kksnowbear

This discussion was about the artificially low price (Nvidia MSRP) cited for the 50 series, and the artificially high performance, thus making any price/performance analysis invalid for any meaningful purpose (at least until more/accurate data is available). The point is that even at Nvidia MSRP (the lowest price cited anywhere I've seen) there is significant doubt as to proportional performance. (Higher MSRP from AIBs etc would only make the prospects worse, not better.) Please stop trying to muddy the water. No one but you is trying to confuse MSRP from Nvidia with AIBs.

The confusion isn't mine. I see the same thing that the majority sees here, including plenty of experts who have already said it's baloney. And not one of my clients had to come out of pocket to replace a melted 4090 or failed CPU that I sold them ...unlike some people apparently did

Nope. I'm not confused about anything, nor are a lot of other people who are beginning to realize what's going on. The confusion seems to be among those who still can't see what's happening. As I said already, I'll try to make it easier, for those having trouble seeing it: The performance being represented (by Nvidia) is artificially high, while the prices being cited (by Nvidia) are artificially low. This leads to supposedly valid price/performance comparisons that simply have no basis in reality. Of course, I'm sure a couple of forum "experts" know better than the growing number of very reputable sources online, all seeing what I do...even most intelligent laypeople aren't buying the "5070 outperforms a 4090" nonsense... And why would Nvidia even try to pass that off if they're *not* full of crap?

Pretty sure that the website will update the prices even if it's on your wishlist. Regardless, "because I said so" isn't really proof.

Again, the discussion as it applies here concerns Nvidia MSRP, not Asus. You didn't get your card for the MSRP that Nvidia specifies, which furthers the concept that Nvidia's MSRP is just part of their bullsh*t marketing tactics. (BTW you really aren't making it look better by insisting you paid more than Nvidia MSRP for what is still just a 4090...the more you paid, the less performance for price you're actually getting). It doesn't matter if vendor X sets an MSRP at $4000, what matters is Nvidia's trying to represent that a $550 GPU will equal a 4090. I'll try to make it easier, for those having trouble comprehending: The performance being represented is artificially high, while the prices being cited are artificially low. This leads to supposedly valid price/performance comparisons that simply have no basis in reality. What genius marketing! Claiming a product is better than it is, while saying it will cost less than it almost certainly will. (Never mind not mentioning you probably won't get one for at least several months after they say you can...) But hey, I can understand that some people will just believe that a $550 5070 will outperform a 4090. Because, you know, Nvidia said so.

The Asus screenshot isn't dated in any way. You have proof they were charging 1800 plus at intro (or that it's ever been Asus' stated MSRP, not what they happen to be charging now)? Not to mention manufacturer's sites will always be among the highest pricesl...so it's not hard to claim you paid less than that. The discussion as it applies here concerns Nvidia MSRP, not Asus. You didn't get your card for the MSRP that Nvidia specifies, which furthers the concept that Nvidia's MSRP is just part of their bullsh*t marketing tactics. The discussion here is that it is not an accurate representation of cost that the vast majority will be looking at - and thus a price/performance analysis based on that MSRP is also inaccurate. At the end of the day, we can go back and forth, but most people are smart enough (and have seen enough out of Nvidia and the GPU market) to recognize what really goes on. And many of the experienced, reputable reviewers have already said the 50 series is "meh". Fake frames, less availability than any other release, making the odds of getting one anywhere near (ridiculous) MSRP for who-knows-how-long after release...lack of real data on 'non-magic' performance...on and on...

Here's the relevant part of your actual post. You asserted two things: One, that you really didn't think you paid more than Nvidia's MSRP (though the 4090 receipt actually shows you did; AIB cards are almost always more than Nvidia's MSRP). Two, you said "these cards were all tax free" after having mentioned 4090s specifically, and more than once. I don't think it's much of a stretch to say that the comments imply you didn't pay tax for the 4090s, even if you arguably didn't state that outright. Looks a lot like you were trying to say you didn't pay taxes, without actually saying it. Maybe you want to make things more clear when you post.

Uhhhh...my comments were directed at the post I actually quoted. In that post, it was obvious that the card was not purchased at MSRP - as I said. Not sure how you mistook that as directed at you (I also actually mentioned a receipt, which I don't think you provided.) I don't recall saying you did not get a 4090 at MSRP. PS I am fully aware tax is not part of MSRP. No need to explain on my part.

Incidentally, I guess the claim you got a 4090 for MSRP winds up in the same bin as you claiming you didn't pay taxes on the 4090... the factual timeline on laws regarding taxes on purchases online would seem to dispute that claim. Fun (actual, real) fact, in context: The tax laws were changed (c.2016-18); several years before the 4090 was released (2022). Allowing for possible exception (as one example, states that still don't charge taxes), it seems chronologically impossible that anyone got a 4090 without paying sales tax. Laying that aside for a moment, the receipt you posted shows taxes were charged.

And all your mountain of speculation is just exactly that. Your "educated guess" doesn't prove anything, any more than your assertion about price/performance did. Even if someone uses all the tools on the planet, it doesn't guarantee they're gonna get a 50 series GPU for MSRP. It just...doesn't. According to several reputable sources, 50 series cards will be in far more limited supply than prior releases. And whaddya know, along comes your next post, to prove that very point (which I'd already made, some time ago): (Worth noting that, for all your brilliant "understanding", this still seems to have caught you off guard lol). I can say at least *I'm* not surprised, at all. And, of course, this will make it even less likely for anyone to get one for anything like MSRP. Mind you, even if they did, there's still the issue that at least one qualified reviewer has said even at MSRP, even with something like 30% performance increase, the 5090 will be "disappointing". There are others (like you) who have suggested the 5080 might "give a very poor increase in performance". The hits just keep on coming lol...

LMAO... The MSRP on a 4090 was 1599. https://en.m.wikipedia.org/wiki/GeForce_40_series#:~:text=(FP32) performance.-,Release,16%2C 2022 for %241%2C199 US. "The RTX 4090 was released as the first model of the series on October 12, 2022, launched for $1,599 US" You paid $200 more, thus you did not get a 4090 for MSRP. As your image shows, they're still typically above MSRP, more than two years later. 1800+ is not uncommon, *if* you can even get one. And that represents more than $200 over MSRP. Seriously. Fact check.

Does anyone believe they'll actually get a 5080 for MSRP? I thought we pretty much covered "MSRP is bullsh*t"...and it has diddly to do with Europe or the US. There were many places you could place an online order for 30- and 40-series cards when they had no stock. Then people sat and waited until the order was finally cancelled, and some of those (who just had to have them) then went and paid 200% MSRP. I know a young gamer who paid nearly $2000 for a 3080Ti, because he had to have it (even though I *begged* him not to do it). That was in 2022, fully six months after release of the card at an MSRP of $1199. That's over 165% MSRP. During the first two years after release, those cards spent far more time way above MSRP than at or below. And that's when you could actually get one. We have to consider here that it's already fairly common knowledge Nvidia is controlling who-gets-what, and isn't releasing many of the 50 series at all. Availability is expected to be much worse than the 40 or 30 series, as I understand it. If the past is any indicator (and I'm sure it is) then I won't be surprised at all if I see 5080s for $1500+(maybe more) and 5090s approaching $3000. Now, if someone does manage to get a 5080 at MSRP right at release, hey - good for them. I imagine they're going to be in a very small minority. We still don't know yet whether the performance will be worth the cost, and some fairly reputable sources are already saying "meh". But whether it's eventually in stock or not, and regardless of price, seeing a 50-series card advertised for sale right now, especially at MSRP, effectively means zero. Zilch. Nada, nil, nothing.

Baaaaaaahahaha! LOL Funniest thing I've heard since the suggestion to pay Nvidia with "fake bills".

Aye. Maybe we should start a pool or something, as to who will be among the first here.

I did say that I could hear the sigs being edited...perhaps that wasn't clear enough. But you're right. I wasn't considering the people who you mentioned, in that specific case. I don't consider the "stupid-wealthy", nor any type of "professional" gamers. Not referring to the type Huang referred to with $10,000 computers. My own professional experience over 40+ years indicates those are far and away in the minority; even if you combined them all, I doubt you'd touch a significant percentage of "mainstream" gamers. So, no, I don't consider a subset of the whole that's so tiny as to be insignificant. Can't say I care enough about that lot to include them in any meaningful perspective of typical gamers. Not relevant to my perspective, and not generally the focus of my concerns. At least I was clear about expressing it as my own opinion, (i.e. "In my own estimation") and not trying to represent it as fact. (The post has been edited to ensure this is doubly clear, in case someone else fails to recognize it.)

Just watching a HUB video where Steve says that, even at MSRP *and* assuming a 30% increase in rasterization, the 5090 would be (his word) "disappointing"... Just don't do it. (Even though I know, the moment I say that, there are people who will line up to pay much more than MSRP). Why, you might reasonably ask, would someone do that? And it's a great question. In my own estimation, there is essentially one way this makes any sense: Bragging rights**. I can already see the sigs being edited... ** see below (although I am clearly expressing an opinion here, and have identified it as such).

Yup. Aaaaaaaaaaannnd this is why the "MSRP argument" is complete and utter bullsh*t. Next up: Lack of any data means no one here knows jack (until 1/24 at earliest) about rasterization performance - in spite of trying really hard to claim they do. Let's look one more time at that quote, shall we? We all see it, right there...lol Pretty sure that, so far, NONE of the drivel Nvidia has published shows pure rasterization improvements (because...marketing). Thus making the above quote completely speculative opinion, just as I indicated here: (Note the phrase "not extrapolation of Nvidia marketing drivel"). So: Nvidia MSRP complete fiction? Check. Completely misleading data/complete absence of data showing performance improvement without smoke-and-mirror, "fake frame" BS? Check. Thus making claims of any meaningful price/performance comparison absolute hogwash. This isn't necessarily about the degree of 'price/performance' gain (even though in part it is) - in reality that's up in the air at best, until there is more data. I personally don't think it's going to be all that impressive; I think the info we do have suggests this already, and I think it's just gonna get worse. But, with that said, this is primarily about there being no actual data to support (yet another) uninformed opinion. And to be completely accurate, there is no such data, at least not available to us at present. Meaning the entire assertion as quoted above is speculative. At the time it was stated, there was no actual, reliable data on either real price or performance (ridiculous MSRP and marked-up marketing drivel laid aside). That's what I called out, and it's still true. The performance part we won't know more about until 1/24, but even then, without seeing actual prices, well... I think we're done here PS To be perfectly clear: I have no issue whatsoever with casual discussion, expression of opinions, and speculation. I enjoy it. That's among the things discussion forums are for. The problem is when people state opinions, speculation, etc while trying to make it sound factual or authoritative, without foundation or basis. This is misleading, it is harmful to the readers (who will see this later and may or may not post anything themselves, and may not realize some comments are misleading). It is thus also harmful to the community as a whole. TBH everyone here should understand this - in my view, of course.

Companies can and do get away with a lot. Fabricating ridiculous MSRP is not a crime, nor is 'throttling' supply to artificially increase prices, both of which everyone knows is what Nvidia has done and still does. "The pattern with you (and other forum 'experts') in general" is making statements as fact, when it's just unqualified, bullsh*t speculation and opinion. And when you're shown to be wrong, it's always reduced to the same personal attacks and insults. I cited exactly what you absolutely did say, but here it is again just for the record: Now leave the personal attacks and insults out of it. You have no factual data whatsoever, and therefore what you most assuredly did say is fabrication, presented as fact when you don't have factual knowledge or any proof. That means it's BS speculative opinion, being represented as fact. (If you want to keep denying that, go ahead..but that little blue quote box would appear to prove otherwise.) Sorry if your own inaccurate comment is what's bothering you - I didn't make you say it, nor did anyone else.

I appreciate your effort in replying. I was more hoping for detail about the PSU. What brand/model? As I described previously, the mere wattage rating doesn't really tell us the whole story; there are great 1000W units, and there are 1000W units that wouldn't power a light bulb. Unfortunately this also happens to be one area where unscrupulous builders cut corners: cheap power supplies. Also, please tell me how your 4090 is connected to the PSU. Is it a 12VHPWR connection all the way to the PSU, or are there adapter(s) in the line? Would you be willing to provide a pic of the GPU? (Specifically the power connector as it sits normally with the case/cover in place). I was involved in a thread here a while back where a user had the same kind of issues as you, which I helped ultimately confirm the GPU power connector was at fault. The unfortunate reality is that a lack of experience and/or training has caused tons of issues with those connectors. You'll note I don't say the connector is defective, as it's also been proven that, if it's done properly, the connector is fine. The vast majority of cases don't have problems. Thanks.

IMO: The best place to start here is to list your machine's specs - particularly the PSU. Without specs, it would be difficult to say what's going on, though (to me) it does sound a lot like you have a power/loading issue (possibly heat-related, although I note you do say sometimes it doesn't even start normally). Possibly a bad GPU, but I don't want to get too crazy guessing without more detail. Many people have claimed here that an 850W PSU is enough to run a 4090 - but I have done testing first hand that shows this isn't accurate, and many models actually recommend a 1000W unit. Moreover, I've watched a 4090 shut down two of five high-quality, brand name 850W 80plus Gold units (yes, I am aware the 80plus rating is efficiency not necessarily quality, these were all units from reputable companies, and all known to deliver 850W without failing). 4090's have been proven to draw 700W (just the GPU) in very short duration 'spikes', which is among the reasons the newer PSU specs now require short-duration 'excursion' ratings of 2-3 *times* the rated PSU output. The excursions can trip PSU protection circuits that weren't designed to accommodate that kind of draw, so that even a good quality PSU rated 850W can still fail if hit with a big enough spike. I've done several builds using 4090 GPUs, and provided even more quotes on builds like that. In every case, I *always* specify a 1000W unit at minimum. TBH your builder should've done the same thing IMO. Anyhow, best thing to do is provide as much detail re: specs as possible. Good luck

Law? LMAO...what "law" is that? All I'm "refusing to accept" is your continued BS speculation and opinion. Here's your direct assertion: To which I replied by asking for the data you have that proves your statement. I furthered that by saying having any such data would be remarkable, since reviews of actual hardware aren't going to happen til the 24th. So, where's your data? Seems fairly apparent at this point you have no such data. In other words, it appears you are stating unqualified opinion as if it's fact. This business about MSRP is no more a 'law' than your speculation about performance. There's no 'law' regarding what the MSRP is set at, nor any law that dictates actual prices. Insisting on comparing one fictitious price to another fictitious price is obviously what you *want* to do, because it favors your argument. However, it remains that both are fictitious numbers that, in reality, mean nothing. In reality, MSRP is a fabrication of the manufacturer, for their own reasons (which is another matter). MSRP means absolutely nothing when no product is actually attainable at that price. And I can all-but-assure you it will not be; certainly not initially. (Yes, that's an opinion, but I am hereby expressing it as such). Thus: Insisting on using some absurd, fabricated "MSRP" invalidates the "price" component of "price/performance" value. Furthermore, as it appears at present you have no empirical data to support the "performance" component of your "price/performance" assertion, you completely lack actual, meaningful foundation for either aspect of any such analysis of price/performance. In short, as I said: More uninformed BS speculation, being presented as if it's fact. Of course, I'd love to be wrong here. If you actually have the data I've asked for three times now, by all means...

No, I really don't need to understand all that blather. What I actually do understand is that your assertion was opinion, being represented as fact. I asked for data supporting your own assertion that "There is objectively a gain in price/performance of rasterization, without factoring in the fake frames, for all but the 5090", as reviews won't be available until 1/24. You don't have any such data - "objectively" or otherwise. Which means your assertion was, just as I said, more BS speculation, but being represented as fact. BTW, I'm also not falling for your (ongoing) attempts to play the "MSRP" nonsense. That's simply not reality, and as already covered extensively, it's just another fabrication; a tactic to make things sound somehow better than they are.

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!