It's all your fault, now we are stuck with the AMD monopoly
>>102155972I hope they pull off Battlemage good enough.
ok this is looking terrible
My A770 will become a relic to be featured in retro techtubers channels 20 years from now.
It sounds like the next round of products will be pretty good. lunar lake, arrow lake, xeon 6 p&e cores variants, battlemage. I don't know how that plays into high finance though.
>>102155972They will get bailed out don’t worry about it
>>102156084desu you couldn't pay me to buy a intel CPU now, but I'd be happy to see them nuke the GPU market with some big dick VRAM bukkake, and would actually buy/use one just to not give Nv or AMD money
I knew they were cooked once it was obvious how badly they fucked the cloud providers, their biggest customers.Two generations of CPUs set for early retirement at AWS, GCP, Oracle, Azure, etc. is beyond fucked. Nobody can recover from that.
>>102155972so AMD will have a cpu monopoly, and Nvidia a gpu monopolythe future looks grim for consumers
>>102156204Only China can save us now
>>102156118not if they split the foundries from the rest of the company. or rather, only the foundries would get bailed out. maybe
>>102155972Is it my fault that they made defective chips?
>>102155972Can someone remind me what the catalyst was that triggered this toppling of Intel, it all happened so fast
Dam dog, those 6W Zen 4 Athlons and the subsequent explosion of $100 tablet PCs are NEVER coming, are they? I'm gonna miss when AMD used to be the underdog so much.
>>102156302firing jim keller
>>102155972That is what happens when non technical people lead engineering corporations. Instead of pursuing the technical excellence that made the corporation valuable they go for the short term revenue. This ends like >>102156216 diagnoses.
>>102156118its as if millions of grandmas was suddenly silenced
>>102156302My understanding is that they screwed up the transition badly to EUV. They originally were planing to skip the 1st generation of machines. The machines are bespoke and made in low volume. They expected to be able to push their older machines using multi-patterning to make smaller geometries. Everything went wrong. EUV machines came slowly and their multi-patterning didn't work out nearly as well as they anticipated. Management failure led them to that point, and then management couldn't plot a course out. The exec before Pat purchased a mess of future 3nm production from TSMC just before he left that is being used for Lunar lake now. Pat is doing his 5 nodes in 4 years to get them back to the cutting edge on things mfg in house.Also intel has apparently somehow gotten some non high-na EUV machines that it is using for intel 4 and intel 3. I don't know if these are considered 2nd generation machines, or if they paid billions to jump the line somehow. If anybody knows the answer to this, and exactly when they got the machines, I'd like to know.
>>102156665>>102156329Okay so it had nothing to do with GN exposing the failures of 13th and 14th gen right? Because the timing was pretty much spot on.
Intel tried to have me fired (from an unrelated company) for making a usenet post 30 years ago shitting on Itanium. Based boss told them to fuck off.>Fuck 'em. Let Intel burn.
>>102156665oh yeah. I vaguely remember hearing that intel had a tremendous amount of trouble using tungsten for its wires. Old school chips used aluminum sputtered onto wafers for wires, but as wires got thinner and smaller, it wasn't good enough. That was the switch to copper wires, but copper diffuses into silicon, so it has to be protected by a jacketed layer. This means the entire wire isn't copper, only the center, so the resistance isn't quite as good as pure copper. Intel moved to tungsten because although its resistance isn't as good as copper, it does not need quite as much protection from diffusion, so the net resistance of the wires is better than copper. But tungsten has all sorts of problems, for example apparently it is brittle. This was a 10nm problem iirc.I'm not in the field, and this is all from hazy memory, so if anybody knows better, please chime in.
>>102156302Intel miss on 10nm process, Intel hope don't need EUV machines meanwhile Intel was the first company on use EUV for chip production as experiment scale.AMD start in 2017 with Ryzen launch one chiplet for 8 core and glue one chiplet for I/O and RAM controller allow glue 4 chiplets for Server chips, AMD will produce a lot CPU chiplets and use some for Desktop and the rest for Servers.Big CPU chips are expensive and hard to produce, chiplets allow unify Desktop and Server market with glue cheap and mass production chiplets.AMD move to 32,64,96,128(compact zen) cores in next generations and is moving to 196 and 256 cores per chip, AMD outperformance Intel in 2-3X at lower price for Server chips in one or two CPU socket servers.ARM Holding start to licenses ARM chips for server and Amazon start to build own chips from ARM designs other mega cloud companies will follow it, but AMD murder the ARM servers startups.Intel need heavy reduce margins on servers for big buyers to compete, AMD destroy the main income for Intel, meanwhile intel fail in every other project, become unable to compete Nvidia on GPU or AI.The problem is Intel destroy margins for survive on market share on servers against AMD.Apple moving to ARM or Windows ARM laptops don't impact it or AMD on Desktop, but AMD murder Intel on Servers.
>>102155972now if only the same shit would happen to boeing
>>102156697I don't think so. Intel's stock tanked immediately after a shareholder presentation in which intel said their outlook was bad and they were missing their previously stated projections. (In business this can be bad, particularly for a company that is trying to turn itself around). In that presentation nobody even asked about 13th or 14th gen problems. This was after the GN stuff was out and known.
>>102156302>what the catalyst was that triggered this toppling of Intelsanctions on China
>>102156325PH2 scales down to 5w, but why would they ever be in $100 devices?
>>102156302Intel was such a behemoth that it'd taken years to get into crisis… had they didn't dive head first into the concrete by blindly jumping into the GPU market.What other anons said about failed transition to 10nm and AMD competition are two of the three biggest reasons why they are in their state right now, the third one is their failed GPU market.Their CPU, even with all of those R&Ds, are still profitable. But their GPU division was such a massive money sink it's not even funny. And despite all of the efforts, they STILL failed to capture the market because scalpers flooded the market with secondhand GPUs.
>>102156084Still have my 750 in the original box. Gonna sell it in 20 years and buy a house.
they need to hire more indians to redeem the company
i think you could probably take this all the way back to intel completely shitting the bed on mobile and just giving updidn't see it coming, barely tried as they saw it grow into one of the biggest markets on the planettotal mgmt failure
>>102155972>Intel to dissolvethat isn't what your own screencap says, you fucking idiot.> no linkplease kill yourself
>>102156697its actually more to do with the AI boom and intel 100% missing out on all of it due to lack of vision and nvidia getting 10x intel stock valueTheir foundry business has massive failures, amd catched up and overcame intel thanks to tsmc. Like theres no investor confidence that this company is going to succeed in this market when their competition has so thoroughly beaten their asses.
>>102157042>Intel was such a behemoth was? still is, retard. i really enjoy how you fucking faggots that were born yesterday have no idea how low intel's share price was during the 1980s and 1990s.
>>102155972I feel absolutely no sympathy for Intlel, but I dread a world where CPUs become a monopoly. Jesus Christ, GPUs aren't even a monopoly and yet Nvidia more than doubled the price for their vram starved cards because they control the market, imagine what AMD will do if they become a literal monopoly. Get ready to pay $3k for your next x86 CPU if you don't want a shitty phone processor.
>>102157183They still have to compete with themselves so it takes time for prices to rise. They can't release a 3k CPU that's 10% better than their previous gen $400 CPU.
>>102157183intel is still going to make CPUs. the absolute state of tech illiterate and ignorant primates of this board.
>>102157038Because even a 2c Zen 4 CPU/2 CU RDNA 2 APU would mercilessly CURBSTOMP virtually all low end ARM SoCs (ie dogshit one in raspberry pi 5).Being such a low end chip to bin I imagine it wouldn't cost AMD more than ~$20 to produce. Use guts from a Fire HD 8 tablet and it seems reasonable that the whole thing would sell for around $100.
>>102155972nothing will happen, ayymd can't produce even close to the volume needed to compete
>>102156204And on top of that AMDs CEO is related by blood to NVDAs (they're cousins). It's all been a ruse all along
>>102157452They aren't direct cousins, and pretty much all of Taiwan are relatively closely related because the island was populated by a small initial group that fled the mainland
>>102155972>apple stopped using powerpc>powerpc dies>apple stopped using intel>intel diesbased tim apple
>>102156325I remember the AMD Doom threads in 2016/2017 when their stock was sub $1.50 - $2.00 and they were in danger of being delistedHow the tables turn
>>102158255fuuuuuuuckwhy didnt i care about finance at that pointfor fucks sakei even got bitcoin shilled to me when it was at 8 usdi went >naah, its an inflation pump. even if its the real deal its IMMORAL to invest in itff ~12 yearsand ill be the first to say fuck morality, gimme the cashoh well
>>102156204>Meanwhile Apple bros comfy with popcorn
>>102156329no, you fucking retard
>>102158483Apple increment heavy the prices for own chips and do more heavy price segmentation for every generation.
>>102158255They even sold their headquarters' land to lease back just to get that capital to survive and someone somewhere from AMD looking back said they had less than 30 days of operating cash at any one point before the release of Zen, it was quite precarious. Intel isn't even close there but it is no longer in a great state. I really hope Pat manages to put it together with Lunar Lake's launch and that will start their comeback.
>>102158533Can you explain that in plain English?
>>102155972If Ayymd managed to return from death's door, Incel will get through this minor kerfuffle. They are just now paying the tax on the 4-core-4-ever era.
>Intel's promising GPU effort might get slashed because their CPUtards couldn't be arsed to release a product that doesn't kill itselfkill me
>>102158483did her nipples always poke out in that picture or is that an edit?
intel has no new working productslinux is building the ecosystem to which anyone can go download it and run their games, and emulate everything else fast enoughintel can go die in a fire
>>102155972>now we are stuck with the AMD monopolyGood.Does anyone else notice the>3-Year duration 4chan Passes are now available for $45at the top of the page now? Is this really how 4chan makes its money? Shitposters avoiding the spam delay?
>>102158822shitposting is serious business
>>102158837>shitposting is serious businessyou bought the poorfag ad function
>>102158888my money goes to tendies, not gookmoot
>>102155991That really depends on your expectations. Everything (including what Tom Petersen said recently) suggests that we'll see around a 50% performance improvement over Alchemist. That'd put the B770 at roughly RX 6800 non-XT performance. Hardly impressive, but it'd be okay if the price was low enough. Yet I see people suggesting it's going to compete with the 4070 Ti and maybe even the 4080, which is just setting yourself up for a massive disappointment.
>>102155972aw fuck, i better sell that stock sooner than later. I don't want another situation on my hands where im stuck holding a worthless stock.
>>102156129Are consumer grade CPUs really such a big deal on cloud market?
intelaviv finished and bankrupt
>>102158581I gotchu brother. English syntax isn't always straightforward, but anon-sams's sentence is a bit mentally taxing to read as a native speaker. We can re-arrange the sentence structure a bit and add/del a few words to make it easy to understand:>Apple [subject]>increment [action verb]>"the prices for their chips" [object] > +> "and do more heavy price segmentation"now move the "for every generation" to the start of the sentence instead of at the end,[Result]:>Every generation, Apple heavily increments the prices on their chips & keeps doing more price segmentation within those new segments.>>102158533nice dubstranslating for you made me feel human today
>>102158822>>3-Year duration 4chan Passes are now available for $45I hate that I'm actually considering it.
so Arc is dead ? I love it when the so-called free market is just two cousins fixing the prices, just like the founding fathers had intended
>>102155972isn't amd forced to help intel?
>>102160406Are you against that? I bet you're a commie.
>>102155972Intel can't fail, they would literally force amd to share technical info to help intel develop a new product than let it go bankrupt, this is literally the same thing they made microsoft do when apple was on the bring of bankruptcy thanks to M$ monopolistic practices.
>>102159722This was absurdly kind and helpful of you, anon. I'm not him but wanted to praise your kindness anyway. Nice one!
>>102155972why don't they just make processors then? no one needs all the troons nigger and e cores.
>>102157628>>102157452Asian take over u will submit to ASIAN COCKSAY U LIKE MY ASIAN PENIS YT BOI
>>102156216AMD has ayy-s working for them.
>>102160768It has nothing to do with amd. It's the failure of foundry. Now all American companies need TSMC to make <7nm chips. 1 missile can halt the whole industry.
>>102156216You really need to kill yourself, /pol/ subhuman.
>>102158483>>Meanwhile Apple curryniggers shit the street
>>102155972They will be bail out by the US gov, just watch. I will seethe when this happen but I know they will just get free tax money.
>>102155972>now we are stuck with the AMD monopolybad news for x86 is good news for RISC-V and OpenPOWER
>>102156302Pawning off their long-term assets for short-term gains and gambled everything on 10nm node on the foundry side. They didn't expect TSMC fueled by Apple's capital to leapfrog them. They got too complacent on sticking with monolithic chips on server/HPC side when tehy should have made the shift to chiplets back with 14nm node had a troubled start.
>>102163884>OpenPOWERQrd?
>>102164017>Qrd?a mature open ISA that can do heavy workloads where RISC-V comes short, these are processors that get used in data centers
>>102156422kek, i understood that reference
>>102156216probably not. x86 isn't a popular platform anymore. everything is arm now
>>102157183they compete with the half a dozen arm chipmakers, way more if riscv comes to be.
>>102156329Nah, It was more like Keller peaked his head through the door after he worked on zen.Got a whiff of the stinky mess Intel was in.Then he left.
>>102156665Probably friends in the US government. Regardless of what people think here, ASML is basically a US company. They may make the machines, but EUV technology is licensed to ASML by the US government because it was developed at LLNL. I get the distinct feeling that at some point Intel might be the sole beneficiary of the DoE's work in "plasmonic nano lithography", aka lithography below the diffraction limit. Given the stage of Taiwan now, I doubt the US government would license this to anyone that's not an American company. The problem with plasmonic lithography is that it has potentially all kinds of classified applications because plasmonics can be used to do all kinds of fun things on the defense contractor front. I imagine there is a contingent within the DoE that is attempting to figure out how to bring it to the consumer world in a manner that won't letting the Chinks on that you can build a gamma ray laser with the same methods
>>102162270The us government won't allow intel to fail, they need their foundry to remain active. God imagine what happens if china invades taiwan
>>102156129>>102159348No, Xeons aren't affected.However all the bigger fintech players got burned, since they do use consumer parts (better single threaded performance than enterprise platforms and much higher single core clock speeds to boot) for high frequency trading.
>>102167446He was at Intel longer than he was at AMD.
>>102169271he publicly has said Intel's corporate structure and work culture is utter shit so he wasn't able to do much
how the turntables
>>102169332Still designed Royal Core and set the wheels in motion for the third attempt at reverse-hyperthreading (rentable units)
>>102157355>Because even a 2c Zen 4 CPU/2 CU RDNA 2 APU would mercilessly CURBSTOMP virtually all low end ARM SoCs (ie dogshit one in raspberry pi 5).The point of the Pi isn't that it holds a candle in performance, it's that it does what it does at like 3W for the entire system. Ryzen sucks dick at scaling downwards in power, even their monolithic APUs require ~12W idle just for the CPU and double that for the motherboard assuming you have the lowest end mobo with the least components, least VRMs, and a platinum rated PSU. And that's idle, not load, where the 65W TDP Ryzen actually uses 110W measured from the wall.yeah, good luck putting that in a mobile device.
>>102155972I have a PC with the following parts laying about i5 2500k ASRock Z68 mobo8gb ramRadeon 6950Is it worth putting in money to make it viable? If so what should I put to make it a general snappy browser/media machine. And maybe some basic gaming for the kids like Minecraft and terraria
>>102155972Oh no it's so unfair and AMD's fault that they also had to do this 16 years ago
>>102169837>even their monolithic APUs require ~12W idle just for the CPU aEvery single current UMPC proves you're a shit eating retard talking out of his ass
>>102155972Kek will they pull out of Tel aviv or will they just take US subsidies and leave altogether ?
>>102155972I'm not the intel engineer that fucked up and made their CPUs rust.Go complain to that guy.
>>102156302>Can someone remind me what the catalyst was that triggered this toppling of IntelMaking fun of AMD for gluing CPUs together. Or I guess even earlier than that was refusing to sell iPhone CPUs to Jobs because they wanted higher margins.
>>102155972is this 110?
>>102162898he's right you demented troon, intel had it in the bag and kikes and shabbos goys with MBAs ruined it.it will go under faster than a thousand trannies necking themselves
>>102163884>tripfagfiltered
that Saudi dude who invested in AMD during their terrible years, and was basically the only reason it still existed, must be laughing to the bank right now.
>>102170342>Every single current UMPC proves you're a shit eating retard talking out of his assThose use laptop SoCs with no chipset, minimal i/o, no pcie beyond m.2 slots, LPDDR RAM, etc. I was talking desktop computers and the figures come from my own measurements from a 5600G system and a 2200GE system (both are monolithic dies).You might argue that that it's unfair to compare a full ATX board to a SoC and I should use the SoC ryzen instead, but the thing is, an i3-13100 in a H610 can achieve sub-10W idle power. So it's just AMD sucking dick in that category. The 2200GE I have can't even go lower than C2 core residency while Intel can go down to C10, this is one area where Ryzens are fucking pathetic, and the chiplet based cpus are even worse in this regard as the separate IO chiplet uses twice as much power.
>>102155972Another company being ruined by a fucking boomer nepo baby.
WOW
>>102155972NothingburgerIntel will be fineNothing ever happens