Aapje

@LucShep Sigh. It is a new chip design, you were incorrect. The entire reason why 13th and 14th gen are affected, but not 12th gen is because the design is different. Just accept it and move on.

I am amazed that it does, but I guess that you've found some settings that work for you somehow. In theory, the 2080 Ti is twice as fast as the 1070, but that is when tested with a much faster system. The issue is that the GPU needs to be fed by the CPU. So the slower the CPU, the more a fast GPU will be slowed down by the CPU. And especially in simming, the game already puts a big burden on the CPU, so the chance is very high that the 2080 Ti isn't going to do much for you. What I would do is to keep saving and to look out for a good 2nd hand deal. Perhaps a second hand AM4 system with a slower processor, that can get a huge boost with a 5700X3D upgrade.

All AM5 board support DDR5 only. And 6000 speed DDR5 should work with all of them.

No, this is incorrect. Raptor Lake is actually a new chip design. It's a bit complicated though, since the 13th gen processors are not all Raptor Lake. Everything from the 13600K to the 13900K is Raptor Lake, but everything below that is rebadged Alder Lake (so 12th gen). Then for the 14th gen, the entire line up is just rebadged and overclocked Raptor Lake. Note that they've been bragging in the past how quickly they designed Raptor Lake, but they probably regret that now.

Keep in mind that productivity is a major focus of his system, so it should work well for that too.

Probably.

The 13th and 14th Intel CPUs have degradation issues. I would instead get a 12700k or go for an AMD solution. AMD is now the most popular, although with your use case you have a bit of a hard choice. The 7800X3D is the top tier gaming CPU, but is a bit lacking in cores for photo editing/post-processing. The 7950X3D is top tier for both, although quite expensive. A good compromise can be the 7900X3D, which sacrifices a bit of gaming performance, but has 12 cores instead of 8 for the 7800X3D. I would personally go for the 7900X3D. Note that you will need an AM5 motherboard for it. Also, you shouldn't get the 5200 memory (note that this is not Mhz, but transfers per second. Since it is Double Data Rate memory, it sends data twice as fast as the speed in Mhz). Instead, get DDR5 6000 memory.

Companies have a strategy of giving you something a bit better if you spend more, but if you keep doing that, you spend way more. So I can see drawing a line somewhere.

Yes, it looks like a competent design based on the internals and the very early reviews are good. Haven't seen any longer term reviews and it is obviously too early to know about long term reliability. It seems a little overpriced compared to the competition, but for you that shouldn't be an issue since you can reuse stuff. I'd personally be a bit wary about spending so much for a non-FFB base right now. Perhaps the Moza or the Winwing FFB bases can be made to work with a desktop-mounting solution and then that seems way nicer than any mechanical base. And even if you are not up for that, you may see a flood of mechanical bases on the 2nd hand market for fairly low prices in the future. But on the other hand, you can always justify waiting and then you never get anything nicer.

It wasn't so much lack of demand for FFB gear specifically, but lack of demand for flight sim gear in general. The entire hardware market stagnated, which allowed new brands like Virpil, VKB and Winwing to compete, despite very little marketing and fairly small R&D departments. Established brands stopped caring for the most part and at best just kept selling the same old tech. MSFS 2020 revived the flight sim market, especially with the good fortune of it coming out during Covid-times, where it allowed people stuck indoors to travel anyway, albeit virtually.

Stuttering is not a black and white issue. And perception is not objective. An objective way to measure it, is 1% lows, or even better a frametime graph, that shows how long the frames take to produce. For example: This graph shows very bad stuttering, as you can see huge spikes. Some frames take 3 or 4 times the average time, and those frames cause the old frame to be shown for a very long time, which is a stutter. This is a much better graph: But even in this graph, you can see that there are little spikes. These spikes are so small that no one would notice them, most likely, but they are still there. Anyway, the goal is not to get rid of stutter, by getting to a graph without spikes, but getting the spikes low enough that you won't notice anymore. And what you notice depends on the individual. And of course, it's not just about the spikes anyway, because the easiest way to reduce the spikes is to lower the graphics settings. So the actual goal is to have the highest perceived graphical quality (which is subjective) with an acceptable level of stutter (which is subjective) and an acceptable average framerate (which is subjective). Because of that all subjectivity, it is very hard to give really good advice. Some people are very happy with a result that others considers unplayable. But there are common factors and things we can suggest. The first thing I notice is that running on 4k with these specs is extremely ambitious. Although a higher resolution does put less load on the CPU and more on the GPU, and this system has a better GPU than CPU, it's still highly questionable whether even the video card can keep up. The 4060 Ti is a strong 1080p card, a mediocre 1440p card and a poor 4k card. If it is the 8 GB version, then it is especially ill suited for 4k. But the platform is very old and I would suggest considering an upgrade, but I don't know what the financial possibilities are. You can always trying experimenting with the settings. In that case, I would strongly suggest also testing what happens if you lower the resolution.

There is something that currently limits the market and that is a lack of official support for FFB in MSFS and non-optimal support in certain DCS modules and perhaps other games as well. So from that perspective it makes sense to first release higher-end products, as the buyers of those product tend to be more willing to tinker and buy add-on tools. Then the game/module developers can use those higher-end products for development, and the manufacturers of the higher-end FFB devices can simplify and improve their hardware and software, before attempting to go for the smaller-margins, larger volume market. As for the demand being there, it seems to me that the players with a rig are only a fairly tiny fraction of players. So there should be quite a bit of demand for a desktop version, assuming that the demand of FFB is there in the first place. Of course, to truly create big demand for FFB, we need quite a bit of marketing. Not just some old geezers who remember the olden days, or weird forum-dwellers who know everything (or think they do). But most people don't just magically become aware of, or excited about FFB. But as we've seen with simracing, FFB can become the way to sim.

Could be, I never made any claims about anything coming to market. But with the Moza base getting sold for $550, it is very easy to imagine someone making something 10% cheaper. That is a difference that can easily be covered with a smarter design, a saving on materials, using smaller motors, or such. And if the FFB market turns out to be much smaller than anticipated, Moza might even drop theirs to $499 to get rid of them. Like I said before, that is trivially possibly by making a copy of the MSFFB2, which sold for less than that, in 2024 dollars. It's simply ridiculous to assume that we could not make something similarly cheap, with improved technology. But perhaps you believe that this generation of FFB will be a flop and companies will deinvest, rather than continue on this path.

Yes, so the companies most likely first want to see whether the expensive products are a success (and whether MSFS will support it natively), before they commit to a cheaper variant. But that still doesn't make it impossible to make a much cheaper variant.

@propeler There are a bunch of reasons why we don't have such a device yet, but it's definitely not the case that it is impossible to make a substantially cheaper FFB joystick. The Ursa Minor is not a direct copy. And apparently VKB chose not to or couldn't patent the gimbal. That's how it goes.

This is way too simplistic and one sided. Smaller saves material, which is a cost saving. Overall, miniaturization can save money or cost money. It depends. Going from the vacuum tube to transistors and shrinking those caused huge savings. In that case, smaller was cheaper! There are a lot of ways to make a FFB stick smaller in a way that saves money. For example, use smaller & weaker motors (which in turn means that you need less material for other parts of the system as well). I respect your work on the Rhino, but from what you've said here and elsewhere, I feel that you take your experience in a low volume machine shop and apply it to large volume, factory manufacturing. However, this causes you to make statements that are very wrong. The very fact that you focus on how easy something is to make, rather than how cheap it is to make, shows an incorrect mindset for volume manufacturing. In large volume manufacturing, you can get huge savings by spending quite a bit on an optimized design and things like plastic molds and specially set up machines. These investments would result in a huge price if applied to small volume products, but pay off hugely with bigger volumes. It's a special skill set to know what large volume manufacturing methods work well in what situation, and to design a product that is optimized for certain manufacturing methods. Key to this skill set is to be able to bring costs down, for example by using less material (including by packaging things more efficiently and thus making things smaller), having fewer separate components, using cheaper materials, having fewer manual steps, reducing variance (and thus fewer products that go out of spec and have to be tossed), etc, etc. The product I envision would not focus on realistic levels of force, which wouldn't work well on the desktop anyway. The MSFFB2 has around .8 Nm of torque, but the motors can do more than that (see the resistor mods), so it should probably be possible to go to at least 2 Nm with similar sized motors. Then it would have a fixed stick, saving money on having a robust connector mechanism on both the stick and base. The lack of an extension and having the stick closer to the motor/gears, means that less motor power is needed anyway (this is also why simracing FFB systems often have relatively small wheels, since then the effective torque to the user is higher). That is why patents exist, so a design cannot simply be copied. Consumers also don't tend to like exact copies, except for designer brands. And there is an advantage to being first to market. Furthermore, the sales calculation for the second entrant is much worse, since they will lose a lot of sales to the first mover. So simply by being first and making it so a new entrant cannot enter that market without competition, you discourage new entrants, at least for the exact same niche. This is why different companies tend to try to differentiate their products, so they are not actually fighting over the same exact demand, but they increase the size of the total market. For example, the styling of Lamborghini and Ferrari tends to attract different buyers, so there are people who are willing to buy a Ferrari, who would never consider a Lamborghini and vice versa. Making a lower-budget desktop stick would actually be a way to distinguish oneself from the Rhino/Moza/Winwing/etc, which are all very similar in use case. A desktop FFB stick would attract a whole new population of buyers, who would never buy one of the other FFB options. By your logic, no new devices would ever be made and there would be a lot of exact copies, but in reality that is not the case.

We may see someone develop something similar to the MSFFB2, but then updated with much more modern tech, but no sign of anything like that so far.

Note that I have an overview of the various bases with the information about them here: https://forum.il2sturmovik.com/topic/88770-ffb-joysticks/ I use that to keep track of the info myself, so I don't have to go looking for it again.

I suggest finding someone with a 3D-printer to make you one of these: https://www.thingiverse.com/thing:4750811 The Combat pedals seem to be derived from the same origin as the pedals in this thread.

I'm sure it'll keep its value quite well, so you can always resell it and switch over.

I'll wait for the actual prices before I get too excited.

That would be surprising, as they seem to use bigger motors. Why do you think that it is cheaper? Do you have any specific facts/insights to share?

1. I don't see any issues. The 4080 is plenty fast for non-VR use 2. The 360 only fits in the front, which limits the size of the GPU you can use. I would suggest using a smaller AIO (240) or going for air cooling (Thermalright Peerless Assassin 120 SE). The latter is more durable and reliable, but you may not like the looks. You don't need a 360 AIO for the 7800X3D. For better clearance, get the non-RGB version of the RAM. 3. Just get a 3-pack of fans and see what extra you can fit in the case. Try to have one more intake fan than exhaust fans, to combat dust.

There is a big difference between motion rigs for VR and non-VR. In the last case, the rig conflicts with the visual cues that you get, while in VR, you can take advantage of the visuals in VR to bridge the gap between what you actually feel and what you ought to feel. The issue with these huge movements, is that you introduce lag. It may look impressive, but for fast flying, I think that a relatively light platform with smaller, precise movements is better than these amusement park rides.

There is little point in waiting for the non-X3D parts. AMD has implied that that the 9800X3D will introduce some improvement beyond what we say in the past. Rumor has it that the new chip will be able to clock higher. And rumor is that the 9800X3D will be released sooner after the non-X3D parts than last time. More like 2 months rather than 6. Then again, the initial price is probably going to be rather high. Note that at least Zen 6 should also be compatible with AM5, so you can also get a 7800X3D and then later upgrade to the last X3D-chip for the AM5 platform.