an example; in reality I quit over 12 years ago). In any event, best of luck to you.

At the risk of speaking for MAXsenna, I believe he's asking is if Windows recognizes the devices when you re-pug them (even if DCS doesn't). He's asking basically as a test: Windows should 'see' the devices when you replug them, and if that doesn't happen, then it points in a different direction from a troubleshooting perspective. If Windows recognizes the devices when you replug them, then it's a problem that's almost certainly confined to DCS itself. If Windows does not see them on replug, then there could be something going on in Windows which is preventing DCS from seeing them. If Windows doesn't 'see' them, then DCS can't. The Windows 'layer' underlies the controller(s) being available to other stuff (games). As for my question about resolving the need to reboot, rather than worry about why DCS won't recognize devices: It isn't common to unplug them (at all, really)...I do understand if you're working with the new stick - but can the stuff you're working on be resolved of itself (outside DCS), before moving on to DCS? Just a thought. As far as forgetting to plug in/switch on the joystick...is this something you do often? Again, it just doesn't seem common (to me) and I'm trying to understand why it happens. For example, do you change joysticks between computers/consoles/etc such that you have to remember to swap it, possibly? Both myself and MAXsenna are trying to figure out what might be done to overcome the problem you're facing - but, of course, solving any problem requires actually understanding the problem (often in detail, I might add)

Indeed lol

Of course. But it would certainly seem that "simple" would be to deal with the problem at the source Maybe it's just the way I think...but my first thought upon reading the first post was "But why do the things disconnect in the first place?" In fact, it seems so obvious that I felt as if I must be overlooking or assuming something I shouldn't... ...but, if that's the case, I can't see what that might be (at least thus far). Perhaps the OP would be good enough to enlighten us.

You're not losing performance per se; It would be more accurate to say you'd be making matters worse (at times, much worse) if you had the OS and the game on the same drive. The key to consider here is that the system uses storage constantly. Files are read by both the operating system and other programs (to do whatever)...and files are written to, for various reasons, as parts of programs running (think program log files, as one example). If you install the game files (any game, not just DCS) and the operating system on the same drive, then at times, that game and the operating system will both be trying to use the single storage resource. (How much is a complicated discussion, but for the purposes of your question, it doesn't matter how much, because when it happens, it's very bad). In technical terms, we call this "resource contention": Multiple processes attempting to use a single resource at the same time. It's easy enough to see, then, that at best, each process is only able to use the resource at a part of 100%; in other words, with two processes you might have a 50/50 split, or it might be 80/20...but because there is only one resource to be utilized (a single storage device), then one way or another, like it or not, somebody's waiting (to be accurate, usually everybody is waiting - more than they'd have had to otherwise). It doesn't matter how much or how little, nor does it matter how often it happens or anything else...the absolute, brutal fact is that a machine can perform better with less resource contention. If it were possible, then ideally, every single thing that happens in the computer would be given 100% of the machine's resources, at any time, for as long and as often as needed. This way, no one's ever waiting on anyone else. Everyone gets what they need in the absolute shortest time possible. But, common sense tells us this isn't possible. It is unquestionably ideal, but it's just not possible. So, the objective becomes trying to get everything as close as possible to that ideal. Forcing resource contention by storing everything on one drive is a very bad idea, especially considering that storage is (relatively) cheap. You might also consider that the folks who design these systems are fairly smart people, and the fact that they see fit to put all the different drive connections and controllers in a design should tell you that they understand it's important to avoid resource contention. So if I understand your question, are you better off having games on a separate drive? Absolutely.

This is a good point... ...but, I might go a step further. If the question is "What should you/we/I be troubleshooting?", perhaps a better answer is "Why do the device(s) disconnect to begin with?". I wouldn't think of that as normal behavior, regardless of circumstances, and if it weren't a problem then you wouldn't need to worry about what to do when it happens. Kind of like playing chess...don't get yourself into a bad position, and you don't have to worry about how to get out of it. Just a thought.

I can assure you, there were no 2G 1660 Super cards. Run GPU-Z, it should give you proper values. You should also be able to look at the Nvidia Control Panel and see the card's properties (see below, albeit for a 3090). https://www.nvidia.com/en-gb/geforce/graphics-cards/compare/?section=compare-16 https://en.wikipedia.org/wiki/GeForce_16_series What is 'demanding' honestly becomes more subjective. People have very different ideas about what is 'demanding', what is 'acceptable', etc. A lot of people complain about using 'synthetic' benchmarks, but this is the real reason they're valuable: It takes the subjective factors out, allowing objective comparisons. The 1650 was and is the go-to for those SFF machines simply because it fits the available physical space and power, that's really it. It is not a high-end card, but it can manage some simpler games at "good" frame rates. There are a few other alternatives, but not really much that makes any sense. Everything else will under-perform, and/or require too much power, and/or costs *many* times more than the SFF unit itself. PS: Found this https://community.ccleaner.com/topic/63672-speccy-not-reporting-graphic-card-gpu-memory-size-correctly/ which indicates it's a known problem dating back a couple years, but without update from Speccy. I don't use it, so I can't say whether it's been updated or not.

The difference in a 1660 Super and a 'vanilla' 1650 4G DDR5 card (which I believe both LP models are) is between 60 and 80% (or more, according to various online sources, depending on who you ask, what games, resolution, settings etc). That's a lot. Even a 1660 (non-Super) is around 50-60% faster than a 1650 (14469 Graphics score in FireStrike, vs 9112 for the MSI LP 1650), so it makes perfect sense that a 1660 Super would be that much better. No one's saying it wouldn't run, but it's obvious that a 1650 would struggle at the same settings and resolution where a 1660 Super does fine. Moreover, given the OPs subject machine (an HP SFF), and that no one makes a 1660 Super in a card that will fit...well...moot point anyhow. BTW not sure what card you're referring to, but unless I'm mistaken, 1660 Super cards had 6G VRAM, not 2.

As discussed above, everything shown in the OPs original post is all Intel hardware. It's not a mystery, and no guessing is required. I realize it adds to the confusion and frustration when things like "Unknown Device" appear, but if you use resources described above (online PCI database) it is actually easy to identify most anything in Device Manager. The unfortunate fact that the Intel driver tool isn't cooperating, etc - all this is perfectly routine for working on PCs that always want to be a pain in the ass. If you don't want to deal with the stuff, you need to seek help from someone who knows what they're doing. If that machine's in my shop, guaranteed I'd have it straightened out - entirely - in less than an hour or so (usually a few minutes). But many people don't want to pay for a professional. They'd rather do it themselves... Up to the individual. Like most anywhere else, it tends to be true here as well: You get what you pay for.

Uhh...that's not even an SFF machine (it's a "Mini PC"), particularly not one like the OP has asked about. Its a cube full of laptop grade hardware, which - as everyone knows - is not nearly the same as the desktop counterparts. Buying into a proprietary design like that is just as much a mistake as buying into Dell/HPs proprietary SFFs. It's overpriced for what you get, and when you get ready later on to upgrade, you'll be SOL. Proprietary designs in personal computing are a joke, period. Always have been, always will be. The single biggest advantage "clone" hardware brought the PC industry - compatibility- is completely lost in *any* proprietaty design, by nature. "Small Form Factor" in this instance (per the OPs original topic) is a proper noun, not a generic description. Yeah, sure, you can buy all sorts of "small" crap, including so-called "gaming laptops". But all of that is way off topic here. Nobody said you can't run DCS on a "small" machine. Mini ITX machines are small and powerful, without being proprietary like that video shows. ITX is a standardized form factor. With SFF at least there is general industry consensus on the size, at any rate - and generally they use FlexATX power supplies (which is also a standard) and typically can house half-height expansion cards, meaning some industry-standard add in cards will fit. "Small" is not a standard, neither is 'small form factor', and although I'm not certain, I don't think "Mini PC" is an actual standardized form factor, either. It matters, because if it's not an actual standard form factor, then the hardware manufacturers aren't going to make components to go in it. Which leaves you screwed at upgrade time. What has been said is that it's not a good idea to try upgrading one of the HP/Dell SFF machines (specifically), to run DCS. That was the topic started by the OP. Oh BTW LOL Desktop 3090 45% faster at 1080p...and those cute little things cost how much? I saw $1000 on one, before taxes etc. Great if you need something portable (I guess)...but for actual desktop gaming? Same Better performance, save money, get a 3060

24H2 = Second half (H2) 2024; not very likely before July, but likely thereafter, and more so as time goes on. (I am not privileged to their deployment schedule, and most anything else would be speculative for now).

OK, best of luck to you.

That only shows drivers for one device being installed, and it has nothing to do with the problem devices in Device Manager. Can you post a pic showing which drivers are actually installed?

Again, the Unknown device is also an Intel driver. If it's not corrected, you're not completing part of what needs to be done. Also, everything I see in my research indicates the Intel Driver Support Assistant should take care of this - if it 'didn't go well' there's a reason. "Didn't go well" doesn't really tell me anything about what went wrong. Try to think of this as talking to a doctor on the phone. He can only offer help as good as the information you give him is.

Well, something changed - the RAID controller is no longer in the list of "whacks" (yellow exclamation points). So there's that. Also, on the others: I can tell you without even looking it up: Vendor code 8086 is definitely Intel. So those devices are definitely Intel devices. Checking the online PCI device database: The 7AA3 device is listed as the Alder Lake-S PCH SMBus Controller - which agrees with Device Manager as to what the device is. ("Alder Lake" is your motherboard's Intel Z690 chipset) 7ACE is a Alder Lake-S PCH Serial IO I2C Controller #2 ACPI\VEN_INTC&DEV_1056 is also an Intel Serial I/O driver. So...collectively, this indicates there is a problem with the Intel drivers for these devices - either they didn't install, or didn't install correctly. In any event, this is what's going to need correcting. Have you tried just deleting these devices and rebooting? This will sometimes force incorrectly identified/installed devices/drivers to correctly install. What, exactly, 'didn't go well' with the Intel tool? Have any Windows updates been done on the machine? Often, drivers are released to work with an updated Windwos install, and if your WIndows isn't up to date, the driver install will fail. When you "downloaded MB software from asus", which exact file(s)/driver(s) did you download/install (by name, if you please)? There are several that have to be installed, and typically not obvious just looking at their drive page. Also, you mentioned "This happened once before, & I found a fix, don't remember where from." Please tell me what the fix is (never mind where it came from); looking for something like file name(s) or program name. When it worked before, was it also a clean W11 installation? Kindly note you need to be as specific and detailed as possible when answering. Guaranteed, the answer to your problem will be among these details. There are quite a few things that can go wrong with driver installation. Most of the time it just works, but sometimes...well, it can be a mess.

I don't blame you necessarily, and I wouldn't recommend that. Why not at least try the Intel tool I linked above? I wouldn't think you'd have trouble trusting Intel (any more so than any other major industry name, that is) Can you at least post pics following the steps I gave for getting the Device ID's? I'll do the legwork required, but can't get the Device IDs for you. Device Manager>(right-click on device)>Details Tab> Use Property drop-box to scroll down to Hardware IDs.

I see. Unfortunately it would seem you're not going to get too far, then. Perhaps your best bet is to consider professional help. I sincerely wish you the best of luck.

Yeah...the problem with that command line is that it can only return something if there's something programmed there (by the manufacturer) to be returned - obviously not always the case. Besides, you already gave the motherboard model number above (Asus TUF GAMING Z690-PLUS WIFI), so that's done already. Have you tried getting the device IDs from Device Manager as I described earlier? What about the Intel tool? If I may ask, where did the machine come from? Usually, I'd expect the seller/builder to provide support.

Dude, you're wrong. It's not water, period. I said nothing about what's in tap water or whatever; those are things you brought up to try and make your point. It doesn't make any difference about those, what I said was what is in the AIO coolers is not water, and it's not. Not H20, thus by definition not water. Any idiot can understand that. All that other crap you're trying to introduce, I never mentioned at all. Straw man.

Can you identify your motherboard model, please? For each of the devices in Device Manager, there is a way to obtain the vendor name via an online PCI database: Device Manager>(right-click on device)>Details Tab> Use Property drop-box to scroll down to Hardware IDs. Example below is for an Nvidia 3090 (Vendor 10DE, Device 2204). You can do this for those devices and use the online database to determine if they trace back to Intel. That said, I also think they may be Intel drivers. Sometimes, the manufacturers are bad about assuming you'll be connecting online anyway, and when you do, Windows Updates should 'pick up' all your drivers. Incidentally, if you haven't done that yet (and have no objection to it) you can try Windows Update. It will often update hardware drivers. But, if you want to try the Intel route, you can use this: https://www.intel.com/content/www/us/en/support/detect.html (Intel Driver & Support Assistant); a tool that you can install/run to identify Intel devices/drivers your system needs.