Aapje

No, it's the opposite, and even then it is a combination of the bus size and how many bits are used per GB. Once the 3 GB modules get used, cards can have 50% more VRAM for the same bus size. And it is also possible to run modules with half the bus width, so they could make a 24 GB 4070-variant right now, although the bandwidth to each memory module would be halved.

About bus size, what actually matters is how fast the data can be transferred to and from the VRAM. This is a combination of the speed of the memory and the size of the bus. I expect a big uplift on this front for the Nvidia 50x0 cards because the new GDDR7 memory has much higher speeds, even if the bus size is not increased. And the 4080 Super uses faster memory, so the actual difference with the 7900 XTX in memory bandwith is a bit less than the difference in bus size would suggest (736.3 GB/s vs 960.0 GB/s).

Hmm, that is very naughty (also, legally so).

Keep in mind that the 5080 is rumored to be announced at CES very early in Januari. I would personally wait for that, also since the 9800X3D should drop in price by then. And the Reverb G2 is end of life, so you might want to start planning a replacement.

There is no issue if they copy the functionality, rather than the actual code.

No, 7800X3D's come in two variants, with 1 CCD and those with 2. Basically, they create a bunch of CPUs with two CCDs and 3D V-Cache. The ones that are fully functional get sold as 7950X3D. Those that are a little bit broken get sold as 7900X3D, where 2 cores are disabled on each CCD. Then other chips have a completely broken CCD or connection between the CCDs and they get sold as 7800X3D. But demand is greater than what can be supplied with these partially broken chips, so they also create CPUs with 1 CCD and 3D V-Cache. These get sold as the 7800X3D if they are fully functional and if 1 or 2 cores are broken, they get sold as a 7600X3D with two cores disabled. In general, a crucial aspect of manufacturing CPU and GPU chips is to make sure that those with defects can still be sold, for a lower price. It is much more expensive to have to toss a CPU/GPU chip in the trash, rather than sell them for a lower price.

So do you consider map releases successful that see almost no use in MP? At that point that map is not just unappealing to a subset of players, but the evidence suggests that the map is not usable for an entire game mode of DCS. I would suggest that such a situation, with many players remarking that buying maps gives very little return on what they cost, might require a bit of a rethink. Yes, but making lots of products that stay in early access for a very long time, or have serious issues after release, is also a shame. At the end of the day, you can't do everything and choices have to be made between different alternatives that all have their pros and cons. In this thread a lot of people are making suggestions that some choices should perhaps be made differently. And spreading yourselves thin doesn't necessarily mean that you deliver more, since there is a cost to getting back into something you set aside for a while, and for keeping track of a lot of half-finished projects. I assume you mean B2B, because the consumer gaming market is also professional, as most of those gamers do pay you.

The question is whether this is a sustainable approach. If people start to feel that their spending is not giving them what they want, or even feel deceived, then they may stop spending. And it also generates bad PR when lots of people have negative things to say about the product. It can even result in a death spiral where to maintain income when people leave, more choices are made that maximize short term income, over long term health of the project. Futhermore, should the game focus on a 'specialized community', or should it focus a bit more on growing the community? Perhaps that requires choices that return a bit less income in the short term, but grow the player base and in the long term, result in more players that are happy to keep spending. As some of the posters have shown, statements have been made by ED employees that create expectations, which then haven't been met. I think that not acknowledging this, but implying that customers have unrealistic expectations through no fault of ED, is a bit of a cop-out. It's not really fair to set expectations by making statements and then chastising people for having those expectations. You keep repeating this, but I see people complaining that some of the goals that you are working towards don't seem that appealing, like making maps that have relatively little use or making a questionable dot system. Getting to the finish line fast is as much about speed as it is about going in the right direction.

It seems like a missed opportunity to not release a low-resolution base map for free and then sell the high-res version/sections. That way the value is much greater for those who buy, since they can actually use the map to play against their friends and/or in servers with enough players. And players who don't buy the high-res map also get more value out of their other purchases. And it is also highly beneficial to campaign makers, who have a much bigger audience if their campaign features these maps.

Is that a guess, or do you actually have it working? That extension is intended for a Winwing base, while the Moza base has a Trustmaster connector.

In my opinion, Moza's software should handle the switch of the axes. It doesn't make sense to have to change it in every game separately, if that is even an option.

Keep in mind that you will stop getting security updates after the end of next year, so at that point you are risking getting hacked.

The worst part is that they seem to be befuddled by the issues. How can it be that reviewer after reviewer runs into issues, that Intel themselves didn't find? Don't they test their own product like a reviewer would? It's so embarrassing for such a big company.

I have other priorities first (including an FFB base, but it does look very interesting).

The 5800X3D is now discontinued, but the 5700X3D is a good alternative.

Interesting. Do you plan to make a kit for those who already have a Crosswind?

A 5% discount is their standard discount that anyone can get by leaving stuff in the cart for a while, so it's not really compensation.

Technically that is true, but if you enable XMP/EXPO, then the motherboard uses the timings that are defined on the RAM stick. So if you have RAM sticks with different XMP/EXPO settings, you get a mess. Of course you can override that, but very few people run custom timings. Figuring out what timings work by yourself is beyond most users, and my advice to replace the entire kit is exactly to prevent people from getting into a situation that they are unable to get themselves out of, because of a lack of skills, time, or willingness to take the effort.

@kksnowbear Building a new system with 4 populated RAM slots is far less error prone than going from 2 to 4. Any difference in timings can wreck havoc with stability, but most people only understand the speeds and think that any 3200 kit (for example) is the same as any other 3200 kit. Even if you buy the same product name at different moments, you might actually be getting memory from different manufacturers with different timings (almost all memory is made by a different company than the one that is selling it and they can switch manufacturers to save money). Only when buying the 4 sticks of the same product name at the same time, are you almost certain to get 4 times the exact same thing, that comes from the same batch. This is why I would always advise replacing/installing all memory in 1 go, if reasonably possible. For a gamer with relatively low RAM requirements, that would generally mean something like removing the existing 2x8 or 2x16 kit and replacing it with a new 2x16 or 2x32 kit.

With the Rhino, we know that telemetry-based FFB in IL-2 is much better than the native FFB. Here someone who tried both, called the native FFB anemic in comparison to using the Rhino software: https://forum.il2sturmovik.com/topic/87017-vpforce-rhino-modern-ffb-joystick/?do=findComment&comment=1329498 But my expectation is that it will take quite a while before Moza is up to par with the Rhino software, although they do have the benefit of being able to learn from it.

Interesting choice of music.

Yes, there is a benefit, since you are also pumping less heat into your house, need less cooling in your PC and will have a less noisy PC. A 7800X3D seems to save roughly 100 Watts compared to a 13700k, which probably uses about as much as a 12900k: https://www.youtube.com/watch?v=78lp1TGFvKc So that is 0.1 KWh saved per hour of gaming. With the current average prices of of 32 eurocents per KWh in my country, that is 3.2 cents per hour of gaming. If you game 20 hours a week, that is 64 cents a week, and 33 euro and 28 cents a year. So if that CPU lasts 3 years until the next upgrade, then you save 100 euros on electricity during that period. And again, because you guys keep misreading me, I am not claiming that you will earn it all back. How much you earn back depends on all kinds of factors, like the pricing of components at that time, your electricity prices, how much you game, at what settings, etc. And you also need to factor in that the depreciation of the 7800X3D or 9800X3D is probably less than for the 12900K, if you resell it. What actually matters is the TCO. Of course, right now the situation is pretty much as bad as it can be, with the current inflated prices for the 7800X3D. But I've seen the 7800X3D go as low as 329 euro in my country, so with that kind of pricing you would in fact earn back the difference completely in 3 years of 20 h/week gaming.

Well, you will probably use it for years, and I never said that you would earn back the difference, but that it reduces the gap. There are a bunch of other benefits, like having less heat and being able to get a simpler cooling solution.

@pierscockey Yes, that is the exact same text as in the racing wheel software.