This Is Gonna Be Good.

[@techlinked]

hey guys its riley from the video, hey do you know I type these comments with powerful and confident fingers strengthened by me being so well-nourished by delicious ready-to-eat meals? You simply can't get tight, athletic fingers like the ones slamming my keys right now by drinking gamer fuel, or gamer juice, or whatever. Upgrade your fingers and get 50% off your first order of CookUnity meals - go to cookunity.com/tech50 and use my code TECH50 at checkout to try them out for yourself! Thanks to CookUnity for sponsoring this video, and for giving me swole fingers

[@layk]

But what if I want dainty fingers, Riley? What if I want to play the piano I don't own?

[@emmanuelgoldstein3682]

Deleting comments isn't going to get you a clean shirt

[@bomxacalaka2033]

nice try diddy

[@lomelyo]

You got me to read an ad, I'm both upset and impressed.

[@crispy_09]

Your grammar and capitalization show how good your fingers are.

[@sharpenednoodles]

They're gimping VRAM to stop people running AI models on cheaper gaming hardware instead of enterprise chips

[@kennyoffhenny]

That would barely make a difference lmao. People that actually want to run AI models would just get 2 3090’s or 2 enterprise GPU’s.

[@jamesjonnes]

​@@kennyoffhennyMultiple GPUs only work for LLMs if one uses a server grade motherboard.

[@dhillaz]

Mr. Noodles hit the nail on the head Why sell a $700 GPU to a consumer when they can subscribe to a cloud provider, who in turn is willing to pay 5 figures for nearly the same chip?

[@mrblurleighton]

Yup. That's why I went AMD.

[@rafaelbelo]

@@jamesjonnes not true. You can run 6 gpus at full speed using a threadripper and a WRX80 Motherboard. you can run 8 gpus but @ PCI-E x8, gonna lose 5% performance Vs. x16

[@SomeGuy-qd3li]

I am completely "unjazz'd" by the 5090's 600W power consumption. Shit will be a radiator in winter, a sauna in summer and an american medical bill sized pain on my utilities

[@reegyreegz]

My 3080 and 11700k in my bedroom during summer months playing aaa titles at 4k or rendering 8k images with ray tracing turns my room into a sauna. Measured the power consumption and its an average of 650w at peak load 😅

[@silvy7394]

I literally use my 12900K and 3060ti as a space heater in the winter. If im not on it, then its mining crypto.

[@Rkcuddles]

That’s peak power draw. Wait for gaming benchmarks. You not gonna push the card to 600w 24/7

[@reegyreegz]

@@Rkcuddles well i can game easily for 8+ hours non stop, want to render a video at 4k? Thats gona take a looooong time so that peak power draw becomes sustained draw for some of us.

[@xlr555usa]

You can undervolt it and invest in to a small nuclear reactor, I'm on a waiting list

[@adityaharryson]

"You need to get comfortable with not owning vram"

[@ilbro7874]

You wont own any vram and you will be happy

[@dustinyoung3069]

less is more, you see...

[@darrengreenacre1142]

"Don't you guys have vans?"

[@aqualtor9696]

I wish there is a lot of multiple choice on good gpu on the market just like gameing market. Then nvidia will not be this arrogant

[@Alpha13yt]

Applying apple math, where 8GB RAM is actually 16GB, then the RTX 5070 does actually have 24GB of VRAM! Incredible achievements by NVIDIA

[@brytonmassie]

Really should be 5050 8gb, 5060 12gb, 5070 16gb, 5080 24gb, 5090 32gb.

[@marcosmauricio7805]

Man can dream

[@abb0tt]

Your sound logic is unreasonable.

[@legendaryz_ch]

AMD 8600xt 12GB 8700xt 16GB

[@eisenklad]

Nvidia winking to AMD to handle the lower GPU market

[@Juanpopspacks]

Heresy! You can’t have nice things

[@JonnyJKF]

I think I understand what is going on with the naming: Robotaxi is the platform. Cybercab is the specific style of body. Teslabot is the project. Optimus is this style of body. T800 is the name of the endoskeleton chassis and Cyberdyne model 101 is the name of the infiltrator skin. Oh shi...

[@theblitzminecraft]

Legit one of the funniest comments that I have seen in a while

[@_Digitalguy]

this was both informative and funny

[@Ian.Murray]

@@theblitzminecraft I didn't even exhale slightly more forcefully than usual. The humor was incidental at best, get some taste.

[@Adama.1]

@@Ian.Murray Get a girlfriend

[@Eikomaniac]

@@Ian.Murray when's the last time you went outside

[@crazyMLC]

12GB ain't enough if you have programming hobbies that use VRAM. Doesn't matter how much cache there is.

[@StefanRial-i4f]

Why don't you just download more?

[@kirby21-xz4rx]

​@@StefanRial-i4fdownloading ram is a scam 😂

[@Tennyson999]

​@@kirby21-xz4rxthat's the joke

[@TikkaQrow]

heck that, Substance painter and Blender are slideshows now just trying to keep up. VR titles regularly demand 12+ with some UGC games claiming up 24gb vram, get a slideshow when ya run out.

[@gelgel4823]

How much VRAM should the RTX 5070, and 5080 have then? Genuine question.

[@les_crow]

Those vram differences are crazy though. It would make sense if the 12GB version was 5050, not 5070 man.

[@z3ke002]

fitty fitty, can’t wait for the 6060 & the 7070 etc.

[@charginginprogresss]

lmao 12 gb 5050 one can be only that much delusional, thank the gods if it will have 8

[@nesyboi9421]

OK I get that this is next Gen and what not but It just seems like devs are demanding more and more VRAM, optimizing less and less, and we as consumers are just going along with it demanding more VRAM for our cards. And to be honest, I haven't personally found a use for all the extra VRAM on the new card I just bought. I bought a 16GB RX6800 which true is last Gen but I haven't found a single game in which 16GB is required for my 1440p high/ultra I play at. Though to be fair, I don't play the newest triple A games.

[@pcmasterracetechgod5660]

"Hey, my 11gb 1080ti is still doing well, how is your 12gb 2060 doing? Decent? Yeah same, I can't wait to see how much vram we get in the future if your 60 series has the same as my last gen flagship!" 4 generations later... 70 series has the same amount...60 series probably gonna have LESS like with 4060..

[@pcmasterracetechgod5660]

@@z3ke002 I too, can't wait for the 8gb 7070. And the 24gb 7060. Doesn't make sense? Yeah nothing Nvidia does makes sense

[@galgrunfeld9954]

With Nvidia's behavior and pricing, the event now stands for Consumers, Eat Shit

[@MrPruske]

Come and expect scams

[@michaelfishman6082]

Remember when 90 series was mostly for heavy users and came out the latest out of the cards and 80 was the standard? How things change now nvidia wants to make you feel poor if you don't buy 90 series and 80 feels like a compromise because you don't want to spend ridiculous amount of money on a gpu

[@DefianceOrDeath]

Meanwhile the cost of the 80 series is "ridiculous" compared to past standards. So they've successfully mindfucked you into thinking that the 80 series currently inflated price is actually somewhat reasonable.

[@Leffrey]

90=highest end but highest price 80=like 90 but better price to performance 70=midrange budget option 60=cheap 50=crap When in reality: Nvidia=not worth it (To clarify, ‘tis a joke. None of the 40 series cards are actually budget cards)

[@789know]

​@@LeffreyIf not Nvidia? Who would u use? For some workload, Nvidia gpu is still the only options/way more superior options. Unless u r gamers or u do things that other gpu brand also have accelerator/good software support/don't have Nvidia tech for acceleration etc

[@explodingpotato6448]

80 was the high-end chip, realistically more people would be buying 70 and 60 cards, with the 20 and 30 series the 70 was considered high-end as well. The 90 and titan cards were meant for enthusiasts willing to spend ludicrous amounts of money and not really for heavy users, those guys would still probably get an 80 card.

[@explodingpotato6448]

@@Leffrey The 70 wasn't mid-range before the 40 series, it was high-end, the weakest high-end chip sure but mid-range would have been the 60 and 60ti cards, in other words, yes the 500$ 3070 was considered high-end while the 600$ 4070 was considered mid-range. If you want high-end nowadays you have to spend 1,000$ which is ridiculous.

[@dosbox1]

That Tesla robotaxi will NOT cost $30k. They initially gave the same price for the cybertruck and couldn't deliver.

[@aparadoxicalone]

idk, $30k a ride sounds about right if you adjust for inflation when it finally launches in 2050 or so.

[@dovesr0478]

$30K plus a $15K per year subscription to the full self driving service 😉

[@zimbu_]

It will cost 30k as long as Musk has a presidential buddy ready to give 120k/car in subsidies. And they can deliver at least a few dozen each year in four select cities as long as they hire a bunch of people in rural Peru to take over the controls whenever the glorified lane assist loses the plot.

[@slopedarmor]

initial price for cybertruck was from 40 to 70k. now its from 60 (supposedly the future base model) to 100k (produced right now)

[@einstijn138]

Yeah they also put a disclaimer that any statements made are "forward looking" and based on "assumptions" so they won't be liable for not delivering. He could have said it would be $1 if he wanted to. This thing is not going to be sold for $30k.

[@DankGank]

The $2500 price tag on the 5090 is going to be awesome

[@bmorg7244]

And only $4000 cdn! what a bargain!

[@pcmasterracetechgod5660]

I'm predicting $1999, mainly because I want to buy one. But in reality I can see it being well over $2k

[@fohhee]

RTX 5080 is just a rename RTX 5070, Nvidia learn their lesson this time.

[@PraetorBeast]

Did they, though?

[@Anudist]

No they didn't. They're banking on consumers not having learned their lesson from the 40 series launch.

[@PraetorBeast]

Imagine spending 1000+ currency on a GPU with 16 or less GB of VRAM and a 256 bit bus.

[@iamnibee]

yup, its half of a 5090... they will release an 80 Ti at a later date most likely

[@CyberneticArgumentCreator]

@@PraetorBeast Imagine making fun of a Ferrari because it only has a 20 gallon gas tank and asserting that the Suburban's 30 gallon gas tank is required for modern driving.

[@thanksyoutubefortakingmyhandle]

I can't believe someone would hack the internet archive

[@garystinten9339]

It's a pretty shitty thing to do.. I mean it does nothing to anyone and is literally a cache of info.

[@IvoPavlik]

Probably those who want it go down 🤷🏻

[@Negativerapier]

Someone might've wanted to make something disappear

[@garystinten9339]

@@IvoPavlik i guess after an investigation to see if anything has gone missing.

[@BlackDub21]

To change history

[@TheLexikitty]

“Cook Unity” sounds like a response to their whole installation fee nonsense.

[@yourlocalit1701]

16gb of vram for 5070 should be standard. 24 for 5080 48gb for 5090.

[@Razear]

The 600W TDP on the 5090 might be the reason 1200W PSUs might become fashionable again. And that's not necessarily a good thing...

[@AffectionateLocomotive]

🔥🔥🔥🔥fire!..........!!

[@Kuli24000]

I'm laughing because I bought a used pc not knowing what psu it had. I saw and was like "oh 600w, not terrible...." and then I saw the extra 1 before the 6. 1600w?!?!? So I'm set. Me and my gtx 1050. XD. I kid you not.

[@ALmaN11223344]

If you think anything close to a 1200w PSU is necessary for a 600w GPU and a standard gaming build otherwise then you're just into wasting money. A 600w GPU will be content on a quality 800w PSU with the extremely efficient 7800x3d.

[@AffectionateLocomotive]

@@Kuli24000 🔥🔥🔥🔥

[@tekeagle2136]

My roommate thought that my 1200w PSU would be used for mining. No. Power levels: 7950X3D: 120w, 3060ti: 240w, 7900XTX: 450w, and other unknown power levels.

[@TrentonMatthews]

<a href="#" class="seekto" data-time="24">0:24</a> "Where Consumers Come First." it's around <a href="#" class="seekto" data-time="570">9:30</a> PM mountain right now, I'm so glad I went back to re-hear what Riley said...

[@mrcoolguy0591]

What did you think?

[@TheHuskyK9]

Oh I'll make them come first alright 😠😠

[@JakobRush]

I had to hold my tongue when we were filming

[@LyndanTylor]

Why ever buy a GPU when they're always gonna make a new and better one? I'll just wait til they're done and then buy the best one!

[@ehenningsen]

CPUs took around 50 years to get towards the top of the S curve, but is still slowly grinding upward. You may have to wait a while 😂

[@DefianceOrDeath]

lol

[@croozerdog]

always go a generation or 2 behind unless you got cash to waste

[@Hybris51129]

Buy a 5090 now and use it for 5, 7, maybe even 10 years and you'll spend less money than the carrot chasers building to the minimum viable spec and have a better experience in the meantime.

[@DefianceOrDeath]

@@Hybris51129 That's generally what I do, but with the 80 series lol. I used my 1080 from pretty early in their launch up until I got a 3080 when the 4000 series was a few months out. So I got about six years out of it, figuring I'll get six years out of the 3080 as well. If I could afford the up front cost of a 5090 I'd probably go for it.

[@mr.walrus2107]

The leaked price for these things are criminal

[@arowhite]

The 4080 12GB problem was not the amount of VRAM, but the fact that it shared a name with the 4080 16GB while having fewer cores and not simply less vram

[@pcmasterracetechgod5660]

Feel like it's worse this time around, half the vram of 5090 and insanely cut down.. It's literally a 5070, named 5080 just like the 4080 was.. They are so braindead

[@arowhite]

@@pcmasterracetechgod5660 Nah, it could even be 10% of a 5090 I wouldn't mind. Different name, different specs. The only two things that should be taken for granted are that two GPUs with the same name have the same architecture/cores etc, and secondly that bigger number equals better performance (with a small pass for GPUs of different gens)

[@pcmasterracetechgod5660]

@@arowhite If it were 10% of a 5090, then it would be worse than a 4080, probably even 4070. So I"m not sure how you wouldn't mind that considering the 5080 will probably cost $1499 if I had to guess

[@ashermerrill]

<a href="#" class="seekto" data-time="595">9:55</a> Riley makes our boring dystopia just a little more bearable. ♥️

[@conormurphy4328]

He’s my favourite AI generated human entertainment program

[@ShawnHCorey]

Padded walls in a space station? No handholds, no footholds, but as you float around helplessly, you won't get hurt bumping into things.

[@joechavez1761]

<a href="#" class="seekto" data-time="405">6:45</a> I just think Riley is incapable of delivering any commentary that doesn't sound sarcastic, so even the ads sound like parodies, and I love it.

[@565redx]

<a href="#" class="seekto" data-time="307">5:07</a> the subtle gesture killed me.

[@inventorofmachines]

Lol was wondering if people caught that

[@Voltaic_Fire]

What the hell is wrong with Nvidia? 12GB ain't enough! My laptop RTX3080 has 16GB for crying out loud, Nvidia are going backwards!

[@knightfolk]

Greed for greeds sake.

[@meyatetana2973]

Don't buy nvidia then they are apple of GPU except Nvidia has good products LOL but AMD isn't bad unless you are editor or maybe AI stuff? the cost and performance just isn't worth the limiting of ram.

[@JTKGAMING14]

I think you mistook system ram with gpu ram . Rtx 3080 has 12gb

[@gorkskoal9315]

AND if Aaaaaaaaaeye I is how musk gets wood: you have to have 30GB minimum for at least one neuronetwork model

[@theseb1979]

@@meyatetana2973what are you going on about? Your ‘jokes’ and ‘metaphors’ are inane and nonsensical.

[@drexfall]

<a href="#" class="seekto" data-time="52">0:52</a> I'll only be buying the 6090 with 42 gb VRAM

[@Tiwill]

nice

[@smallpeople172]

Only 42? Only losers dip below 16 terabytes

[@paulct91]

Hah! So lil VRAM, my good sir... are you now in the poor house... get away... poor person! 🧐

[@Hiddenus1]

RTX 5th gen "highest" card needing 600W reminds us that 6th gen will require our own ARC Reactor for just GPU alone. And we don't have a cave to build them reactors!

[@danieloberhofer9035]

But you can be sure that Elon "Tony Stark" Musk is going to promise investors that he'll have that ready in two years to power their 6090s.

[@rhoharane]

Why are you calling it that? It's not 5th gen. Not even of RTX.

[@XGD5layer]

If they provide those integrated with the GPU, they could call them ARC GPUs... Wait

[@danieloberhofer9035]

@@XGD5layer 🤣

[@gnobba]

Rtx 4th gen*

[@magikern_1k]

Techno-Rickshaw got me 😂😂

[@rainessex]

I swear every time I watch these, real life feels more and more like a black mirror episode. The other day at my dentist, the dental xray robot told me to stay calm and still while they took a picture of my teeth in an unsettling calm voice. Now, I'm not trying to be robotist but .... idk if I want robots babysitting my kids when its a dude in a back office controlling a robot lol. On the other hand though, sleeping out the dystopia in space seems cool!

[@Ian.Murray]

Reject the cringe modern nonsense. You already intrinsically know that it's not how we're meant to live.

[@JanAnonim-t3z]

Robots should stay robots, trying to make the human-like always just ends up in the uncanny valley. Automation that does not replace extremley tedious or dangerous task nor increse nation's productivity in a significant way is just a tech-bro's gimmicky way of separating you from your money. And possibly personal information, and a job and family...

[@MrPruske]

The machine read out text to you. If it was a screen, you'd have to read it. Not everyone can do that. 🤦

[@ich_16]

"the best way to hide corporate incompetence is comparison " xD

[@JakobRush]

❤

[@WilliamHaisch]

How long until some company will take your 5000 series GPU and solder it to a custom board with a real amount of RAM? 😂

[@zyeborm]

The issue is the BIOS in the card. They can make 24gb 3080s because there was a bios released for a specific mining card. We can solder the ram on but it won't use it without the BIOS and driver support. At least that's where it was at least time I looked (That's assuming they are using low capacity chips not just having fewer high capacity ones and you can get high capacity ones to solder on, like replacing the 1gb with 2gb chips on the 30 series)

[@ggwp638BC]

Honestly, the best thing people can do is skip the 50series launch, wait a bit and buy the 60 series which will have the same performance at half the TDP.

[@Sunshrine2]

The “VAST” space station looks like a floating fire hazzard. I am pretty sure the station, if launched, will look completely different and much more sterile.

[@ManleyReviews]

Can't wait for the Nvidia VRAM monthly subscription

[@DrySushi]

Fun Fact: Riley 1.0 started out as LTT's first push into artificial intelligence cab drivers. Riley didn't like driving but showed aptitude for social media engagement. Riley is now cloud based in its current version.

[@belgianstar]

Every year the tech news reminds me of Canadian thanksgiving

[@IanHobday]

Even the 3060 ended up with 12gigs in the end. Crazy they're still putting that in a 5070.

[@fleurdewin7958]

But the laptop variant 3060 ended up with pathetic 6GB of VRAM despite having abit more shaders.

[@yfnratchet]

"We're not focusing on VRAM because our cache provides the necessary performance" I'm sorry, does the 5070 have 4 GIGABYTES of L3 cache?! I didn't think so!

[@_analog_photo]

Spiderman: with the great power comes great responsibility. Intel: let me decrease power...

[@Stratos1988]

Intel is pulling past AMD move and tries to be king of low to mid-end

[@JoshuaWayt]

You mean Nvidia is more into cache than it is giving consumers what they want? No way.

[@fleurdewin7958]

Nvidia wants to convince buyers that their cache can justify more cash.

[@xpforevergaming8609]

It's how Apple thinks 8 GB on a Mac is analogous to 16 GB on a (Windows) PC. Doesn't make any sense, but there will be plenty of people who believe that marketing crap.

[@brytonmassie]

3060, 4070, 4080 (Unlaunched), and 5070 all have something in common... 12GB of VRAM.

[@EmanuelLopesS2]

<a href="#" class="seekto" data-time="307">5:07</a> The hand movement for that sentence is hilarious.

[@TomaszBrodecki]

Someone forgot to delete the buzzing audio track

[@CautiosulyOptimistic1440]

Apparently by some rumors from moore's law is dead, the 5090 could be priced at 2000-2500 USD.

[@ArgueWithTheMajority]

My take on the Intel CPU might be an unpopular, boomer take, but I like it a lot. I was thinking about building a new computer and 13 & 14th gen were out of the question for me, because of how hot they run, even with 420mm closed loops. Most people can't even get their full performance, because they constantly throttle without very solid cooling + undervolting. About the same performance with much more reasonable temps that run stable and make longevity more likely, is really appealing to me. Also, as was kinda mentioned in the video, If they are now able to get similar performance at somewhere around 50% of the wattage, they should be able to later release more power hungry, but stable CPUs with solid performance gains. I actually think this is a step in the right direction, because they realized what the biggest problem of their recent high-end CPUs was. Of course, this was clumsy marketing, as they should know their "more power, more better" NPC consumers.

[@IvoPavlik]

Agreed. Efficiency is not as sexy as more power, but it's much needed at this point. Power consumption of modern high-end CPUs is just ridiculous.

[@musaran2]

Except it should have been that way from the start.

[@ArgueWithTheMajority]

@@musaran2 Yes, again, clumsy marketing. They should have also announced a 70-80% watt option with non-embarrassing performance gains to appease the enthusiasts, if they have such a thing planned.

[@ArgueWithTheMajority]

@@IvoPavlik Yup. Also, let's be honest, people only react like this, because they are so very used to the new generation = performance boost cycle, which is fine. However, in an actual use case, if you have some high-end i9 12-14th gen CPU and use it just for gaming, you are basically crying "oh no, in some games I'm now getting 340 fps instead of 350, but without cooking my components". At that level, the CPU doesn't have a huge effect on gaming performance, because you are already well within overkill. GPU is obviously much more important here. And in many professional use cases, in which the CPU can actually be the most important factor, I'm sure most businesses would be very happy to do their work, no problems, more reliably, while paying half the energy bill (CPU-wise). So, except for E-Penis, this is simply not a bad idea. (Even there, with 50% energy consumption, overclocking MIGHT be back on the table.) BUT pricing is the thing that will actually decide whether this announcement is trash or not.

[@ggwp638BC]

Yep, I have definitely switched from the "more power" to "better efficiency and lower cost" camp. Cost of electricity, heat, longevity, there is so much to weight in now. Also, this is a bit controversial, but I think we are reaching a point where it's fair to argue "how much processing power does average people, especially gamers, really need?". It's been almost a decade that we are seeing only marginal perceived improvements to graphics, despite the massive jumps in power. Do we seriously need an extra 200W per person for an extra 10% particles in the latest ubiflop? Do we really need a more advanced GPU consuming twice as much power to mine shitcoins and NFTs 8% faster? Not saying we can't have those things, but there is only so much energy we can produce and so much CO2 the atmosphere can take.

[@Guru_1092]

That space station is going to smell SO BAD.

[@pacmon5285]

You could have kept the alliteration with Robo-rickshaw

[@NightFoxZero]

I’d take a 50% power efficiency boost over a 10% performance boost at significantly higher power draw any day of the week. For one, that’s a lot less thermal load for coolers to deal with and lightens the electricity bill contribution a little bit

[@ram64man]

The 5080 should be called a 5070ti demand nvidia change it’s name till a 386 bus and 24gb ram is forced to be released

[@ragepoweredgamer]

Another TechLinked with the BACKGROUND HUM. It goes away at <a href="#" class="seekto" data-time="371">6:11</a> when the sponsor spot starts, but then it COMES BACK at <a href="#" class="seekto" data-time="440">7:20</a> when it goes back to the episode. The silence at <a href="#" class="seekto" data-time="445">7:25</a> makes it super noticeable.

[@narwin2477]

12 gigs of Vram being complained about is something I am deeply saddened by, my 2080 super is a bit outdated :(

[@brucewilliams6292]

"More, different ones." That was good. It shows great preparation. Exactly what I expect from you guys. Thanks for the giggle.

[@friedgpu]

<a href="#" class="seekto" data-time="482">8:02</a> THE BEST WAY TO HIDE CORPORATE INCOMPETENCE, IS COMPARISON HOLY SHIT I DIED RIGHT THERE

[@GRIMHOOD99]

I haven't touched 40xx series with ten-foot pole and I intend to add extra 10 feet to it in regards to 50xx series.

[@kaidiemulan6061]

<a href="#" class="seekto" data-time="167">2:47</a> greater power comes with greater responsibility 🕷️

[@wileysneak]

is there a HVAC or something in the background making a low freq hum? could maybe do with chucking the audio through a pass filter next time

[@mateoconk]

<a href="#" class="seekto" data-time="136">2:16</a> “I meant no offense, young hobbit… can’t be too careful”

[@00Cubic]

<a href="#" class="seekto" data-time="197">3:17</a> ryzen SEVEN* 9800X3D

[@Conqueror25]

Elon Musk in 2035: Your honour, my claims about robotaxi and cyberbuses were just corporate puffery and any serious investor wouldn't have been misled by them. 😶‍🌫️

[@Name-gi8dr]

<a href="#" class="seekto" data-time="227">3:47</a> yes please. Techno Rickshaw is the perfect name.

[@SaadKidwai]

The full house family had me dying. This man is a gem.

[@johncandelario294]

considering my electric bill is at 250 now in PR im liking the half off stuff

[@SubatomicPlanets]

The 3060 had 12gb... How is that the same as the 5070...

[@ShaferHart]

The only good thing about the 5090 is that it will make the 4090 cheaper.

[@tschorsch]

If they don't keep supply limited.

[@no-one3795]

We got Delamain before GTA VI 😭

[@constantinosschinas4503]

*50<a href="#" class="seekto" data-time="3012">50:12</a>GB, 5060: 16GB, 5070: 24GB, 5080: 32GB, 5090: 48GB.* Enough with the bshit ram and prices.

[@PlanetXtreme]

How do they plan on cleaning the space hotel?

[@thedudely1]

boosting GPU caches instead of VRAM is a great new way to segment gaming cards from enterprise/Ai hardware since Ai workloads pretty much just needs as much vram as you can give it. It's a cheeky way to make the gaming cards faster without letting people get away with GeForce cards for ai. shady.

[@clashwithkeen]

I mean, was it Amazon or another company that used "AI" for their no-checkout stores and it turned out that it was just kept track of by people in India (?) through cameras. If it was cheaper to use task rabbits for that then it would most assuredly be cheaper to hire them to control those robots remotely vs the cost of actual AI. Please note, I am not making any kind of ethical argument, only noting that it's in the realm of possibility. With latency involved I wouldn't let that thing near a child, or even one with AI for that matter.

[@zyeborm]

They would presumably be running at least all the "interactive" elements locally. With cloud elements only for "higher" functions. Average human reaction time is 200ms or more, 6ms to a server farm isn't much. Tesla's are already driving around with pretty modest computing power. You can run a small llm on your phone. A 5090 would run a decent sized one with decent context in real time. However what will likey happen is in about 2 years the plethora of startups making "ai" focused hardware will hit the mainstream and you'll have huge vram chips for ~$10k that'll run gpt4 level models locally and in real time. Cheap on a $100k robot.

[@liamleech4748]

Elons saying autopilot being two years away is like Dutch saying "just one more job" lmao

[@colbster12302]

timed release right after wan show: 100% accuracy

[@Haaris.Qureshi]

I like that Jakob co-wrote this video twice!

[@matthew.stevick]

<a href="#" class="seekto" data-time="304">5:04</a> if you can’t trust a 500 lbs awkward robot that elon is selling to babysit your child, who can you trust?

[@vincepalaa]

My current power supply is 650W. RIP

[@davincij15]

I cannot believe we are here saying 12 gigs is not enough! Next year 16 will not cut it!

[@matthewjbauer1990]

16 is almost not enough if you want to go ultra settings 4k 120Hz RTX HDR.

[@charginginprogresss]

@@matthewjbauer1990 which is not needed at all unless you want to game on a 80" TV from 10 meters away

[@chrismay2298]

@@charginginprogresss Maybe he does. Why do you weirdos think that YOU get to decide what others need or want? Sick to death of you muppets.

[@matthewjbauer1990]

@@charginginprogresss Totally agree. That's why I still rock a 5700xt. Serves my needs for medium high settings gaming at 1080p without RTX or HDR on a regular 21" LG business monitor.

[@andrekz9138]

Happy early Canadian Thanksgiving!

[@littlewillie65]

Call it "JOHNNY CAB"

[@PropaneWP]

It better not hum the Norwegian national anthem, though. Your standards don't meet ours anymore, Elon.

[@phantomabid]

Even my 3060 has 12 GB of VRAM!

[@D00m3dHitm4n]

The Tesla Bots remind me of the evil robots in the movie I, Robot with Will Smith

[@underdonetoast3335]

this was the first comment congrats

[@vendetta9087]

So Tesla bot gonna slap you when you mention his wife

[@vendetta9087]

So Tesla bot gonna slap you when you mention his wife

[@joefrog1996]

My 4080 uses over 12GBs just for 3440×1440 when playing Star Wars Outlaws.

[@archerstown]

I just caught a live WAN show for the first time in literal years and then a Tech News launch at the exact end? Too kind, too kind

[@brunoge26]

You're welcome

[@jefez75]

they almost always launch exactly at the end of WAN, if not a few moments later

[@Sinner487]

This is my first time seeing this channel and I just love it. You guys keep it real. Keep it up.

[@ybing]

full self driving only works when the whole road is all self driving...

[@DeanoDrives]

Well said, To be fair though. I wouldn't mind sharing the road with overly-cautious AI granny drivers. At least they aren't going to cut you up.

[@XGD5layer]

AI lanes/roads when

[@wild.legend-music]

You guys need to watch some videos on Full Self Driving. There are channels like AI DRVR. Watch the latest video and your jaw will drop. No grannies, no overly cautious AI. Drives better than humans. P.s. Not a Tesla fanboy, but what they have achieved with FSD is incredible.

[@chrismay2298]

@@wild.legend-music "Better than humans"... LOL NO. Better than some humans. I can also do a myriad of other things while driving way better. You guys are being fooled into admitting inferiority to something that can barely do one simple job.

[@JanAnonim-t3z]

Have you completley forgotten what Driverless taxis did in San Francisco? Even with autopilot potentially being better there is still a huge possibility for self driving cars interlocking each other just like NPCs do in doorways/stairs in unfinished games. Musk already had taxis and a hyperloop. They all ended being a mostly useless and expensive science fair projects, already begin taken apart for scrap. But yeah, I don't see an AI taxi going aywhere in big cities without others letting them through. And I guarantee while human drivers might let a normal taxi, food truck or a bus go they will intentionally block theese things.

[@fran560]

in the Last days my speaker keep makinhg buzzing noise from any LTT video even this! all other its fine

[@lolpl0000]

wait, they invented actual ocd micro chips thingies? that could change my life, as someone suffering from the illness for decades.

[@Finalizor]

My OCD would not allow something to be put in my brain

[@lolpl0000]

@@Finalizor Luckily, or not, ocd differs from one person to another.

[@nkay013]

This episode was the best ever. I usually cringe at the jokes but today it was a banger. No ball dropped. Absolutely funny. 😂😂

[@lomelyo]

starting at 11 dollars a meal for frozen food??? What the heck how much is take out in the US lol.

[@daniegamin]

I work at a fast casual Thai restaurant that does a lot of take out and it's actually about $9-$11 after taxes depending on the meal. and that's with out a drink.

[@o0Donuts0o]

Has America not figured out how to add tax to the displayed sale price?

[@XA--pb9ni]

​@@o0Donuts0othey still have not...

[@garystinten9339]

I think most countries have. I also think the US leverage their food tax so it doesn't get added to the price.

[@TheDragonfriday]

Most of the time 15$ food takes out or 20$

[@TheDarkSoul97]

Yay! Saturday’s techlinked is out!

[@Paelmoon]

Hey you left "for Nvidia shareholders" off the end of the video title. THE 4080 SCAM 2.0 - Nvidia learned their lesson. Don't release two versions. Release the gimped one first then do a "refresh" and release the real one later.

[@sparky6757]

When I was building a PC for the first time ~7 months ago, I was having a hard time picking between AMD and Nvidia. At the ~400-600 USD price point I was shopping in, AMD had a moderate lead in raster performance, while Nvidia had a moderate lead in raytracing and power efficiency. Decided on AMD in the end, and I'm happy with my 7700xt. But from the rumors of the next gen AMD and Nvidia products and prices? That's a yikes from me, Nvidia wouldn't even be a contender

[@codycast]

<a href="#" class="seekto" data-time="490">8:10</a> lol who uses “threads” ?

[@chrismay2298]

Another small thinker who thinks that gaming is why CPUs exist... I use all 56 threads 24/7 and will be getting more.

[@KiegKillsReality]

@@chrismay2298He is talking about the app genius lmao

[@azzor4134]

@@chrismay2298 ???

[@RushGamingDK]

The background is flickering. Bad bitrate encoding or wrong camera settings?

[@TheOneAndOnlyTeknocat]

I want robotaxis like the ones in the classic Total Recall movie (the one with Arnold).

[@GizmoMaltese]

Johnny Cab! Now if he brought out a bunch of Johnny cabs I would have been really excited.

[@rmadzharov]

"This is nice!" Riley is a national treasure!

[@Onyxle]

The Tesla robotaxi is the next step in the Ai uprising, they can now transport themselves.

[@mdhasiburrahman8806]

This is called what is a greedy corporate sh**. Why on Earth you would buy a car with 30k $ that has just only two 2 seats, I think with that price you can buy even a suv ,(if not new at least 2nd hand). This kind of greedy corporation bull sh** should stop. have to stop.

[@overtoke]

mom mom, i need a less powerful computer for the holidays

[@brendanhoffmann8402]

"Look at him yearning Desirous His craving, his hunger Constant needs, constant wants Nothing will quell Nothing will slake Insatiable vacuous man Hollow and see through His emptiness has made him frail That dying thing still believes Others drive the nails Incendiary god complex Such gift to flammable man This is what it made of him This is what remains Behold, how he devours Unappeasable and cursed That arid soul, famished and desiccated I am that thirst His gluttony can not be stilled This state can never be reversed Hollow and empty Endlessly"

[@smellthel]

Imagine the 6090 jokes