CEO Pat Gelsinger retired from the company after a distinguished 40-plus-year career and has stepped down from the board of directors, effective Dec. 1, 2024.
and
The board has formed a search committee and will work diligently and expeditiously to find a permanent successor to Gelsinger.
Wow, this is a really bad look for Intel. Gelsinger stepping down without Intel having a replacement! I always wonder when it doesn’t say why a CEO is stepping down suddenly without warning.
It’s notable that the announcement says nothing about Gelsinger having finished the part of the task he started on. It looks like they’ve lost confidence in Gelsinger (speculation). If that’s true, it also means they’ve probably lost confidence in the entire rescue plan he started on?
This is a huge bombshell, and not very elegantly executed IMO. Not just effective immediately, but effective YESTERDAY!?
Tbh it’s not 100% his fault the engineering competence began to visibly crumble under his leadership, but at the same time he absolutely stayed the course that his predecessors chose, which is what got them here in the first place. So yeah, he deserves to be excoriated for this stuff, but so do his predecessors.
he absolutely stayed the course that his predecessors chose,
Yes that part was always a bit confusing to me, because I couldn’t really see anything new in his strategy, except he was doing it harder. But isn’t that what it takes when you fall behind?
As much as I hate Gelsinger’s pompous bragging style, it’s hard to see what else they could do?What they could have done is to try to reverse the hollowing out of their engineering divisions, and give them more agency and control in leadership. Finance types trying to min/max the P/E ratio is what got them where they are. Serious tech companies that do REAL engineering can’t really follow the norms that Wall Street loves these days and expect to remain technically cutting-edge.
Engineers are not really plug-and-play. Institutional expertise is a real and meaningful thing. They got here because their leadership has ignored those facts for at least a couple decades now.
But think of the KPI’s and all other three letter acronyms!
Stellantis CEO just abruptly quit yesterday or day before too.
True, Stellantis has done extremely poorly the past couple of years IMO. They have several brands that usually feature in the top 10 most sold car models here (Denmark). But currently they have zero models even in top 20. Their EV cars are underwhelming generally offered with too small batteries, and are almost complete failures in the market. There have been some seriously wrong administrative decisions at the top.
Some of the same can possibly be said about Intel, except Intel was already in trouble when Gelsinger took over, and he is still working on the plan he set out to execute. Switching him out now looks really bad. Of course it may be they have to, disregarding how it looks.
With Stellantis it seems more the logical thing to do. Because Stellantis is bleeding, and losing market share fast.
Edit:
Huge difference with Stellantis is that they are quite open about the performance of the company has been poor, and that’s the reason he retires abruptly.It might be bad but their stocks went up significantly today…
Intel’s stock, not Stellantis they are down bad.
Wow, so an unknown is better than Gelsinger!!
Isn’t stellantis Chrysler? Listen I’m not a betting woman, but I feel like betting against Chrysler has been safe for most of my life, no matter how much I loved the 300M I inherited as a teenager.
Yes Chrysler is part of Stellantis, and once upon a time, way way back in what has since been called the 80’s, Chrysler was near bankrupt, but a savior came to Chrysler with the name Lee Iacocca. The mastermind behind Ford Mustang. And he came to Chrysler and saw all that was bad and fixed it. He undertook to finish a bold new type of car in the Dodge Caravan, which became hugely successful and saved Chrysler. Chrysler went on to become so successful they were even able to buy up other brands like AMC that also owned Jeep.
Ah well, as a European I know little about Chrysler today, but I have fond memories of once admiring mostly everything American, and Lee Iacocca and Jack Tramiel are probably the two business leaders I respect the most of all time.
Sorry to hear Chrysler is now considered safe to bet against. But sadly Stellantis has been shit for some years now.
Stellantis has many traditionally popular European brands, like Citroen, Peugeot, FIAT, Opel (Vauxhall in UK), Alpha Romeo and Lancia. And AFAIK all the brands are doing poorly.Yeah Chrysler shit the bed in the 00s and has more or less just been there for a while now
Yesterday was Sunday and the beginning of the month
Which means it was kind of effective Friday, either way it doesn’t change that it is very sudden.
If this was done properly, it should have been announced Friday at the latest.They would not want to break the stock game for their friends
What was I thinking? 😋
But their stock is actually up a bit today!
Maybe Lisa Su goes to Intel 🙃
Maybe intel gets absorbed into amd after Lisa Su becomes ceo.
Every time Intel makes a release when AMD is about to release their own competitive product, Lisa is seen sitting in a big comfy chair, next to a fireplace, calmly sipping tea at them.
I don’t watch LTT anymore, but they do have my favorite example of this: https://www.youtube.com/watch?v=vuaiqcjf0bs
tl;dw: Intel did a last minute reschedule of the embargo on their new HEDT lineup to lift hours before the new AMD Threadripper chips did the same. Those chips could barely keep up with AMD’s current offerings, and everyone involved knew that the new AMD stuff was going to crush it (which it did). Intel bumps the timeline so reviews have to go out (because reviewers have to get them clicks) without being compared to what’s actually going to be sitting on shelves alongside it. Linus sees what they’re doing and absolutely rips into them at the start of the video.
IDK if that’s meant as a joke, but I don’t see a single reason why she would do that. She is doing very well at AMD, and the pay is better at AMD.
Intel is receiving massive subsidies. If you think about it it is miraculous they managed to lose so hard to AMD with all those subsidies.
Absolutely, AMD was able to make Ryzen on the brink of bankruptcy, I fully expected Intel to make a comeback, with all the resources at their disposal.
But instead it’s been a long string of failures and at most half successes since 2016.
I have a bit of AMD stock, but still I don’t really want to see Intel fail.Aren’t those subsidies for building new fabs? They aren’t able to use those funds for general operations.
Well, he tried doing nothing, now he’s all out of options?
Stock didn’t sink so at least investors and Wall Street think they’re headed in the right direction.
The biggest problem is that any change that would come from an engineering effort is going to take many many years to even have a shot at changing anything. Speeds can’t really get much higher and they can’t seem to crack making stuff smaller. There are limits to making stuff bigger.
Their video card division are essentially making $5 Walmart rotisserie chicken and $1.50 Costco hot dogs. They’re not fantastic, but they’re not bad and they’re extremely cheap.
They needed to make the next big thing three or four years ago to have it on the plate by now, assuming they don’t have anything viable in their skunk works at the moment that’s a very big ship to turn around.
So even if someone else walks in, what do they do? Fire sale inventory, put a bunch of dreamer engineers in places, hire a bunch of rock stars. Produce a new unicorn after operating for about 5 years during losses and a possible economic downturn.
I think it’s looking pretty grim even with the subsidies and a bunch of people who know what they’re doing.
I think it’s looking pretty grim
Absolutely, but for some reason Intel has a history of failing in new areas. Their attempt with Itanium for high end was really bad, their attempt at RISC which mostly ended up in SCSI controllers was a failure too. Their failure with Atom not being competitive against Arm. Their attempts at compute for data-center has failed for decades against Nvidia, it’s not something that just happened recently. And they tried in the 90’s with a GPU that was embarrassingly bad and failed too.
They actually failed against AMD Athlon too, but back then, they controlled the market, and managed to keep AMD mostly out of the market.
When the Intel 80386 came out it was actually slower than the 80286!, When Pentium came out, it was slower than i486. When Pentium 4 came out, it was not nearly as efficient as Pentium 3. Intel has a long history of sub par products. Typically every second design by Intel had much worse IPC, so much so that it was barely compensated by the higher clocks of better production process. So in principle every second Intel generation was a bit like the AMD Bulldozer, but where for AMD 1 mistake almost crashed the company, Intel managed to keep profiting even from sub par products.So it’s not really a recent problem, Intel has a long history of intermittently not being a very strong competitor or very good at designing new products and innovating. And now they’ve lost the throne even on X86! Because AMD beat the crap out of them, with chiplets, despite the per core speed of the original Ryzen was a bit lower than what Intel had.
What kept Intel afloat and hugely profitable when their designs were inferior, was that they were always ahead on the production process, that was until around 2016. Where Intel lost the lead, because their 10nm process never really worked and had multiple years of delays.
Still Intel back then always managed to come back like they did with Core2, and the brand and the X86 monopoly was enough to keep Intel very profitable, even through major strategic failures like Itanium.
Pat came back to intel to help turn them around. I doubt he was planning to stay there.
I am pretty sure the plan wasn’t to quit at 63, at what looks like halfway through his plan.
Losing 16 billion dollars totally has nothing to do with this at all. Its time for Pat to pursue some creative hobbies at home, enjoy his retirement, and be with his family. There are no American troops in Baghdad, everything is fine, Intel is fine, and will soon be back to doing great things. Just ask Userbenchmark, Intel products are the best in class and highly sought after. nVidia has no real advantage in the AI race, and Intel is just dominating.
That 16 billion is just a brief hiccup, company is totally about to do great.
Maybe now they can forget all the expensive chipmaking and get back to their core business of stock buybacks.
That’s probably the real reason. He was going to invest all that money instead of doing more stock buybacks. What an idiot!
With the CHIPS Act funding they can’t do stock buybacks.
It just frees up other money for buybacks that they would have had to waste on R&D.
the fact that you have to explain to people that money is fungible…
Well, at least one stupid fucker disagrees with that idea…
This simultaneously made me laugh and sad. Sad because I have heavy investments in Intel.
Well their stock price just jumped up after this announcement.
To less than what it was a month ago.
And more than what it will be next month.
This is the real lesson here and US taxpayer has to now pay for Intel CapEx.
These parasites are able to make “business” decisions that impact all of us with zero accountability.
Clown capitalism and no lessons learned.
Disgusting parasites are enabled here IMHO
The sentiment was not bad. TSMC is a shining example of how fab subsidies can be a good idea, and Intel’s fabs going under is bad and basically irreplaceable. Like… I am still happy with my tax dollars taking the risk, and Intel was clearly trying to right the ship when CHIPS was conceived.
But theres clearly rot in Intel. Thats a big difference I guess, as TSMC was built from the ground up (in a time where that was possible) while Intel is already weighed down with its sins.
If we give them billions of dollars, then why are we not taking equity position?
You do understand that shareholders were transferred 100billion dollars over last 20 years?
Why is us taxpayers bailing out their position?
Why Intel needs cash, why doesn’t intel issue shares and gut the shareholder?
Eitherway, I am happy that you are satisfied with this transfer. I am not.
Where do you get the 100 billion USD amount from?
AFAIK Intel has received less than $10 billion.sharebuy back is cash transfer to shareholders.
CHIPS act allocated about 34b to be transferred to the corporation.
sharebuy back is cash transfer to shareholders.
That is simply not true, and Intel has only gotten less than $10 billion from CHIPS.
https://www.intel.com/content/www/us/en/newsroom/news/us-chips-act-intel-direct-funding.html
AFAIK they’ve actually only received 1 billion of that.
While we’re at it, let’s go back to 10nm chips too. That’s Intel’s bread and butter. Phones get bigger every year. Why not transistors too?
userbenchmark should not be trusted.
My comment was dripping with sarcasm, I refer to them as “Loserbenchmark” most anytime they come up. Complete toolbag shill assholes over there, lol.
Are they still posting salt over AMD cpus spanking Intel haha
That shit started in 2017 and it got progressively more pathetic.
AMD wasting money on marketing… sure buddy, cope
So what IS their strategy now?
Some of Pat’s initiatives were good (stay the course with Xe and fabs, which take a long time to pan out), but they kept delaying everything!
Yet Intel is kind of screwed without good graphics or ML IP.
If they spin off the fabs, I feel like they are really screwed, as they will be left with nothing but shrinking businesses and no multi year efforts to get out of it.
Like… Even theoretically, I dont know what I would do to right Intel as CEO unless they can fix whatever is causing consistent delays, and clearly thats not happening. What is their path?
If they spin off the fabs, I feel like they are really screwed
One of the stipulations of the $8B in CHIPS Act funding that they just recieved last month was that they not separate their fabrication business from the parent company. That’s unlikely to happen now unless it gets separated at a bankruptcy auction.
Flail until the light leaves their eyes.
Mmm, I hope the fab survives on their own…
unless they can fix whatever is causing consistent delays
Yup, that’s their #1 goal right now. If I were CEO, I’d cut/sell any part of the business that doesn’t directly support CPU and GPU sales, which is basically what Intel is doing. My priorities would be:
- rescue server CPU business - this is their main money printing machine, and while they may lose to ARM in the future, they need a cash cow in the medium term
- get a competent server-oriented GPU product out - they’re late to the game, but they can bundle them w/ their server CPU contracts to get some market share; overwhelm these corporate customers with first class driver support
- get something to compete w/ Apple’s M1 - this means super low-power CPU that can scale to gaming workloads, and capable-enough graphics (something a bit better than AMD’s APUs); sell this near cost to keep a foot in the door in the mobile space
- sell domestic fab capacity - now is the time to get Sony and Microsoft on board with their next gen consoles, and it might not be too late for Nintendo
I would essentially ignore desktop workloads and solve workstation workloads w/ server chips. To me, those sound like the highest margin businesses that they could potentially still capture, and at least 1 & 2 are a bit less sensitive to being behind on their fab process (corporate contracts respond pretty well to bundle discounts).
This probably wouldn’t work though, especially since I’m an outside observer with zero industry experience. But I think a good CEO would do something along those lines, which seems to be what Pat Gelsinger was going for as well.
They tried all this:
-
Not sure about this, but it appears AMD is simply out designing them. Some concepts like the many-little-core SKUs seem promising, but ultimately the EPYC MCM design is fundamentally very good here. And… Delays. Delays are killing them here.
-
This was Xe-HPC, the Falcon Shores APU, the Falcon Shores GPU, Gaudi… They’re so late to everything it didn’t work and it appears they’ve basically given up on the whole line besides consumer inference products, which is also kinda meager atm. And even AMD is mightily struggling here, with hardware that is straight up bigger/faster than Nvidia.
-
An M Pro esque chip was also in the plans, but seemingly canceled? Or way behind AMD, at least. And OEMs have repeatedly rejected their GPU heavy designs like Broadwell eDRAM and the AMD collab chip, as they’re kinda idiots and Intel is at their mercy. And the laptop chips they are selling now are basically their best shot at an “M” chip and arguably one of their most decent products.
-
They tried, and no one bit. Who can blame them, given Intel’s history of delays?
Its all the delays! Its destroying them.
I mean I’d guess I’d press on with Xe if I were CEO, but if they can’t launch anything on time what does it matter?
And even AMD is mightily struggling here, with hardware that is straight up bigger/faster than Nvidia.
The problem has always been software support. If Intel wants a piece of the AI pie, they need fantastic software support. AMD has always been a bit lackluster here, whereas Intel has done a pretty decent job in the past (esp. on Linux, their drivers rock), so they would need to double down if they truly want to get after it.
Intel is at their mercy
Then Intel should make their own laptops and prove the model.
it appears AMD is simply out designing them
I don’t think so, they’re just better at improving margins. Intel was able to keep up for a while despite not keeping up w/ the fabs, so I think their designs are absolutely fine. They’re not cheap to manufacture like AMD’s are, but they are really good.
Its all the delays! Its destroying them.
Exactly. They need to double down on something instead of faffing about with different ideas. Their money maker is server chips, so that should be top priority. Their next biggest is probably laptops, and AMD is getting massive inroads here due to Intel sucking on their fabs. Catching up on servers should be easier than catching up on laptops, because corps can be bought w/ value, whereas the CPU makes up a much smaller portion of overall laptop price, so they have less leeway here.
But yeah, they need to fix the delays. Get the fabs on track and get steady CPU production in their core markets. And do that without giving up on GPUs, because that needs to be in the future plans since people are generally moving away from CPUs to GPUs for compute.
Everything else Intel does can be scrapped for better software. Really good software can do a lot to make up for lagging hardware, so make sure that is top notch while you’re fixing the hardware delivery.
The problem has always been software support. If Intel wants a piece of the AI pie, they need fantastic software support. AMD has always been a bit lackluster here, whereas Intel has done a pretty decent job in the past (esp. on Linux, their drivers rock), so they would need to double down if they truly want to get after it.
Actually AMD is pretty okay for running LLMs and other ML workloads. Many libraries now explicitly target rocm, you can just plop down vllm or the llama.cpp server and have it work with big models out of the box. There are some major issues (like flash-attention), but its quite usable.
Intel though? Their software is a mess. You have to jump throigh all sorts of hoops, use ancient builds of pytorch, use their own quantizations and such to get anything working, fix Python errors, and forget about batched enterprise backends like vllm. And this is just their IGPs and Arc, forget trying to use the vaunted NPUs for anything.
This could change if they actually had a cheap 48GB GPU (or a big APU) for AI devs to target… But they don’t. And no one is renting Gaudi to build in support because its not even availible anywhere.
EDIT: oh, and one weird thing is the volume of Intel software support is high. Like they have all sorts of cool libraries, they make contributions to open projects… But its all disjointed and fragmented. Like theres no leadership or unified push, just random efforts flailing around.
Exactly.
Intel is shooting itself in the foot by going halfway. If they want to compete in the AI space, they need to go all-in w/ a solid software and hardware combo. But they don’t.
They have the capability, they’re just not focused. A good CEO should be able to provide that focus. Maybe they should hire Lisa Su. 😆
Speaking as an holder of AMD stock since ot was $8, and an all AMD CPU user, IMO Lisa Su is either an absolute idiot or colliding with her cousin, the CEO of Nvidia.
All they had to do was lift vram restrictions on consumer GPUs (so their OEMs could double the VRAM up) and sick like four engineers on bugs blocking the AI space, and they would be dominating the AI space and eating Nvidia’s pie…
And they didn’t. Like, its two phonecalls, thats it.
Intel had monumental problems it has to solve and struggles, but AMD has tiny ones they inexplicably ignore. Its mind boggling.
I work in CV and I have to agree that AMD is kind of OK-ish at best there. The core DL libraries like torch will play nice with ROCm, but you don’t have to look far to find third party libraries explicitly designed around CUDA or NVIDIA hardware in general. Some examples are the super popular OpenMMLab/mmcv framework, tiny-cuda-nn and nerfstudio for NeRFs, and Gaussian splatting. You could probably get these to work on ROCm with HIP but it’s a lot more of a hassle than configuring them on CUDA.
- An M Pro esque chip was also in the plans, but seemingly canceled? Or way behind AMD, at least. And OEMs have repeatedly rejected their GPU heavy designs like Broadwell eDRAM and the AMD collab chip, as they’re kinda idiots and Intel is at their mercy. And the laptop chips they are selling now are basically their best shot at an “M” chip and arguably one of their most decent products.
Wasn’t Lunar Lake supposed to be this?
It’s 128 bit. I’d say it needs a bigger GPU and a 256 bit bus to be “M2 Pro” class.
-
Almost certainly too late to get Nintendo. According to Nvidia insiders, their work for the Switch followup SoC has been done for ages, and they’re a bit puzzled that Nintendo hasn’t released it yet. The reason seems to be unfavorable exchange rates between the Yen and USD, and Nintendo’s board of directors has worked themselves into analysis paralysis over the “best” time to release.
Heh this is so Nintendo.
They must be pulling their hair out trying to make predictions now.
If I were CEO, I’d cut/sell any part of the business that doesn’t directly support CPU and GPU sales, which is basically what Intel is doing.
That’s pretty much what they did. They sold off most of the “other” stuff, like their modem division, shut down their SSD division, sold part of Mobileye shares in the IPO, and reportedly Intel is looking to sell part of Altera, their FPGA division.
Yeah, and I generally agree with Gelsinger’s direction. I’m interested in the reason for him retiring, as well as who is likely to replace him.
It would be really funny though if Intel tanks and AMD buys their fabs from them.
So what IS their strategy now?
I think they need to bet the company on regaining their previous lead in actual cutting edge fabrication of semiconductors.
TSMC basically prints money, but the next stage is a new paradigm where TSMC doesn’t necessarily have a built-in advantage. Samsung and Intel are gunning for that top spot with their own technologies in actually manufacturing and packaging chips, hoping to leapfrog TSMC as the industry tries to scale up mass production of chips using backside power and gate all around FETs (GAAFETs).
If Intel 18A doesn’t succeed, the company is done.
Damn. I actually thought he might turn things around back when he was brought in. Their engineers have let them down, how did they fall so far behind after being so far ahead just 15 years ago?
They released the same 4 core CPUs for like 6 years in a row
Why innovate when line go up?
Profits go up if you don’t invest in your company but instead pay your shareholders. But at some point you atrophied your R&D so much you have fallen behind and then the question becomes, do you bleed it dry and sell it for parts, or revitalize by investing. And if revitalizing is still viable.
It’s almost like maximizing profits over everything isn’t a great idea.
If you are a shareholder that gets the payout and make a profit… it sure is. For everyone else… Who are we kidding. Fuck everyone else.
And only if you’re an old shareholder that wants to cash out.
Or a daytrader.
Small time, long term holders that just want the company to be sustainable… Pfft.
Yup, I was being sarcastic, but I appreciate the response
revitalize by investing
they are not investing… they need to raise capital to do that. currently they are praying US and german taxpayer will fund their CapEx.
Looks like US took note and this clown got sacked, which is good but enough.
Between boeing and intel… you corpo practices are being exposed for the pathetic extraction racket it is. at least boeing went out and raised 20b via share offering and gutted their shareholders.
Intel stock prolly too gutted to do that now tho lol
pathetic.
It was 8 years ago. AMD was on the verge of bankruptcy and Intel had been propping them up for years so they wouldn’t have to deal with the government going after them for having a total monopoly. If Zen had been a failure AMD wouldn’t have survived. I figured Intel had advanced stuff in the pipeline that they were just sitting on waiting for AMD to force their hand (because they were dicks like that). I was wrong.
Hey man, they just renamed their newest chips to have a completely different confusing naming scheme! What more innovation do you want?
I think we all thought that tbh. Intel let their hubris get them and this is the result.
They don’t have innovation anymore, I don’t know what they’re doing and I don’t think they do either.
I wish AMD would catch up in the GPU side of things so it wasn’t such a monopoly with NVIDIA but I guess we’ll see, I mean they did knock Intel down eventually so who knows maybe it’s possible.
That would have gone truly horrible if AMD did go bankrupt, that would have been a really dark timeline for all of us.
Honestly, I hope AMD-s shift to focus on lower end cards is successful. It should be considering the xx60 series (and performance equivalent) cards make up like 50% of the entire consumer GPU hardware? At least I think it was around 50 the last time I tried to sum up all the percentages of the Steam hardware survey. There’s definitely a huge market they can tap if they can bang-per-buck outprice Nvidia (and I guess also Intel). Maybe even bring down the ridiculous pricing of modern GPU-s.
Even if their cards offer better value than Nvidia at the low end, AMD still have an uphill battle to get people to switch. The brand recognition they have is insane, and for some reason people value dlss and frame gen very highly (wether it works well for their card and game or not).
Steam hardware survey is questionable. There’s a lot of computer cafes in east Asia where people login to their Steam accounts and happen to hit the survey. Those machines are often running very low end cards like the Nvidia 1050. This is common enough that the results are heavily skewed.
You can check yourself. I’m pretty sure the “cafe cards” amount to around 3-8% of the lowest end cards depending on whether we consider 1650 and 1060 as cafe cards. Obviously also excluding integrated cards because those I didn’t consider in the first place. On the other hand the current gen and last gen low end cards (xx50 and xx60) make up 25-28% of the market.
Also I don’t understand why you’d want to exclude cafe’s from the potential market? It’s not like internet cafes don’t upgrade their hardware. When they do upgrade they’re definitely going with the low end cards.
The issue is that they’re being counted extra times because of multiple people logging into the same machine.
How much RAM do you imagine internet cafe machines use? 8gb? 16gb? 32gb?
They’ve been smart in continuing to invest in datacenter cards and investing in open compute tooling to support them. Nvidia is at the top of the world and has a long way to fall, so if they start restricting supply of datacenter GPUs or simply charging too much that leaves plenty of market for Intel and AMD both to feast on and build up healthy product stacks to eventually surpass Nvidia.
On the flipside Nvidia is smart to be diversifying right now. Their forays into GPU servers with custom ARM CPUs might become fruitful in the long term, plus their networking investments really allow them to build a unique and compelling datacenter package
It’s a story that’s been repeating for decades now. Company creates a new market with new useful tech, run by engineers passionate about the tech, experiences exceptional growth, becomes large corporation, much larger than any competition. Uses relative wealth to keep competition from catching up. Eventually saturates market to the point where market growth doesn’t finance the growing R&D expenses (which were tuned assuming previous rate of growth would just continue). At some point, profit increases start coming from business/marketing side of things more than engineering side, resulting in MBAs and marketers getting more promotions and eventually control of the company. Then tech stagnates because they don’t think investing in R&D is as worthwhile. Also aren’t able to prioritize what R&D is still happening effectively because they don’t really understand the tech as well as engineers. But they tread water and even increase profits because they dominate the market.
Until competition that is engineering focused (often also made up of former engineers from the dominant company) catches up or creates a new market that makes theirs start going obsolete. Suddenly trouble, then they either pivot to quietly supporting businesses that continue using their products, or gets in trouble with the law because of increasingly anticompetitive practices.
Xerox could have owned the PC market but thought they could continue being a household name sticking with copiers. IBM outsourced everything and people eventually realized they didn’t need IBM.
FoxconnFairchild had two groups of engineers leave and create Intel and AMD when they were dissatisfied with how management was running the company. And now Intel coasted while AMD floundered and was completely unprepared for TSMC and AMD to make large technical leaps and surpass them.Foxconn had two groups of engineers leave and create Intel and AMD when they were dissatisfied with how management was running the company.
You’re thinking of Fairchild, not Foxconn.
William Shockley led the team that invented the transistor while at Bell Labs, and then went on to move back to his home state of California to found his own company developing silicon transistors, ultimately resulting in the geographical area becoming known as Silicon Valley. Although a brilliant scientist and engineer, he was an abrasive manager, so 8 of his key researchers left the company to form Fairchild Semiconductor, a division of a camera and imaging company with close ties to military contracting.
The researchers at Fairchild developed the silicon integrated circuit (Texas Instruments developed the first integrated circuit with germanium, but it turns out that semiconductor material wasn’t good for scaling and hit a dead end early on), and grew the company into a powerhouse. Infighting between engineers and management (especially east coast based management dictating what the west coast lab was doing) and Fairchild’s policy of not sharing equity with employees, led Gordon Moore and Robert Noyce (who had been 2 of the 8 who left Shockley for Fairchild) to go and found Intel, poaching a talented young engineer named Andy Grove.
Intel originally focused on memory, but Grove recognized that the future value would be in processors, so they bet the company on that transition to logic chips, just in time for the computer memory market to get commoditized and for Japanese competition to crush the profit margins in that sector. By the 90’s, Intel became known as the dominant company in CPUs. Intel survived more than one generation on top because they knew when to pivot.
Ah right, thanks for the correction!
We’re raking in money, why spend on R&D?
Man I hope Battlemage is an actually profitable launch, or at least not a massive loss. Otherwise who knows if the next CEO will axe their GPU line. People liked to fearmonger them killing Arc before, with with a new change in management I can actually see that happening.
Its a lower midrange only launch like it appears to be, it will be extremely unprofitable. AMD may even eat large chunks of this market with the Strix Halo APU, which could be similar to the B570 with no need for a discrete GPU.
Theres actually a big and growing demand for ANY high VRAM GPU for the LLM crowd (that AMD is ignoring for inexplicable reasons beyond Strix Halo) but it appears Intel can’t even compete there. No 256 bit APU, their GPU is 192 bit so capped at like 24GB…
This is why I got a 4070 ti super because it has a 256 bit bus.
Eh actually the 4060 TI is way better for LLMs :P With Nvidia its all about VRAM capacity.
I only game and a larger bus is better for 4K.
Intel is totally missing the boat honestly. Their mobile i9 with the built-in GPU can share DDR5 with the video card.
You can put 96 gigs of RAM in a small form factor and load in a monster model. It’s not super fast, But it works, and it’s a lot faster than not offloading layers off the CPU.
They should be selling nuk sized PCs with built-in graphics and 128 gigs of the fastest RAM they can put on the boards.
IMO its not really “enough” until the bus is 256 bit. Thats when 32B-72B class models start to look even theoretically runnable at decent speeds.
he was getting 1.4 tokens on a 70B model. Not setting the world on fire, but enough to load and script against 70b
Also that is a very low context test. A longer context will bog it down, even setting aside the prompt processing time.
…On the other hand, you could probably squeeze a bit more running openvino instead of llama.cpp, so that is still respectable.
text test. A longer co
yeah, it’s definitely not good enough for user-facing work, but if I’m working on development for something like translations, being able to see the 70b output to compare it to other models, it’s super useful before I send it off to something that costs more money to run.
9/10 times, the bigger model isn’t significantly better for what I’m trying to do, but it’s really nice to confirm that.
I hope Intel gets their act together soon. We can’t have a monopoly on chips on the CPU or GPU space.
It’s baffling how fast they fell, since they had a monopoly for ~20 years.
They had untouchable market dominance from the mid 80’s through the mid 2010’s, so probably closer to 30 years.
AMD and Apple caught up on consumer PC processors, as the consumer PC market as a whole kinda started to fall behind tablets and phones as the preferred method of computing. Even in the data center, the importance of the CPU has lost ground to GPU and AI chips in the past 5 years, too. We’ll see how Intel protects its current position in the data center.
I’m personally excited about the actual engineering challenges that come next and think that all 3 big foundries have roughly equal probability of coming out on top in the next stage, as the transistors become more complex three dimensional structures, and as the companies try to deliver power from the back side of the wafer rather than the crowded front side.
Samsung and Intel have always struggled with manufacturing finFETs with the yields/performance of TSMC. Intel’s struggles to move on from 14nm led to some fun memes, but also reflected the fact that they hit a plateau they couldn’t get around. Samsung and Intel have been eager to get off of the finFET paradigm and tried to jump early to Gate All Around FETs (GAAFETs, which Samsung calls MBCFET and Intel calls RibbonFET), while TSMC sticks around on finFET for another generation.
Samsung switched to GAAFET for its 3nm node, which began production in 2022, but the reports are that it took a while to get yields up to an acceptable level. Intel introduced GAAFET in its 20A node, but basically abandoned it before commercial production and put all of its resources into 18A, which they last reported should be ready for mass production in the first half of 2025 and will be ready for external customers to start taping out their own designs.
Meanwhile, TSMC’s 3nm node is still all finFET. Basically the end of the line for this technology that catapulted TSMC way ahead of its peers. Its 2nm node will be the first TSMC node to use GAAFET, and they have quietly abandoned plans to introduce backside power in the generation after that, for their N2P. Their 1.6 nm node is going to have backside power, though. They’ll be the last to marker with these two technologies, but maybe they’re going to release a more polished process that still produces better results.
So you have the three competitors, with Samsung being the first to market, Intel likely being second, and TSMC being third, but with no guarantees that they’ll all solve the next generation challenges in the same amount of lead time. It’s a new season, and although past success does show some advantages and disadvantages that may still be there, none of it is a guarantee that the leader right now will remain a leader into the next few generations.
Packaging/interconnect tech is starting to be a big factor though, and TSMC is very strong in this area, no? They can lean on that.
Also its weird to even imagine Intel with big external customers…
Intel’s packaging doesn’t seem to be that far behind TSMC’s, just with different strengths and weaknesses, at least on the foundry side. On the design side they were slow to actually implement chiplet based design in the actual chips, compared to AMD who embraced it full force early on, and Apple who rely almost exclusively on System-in-a-Package designs (including their “ultra” line of M-series chips that are two massive Max chips stitched together) where memory and storage are all in one package.
The competition for CPUs can be AMD vs ARM vs RISC-V. It doesn’t have to be between two x86 giants.
That’s better, not necessarily for instruction set reasons, but because ARM and RISC-V are more open to multiple companies stepping in to produce chips.
Eh, a lot of big players have backed off from custom ARM CPU cores. So the question is how many even have the muscle to compete?
Double so for RISC-V.
Especially one just a small naval visit from China.
Damn, this guy is utterly fantastic at ruining huge tech firms.
Eh, he was handed a company in a bad strategic place and he did not fix it.
Lisa Su was in a similar position when she took over AMD, but she managed it. While I don’t want to put too much emphasis on the CEO alone, AMD’s turnaround is quite remarkable. They very easily could have collapsed at one point.
He was handed a company in a horrible strategic place and he did the right things to fix it. Reinvest in process technology mainly. Those investments do not bear fruit overnight. They take years. Whoever replaces him could basically be a stuffed suit and will probably have some success if only from his investments starting to pay off. It’s too bad he didn’t get a few more quarters to see it happen.
Nah, they’re stuck. The most recent 2xx series Intel chips are actually on a better TSMC fab than what AMD’s 9000 series chips are using, but you wouldn’t know it from almost any benchmark available. Their architecture is just bad, and a fab improvement can’t even save it.
All they need to do is hold out and survive until China invades Taiwan and the chip foundry game will change overnight. I bet they’ll even get free access to TSMC patents just to try to get the west back into the chip lead. They won’t be allowed to fail at that point.
Though I don’t see the consumer semiconductor industry thriving after that.
China will never invade Taiwan. Taiwan has a backchannel protection deal with the US and China knows it.
China’s too smart to ‘invade’ Taiwan. There will be no tanks and helicopters invading. China / CCP may be assholes but they are also fucking smart.
Look at Hong Kong. There were no tanks or helicopters. Just steadily increasing political control. More or less the entirety of HK protested for weeks/months. It did fuck all.
That will be what happens with Taiwan. It won’t be an invasion. It will be a gradual slide.
Right now, USA officially supports the ‘One China’ policy to appease China even though we want Taiwan to be independent. It’s let us keep huge trade with China (which the Chinese also want/need) while we depend (and NEED) Taiwan for a lot of tech manufacturing especially computer chips.
Thing is, China has no desire to be dependent on us. They want us dependent on them for manufacturing, but don’t want to need that business. That’s why China is doing aggressive R&D on pretty much every high tech area they depend on the West for, trying to ensure that everything China needs can be made in China from Chinese tech. To do that they need to be able to design and manufacture the latest computer chips, which they currently can’t. But they’re pouring billions into figuring it out.
If China takes over Taiwan, either openly or covertly, they get TSMC. And that gives them all the chipmaking tech they need.
Don’t expect tanks. Expect state sponsored industrial espionage at TSMC and their own suppliers. Then expect Chinese chipmakers to flood the market with top-line or near-top-line hardware at low prices, which US won’t embargo and thus we’ll get even more dependent on China.
Lol the biggest reason you’re wrong to make that comparison is that Hong Kong was never its won country. Hong Kong was a British colony and then a Chinese special administrative region (SAR) given a degree of administrative autonomy by the Chinese government voluntarily as part of a treaty with the British. The treaty expired and then China decided to change the rules for Hong Kong.
Taiwan meanwhile was the territory that the Republic of China (RoC aka Nationalist China) held on to when it lost the Chinese Civil War against the People’s Republic of China (PRC aka Communist China) who now control the mainland. The PRC never controlled Taiwan and the RoC government which rules there does not answer to the PRC nor has it ever. The PRC and its Communist Party can claim that Taiwan is a rogue province all they want but that’s a lie. Taiwan is not theirs it was and still is under the government of the ROC even if the ROC has lost the rest of its territory to the PRC since the Civil War and World War 2.
Hong Kong’s city government allowed China to take more direct control because it always answered to China since the British gave it to China. Meanwhile the ROC government in Taiwan has never answered to the PRC and it never will. Opposing the PRC is literally one of the main goals of that government and country and I don’t think there are any major politicians there who want to join the PRC willingly nor would amy such politician be popular there.
Long story short the ROC (Taiwan) and Hong Kong are not even remotely comparable and the former won’t just accept any attempted takeover by the communists.
That depends on Trump giving a shit about protecting anyone. I wouldn’t be surprised at any outcome.
Trump hates China so he would do it just to show that he’s opposing it.
I agree and my comment had obviously no nuance. I’m still dealing with VMware fallout in my professional life which is on Broadcom but still, this dude had control of another huge sinking ship previously…
Yea, he was CEO of VMware from 2012 to early 2021. All the issues VMware has now came from broadcom buying them which happened well after he left.
VMware had some pretty cool stuff in the pipeline related to DPUs that would’ve been killer in hypervisor networking but I’m pretty sure that’s out the window post-acquisition.
Honestly with how good kvm and qemu have been getting and the number of competitors building hypervisor off of open source virtualization technologies it was probably a ticking time bomb before it fell to cheaper, freer competition. This way we have a bad guy to blame and not just pure corporate hubris
A I!
