[Music] So, in the final run-up to 2025, there have been no shortages of rumors emerging

regarding Nvidia, AMD, and even Nintendo's upcoming products. I want to mostly talk about Nvidia in this video,

but I would be remiss if I didn't cover those Nintendo rumors quickly; namely, the rumors that have emerged of the Switch 2

having 4K output and being able to do so because, in docked mode, it can boost

much, much higher than a lot of people were expecting because of the power supply being more powerful than the previous generation,

and also a dock that allows for better cooling. Both of those things, yes, even the 4K output thing, were exclusively

leaked by this channel months ago, some of those things near the beginning of the year. And so I would be remiss if I

didn't point that out, not just to brag that it seems like now we have complete confirmation of everything this channel

has been saying about the Switch 2 at this point, but also, there's a lot of people interested in the Switch 2,

and there's a series of a few videos that you could watch in order if you want to know everything else about the

hardware, as there's more details there that a lot of people have seemingly missed or forgotten. But anyways, besides

that, the main set of rumors I actually want to cover today is to talk about the RTX 5000 series, namely, for the most part,

although I do have a leak at the end. I want to logically lay out—and when I say logically, that means not just relying on

sources—but using publicly available information and then analyzing it, to go into where I believe Nvidia will price their upcoming graphics cards, which is

not what I did a couple of months ago when I put out a leak detailing what my sources were telling me the prices of

The RTX 5000 series would likely be... Now look, I still think you should take that leak very seriously. After all, those are

the same sources that were dead on with RTX 4000 Super pricing, despite basically everyone at the time saying things like,

there's no way Nvidia would launch a 4080 Super for $200 less than the 4080 they

did. So yeah, you should probably take that leak seriously, but I think that there are very real, specific, and

explainable reasons that Nvidia would raise prices on some of their upcoming cards compared to their previous predecessors from the 4000 series, but

not on some of the others, and I want to just kind of prove it with numbers today. All right, so first let's talk about the

RTX 5060. A card that's main debate online seems to be over whether it will have 8GB or 12GB of GDDR7, and

honestly, look, this is mostly just me doing some critical thinking here and giving an AI-informed, but still overall

just my opinion, but I think there's basically no doubt whatsoever that Nvidia is going to give 8GB to the

laptop—underline that word, laptop—5060, but then that there is also no reason for Nvidia to launch their desktop 5060

anytime soon. You see, when I speak to people behind the scenes, and I—I even spoke with a couple of people for a gut

check before recording this video today—nobody is mentioning the 5060 desktop card is on the horizon. Like, all of

Nvidia's partners are mentioning the 5090, the 5080; most are mentioning either a 5070 or 5070 Ti. No one is mentioning a

5060 coming to desktop anytime soon. At the same time, they also tell me, like, why would they? I mean, tons of RTX 4060s and

even, yes, 3060s are still flooding the market right now, and they're doing just fine. So why would Nvidia launch some

replacements for them until they've gotten rid of the previous generation of that tier, or at least alleviated that supply

issue a bit first. Furthermore, I've also noticed that reporting by outlets that seem to be sure the 5060 will only have

8 GB of RAM on desktop seem to be based on omission, not suggestion or like proof in writing, but just the

omission of the 5060 showing up in some lists with more than 8 gigabytes of RAM.

So think about what I'm saying here: if the 5060 isn't launching soon along with the rest of the RTX 5000 series above it,

and it doesn't seem like it is, and if it isn't listed yet as having more than 8

gigabytes, but it is listed as having 8 gigabytes for a laptop, I don't see why

that doesn't suggest that it's not being listed for desktop right now because it's not coming out soon, and maybe it

will come out later when the 12-gigabyte modules of GDDR7 are fully available. Which means they could just

wait until they can launch 12 GB models over a 128-bit bus. Now, if Nvidia does that—you know, launches the 5060

later next year with 12 GB—does that mean that I think they're going to charge $299 for it? No, I don't think it's

going to get the same MSRP as the 4060. I'd be very surprised if it did. At the

same time, I would be surprised if they did, you know, something insane like $500. My guess, and it is a guess, is that it

will be $349 to $399. But considering the fact that it will come with GDDR7,

which should bring massively more bandwidth, I think this thing could end up a solid 20 to 40% faster than the

4060, and so if Nvidia were to charge 15 to 30% more but give you 20 to 40% more

performance and 50% more VRAM than the prior generation, I think what they'll do is try to

argue it's a 20% increase in price-performance with more VRAM, and that you should just accept the higher price than

the previous generation 60s. Uh oh, and one more thing about this. Yeah, look, I think—I'm not 100% sure—but I think the

5060 desktop will have a 12-gigabyte variant. Absolutely. It would not surprise me if there was also an 8-gigabyte

variant. Let's be honest here, Nvidia loves having 60-series cards that are basically different cards with the same

name. So don't take this as me saying there will never be an 8-gigabyte model. I think there will be a 12 GB model, but

I would also be surprised if they didn't try to trick some people into getting an 8 GB version that was cheaper on Amazon,

but you know, on that 20% note—not like more performance, but a higher price evening out to, in Nvidia's eyes, 20%

better value. Again, Nvidia's words, not mine. I want to take that 20% better value argument and run with it for what

I expect out of 5090, 5080, and 5070 pricing as well, and I also want to link

details about 5080 16 GB laptop specs, but first, an ad from a sponsor. This piece of content is brought to you by

the KTC G27P6S 240Hz OLED gaming monitor. This monitor's specs include a

27-inch OLED 1440p screen that goes up to 240 frames per second refresh rate and also has adaptive sync support and 0.3

millisecond response times. But then, you know, it also has built-in speakers, multiple HDMI ports, a DisplayPort, and a Type-C port.

and USBs for powering devices from the monitor, as well as HDR support. And then, importantly for me, it

has a 136% sRGB color gamut with factory calibration. That was essential to me deciding to use this monitor past

testing it, because I did test it, but now I'm still using it as my daily driver. You know, I used to have a 4K 120Hz

as the main daily driver, but it was an IPS screen and it was really color-accurate. So it's still useful as a side monitor

in that regard. But I have to say that, despite dropping the resolution, man, OLED really does punch above its weight in

sharpness with that per-pixel dimming. It just looks better at the same resolution. And so it's not 4K, but it's

good enough. And of course, the color is just superior to anything that's not OLED. And because it's color-accurate,

I've chosen to use it. And of course, the color itself—not just as accurate, it just looks amazing with OLED. So

seriously, you can often find this thing for around $500, and at that price, I just think it's an absolute steal. So if you

want to support Moore's Law, dead, follow the links in the description below to get this awesome monitor, the KTC G27P6S,

today. All right, so I talked about the RTX 5060 in the first part of this video.

Now let's move on to the RTX 5090, a card that I've seen people online claim

should cost the same as the 4090 or close, which seems really silly to me. I mean, I leaked what I was told in videos

telling the reps it should cost months ago, and I thought those prices were like—I mean, come on—like, is that really hard

to believe Nvidia is going to charge more money for a bigger card with more RAM when they don't have AMD competing?

In the top end, but to the people who doubt they will, I want to actually break down the numbers a little bit here so

let's start with the RAM. I would estimate the RAM costs alone; the 90s memory is $200, around more expensive

than the 4090s. I mean, we're talking about GDDR7; it's new. Other people might want

it. It's not like GDDR6X, or basically only Nvidia is using it. No, there will

be demand for GDDR7 from other people, possibly, and it will cost more, just like GDDR6 did when Nvidia launched Turing.

And my napkin estimate would be it's 50% more expensive per gigabyte, or around there. And then, of course, they also

have another 8 GB of RAM. So yeah, $200 more for the RAM alone. So that already gets us to at least $1800 for the price

of the 5090, if Nvidia wants to keep the same margins, and why wouldn't Jensen? But then also it has a 744-millimeter squared die, or

at least that's what's rumored. I can't independently verify it, but let's run with that. Let's say it is that much

larger of a die. And if I account for the fact that TSMC has lowered prices on the 5-nanometer family of nodes a little

bit compared to the previous year, even I would roughly estimate, using a calculator online, that the die for the

5090 is around $30 more expensive than the 4090. And so let's add that together: $200 more of RAM,

another $30 more for the silicon, and then let's add that 60% margin that

I know Jensen, at least, will want. For those who don't know, Jensen typically drives for 60% margins on their graphics

cards most of the time these days, if not more. So you take that, multiply by 1.6.

and put that on top of the 4090's price and you get to almost two grand, so that's not even accounting for the fact

that the board's costs will probably be more, that shipping costs are up this year versus the previous generations'

launch, according to people I talk to in logistics, and also that there's just more expensive components possibly in

the cooler and for the connector ports of the card. I mean, the minimum I'm seeing here is likely to be $2,000, and even

a price like $1,900, which is the lowest I could believe, really. That would be with Jensen

trying to get a marketing win. So that's what I'm just saying to you all: I think $2,000, around there, is realistically

what we should expect, and they could go to $2,500. But alright, now let's go into

why Nvidia might think $2,000 is fair. Well, if the 90 is 50% faster than the

4090—I'm just using round numbers to make a point—but let's say it's just, you

know, 1.5 times the performance of the 4090. We'll take $2,000 divided by $1,600,

Nvidia RTX 5080 16GB Specs Leak | 5090 & 5060 Price Analysis | Nintendo Switch 2 Update

The Moore's Law Is Dead Team creates videos, records podcasts, and writes articles containing in-depth commentary and analysis of what's going on in Technology, Gaming, and Computer Hardware.  We aim to make our opinions not just informed, but also full of unfiltered insights that stay unencumbered by overbearing sponsorships.

Broken Silicon Podcast:
https://soundcloud.com/mooreslawisdead
https://www.stitcher.com/s?fid=421209&refid=stpr

Address (Use USPS if a US Shipper):
Moore's Law Is Dead,
PO Box 60632,
Nashville, TN 37206

Amazon Affiliate Link: https://amzn.to/2KpXaTm
Patreon: https://www.patreon.com/MooresLawIsDead/overview
Cash App Donations: $MooresLawisDead
Bitcoin: bc1qz6aqrr4syhl8nh4dwr4x24q6a67r3jts0n86x7

All Music by SAHARA: 
https://soundcloud.com/saharasuck

Nvidia RTX 5080 16GB Specs Leak | 5090 & 5060 Price Analysis | Nintendo Switch 2 Update

Nvidia RTX 5080 16GB Specs Leak | 5090 & 5060 Price Analysis | Nintendo Switch 2 Update

Scripts