kksnowbear

Balony.

According to whom? Where is your data? Having any actual data (and not extrapolation of Nvidia marketing drivel) would be remarkable since, AFAIK, reviews of actual hardware aren't going to happen til the 24th. Sounds like more BS speculation, but being represented as fact.

Now *that* is the best idea I've heard yet!

And you keep insisting that somehow this is a simple misunderstanding about a word. It is not.

Having been close to addiction at a far less responsible point in my life decades ago, I can see your point And yes, with games demanding ever more performance, it seems like this business with letting AI do all the work, and 'fake frames' artificially inflating performance to an absurd level...I dunno, the whole thing is really disappointing to me, honestly. I actually get it, the physics and science that goes into all these components, I spent more than four years in formal study about all that...die size, shrinkage, current, heat transfer... ...but knowing *why* it is - although I do genuinely see the reasoning in the approach - it may even be looked upon as inevitable...just doesn't seem to make it right. It still seems disingenuous and...wrong. That they would make such a ridiculous claim as they did about the 5070 just absolutely proves my point. They're simply taking unfair advantage of the fact that people still want to believe you can get something for nothing.

Well, I guess that ought to address the MSRP argument, then... (And more specifically the point about cheaper bottom of the barrel models) All 4090s/3070s/.../(insert GPU chip number here) are *not* created equal; this has been shown many times over the years. So getting a card of decent quality often winds up costing even more.

What makes the MSRP argument bullsh*t is that the cards don't typically sell at that fictitious price, especially not at release. It has nothing to do with exchange rates. Sure, if you wait long enough, kneel down and pray facing a certain way, wear your lucky socks and rub the rabbit's foot lol you can find better prices (duh). I've found deals, I do it all the time. But that doesn't mean that all cards will eventually sell at MSRP, nor that they release at MSRP, nor that the actual value they bring is worth the cost to begin with. LMAO if all the were true, Nvidia wouldn't need BS marketing tactics claiming a 5070 can outperform a 4090...yet they've actually tried to sell that cart load of BS to the consumer. I own a 4090. It's a freaking awesome GPU and I got a better than average price...but I still tell people they're too stupid expensive and not really worth the cost. Also, you can't compare best-case special pricing on a third-rate GPU released years ago to release prices for the model someone might actually want today. Sorry but I don't want to be forced into buying something I don't want, just because of a 'good' price. And again, none of this has anything to do with people being tired of getting ripped off. What happened before has nothing to do with it. It's taken time, but people are starting to see how ridiculous it is - unlike your assertion that people somehow 'don't understand'. "Getting worse" as applies in any meaningful human context is a matter entirely of individual perception and perspective. "Worse" itself is subjective. Worse than what, according to whom? And arguing that a Ferrari isn't sold as a value proposition is pointless; we're not talking about Ferraris, and there are far (far) more people who buy GPUs than Italian sports cars.

Yes, I live near MicroCenter too. You've been fortunate. I've been fortunate. What I'm saying is two guys, a half-dozen exceptional deals between them, does not a rule make, that's all. Among other things, as great as MicroCenter is, the reality is that less people (by far) have that option than the comparatively few of us lucky enough to have it. (And I believe it accurate to say they don't ship, at least most things anyway).

There are always exceptions. I've bought three 4090s for less than MSRP. But, as @LucShep indicated, these were exceptional deals/circumstances that really can't be applied as rule.

Uhhh... The laws regarding charging sales tax were changed beginning in 2016. Most online sellers were required to collect sales tax by 2017 (Amazon started in April 2017, strictly as one example). The 4090 wasn't released until some 5 years later, in 2022. (I believe it accurate to say that some states still don't charge sales tax, but the states that do charge it have been doing so since well before the 4090 was released, thus making it impossible that any system containing a 4090 [or a 4090 by itself] could have been bought without paying sales tax, in any state that pays sales tax). It would seem some claims don't stand up to fact-checking.

"Things getting worse" factually includes when expectations are not being met. It is neither required nor necessary to distinguish between the two for the purposes of this discussion. You're just trying to make it sound like someone's being unreasonable - but that's only according to your opinion. My opinion, as well as others here and elsewhere, is that he's not being unreasonable by being fed up with Nvidia's marketing BS and greed. And, without going into this (very thorough and insightful) level of detail, I'll just second the assertion that the "MSRP" argument is also BS. As @LucShep has indicated, the 'real' price of these cards is typically 25-50% more (and doesn't necessarily get back down to MRP over time).

You are not entitled to any such expectation. No one owes you an explanation, and anyone is free to make their own recommendations to others, regardless of whether you agree.

lol The wait list to buy these things just as easily indicates how much money people are willing to spend for nothing more than being able to brag about being the first to get one. Making a profit is part of business. Excessive greed is bad business, and eventually people start to realize it. Of course, there will always still be those who want to post in a sig on a forum somewhere that they have a 5090. These exceptions are just exactly that.

Whether/by how much Nvidia may or may not have misrepresented in the past isn't at issue. People can be lied to for any length of time before they decide they're done with it. What's at issue is people finally deciding now they're done with it due to the lies and greed, and no one is required to provide you with any answer at all as to why they decide as they do - whether you consider it rational or not. Keep waiting if you want, but I don't suggest you hold your breath. I personally see enough of their reasoning to see it as a reasonable decision. As always, what's "worth it" is absolutely and entirely up to the individual.

I can defend whatever I see fit, thanks. He's not wrong for feeling the way he does, and it's wrong to imply that his statements are silly and lead to people doing irrational things. What is also wrong is Nvidia's ongoing and increasing greed. For most people, the fact that eVGA decided to bail on the entire thing, after years of being king of the GPU hill, is evidence enough. I don't think people are actually as bad at understanding this as you do. I think in large measure, they're starting to understand fairly well. If each successive generation of GPUs were actually that much an improvement/value proposition, they wouldn't need to force obsolescence like they've done forever. If they were on the level and treated customers with respect, then why even try to fly this BS about things like the 5070? I'll tell you why: Because they know some people are still uninformed enough to fall for it. There are still an alarming number of people who believe that any 40xx card is inherently better than any 30xx card. I run into this constantly, even among 'experts'. The reality is that the majority of people who use PCs for gaming don't really posses that level of expertise; I'd say about 1 in 10 that I do builds for have something beyond even basic knowledge. A couple builds over 5-10 years does not equate in any way to a lifetime of studying and actual production. The xx70s have always been among Nvidia's most lucrative and popular segments. So why not use that popularity against the consumer, by implying (if not stating outright) the latest generation xx70 will outperform the prior gen flagship? It's wrong, plain and simple.

Stupid. Stupid, stupid, stupid. All this is incredibly easy to avoid: Deal with reputable people who offer decent, clean, professionally tested stuff at a fair price, who back it up, and who have a demonstrable, verifiable history of taking care of their clients. Simple. Yes, I am planning to sell a 4090. No, not because I am buying *anything* in the 50 series. Yes, I can provide references.

Not to speak for others, but I think he means he's tired of Nvidia's greed and baloney sales pitch...like the concept of a 5070 delivering the performance of a 4090, as Nvidia marketing BS would have you believe. The effect is that, while they're hawking this blather, the reality is that in order to exceed performance of a 4090, you'd have to go with a 5090 - which at release has an MSRP that is ~25% higher than the 4090 was (2000 v 1600). And I think it's a pretty safe bet you won't find one for MSRP for quite some time (if at all). I wouldn't be surprised to see them at $2500+, or even approaching $3000. 4090s are still over MSRP some two years after release. And the actual, hardware/rasterization "horsepower" - while admittedly yet to be confirmed - is expected to be around the same margin of increase; some reputable reviewers have already said they anticipate something like 25% performance increase for 25% price increase. No big whoop. Not this drivel that you can spend <$600 and get performance on par with a 4090. Balderdash. Smoke and mirrors. Maybe you believe that nonsense. I'm kinda like @The_Nephilim - I hope this blows up in Nvidia's faces. In the DerBauer video above at 6:40 he discusses specifically the 5070 vs 4090 thing, and ends by saying "Just don't do stupid actions; a 4090 is still a very good card", and then he repeats this at the end of the video. He also says "the 5080 is already slower than the 4090, so..." His English isn't perfect, but I think what's he's saying is crystal clear.

ROFLMAO Holy gumballs, I almost peed myself laughing. Ain't it the truth. TY TY TY

I would readily concede that the manuals leave a lot to be desired at times But again, don't underestimate the information that can be gleaned from utilities like GPU-Z and the manufacturer's SSD software.

Nah, it doesn't work that way The usage/assignment of PCIe lanes is covered in your motherboard manual, p 15-16 The M2s (with one exception) use the chipset PCIe lanes which are not shared with CPU lanes. The GPU uses CPU lanes, as indicated, as does the M2_1 slot. The 12700K you're using provides 20 "user-accessible" PCIe lanes, 16 for the GPU and 4 for the M.2. Although some boards will bifurcate the 16 CPU PCIe lanes when populating two PCIe slots (eg x8x8), they don't do this with the 4 CPU M2 lanes. The chipset M2 slot lanes are sometimes shared with PCIe slots, but an (admittedly quick) look at the manual doesn't indicated any of this type sharing. I believe it is accurate to say that chipset lanes are never shared with CPU lanes, and vice-versa. At least, I don't think I've ever seen it done that way, and I don't think it's possible with motherboards' design as it is now. Also, using Samsung's magician software will show you PCIe lanes used by the M2 drive (or other mfr utils for other drives)., while GPU-Z will show you lanes being used by the GPU. All of this is reliable information, no need to wonder if it's accurate (assuming, of course, you understand what to look at). This makes it fairly easy to spot if anyone's not getting the lanes they should. HTH

Nope, and I said "I don't want to lecture you or anything of the sort at all." as well as that I wished you the best of luck. Sorry if my comments were taken the wrong way. As I explained, my concern is for accurate understanding of the way this stuff behaves, for both present and any potential future readers. Power supplies (at present) simply do not work as you speculated. That's just not the way it works, and I feel it's important to avoid misunderstandings. Factually, that has caused far worse problems than anyone having their feelings hurt. I've read, fairly shocked, where one "expert" on this site recommended (and linked) a power supply adapter cable that was not compatible with the OP's PSU. Had the OP followed this advice, it would have undoubtedly fried his motherboard. I know this because I've had to repair a few machines where this happened due to people not knowing what they were doing. (Incidentally, the cable in the example I cited above wasn't even the problem, after all....so a single uninformed opinion yielded not only an incorrect diagnosis, but a potentially disastrous 'solution') I realize you were just 'throwing darts' but it's important to keep in mind that people read this stuff - sometimes years after it was posted - and walk away with some pretty incredible ideas. Thus, I believe it's important to take the time to offer accurate explanations, to discourage inaccurate conclusions/ideas. You wouldn't believe some of the messes I've cleaned up, all avoidable if it weren't for some guy's advice on Reddit, etc. Rather than just sit around wishing I could change this to help people, I prefer to take a more active part in trying to prevent the problems. Again, sorry if that offends.

Well...no. As was stated above, I don't want to lecture you or anything of the sort at all...but an underpowered PSU wouldn't exhibit the kind of behavior you've indicated. The GPU has no inherent awareness of how much power the PSU can provide, and therefore it's not going to reduce it's capability/performance in any sort of "low power" mode. The GPU will simply draw whatever current it wants, and if/when that's more than the PSU can provide, something will go wrong. (The newest GPUs/PSUs have some ability to avoid this, but it is still very imperfect; there is no 'intelligent' communication between a PSU and the components it powers). What the machine very likely will do at that point is lock up and/or crash frequently, and more so when it's hotter. I am also glad to hear you're happy with the new setup - good for you But it's probably better that we keep in mind accurate ideas about how these things actually do work, so as not to confuse/misdiagnose or mislead others who will at some point see this. I hope that makes sense, and wish you the best of luck.

Profile says Fairhope AL @tribundf check PM

This is among the problems with not buying memory as a set. It can absolutely be done and work just fine, but it sometimes requires extra effort. Unfortunately, all the memory manufacturers have at times changed the guts of their modules, with no prior notice. They buy what they can find in the market, and cost is always a factor as well. It is exceptionally difficult to 'specify' what devices (the actual chips) you want when buying a new set, because the manufacturers don't disclose it. One of the (very) few ways to do so is if you can find a seller who can tell you (but again, this is very uncommon). I'd consider the 'secondary market' because you can still find new components (if you must have new) and there's a much better chance you can determine what actual devices are used on modules. Also, Corsair (used to) put version numbers on the module labels, and I believe it accurate to say the version numbers would match if the same guts were used. One thing I do is, if I'm adding modules to a build, I make sure the actual version numbers match. Note not all manufacturers do this, and I'm not sure if Corsair still does or not. I don't generally use their memory these days (and as you've learned, they don't actually manufacture the chips anyhow - never did). Did you check the BIOS to make sure you're not running a profile that actually runs at CL40? It's fairly common these days to see at least two profiles in BIOS, and often only one of these is what the memory is actually rated for. TBH if it's rated for 32 etc, then the SPD should contain the proper data. You indicated CPUZ shows Hynix devices, but you don't say if it shows an XMP with 32 vs 40.

Late to the discussion, but yes: 6700K = Z170 = PCIe Gen 3 support. Thus, any drive you put into the M2 slot will run (at most) Gen3 speeds. The 990Pro won't perform in proportion to it's higher cost in that case; it would be just as well to buy a slower Gen 3 drive and save yourself some money because the slot is the limiting factor, not the drive. (If you plan to change to a newer board reasonably soon, it might be worth considering) However, with that being said: Unless I'm mistaken the 860 EVO is a SATA M.2 drive, meaning it's not operating at PCIe speeds. It's running (at most), ~550Mb/s - just like a 2.5" SATA SSD would. (The Z170 motherboard typically supports both SATA and PCIe drives; M2 is the drive form factor where SATA and PCIe are each a different type of interface.) That being the case, the 990 will definitely be faster - and quite a bit at that. Even at PCIe Gen3 speeds a Z170 motherboard supports (~3500Mb/s), the drive is around 7 times faster than a SATA M.2 (as the 860EVO). As mentioned above, the likely difference of noticeable improvement while running the game is probably small, but that depends. Specifically, there will almost certainly not be any increase in FPS, which is typically how people gauge performance (albeit misleading to do so). HTH