Intel have just made their BIGGEST MISTAKE yet

Urcdkeys.Com 25% code: C25 【11.11 BIG SALE】
Win10 pro key（$15）：biitt.ly/pP7RN
Win10 home key（$14）：biitt.ly/nOmyP
Win11 pro key（$21）：biitt.ly/f3ojw
office2021 pro key（$74）：biitt.ly/DToFr
MS SQL Server 2019 Standard 2 Core CD Key Global（$93）：biitt.ly/oUjiR
Affiliate links (I get a commission):
BUY an LG C1 OLED 48'': amzn.to/3DGI33I
BUY an LG C1 OLED 55'': amzn.to/3TTpQp9
Support me on Patreon: / coreteks
Buy a mug: teespring.com/stores/coreteks
My channel on Odysee: odysee.com/@coreteks
I now stream at:​​
/ coreteks_youtube
Follow me on Twitter: / coreteks
And Instagram: / hellocoreteks
Footage from various sources including official youtube channels from AMD, Intel, NVidia, Samsung, etc, as well as other creators are used for educational purposes, in a transformative manner. If you'd like to be credited please contact me
#intel #apple #arm

[@pabloquijadasalazar7507]

If games work well on ARM chips, they'll definitely be a threat to Intel

[@platin2148]

I don’t know yet of any decent arm chip which can do very good PCIe anything even apples PCIe is sort of pointless. And the internal gpu‘s are a joke no matter how one turns it.

[@blkspade23]

PC gaming is such a relatively minuscule element to the space that it won't much matter one way or the other. For example, in 2022 286.2 million PCs were sold, but only 30.34 million discrete GPUs were. That's basically 10.6%, assuming all those dGPU target gamers, that is already being split with AMD CPU competition. Never mind the fact that AMD has APUs competent for casual gaming needs. That other 89.4% having their needs met in better ways that don't involve gaming is the real problem. There is no reason games wouldn't work well on ARM, other than games not being coded to run natively on them. The impetus for that native programming is a shift in market share.

[@platin2148]

@@blkspade23 It doesn’t matter really if the maximum that you can reach is already hardcoded.

[@mddunlap03]

​@@blkspade2361 million dgpu just in q2 2023.

[@carolinas8886]

You don’t need all working well. Only the casual ones in the start, which they are. These will help create critical mass, and then all studios will have to at least consider them.

[@epobirs]

The loss of Apple was more of a hit to prestige than financial. The volume of Intel-based Macs was always a small slice of the market for x86 CPUs.

[@Techaktien]

True

[@mddunlap03]

Apple silicon is really just for nich fans most people won't pay 5k for a desktop that loses to last Gen chips

[@chidorirasenganz]

@@mddunlap03Considering Apple is industry leading in CPU and intergrated GPUs and can be had for as little as 499$ your full of it

[@chidorirasenganz]

Intel also missed out on designing the A series in addition to the M series. Phones and Tablets far outsell x86 PCs

[@IonorRea]

Intel's share of CPUs in Apple laptops was indeed small but since Apple switched to own CPUs in their laptops their market share rose rather significantly while Qualcomm has been trying to enter the same space on the Windows platform for a couple of years too (like Lenovo Yoga/Flex 5G), thus like Apple it has necessary previous experience to do it well enough eventually for presenting a significant threat to Intel's share of the laptop market. So, at the moment the most likely outcome where Intel wins eventually anyway is in a Google Stadia-like online service-based model where you will own (nothing) just a screen connected to the internet while servers do the heavy work, at least as long as Intel can hold its dominant position in servers and at the prerequisite that internet tech for reducing latency improve so significantly that near instantaneous response times once associated for display latency become standard for internet latency thus make remote processing power as service possible. There are already Arm-based supercomputers while graphics cards have been used for crypto mining for a long time, so Intel's dominance in any space is far from assured, unless Intel's manufacturing technology becomes superior to what offers Taiwan or China in the future...

[@TheEVEInspiration]

If all those new ARM chips are not using the much stricter x86 memory access ordering (which Apple also adopted in the M1), then I predict loads of compatibility issues. Apple had good reasons to choose this path and as they make their own chips, compatibility is implicit. Which is not the case when there are 4+ different ARM based chips running a version of Windows, each working just a bit different in this area.

[@lucabartolucci919]

Maybe for the personal computing market, not pro or company, Nvidia, Qualcomm, Samsung will develop a in house distro of Linux the major downside is AI right now, but on the other side Win for ARM is very poor.

[@anarex0929]

Yeah we never seen that happen before cyrix, AMD even intel at one time. What do you think the x86 processor had to deal with? X86 processors were all over the place some didnt even have math co processors. Seems you forgot or where born to late to know better. Judging by how you speak it seems you forgot that. 😢 Remember when Cyrix quake problem due to the cpu missing some instructions, intels cpus had. So ya you forgot about a lot of history and it seems your repeating. Get Windows seem to get along just fine without a math co processors. So your argument is invalid windows and other operating systems will adapt just like Android os did.

[@nikos_Man]

@@lucabartolucci919 why would they do that ? linux is better for some uses but not the desktop It's more likely that android might be able to overtake windows

[@lucabartolucci919]

@@nikos_Man Android is very crappy for desktop even for tablets. For average use Chromebooks are enought for a lot of consumers, why not Linux? Anyway as in the Apple ecosistem tenere will be a merging between mobile (smartphone) OS and desktop OS due to’ the fact that everybody use mostly phone. We start with PC, now everybody starts with phone, the interface should be similar in future, we are the past.

[@maxwellsmart3156]

@@nikos_Man Linux has an excellent desktop experience as I use it everyday. What most people mean about the Linux desktop is that it can't run MS Office or some other program written for Windows application (mostly gaming). Arm hardware, at the moment, running Windows have the same issue, they can't run most of it.

[@davidbuddy]

I really wish we could skip the ARM era and go straight to RISC-V since I think from a consumer's perspective having an open ecosystem where anyone can throw their hat into the ring without having to deal with licensing, and being able to add custom instructions/extensions I think is a for more interesting proposition than moving from something completely closed (x86) to something less-closed but needs licensing (ARM)

[@IakobusAtreides]

Well said

[@mickmoon6887]

Every era has its beginning and end RISC-V has barely begun in the west but in the east its the only way due to balkanization or less globalized version of upcoming world order Open source nature wouldn't work with corporate systems only licencing model would work

[@TheReferrer72]

Sounds like a disaster to me, can you imagine vendors doing a Intel and starting to add instructions that are patented, making not only hard for competitors, but hard for compiler writers. I think sometimes its good to have a dictator that can keep everyone in line.

[@FlorinArjocu]

We are at least 10 years until we see a real desktop competitor.

[@quantumdot7393]

I don't know when risc v fans are going to stop being so delusional

[@maxwell2288pspgo]

Microsoft would need to make a great translation layer like Apple did to make ARM a viable option on windows. The reason Apple was able to move to ARM is their vertical integration. They also spend lots of money for developers to make their software compatible such as photoshop, Final Cut Pro, and other creator based software.

[@epobirs]

Microsoft has a group dedicated to that purpose, starting with when they acquired Connectix way back. Connectix was best known for emulation products to use Windows on PPC Macs and a product for running PS1 games on PC. This is where the ability to run a large library of PPC Xbox 360 games on x86 Xbox One and Series consoles was produced. Windows on ARM already has x86 emulation. It was limited to 32-bit for the first few generations but is 64-bit now.

[@TamasKiss-yk4st]

Microsoft thinks that their name will solve everything (developers will makenprograms for their low market share product), this was the reason why they throwed away more than $6billion (Nokia), their app store was empty even basic app like social network apps wasn't avaliable on Windows Phone for 3 years, and that time the most of the users alredy bought a new non Windows phone, where they can download apps from the app store.

[@theronwolf3296]

@@epobirs The basic philosophy of RISC should make interpreters for it much easier than from going from one CISC to another. Translating instructions into the streamlined RISC type is a lot easier and more predictable than going from one CISC to another.

[@epobirs]

@@theronwolf3296 That's nice but has no bearing on the situations faced by the companies at the time of need. Apple wasn't going from PPC to ARM, it went from MC680x0 to PPC to X86 to ARM. Likewise, Microsoft consoles backward compatibility first had to go from x86 to PPC, then back to x86, with more x86 following. Further, for the consoles, Microsoft didn't rely on an interpreter for the end product. It used iterative translation to produce the binary the user was required to download to run an Xbox 360 game on the Xbox One and later. This goes back much further to the DEC Alpha days and a product called FX!32, which got x86 Windows apps running on NT on the Alpha, with improved performance with extended use by the iterative process. This was considered viable for PC apps but not for console games.

[@stevewilson6193]

"5 years ago I predicted apple would ditch intel and create their own processors..." dude, 5 years ago EVERyONE was making that prediction.

[@Hugo-yp9dt]

Tbh, I certainly wasn't

[@slmille4]

Intel wasn’t 😅

[@POVwithRC]

Wow that's crazy bro but nobody asked.

[@alejandrorisco4201]

Same as when he thought Nvidia was going to use an independent rt processor just because of the card visual design

[@mddunlap03]

Ah yes really showed them a 5k desktop that's slower than last Gen cpu and worse than 2 Gen old gpu..... but for the money at least they made sure it was not upgradable.

[@guderian557]

Competition is always good for the consumer. Let's hope Qualcomm and Intel step up in the GPU market so that we can get some competition there as well.

[@denverbasshead]

Intel won't, battlemage will suck, and I doubt if the generations after even release except for super low end

[@attractivegd9531]

Can you explain how competition maters if the inverstors behind intel/amd/nvidia/arm are the same?

[@MaTtRoSiTy]

Nvidia are too far ahead at this point, wont happen sadly

[@fepethepenguin8287]

New GPU makers won't help prices... See gas prices How many different gas companies are there. And why is every gas station the same price, with in 2 cents

[@dac33nr58]

AMD, if they don't abandon high end gpu in the future.

[@Voidkitty_]

I don't think intel is underestimating arm, intel is trying to stop bleeding rn and x86 is what they are best at, so they're sticking to it, I think Pat is most likely saying what he said in order to calm share holders and prevent intel's stock from falling farther. I also think he understands that x86 has a significant advantage over arm simply due to most windows binaries being compiled for x86 meaning for a while arm is going to suffer from performance loss from translation as well as a general lack of software support, in addition to this, if I'm not mistaken has a rv program in the works as well. As a final point the intel title system as well as their push towards becoming a fab means they are going to be able to work with arm manufacturers as opposed to against them Intel's biggest enemy is their anaphylactic allergy to releasing a product on time

[@theexoreviewer5699]

I been lifetime and user. I would quit the day they said “sorry, all your previous software won’t work” I will stick to X86

[@coladict]

Dethroning x86 in consumer PCs will face a very hard chicken-and-egg problem. If you're a software developer, why compile your software for a platform with no user base? If you're a user, why buy the platform with no software ecosystem? Remember IA64? Very few people do.

[@quantumdot7393]

I think this will start with the most widely used applications that are usually developed by big companies . These companies can afford to port their stuff to arm without it making a dent in their wallets. Especially since many of them already did that for apple silicon. And if most people switch to arm because it has better battery life and the applications they use everyday are native then you don't really have a chicken and an egg situation Gaming is a big question mark tho.

[@axe863]

There are so many highly optimized libraries for x86 that its insane. People dont realize the extreme effort that will be required.

[@cbuosi]

Nothing that a good (and fast) x86 emulator could not handle

[@axe863]

@cbuosi At 20-30 % performance cost worse when emulating multicore. ARM vs x86 memory model need I say more

[@SirDVV]

absolutely @@quantumdot7393 the performance differnece is so massive, it makes no sense at all to keep using x86. Microsoft will also develop some kind of "rosetta" if they don't have it yet and just like at apple, softare will perform better even not being ARM native.

[@MoraFermi]

My biggest worry is that this might enable Microsoft to try to lock in the consumer devices market again, via "security" measures, like they tried to do with UEFI, secure boot and similar.

[@_yuri]

surely they will let you install *nix and bsds unless satya is a mythya

[@epobirs]

UEFI was all Intel. Microsoft only had peripheral involvement in that.

[@miaya3898]

funny considering Apple in the 90s was closed resulting in only 5% share of the desktop market

[@madmotorcyclist]

PowerPC was the issue given IBM's real lack of pushing its capabilities. ARM is now swooping in and given its much more processor thermal efficient handling over CISC based systems. The interesting question in the future in this competing market is the unified memory architecture that Apple is using going to compete well with the standard bifurcated memory architecture of CPU and Video interfacing over a bus. Right now, Apple's GPUs are several steps behind the Video card makers in performance.

[@johnwayne-kd1pn]

Nah, don't worry about Microsoft, worry about Arm. Arm is basically a locked down architecture, with all devices pretty much using locked ROM memory. If that comes to PC, you'll be locked out forever. Problem with porting things like GNU/Linux to Arm and use it there is that there aren't any functional drivers, and it seems pretty much mission impossible for devs to develop generic drivers for Arm based hardware. Probably it changes every Arm generation, or drivers made for hardware X is incompatible with drivers for the exact same hardware with name Y. If that's the case, you can't make things like say SATA drivers and expect them to work on all SATA devices, because there are no standards in the ARM world, except Arm, just wild wild west. SOC is the threat to freedom for computer users, not Microsoft (as much anymore). As bad as Microsoft is, they are like a smiling baby compared to the Apples and Googles of the world. Those will lock you out of anything they can, and try to control you and own you (and your mind). Mobile devices of today (made by Apple and Google-submakers) is an existential threat to humanity. They want to own you with their devices.

[@chair3299]

Mediatek is also rumored to enter the PC ARM market. They did have a partnership with NVidia and their CPU's can run alongside RTX cards.

[@stevewilson6193]

Qualcomm will have a hard time breaking into the PC SoC market if they don't open up their peripherals to other OEMs. right now if you want QCOM processor you need to use QCOM PMICs, QCOM battery charger, audio codec and Amplifier, its a closed system. AMD and Nvidia will partner with many silicon vendors and the collective power of the industry will leave QCOM in the dust. I love how you talk about apple's questionable business practices but don't mention QCOM's

[@POVwithRC]

Hey wait, don't both AMD and NVidia both sell GPU packages to board partners with a bunch of locked in supporting hardware?

[@baronvonlimbourgh1716]

That isn't remotly the same t thing.

[@mddunlap03]

​@@POVwithRC? They both sell the silicon and memory together but seeing as the gpu is built with a specific buss its just the way it is. But no partner boards will have different mosfets, layouts,pcb and components but do have to hit minimum signal noise specs

[@lamhkak47]

The software compatibility side would be quite a bit challenging for the non-x86 entries, but so long there's some all-round raw performance advantage the competitors have over the Intel's chip, Team Blue would slowly have their advantageous position slowly chipped away, until the building collapses and not able to bring excess revenue for the company.

[@cjadams7434]

intel is the new ibm i think…too much hubris and too slow to pivot

[@platin2148]

I highly doubt anyone will be more efficient in emulation + memory model than the native programmable hardware that intel already provides and they can’t really do too much software things. CAD for example runs extremely bad/not at all on arm.

[@chidorirasenganz]

@@platin2148Actually CAD runs pretty well on ARM

[@platin2148]

@@chidorirasenganz Which one? Last time i tried was solidworks and that didn’t go well at all.

[@chidorirasenganz]

@@platin2148 Shapr3d

[@levislevitas]

getting non-x86 architectures to be good for x86 applications will slow and bloat them. the movement is just towards more task specific compute rather than generalist solutions that will necessarily have a lot of redundancy and bloat. arm chips when adding more and more instruction sets will have the same bloat problem as decades old x86.

[@axe863]

Even having a hybrid big little structure bricked AVX-512 lol

[@salainen6850]

Hybrid cores should ALWAYS be instruction set compatible to avoid stupid software level shit. I mean they could have just split avx512 operations into two on E-cores

[@odaharry]

I don’t think it will ever be dethroned unless it has really good backwards compatibility for all your old x86apps

[@baronvonlimbourgh1716]

Old x86 stuff can easilly be emulated on new arm silicon.

[@sinephase]

guaranteed M$ is planning on a translation layer like they have for 32-bit compatibility but the performance is going to be trash. They still need devs to port to ARM. I hope it's different this time but it'll take some time either way.

[@kensmith5694]

Box86 emulates X86 on ARM. It works OK

[@aforty1]

Have you tried ARM Mac’s? It’s more than good enough as a holdover until heavier apps are ported.

[@eyeOfAC]

This is an interesting question. Apple’s closed market place was propably a much easier target than PC and all it’s obscure software out there. I’m worried that businesses wont be able to transition if their particular dinosaur era program doesn’t emulate…

[@mannkeithc]

I am great fan Windows 11 arm. My daily driver is currently a Windows Dev Kit 2023 PC and my laptop is a M2 MacBook Pro. I totally agree with what you say about Apple. Apple events are a bit like watching a slick Magician, where you are never quite sure how they do it as they magically present their data, but you have some nagging doubt there may be some sleight of hand at play!

[@FlorinArjocu]

It also depends on how good Intel's next CPU is. If the efficiency is very good, so if one can get a 10-12-15 hours of mixed usage on laptops, there is not that much incentive for searching something else. I am very curious about their next release.

[@scotane7753]

They are currently building a new fab and are the first to receive the next gen euv machines from asml so the next few years will be interesting.

[@mikelay5360]

Did anyone actually expect intel to outright admit that they feel threatened or defeated? Be realistic !

[@baronvonlimbourgh1716]

They would at least signal that they are working on something to face the chalange. And they clearly don't

[@quantumdot7393]

​@@baronvonlimbourgh1716I agree with the op. Also Intel can't pivot to arm now and expect to win. Like it or not their fate is attached to x86. That goes for AMD too they are just too stupid to realize it. These arm SoCs are not just about the CPU you also need a great GPU and AI capabilities. Something that Nvidia and Qualcomm are good at. AMD is ok but you don't win by being just ok and Intel is trash in that department. If x86 fall intell will just be a fab company like TSMC with some research like IBM

[@SP95]

And some day INTEL will come with friggin photonic chips out of nowhere and dwarf the competition for the next 10 years

[@redachhaibi9708]

ARM's biggest issue used to be software for adoption as a PC or workstation. For me, Microsoft is again the last one to move. But the sign was always on the wall. I find it strange that few people ever mention: - How much Raspberry Pi has advanced software. Linux for ARM is now incredibly functional. - That Nvidia has huge experience with their Jetson series. It's hard to say how far they are from a great chip. Jetson designs have generic Cortex CPUs and custom Nvidia CPUs such as Denver/Carmel. But the software stack is ready, especially for server. Again Linux for ARM. - The emulation community does wonders with these.

[@jamescarter8311]

Don't bet against Intel.

[@danielduxmwangi5799]

Intel needs a slap in the face for the past decade of profiting off releasing the same chip in different clothing.

[@ouulriyac4747]

Complacency as a strategy - By Intel

[@bassamatic]

snapdragon cpus sure made a splash a few years ago in samsung phones... until we figured out they manipulated the bios to overclock during benchmarks... wouldn't trust one word out of the chinese

[@davivify]

I should point out that we don't know what Pat G. thinks, only what he said publicly. He may be trying not to panic investors.

[@davidmiller9485]

So how many years do we have to we have to hear about this??? X86 isn't going anywhere until someone comes up with a code base that is backward compatible (without performance loss). This lovely idea has been roaming the halls of peoples biases since the early 2000's.

[@sinephase]

Apple jumping ship from PowerPC was the history you should be comparing. M$ tried to support ARM with windows 8 but we see how well that went over with developers. With Apple, devs have no choice but to either stop development for OSX apps or adapt; this isn't the case on Windows. Unless they sell ARM systems at a subsidized price to generate users or if they can really get a very high performance gain over x86/64, it probably won't be much of a competitor.

[@AschKris]

> "Unless they sell ARM systems at a subsidized price to generate users or if they can really get a very high performance gain over x86/64" Both of those things will probably happen lmao

[@MoireFly]

On the one hand, apple's M1 proved ARM can compete with x86 - it's a fantastic chip. On the other hand, it and in particular the M2 and M3 have also demonstrated they can't really beat x86, at least not easily, not resoundingly. Despite the backing of the literally most valuable company in the world and throwing a huge transistor budget at the problem, Apple's chips merely compete with x86 chips, often losing on threaded workloads, but leading in efficiency. I think it's not at all clear that ARM chips will have an easy time competing with x86; and even if for a short time they outdo intel, that might not mean the advantage will last, nor necessarily will hold vs. AMD's zen which isn't handicapped by using Intel's foundries. Intel's architectures might be in fatal trouble. It's definitely plausible. But I don't see convincing evidence that that possibility is a clear likelihood.

[@allanallansson9532]

Your comment about Apple M-processors being excellent and still dissing Apple as a company for other very valid reasons shows that you are able to make a very balanced analysis. Very rare these days!

[@Crihnoss]

How does RISC-V fits into this? I read that google is now supporting the architecture in Android?

[@orthodoxNPC]

Its physically impossible for an arm chip to have the same speed as an x86 chip, without consuming MORE power than the x86 chip. Arm is slower and cooler, x86 is hotter and faster.

[@TheReferrer72]

Clock speed maybe, not processing power.

[@orthodoxNPC]

@@TheReferrer72 yes, processing instructions per time

[@quantumdot7393]

​@@orthodoxNPCmaybe in the past but It had already been proven that arm can compete in high performance applications with x86. x86 can also be more efficient if some necessary changes were made. We are just seeing a conversation in CPU design. We are reaching maturity where all instruction sets are reaching the same logical end goal after many years of refinement

[@Giorgal]

My guess is Intel is betting on history repeating itself, because in the past Intel itself tried to move into Itanium but then moved back to x86, Apple used the PowerPC architecture but then moved to x86, and then the console manufacturers abandoned the convoluted PowerPC based architectures and moved to x86 architectures (except Nintendo) and AMD had tried to create ARM based server CPUs but eventually abandoned the endeavor. Personally, patents aside, I don't think the ARM architecture has an inherent advantage over x86, the main reason Intel's offerings currently suck is because Intel's fabs are one generation behind TSMC and Samsung, whereas the fabless chip manufacturers have the advantage of choosing the most advanced nodes for their products. ISA-wise x86 has a lot of legacy baggage that might weight it down but Intel recently released an x86S specification, which would drop all legacy x86 modes, and run the CPU in 64-bit only mode from boot, while still retaining backwards compatibility with 32-bit programs and 64-bit programs and OSes. x86 backwards compatibility will be the only thing stopping ARM CPUs from taking off, though that might change if Microsoft releases a good x86 emulator. Also I'm kinda wary on whether these new ARM based computers will be able to play older, x86-based videogames. Newer versions of Windows already have issues with playing10 year old games.

[@gaffo7836]

"I don't think the ARM architecture has an inherent advantage over x86" To me this seems so true, by the time you upclock the chip, make it out of order, etc (can you hyperthread an ARM?) - it will perform GREAT = to an x86!, and take as much (ok maybe 80-90- percent as much, then you are like, "big whoop" is there really a need now?) Agree on the Intel process/node - their x86 chips are fine, and if made by TSMC would match Zen4 ones in all ways including power consuption. But sadly Intel is not TSMC. There was a time 10 yr ago, when Intel was the leading process/node manufacturer of chips, but that was a long time ago. Intel has done well - making all the "effiecency" cores (because they had to in order to compete in performance with AMD while behind by 1-2 generations in manufacturing (IMO that is the only reason the "effeicency" cores exist - i see no real need for them objectively. Let role reverse, Intel has the best node, do that then make "efficency" cores to beat AMD? or do they load down their chiplets with hyperthreading p-cores to chew Zen4 several new ones. I welcome a harware/socket compatible ARM chip so I can build an ARM desktop computer off the shelf, but if Qualcomm want to play the Apple game and sell me thier special golden rimmed toilet shaped desktop PC that is "Special" in that its parts - motherboard/vid card/etc is only made by them -------------um no fking way. open hardware platform all the way - that war was fought by IBM, and THEY lost - get over it Quallcom. IMO

[@michaelashby9654]

The M2 Pro is the first CPU that actually felt like an advancement to me. Intel chips have not felt like an advancement in decades. Intel was happy to keep churning out 4 cores forever.

[@kamenriderblade2099]

There's several things you forgot: 1) The incumbent effect The vast majority of MS Windows PC users are on x86. So why would any software developer waste a single key stroke on ARM? No matter how many ARM CPU's they crank out, who will buy it if the software library for Windows software on ARM is so meager? The Chicken & Egg effect compounded with the incumbent effect! - 2) The software Library is largely more important Just look at Video Game Console History and PC history. He who has the larger Software Library usually wins And x86 has a VERY extensive library. And since Team x86 has Snow-Balled several times over, the x86 library is MASSIVE ARM is starting from nearly scratch on Windows - 3) No amount of Code Translation will account for running the x86 software naturally. The sheer amount of bugs, performance issues, etc.

[@4.0.4]

It's perfectly possible to emulate x86 in ARM though.

[@kamenriderblade2099]

@@4.0.4 I never said it wasn't. It's just never going to be quite the same as running x86 software on a natural x86 CPU.

[@RepsUp100]

Why hasn't Microsoft released their own chip if ARM on Windows is the future?

[@Techaktien]

Work in Progress: Project Athena from AMD

[@quantumdot7393]

Microsoft isn't exactly a chip design company. You saw what happened when Google tried to do that in their phones

[@ItsAkile]

I don't really see trouble as they're all about the same performance wise with efficiency possibility closing in, I like the competitiveness of ARM, clear PPW lead in the now in its clean state. I think its worth seeing X86 through for Intel but they need to keep ARM as a clear backup choice. they know what the next decade entails for them as long as they can execute, the Foundry being top priority is key. So much new and world first tech on their hands 2023-25

[@heickelrrx]

Intel assuring their investor wasn't a mistake, it's just corporate sweetalk 101 Meteor lake is Intel Approach to counter ARM rising on PC Desktop, an ability to completely Shut down Compute die is what set it apart from previous generation. The question is whether this approach is efficient enough to erode ARM relevancy on Windows Desktop, if it is then Intel is winning, else ARM is winning Either way we'll see long battery life from laptop in upcoming generation

[@giacintoboccia9386]

Well, I am interested in ARM PCs, but note how my AMD64 laptop does five hours of notetaking without a carge, wich surprised me so ARM would definitely improve that but not change my habits. Anyway, it will be intereating to see the PC ecosystem becoming architecture agnostic, something much more interesting than just switching like Apple did.

[@arsims1]

I read that INTEL will soon change into the new architecture of CPU's and Chips manufacturing. The 14'th gen was the last INTEL Processor with this kind of architecture. The next one, the 15'th gen CPU will be made with all new INTEL Chips and CUP's architecture and it will be a game changing for everything it will have more AI and less watt consuming. INTEL will beat all, in Chip and Processor manufacturing next year. We have to wait and see until the next generation of INTEL CPU and Chips come out.

[@arthurcuesta6041]

The bottleneck isn't really in the instruction set, but rather in the fab process. As long as TSMC is the only capable of cutting edge lithography, stuff won't really change. That extends to the machines that enable the fabs: ASML machines. Create a decent competitor to ASML and you'll truly disrupt the market.

[@christopherjames9843]

I think it will happen. People probably aren't happy with one company with as much leverage as ASML. Maybe Canon?

[@vmafarah9473]

Do ARM chips has any advantage over x86 IN 3D RENDERING ? GAMING ? or only in smaller tasks.

[@Speak_Out_and_Remove_All_Doubt]

I wonder what Qualcomm's graphics drivers will be like? Will you be able to play games on this new chip? Is there not a market for them to release a discrete GPU if they are going to have to have drivers for their integrated GPU anyway?

[@luckys9249]

In the end the future is APU with integrated RAM. This is why they are all starting to make CPU and GPU

[@sinephase]

starting? This is just an expansion of tech that's already on mobile computers like phones and tablets

[@SMGJohn]

HBM3 and 4 is already being proposed for it.

[@theexoreviewer5699]

Of corse, then the industry can price lock and rig the industry ❤ gotta love companies locking everything down for “your benefit”

[@prague5419]

It's a tremendous race to create the most powerful ARM CPU that's compatible with....nothing. Good luck with that.

[@supernova874]

Yeap i agree with your assumption about ARM , the efficiency gains in comparison to x86 shows at least the need for ARM processors in laptops and so it will make the switch eventually and for Desktops (no company going to split their R&D to develop both arm and x86 processors and be a success , will research for example the ARM and when ready will replace the aged x86.

[@TheHangarHobbit]

I'd say Intel has it right here and this is coming from someone who has been on team Red since the early 00s, for one simple reason...unlike Apple whose machines run majority first party applications and thus can be easily ported the vast majority of Windows programs are written for X86 Win32/64 and emulation of X86 on ARM has always suuuucked. There is simply way too much custom personal and business software, from QuickBooks to Photoshop to the custom charting apps used by the medical industry, not to mention 30 years of games, that just are either not going to work at all or run like 💩 on Windows on ARM and if I can't run my programs why would I stay with Windows instead of running Linux or MacOS?

[@eg4933]

doesn't surprise me, i mean CISC has always been a mess. It was good in early days to some decades after because much the design of chips was still un-folding. Now its very clear of what these designs are and they're ready for a more polished and refined design like RISC.

[@marceelino]

When it comes to the new generation of programmers from Western countries or India, it's important to note that they do not have in-depth understanding of how hardware functions. They are the modern-day equivalent of assembly line workers, although they certainly receive better compensation for their work. They are the Apple's target audience. They're more interested in the convenience of a "works-out-of-the-box" experience. The thought of modifying their system is non existent, and they're more inclined to seek the expertise of IT support whenever they need something to install or configure.

[@AmlanjyotiSaikia]

Script kiddies everywhere.😅

[@AmlanjyotiSaikia]

I have a UI developer in my team, who is using a company issued Linux machine. I think it was running ubuntu. During a demo I asked the guy to open the source directory on the terminal for something....he proceeded to open the gui file explorer, navigating to the dir and clicking on open in terminal. I didn't think much of it at that time. Later I asked him to move to the parent dir and the guy does the same gui maneuver instead of cd .. . It then dawned on me that most of the young folks are absolute idiots when it comes to actually knowing how the basics of how the computer works. FML

[@steemium]

What do you guys recommend for a new x86 laptop CPU, now or in next few months?

[@Aranimda]

As long as it does not run Wintel or 100% compatible, it's not a PC.

[@MozartificeR]

What happens if intel buy a version of ARM instruction set. And they use x86 for their p cores, and Arm instruction set for their E cores? And you end up with two instruction sets on the one chip.

[@shanent5793]

A bunch of silicon gets wasted

[@Abinyah]

2024-5 will see a lot of BUZZ about Windows on ARM. The reality will be VERY PAINFUL for both end users and developers. The fight will be around the middle-ware stack/proto/rosetta like compatibility layers. Fortunes will be made and lost here.

[@MyEconomics101]

What is the leg the Wintel Monopoly stands on, if it is not performance? It is software compatibility. x86, and Intel specifically can afford to be generationally be 10-20% slower, if there is no credible alternative. Thus, if ARM CPU vendors really want to make a dent in the Wintel Monopoly, they have to subsidise software development itself (like Open Source alternatives), and the translation layer as a bridging gap.

[@xymaryai8283]

Qualcomm have stagnated so hard, if they managed to keep up their perfomance gains and make a more sensible naming scheme they could have been ahead of x64 by now, but after the 888 they slowed down, the XR2 is so old now, there was such a big gap. so i didn't stay on the flagship train, i went down to a 600 series and i don't regret it.

[@3ractnodi]

My thinking is that if node sizes stops shrinking, the lowest hanging fruit for better performance is the model Apple is following. Right now the biggest benefit to ditching x86 is performance per watt, but at some point the benefits to power/energy efficiency are the same benefits to power/increased performance. That and tighter integration between the software and hardware that Apple is notorious for forcing at the cost of backwards compatibility. I'm glad there's competition there to keep Apple from being able to get complacent, and pick up the torch if they squander their lead.

[@platin2148]

Is apple even in the same sphere there bundled SoC is still weak in GPU performance compared to almost anything dedicated. And there speed comparison is a joke actually when one looks as they use the last intel mac..

[@3ractnodi]

@@platin2148 I don't see how anything you said here is related to the topic of roadmaps after node size stops shrinking.

[@WSS_the_OG]

It's sort of weird to think about how long ARM has been around. Apple tried the RISC thing so many years ago, and bizarrely went back to x86 Intel, only to go back to ARM again. The more things change, the more they stay the same. I really dislike Apple, perhaps almost as much as I dislike Microsoft. I'd call it a draw. Nvidia, in spite of the fact I keep buying their GPUs like a hypocrite, make me feel sick the more I learn about them. Thank goodness they didn't acquire ARM. And to further underscore my hypocrisy, and perhaps I may give a convulsion or two to Apple fanboys before they go back into "defend" mode," I have to admit that as much as I hate Apple, their laptops are truly fantastic things. They perform decently, but the killer trait is battery life and lack of heat when using them. You can't deny that's a fantastic thing. I'll be a very happy person when I can have that kind of power efficiency without having to buy an Apple product. I really hope it's the way of the future. I have no nostalgic affection for X86.

[@johnwayne-kd1pn]

I agree arm is trouble for Intel and a threat also for the general computing market, in particular for laptops, but potentially also for servers and desktops. However, Arm only runs useless crap, while x86 runs all the important software in the world, and developers are unlikely to move to Arm anytime soon. Not only that, but Arm is pretty much incompatible even with itself, which means that if Intel sees Arm as a threat in the future, they could add any new architecture to their line (ex.risc-v) and it would create no further incompatibility than Arm does, or another new less changing (backward/forward) compatible (like x86) architecture. Or perhaps even an x86-risc, or some x86 mostly compatible lesser instruction architecture of some sort. Perhaps we can sum it like this (?): x86 is for general purpose computing, Arm is not. Arm is almost device and purpose specific.

[@AndersHass]

You did have a video about RISC-V’s potential to break through so it is possible Intel just rather pivot towards that than ARM, lol. Otherwise they could also drop the chip design business and go all in on manufacturering. I do wonder if Microsoft would dare to use Nvidia with ARM CPUs on a new Xbox generation, which could also be a hard blow to x86 with all the games being made for ARM.

[@florinszasz5506]

well and if sony goes for x86 ? amd has partnership with microsoft and apple i dont think that will happen.

[@AndersHass]

@@florinszasz5506 it will definitely help x86 keeping its staying power for gaming if Sony would still use it for their consoles (I doubt Sony will at any time soon loose their dominance). I don't think AMD has any partnership with Apple anymore given Apple make their own chips now. I guess their partnership is active while they still sell Macs with AMD GPUs but that probably won't last for much longer (if that is even an option anymore). I don't think there are any deep partnerships between Microsoft and AMD other than with the Xbox consoles currently. Otherwise it is just making Windows work which is fairly standard with various companies and not something special with AMD. It has been more special with Qualcomm since they were the only ones that could make ARM CPUs for Windows.

[@florinszasz5506]

@@AndersHass it could be true that they may not have any partnership with apple anymore, however i think with microsoft they have for server i might remember wrong but microsoft might use epyc cpus besides xbox and windows stuff.

[@baronvonlimbourgh1716]

Their manufacturing devision has been one of their main issues over the last 5 years. I doubt going all in on that would be wise.

[@AndersHass]

@@florinszasz5506 I don't think there is a deep server partnership. EPYC is just the best and thereby why it is bought. If Intel makes something better then that would be used.

[@Rahul_G.G.]

my only concern with the arm is that it will be backwards incompatible with practically everything except maybe a few apps backed by large companies,

[@realthunder6556]

Tbh, before the modern computer was built on X86, but these moves seem long overdue. Windows 11 already can run a lot of mobile applications based on arm, I believe Windows 12 will make sure you can run both X86 and arm applications regardless of your platform, even at a lower performance. Tbh, I fully expect Linux systems to start actually be really viable in this ARM X86 desktop environment,just because of Chrome OS and Android, plus it will just be plain better for optimization for specific platforms

[@cjadams7434]

basically someone needs a ‘rosetta-like” middleware like apple has

[@thejohnnerparty]

If the "Green Initiative" is real, the power saving world wide with an ARM architecture would use about 1/2 the power of the x86 architecture. That would be a big power saving - a lot less oil consumption.

[@almond-lol]

I think being able to use an arm-based CPU with x86 emulation under Windows core OS it would be really interesting to see without necessarily having x86 attached to the OS.

[@TY1979KA]

I remeber back in the day when I was studying it was prophecised by out professor that RISC processors will overtake cisc processors. I believe we are finally there, ... only a few decades later

[@frankwong9486]

Yes it is fast But question is how much and how windows arm performance after code conversion Good to see more options on market

[@Steamrick]

I wonder if M3 will actually be faster than M2 for heavy AI workloads like Stable Diffusion or running LLMs? Apple slashed memory bandwidth for a lot of their SKUs...

[@axe863]

Apple and Nvidia are both playing games with skimpflation.

[@D3LB0Y]

What ARM is a chip?

[@Nobe_Oddy]

I think I remember hearing something about an Intel roadmap that was going towards their chips not only being x86/x64 but ALSO ARM TOO... like all in one chip it could run both architectures but this is a ways off.... like in 2028.... I could be wrong and remembering something totally differently

[@utubekullanicisi]

With regards to Apple's performance figures, my guess would be that they're certainly using SPEC and taking the median of the performance improvements they get in all subtests. They might be throwing Geekbench in the mix too. When you translate the marketing speak, you will see that in the keynote they separately talked about both the low-level improvements via synthetic benchmarks (the opaque "10% faster, 30% faster" claims) and improvements they get in apps at the OS level (saying the M3 is 40% faster in "Image filters and effects performance" or 2x faster "for scene rendering in Cinema4D"). It's not perfect, but trying to explain to the average consumer that te M3 Pro's memory bandwidth is 150GB/s and how that will affect performance compared to M3 Max which has 400GB/s would probably be a pointless exercise as well, so they just spice the marketing up instead. Personally I think Apple's performance figures generally match the actual performance about as well as marketing figures you get from Intel or AMD or any other company.

[@brianmcguigan4785]

I have been expecting RISC Processors to take over from CISC, for the last 35 years. It has only been held at bay, by the incessant node shrinks that has enabled x86 performance increases. Now that this is becoming more difficult, it looks like the dominance of x86, might be in sight.

[@chebrubin]

Unified memory is bogus. Where is the modularity? Where is the peripheral interface bus? Apple Silicon is a iPhone with more memory and higher voltage.

[@dneary]

Yup - Qualcomm, Apple, and Nvidia Arm64 SOCs will displace x86 from laptop and tablet, and Ampere high core count CPUs and CSP's home-built silicon like Graviton will displace x86 in cloud and server.

[@axe863]

When ARM has highly optimized libraries using NEON more efficiently maybe x86 is out of luck. The amount of libraries highly optimized on x86 is a stumbling block for ARM that will buy it several more years ... really maturing a neuromorphic structure is going to be more of a game changer

[@ennio5763]

I don't see how having strong competition coming your way is a "big mistake". Intel did not invited this competition. It's coming no matter what. Pat's answer to journalists is essentially the only one he could make publicly. That's not a mistake.

[@bobnelsonfr]

The low TDP version should be great in a 4x4 mini-pc.

[@piotrd.4850]

People just don't get one thing. ARM based computer is AN APPLIANCE - usually, with single vendor supported OS, sometimes even locked, with little to none ability to upgrade or fix. Nothing wrong with ARM, that's just how things turned up. X86 PC has been about choice (Win/Lin even pirated macOS and niche stuff). And all this supposedly amazing hardware is massively nerfed by SOFTWARE.

[@davibelo]

I think a lot depends of windows on ARM developing...

[@JonMasters]

Love your videos - this Arm thing might end up being not so insignificant after all 😉

[@kborak]

yeah have fun with that. I want things I can upgrade and fix on my own. I hate the idea of SOC's.

[@notisln]

Don't forget that Intel's CPUs are still using 10nm nodes. If they were 3nm like the competition they could also be 3x more power efficient.

[@iikatinggangsengii2471]

well see 7nm intel ultra in abt a month

[@gaffo7836]

Exactly, and the only reason they have "effeciency" core as well, last gen node makes for more power and more heat. Their P-cores are very good, equal and maybe a little better than Zen4 ones at equal clock (not power of course - hence their process tech being behind, and why they add those "efficiency" (skylake) cores Intel really can't do anything (what they've done has been smart - add eff cores so they can tie AMD in performance (and ignore power use since that is not winnable) - without the eff cores, AMD would be beating Intel soundly on performance, because the power draw of the P-cores alone limits thier number, due to old process node tech compared to TSMC. AMD when through the same thing 10 yrs ago - where their node was behind Intels, but back then AMD CPU design - Bulldozer - was sht and so AMD lost in BOTH power and performance!!!!!! If you go back 10 more, pentium 4 vs Athon, that was somewhere in between, AMD had the better CPU (Athlon), but Intel had the better node (so could clock the inferior Pentium4 to the moon to match the Athlon in performance). anyway I'm rambling now - just replied to say I agree with you - lol.

[@notisln]

Wow, giant reply. Also Apple and Qualcomm claim that they offer similar or better performance while using 3x less power even though they are using 3nm process nodes which could mean that ARM isn't that much more efficient compared to x86. And by the way, 12th 13th 14th gens support both DDR4 and DDR5 which takes up physical space inside the chip. If they reserved it all for DDR5, like AMD , they could have used that extra space for more or bigger cores. Not sure how it works, exactly, to be honest. I'm still hopeful for Intel though.

[@InnocentiusLacrimosa]

You do not understand process nodes at all :-) That is hilarious. Tell me, what the the transistor density at these different manufacturer based "nanometer" nodes you are talking about? Do you even know how they decide what "nanometer" to call a node? It is not connected to transistor gates anymore like it used to be DECADES ago.

[@notisln]

When I said "efficiency" I referred to power consumption, not performance. From what I understand, the nanometers refer to the diameter of the transistors or the little "wires" that connect everything inside a core. Smaller nodes require less electricity in order to work which decreases power consumption. Is this wrong? If the exact same architecture is shrunk in half, wouldn't it consume half power?

[@vfn4i83]

The new huawei celphone, the mate60 has a desktop mode, a PC and mobile phone in one package and Linux size on a RISC CPU.

[@nadtz]

Intel did the same exact thing with AMD and their 'glued together chips' thing back when. One would think Intel would have learned from the encroachment AMD has been making and losing Apple but I guess not. With the recent ARM boom I'm just glad the Nvidia sale fell through.

[@Traumatree]

So AMD will probably come out with another AMDARM86 that will be able to run softwares from both architectures at the same time. That would be huge! And Microsoft has seen the light and is also distancing itself from Intel.

[@christopherstaples6758]

I can only hope so :P then I can use the same Vm's on all machines

[@vendacious]

How do you reckon the Ampere Altra Family rates vs the Qualcomm and M3 CPUs? After all, they already have a 128-core CPU you can buy right now, which can be used with an RTX 4090 dGPU, which destroys the M3 in every metric.Do you think MS might be using one of these to develop an ARM version of Windows? Because people like me who use Torch AI tools with CUDA/CUDNN binaries can't use Apple's weak integrated GPU, but I would definitely consider using a PC that can work with Nvidia cards.And thx for the great video!

[@marceelino]

I wish companies would not follow apple like idiots. I like that Intel is making it's way and it's not swayed.

[@cjadams7434]

cough usb cough firewire cough thunderbolt cough arm…..your welcome!

[@leorickpccenter]

Intel is not stupid. They're not underestimating the ARM threat.

[@dark_shinryu]

well suppose we need to wait until the real product is on market not just company claims and also I'm already sick with test bench software so we need to test with real productivity software. other think also is software availability with the arm base, like the M chip first release.

[@hakanyucel1639]

it's not first time and not last time !

[@te0nani]

Selling off Xscale is one of the mistakes on pair with Itanium.

[@thecooletompie]

While ARM (or another risc based instruction set such as risc V) will probably come for the desktop sometime I wish this video was a bit more self critical. Almost 4 or 5 years ago you predicted the same would happen instead we got one or two surface laptops with an ARM processors and nothing happened. Just because qualcomm announced a new shiny desktop processor does not mean it will result in success just look at the SQ1, SQ2, SQ3 failing to take any significant market share.

[@coolwinder]

I am noticing this corporate overloads trying to hinder other companies developments by any means in order for dinosaur Intel to continue working and artificially being competent.

[@SignalChange]

I wonder if qualcomm will put this chip onto a nice itx motherboard, would be a nice little system build base.

[@No-jb6fy]

I absolutely don't get why Apple doesn't cut their outrageous markup on memory and SSD and win. Why the premium? I'm not very much into Apple products and I don't care for ecosystems but I would maybe buy an Imac if the basic model included a 2TB SSD with 16 GB ram. So competition will be good indeed. Maybe Apple will regret catering only to the elite and lifestyle workflows.

[@christopherstaples6758]

I wish Proxmox would pull their finger out and officially support ARM , most of my cloud servers are ARM , and rebuilding custom multi-platform versions from source on every release is tiring

[@vzuzukin]

Where is Jim Keller working now?

[@mattbosley3531]

Qualcomm still has to work out their legal issues with ARM before there's going to be much uptake of the Snapdragon X Elite.

[@vigamortezadventures7972]

yeah true if intel dose not pivot in the next 3-5 years could potentially miss the arm train .. then again with all the prefabs being built and photonic computing underway could lead to to a different future .

[@giserson2]

Windows on ARM has had a pretty rough start and I don't see that changing much anytime soon. The reason why Apple can transistion from one architecture to another is because they do just that, by dropping support for their older architecture they force their developers to move development to the new architecture. Microsoft can't do that since Intel and AMD would sue them if they tried. This means that developers will keep developing with x86 in mind as it's the default and has the most market share and let performance heavy emulation take care of the ARM side. I don't think most Windows devlopers will bother with ARM unless it's very easy for them to make their app work on it or ARM takes significant enough market share, something that will take a lot of time at best given the lack of support. I really think that the best case scenario for ARM on Windows is for it to slowly gain adoption from clueless consumers that only use MS Office and Google Chrome who won't notice that other applications run awfully slow. I don't think Intel has anything to worry about in the short term, but long term things could look really bad for them if ARM gains market share.

[@ulamss5]

I really hope this was just Pat's CEO act, and that they already secretly have been full blast developing ARM solutions.

[@plamen5358]

The points about the bloated architecture are kind of irrelevant nowadays. After the decoder it doesn't really matter and all cores are more or less the same (big cores = more ALUs, more cache, etc) + Rosetta2 has hardware acceleration for x86, so Apple has the bloat as well. The sad truth is that Apple and TSMC are always 2+ nodes ahead of Intel and AMD is at least one, no design can save you from that. Nvidia got 70% uplift from 8nm to 4nm for less power and sillicon

[@Steel0079]

You're forgetting about the army of shithead bean counters in Intel's marketing team that are good at giving incentives to OEMs to use the Intel's inferior power hungry space heater silicon over AMD's more efficient silicon. This isn't gonna change anytime soon.

[@Teluric2]

Amd cpu dies faster

[@MARKXHWANG]

He has to say this right, otherwise their stock will get crashed