Uh, amdbros?
Some larger entity has to keep intel afloat after they failed to maintain competitive product lines in pc desktop and server spaceSo now that entity is nvidia with us government, softbank providing fundingNvidia will obviously have to keep amd around for anti-competitive reasons.
if i were leather jacket man i would be banging young hotties left and right, no need to divorce
https://newsroom.intel.com/artificial-intelligence/intel-and-nvidia-to-jointly-develop-ai-infrastructure-and-personal-computing-productsFINALLY NVIDIA X86 CPUSIMAGINE
>>106623072Kek, AMD is killing Intel so Nvidia has to get onboard to keep Intel afloat.In a few years AMD is gonna get really competitive on the compute side, and Nvidia knows it.
>>106623072intel d-gpu division is definitely dead
250 billion investment in the UK alongisde 136 billion from salesforcethis hasnt happened in history before right? why are americans investing so much in overseas tech? they also announced 12 new nuclear reactors in the north east of the UK
>>106623357This is pretty much the only real take away from this. Jenson betrayed his cousin (Lisa (AMD)) in order to likely cripple Inte's dGPU development and to keep them out of AI by focusing on NVIDIA support instead.
>>106623470The simple explanation is that (a) they get cheap access to x86, and (b) it curries favor with Trump for Nvidia to "invest" in Intel. 5 billion is dirt cheap for what they get in return. And it will take years to mature so if it really falls flat then they can quietly bail out later. Intel dGPU is definitely dead, but this reeks of desperation from Intel as well. Their licensing is one of the few competitive advantages they have left, and they just gave their largest competitor access for bargain bin prices. This also means that ARM (or at least Nvidia's ARM endeavor) is shitting the bed to some extent.
>>106623072just look granny karen n100 dc boards seem 19v could they run 14.4V
>>106623072>$4 trillion company that cant make a connector teams up with failing cpu company that cant make cpus
>>106623560Why would NVIDIA want x86? It's literally dogshit. x86_64 is owned by AMD and is actually useful and needed to even for the scale NVIDIA has now with their GPUs and AI SOCs systems they sell on ARM
>>106623072so Intel going to explode like Nvidia did? Guess investing the 10b in Israel was worth it...
>>106623587it's unironically over
>>106623315That's part of the deal. Trump and Jensen will be splitroasting hebe girls will gaurenteed pardons and libtards will seethe
>>106623072>the jew company invests in the other jew company to jew the end user.still sticking with amd.
>>106623072>MFW Nvidia x86-64 CPUs are going to be a thing in my lifetimeThe weirdest timeline. Jensen got what he wanted, just 20 years late.Also this means total arm death.
>>106623072>get banned from china>beg for pity from isrealdamn
>>106624193>slap nvidia gpu onto intel soc>slip past all sancitons and sell soc to china will this even work
>he sold?>pump eet.
>>106623693Nvidia wanted full SoC for long whileThey wanted ARM just so they could make full Nvidia+Nvidia PCs(both desktops, workstations, and mini PCs)
Cool, now MS can spy on their users more efficiently using AI.
>>106623587To make Trump happy so that export restrictions get eased. Trumps crippled brain can't remember anything after 1994, so he's hyper focused on Intel and thinks the Pentium is a great new product. This also explains why he somehow didn't know that Nvidia existed at all prior to June of this year despite being the most successful computing hardware company of all time.
>>106623072so the ZOG infiltration into nvidia begins...>stole intel>run it into the ground>steal nvidia>???
>>106623072>>106623587AMD W
>>106624344>>106623938>>106623336Read the article. bros. NVIDIA is giving Intel rights to put RTX GPUs onto their chips instead of Xe iGPUs. It's Intel SOCs, not NVIDIA. NVIDIA is basically nicely telling Intel to stop pursuing the GPU market lmao. I wouldn't be surprise if an announcement to cancel Xe dGPU will come in 2026.
>>106625114>NVIDIA is basically nicely telling Intel to stop pursuing the GPU market lmao.Arc Sistas not like this!!
>>106623072InTel Aviv wins again.
>>106623072Nvidi-Ack
>>106625086AMD will indeed conquer within time.
>>106625466After nearly 4 years AMD still can't compete with Cuda for AI. There's a reason their market share is going to 0
>>106623106I honestly don't think this is a terrible idea. I do think it would have been a better idea is to put Intel's fabrication division under the temporary management of the DoE with someone like LANL or LLNL at the helm to help them get their shit together. It's really, really fucking gay that the Clinton administration didn't have the foresight to realize that maybe US developed licensing EUV technology to exclusively foreign companies, and then letting ASML slide on domestic production plants because they and a couple bad years. ASML's entire technology stack was developed by the DoE.
https://www.tomshardware.com/pc-components/cpus/nvidia-and-intel-announce-jointly-developed-intel-x86-rtx-socs-for-pcs-with-nvidia-graphics-also-custom-nvidia-data-center-x86-processors-nvidia-buys-usd5-billion-in-intel-stock-in-seismic-deal>In either case, Nvidia has been mulling using Intel Foundry since 2022, has fabbed test chips there, and participates in the U.S. Defense Dept.'s RAMP-C project with Intel. The DoD project involves Nvidia already making chips on Intel's 18A process node, so it wouldn't be a total surprise.NVIDIA has been working with Intel in secret since 2022 at least.>Intel will fabricate custom x86 data center CPUs for Nvidia, which Nvidia will then sell as its own products to enterprise and data center customers. However, the entirety and extent of the modification are currently unknown. We do know that Nvidia will employ its NVLink interface, which tells us the chips could leverage Nvidia’s new NVLink Fusion tech that enables custom CPUs and accelerators to enable faster, more efficient communication with Nvidia’s GPUs than found with the PCIe interface.Intel will now also fab x86 chips for NVIDIA to brand and sell as their own.Timing is questionable, but this seems to be long in the making.
>>106623072>custom x86and in what world is this allowed
>>106625695Nvidia isn't directly fabbing them. They are paying Intel to fab chips with rtx embedded on them. Intel then sells the chip to nvidia. nvidia rebrands it and sells it as its own custom SOC.
>>106625114Intel's strength was always in iGPU (not pure performance, but efficiency and capabilities). And that won't change even with Nvidia. Nvidia will likely package xx60 and up with Intel CPUs to counter Strix Halo successors, and Intel will still take care of the lower lineups. This is more of an expanded Kaby Lake G than Nvidia telling Intel to fuck off with GPUs.
>>106625492>Cuda for gAIyWho aside from brain fried gooners with niche fetishes who are fine with uncanny porn and cringe captions and absolute hatred for talented artists in Korea and the Philippines who accept commissions for pennies does the AI Cuda cores actually appeal to?
How long until SMIC mogs the entire western world? I'm betting 2030 at the latest
>>106625772>does the AI Cuda cores actually appeal to?Literally every major AI company that trains models. I get you're in denial about AI being a thing but it is and it's in the tens upon hundreds of billions in revenue. Nvidia has squarely became the market leader of it.
>>106623072nvidia scared about monopoly in AI and getting the US gov up their asshole in an anti-trust suit. china is already greasing up their assholes with gutter oil for the monopoly on gpus.
>>106625803>Nvidia has squarely became the market leader of an over valued market bubble doomed to popOkay, enjoy being a bag holder in the future anon.
>>106625779safe bet because it is a military controlled project so much of it is secret and no one will know until the press release. it will be "SMIC is low quality" until "SMIC is 1nm"
>>106625833le epic troll bro xD!!!!!
>>106625720this is the same situation as with arm nvidia shouldnt have gotten x86 licence in any way
>>106623361they left the EUnow they are more attractive to exploitative US capital weird
>>106623072BUY Intel CPU & Nvidia GPU, CHUD
Jensen will get an x86 license. It is inevitable.
>>106625862Cope, Delusional, Please seek mental help anon. Deepseek alone almost popped the bubble that can run natively on a decent i5. Now imagine if the Chinks release Deepseek 2.0.I was laughing my ass of when Bloomberg starting sperging out about Chicom software and desperately tried to convince Coomsumers to keep investing in Open AI.
>>106625920>Jensen will get an x86 license. It is inevitable.Short term gain for long term profit loss. Intel is being retarded, they've actually got a good thing going for them with Arc and should focus on that.>b-b-b-b-but line go down!!!! Make line go up now!!!!!!!!God I hate investoooooorrrrrs, Intel should just go private and leave the circus that is the US stonks exchange.
>>106623072my thoughts on this technology?sex with monkey woman on right
>>106623072So we're basically getting Intel CPUs with NVIDIA iGPUs? I'm sure the laptopfags will be thrilled.
>>106623072CPUPro chads feasting on AMD's grave
>>106623072Hasn't nvidia wanted the x86 licence for years? Do you think the republicans will look the other way Nvidia buy them
so amd raped intel so hard that they created a monster
>>106626098I bet the Department of Energy is secretly behind some of this because they need a commercial partner to bring plasmonic nano lithography (i.e. lithography below the diffraction limit using atom lasers or polariton lasing systems). After all, EUV was originally a commercial application for SDI-era excimer laser defense programs. There is a reason why the patent applicant for EUV LLC is Sandia National Laboratory.
>>106625772CUDA is used for a lot more than just AI
>>106626122Both Intel and Nvidia didn't really have a comparable Strix Halo competitor even if Intel went balls to the walls with a giant iGPU. DLSS still reigns supreme, but nobody is gaming on ARM. Now in one fell swoop Strix Halo and its successors are dead for good.
>>106623693patents on amd64, except SSE4.2+ have already expired and AMD only owns somr exclusive extensions that (like intel's proprietary extensions), nobody uses.Everything else is cross licensed between both, so extensions that matter like AVX, AVX2 and AVX-512 (even if intel dropped the latter), both can freely use.
>>106626285>DLSSNobody wants fake pixels, no matter how hard nvidia advertising tells you you want them.
>dumbass senile politicians push companies to save other companies>actually it just eliminates the competition in the marketyou love to see it
>Intel announces high VRAM cards for low money>Puts 96GB VRAM within reach of the home enthusiast class>NVIDIA scared it will disrupt their AI business>"Partners" with Intel to keep their cheap VRAM out and have them use NVIDIA insteadFuck sake.
>>10662640180% of gamers are "nobody" indeed. The fact is DLSS makes gaming on laptops viable, even for non-gaming laptops with an underpowered 50-70 series chip.
>>106626285the strix isnt the problem its an indicatorthe mi450 is the problem which will probably birth the second generation of strix haloand it shouldnt be a coincidence that china is now raping nvidia h20...
>>106627015yeap this mi450 will be the first rack scale of amd they claim 5 pflops at fp4 so 2.5 pflops at fp8 it will probably be faster than whatever nvidia will release at the time for the same power
>>106625892she cute
>>10662649780% of ray tracing users you fucking mong
>>106628731No, 80% of people with an RTX card.
>>106623072>>106623106>>106623315>>106623341>>106623357Do you have any idea how much money I instantly made this morning because of this retarded shit?
>>106628971are you the granny inherintance intel investor
>>106628965source?
>>106623357>Only company that makes budget GPUs bought and paid for by the company that buried budget GPUsIt's unequivocally over.
We can stop giving TSMC all the money and Intel will produce chips again. Technologically it's a good thing. Not good for consumers cuz we won't be able to afford this new shit. We just have to get Intel to compete with TSMC at least. Sure 4T eclipses 5B but Nvidia doesn't have the foundry knowledge. Arc is cool and all but meh 5 year old GPUs are better
