How are they going to afford an investment thats ~1/3rd the value of the company? Seems like one of those announcements that no one follows up on to keep them honest?
They wont its another marketing scam by the admin lol, promise shit now, don't deliver reap benefits till the public/admin forgets, you dont actually have to do shit you just have to say you will same as the last time
Much like when the administration announced an investment of 500 Billion USD in AI some months back, not a peep since. Actually shocked nobody here has even referenced that in relation to this, at least at the time of this writing
Much like that “initiative”, I imagine this won’t materialize either, they’re largely hoping government funds get steered toward them and by which point they’ll pocket everything they can while delivering a minimum of what was promised, if anything at all, and they won’t be properly held accountable for it.
Are you referring to the Stargate Project? I guess so, since there weren't any other 500 billion USD AI projects announced recently in the USA.
Construction of the first data center Abilene, Texas site is ongoing. Emily Chang visited the site over a month ago and got video footage [1]. So it's not nothing. Will they actually invest 500 billion in the end? I dunno; I guess we'll see. The main challenges seem to be the availability of electrical power and skilled workers, not demand, investment or ambition.
The overall impression I get of such projects is that politicians celebrate the funding victory, and then forget about it. Then the project's funds get eaten up by administrative costs over time. Ezra Klien [1] recently highlighted two such examples: high speed rail in CA, and rural broadband subsidies. Personally, I haven't seen much, if anything, actually spent out of Biden's infrastructure bill. The money from these programs just seems to...vanish. Allowing this to happen undermines the idea that the USG is a trustworthy spender of our money. I would personally prefer it if political success wasn't just grabbing more budget, but actually overseeing successful execution of public projects.
I've seen a lot of transportation infrastructure projects at least partially funded from the bill around me. I've charged at EV charging locations funded by NEVI, and I've charged at some stations built expecting NEVI funding that is now on hold. I've ridden on busses purchased by grants issued through the IIJA. I've seen the ground broken for building new bike trails around me based on grants from the IIJA, and they're getting close to done. The highways around me have current construction projects underway which were funded through the IIJA.
I see some of the things funded by the IIJA every day of my life.
And that's not including the projects which are still being worked on getting approved including projects to replace sewage lines in my area, reworking stormwater drainage systems, and additional flood control projects, all of which are planned to be funded through the IIJA and all of which I think my city could really use. 60% of the IIJA funding hasn't been spent yet.
I have fiber, as do the 9 counties around mine because of rural broadband investment. My sister moved to downstate Illinois, and most rural counties offer high speed Internet if not straight up fiber now.
Go to flyover States to see the impact of that money.
Are these instances actually funded by 2021 IIJA BEAD program?
I don’t think Illinois has actually awarded any of that funding to providers to build anything yet. It looks like the original schedule was to start awarding grants this summer after planning process from 2021-2025.
It doesn't matter how many locations they spread it over, they aren't spending that much money in under 20 years on chatbots and buggy code writers. Maybe if a massive unforeseen breakthrough comes and breaks us past the wall of diminishing returns we've hit.
>Much like when the administration announced an investment of 500 Billion USD in AI some months back
That was a project by Oracle to build some data centres, many of which had already broke ground and were under construction, and they pulled OpenAI and Softbank in to give it that AI veneer.
So it's stuff that is really happening, and was already happening. Similarly this TI plant was announced under the prior administration as a win of the chips act, and again it's being recycled to current.
This sort of shell game works for a while, but a couple of years the bill comes due.
I have no desire to defend the admin, but that was not what was announced at all. It was announced that OpenAI/Oracle/Softbank were investing $500B, which they are. It had nothing to do with government spending.
I find that the easiest way to suss out corporate political advertisements is to look for how they refer to the federal government. If they say "federal government" or "$NAME_OF_ACT" then there's usually at least an inkling of intent behind whatever they say next. If they say "$PRESIDENT administration" or "$PARTY_NAME," you can be certain you're reading a political advertisement.
Not the 51st State. It is more like our Colonial Metropolis. We just need to pay tribute and keep quiet about it. Occasionally the tribute must be paid with the blood of our young, in order to allow our colonial master to most effectively genocide their neighbors
Akshually I was referring to Qatar, but if you deny it then you're an antiqatarite, and you literally condone the destruction of Tatooine and its 6 million habitants.
Iran has been attacking Israel for 50 years through proxies. I don’t think it’s a clear “attack vs defense”. Also now is the time to deal with Iran and restore a sane government before they have a chance to add nukes to the equation.
General Electric was (just a few decades ago) the largest non-gold asset, and isn't even Top 50 anymore.
Saudi Aramco was (just a few years ago) the largest non-gold asset, and isn't even Top 10 anymore.
If you want to read about similarities in Xerox's own hubris/downfall, Dealers of Lightning is one of my favorite non-fiction reads of the past few years (about their SFbay PARC Labs of the 70s/80s... and its failures to adapt).
Just because you brought up this topic: Which of the top tech companies today do you think are most like Xerox/GE in the sense that they're maximizing financial gain out of what they've built without building for the future?
Also not OP, and I'm not an analyst, but my impression is that each of these have stopped looking to the future:
• Meta (all of it)
• Google (not necessarily the rest of Alphabet)
• Apple — UI updates are not innovation, and there's a lot of bugs
• X (the social media website, no comment on the new parent AI company)
• The car-shaped bits of Tesla — the humanoid robotics bits are at least "future", even if Tesla is not demonstrating anything new in this space
To me, it seems to have the same problems as the Google Glass, in that it's far too expensive and doesn't have a clear idea what its own USP is.
That said, while I've played with a few different VR headsets, I've not had a chance to play with the VP, so perhaps there's something in the quality that would become visible if not for the prohibitive price.
The 128K Mac was not too far off from being a promising toy (it wasn't until the Apple LaserWriter became available that the true promise/utility was made real) --- you have to build the expensive first version for early adopters so that you can later figure out how to make the affordable version folks will actually buy and use.
But you don't have to start expensive: cheaper headsets already existed for several years before it came out, and those are perfectly adequate games consoles; and of the reviews I've seen for home cinema and virtual monitor uses, nobody seems to prefer AVP over other headsets half the price.
It has a lot of drawbacks; I agree completely. I don’t own one, I don’t like it as a product, and I wouldn’t buy one, but the tech was undeniably innovative.
I do not own any AAPL (neutral feelings; don't know enough about their direction).
Unfortunately, my mid-sized US city still doesn't have an Apple Store... so I haven't demoed the VisionPro, yet; but I want to be able to see if its endorseable for my Parkinsons friend (whose eyes still work well); I suspect if I demoed it I'd also purchase one too.
With the caveat that often the only difference between being mad and visionary is if you ultimately succeed: startups.
At least, so long as they're not of the [buzzword]+[existing product] variety of startup. A decade ago we had "Uber for $[dictionary merge all nouns]", and today you can easily find a lot of people trying to shoehorn "blockchain" into things without any apparent understanding that other ledgers exist. Not those kinds of startup, those are all just copycat pander-to-the-investors schemes at best, and magical thinking at worst.
On a similar note about madness and vision: SpaceX, Rocket Lab, and SpinLaunch. SpaceX clearly has Mars as a big futuristic vision, and unlike say Blue Origin they actually make rapid progress in that direction; Rocket Lab may be small fry compared to SpaceX, but are thinking along similar scale for launch vehicles and have the political advantage of not being Elon Musk, so if Brand Musk implodes as I now expect it to, they may end up doing many similar things in practice; SpinLaunch may not be going anywhere fast, but they're aiming for a very different launch system and are therefore will end up in the history books even if they run out of investors.
A few years ago I would have also said Relativity Space, but these days everyone is 3D printing rockets. I've even seen a small 3D printed rocket engine in person, on display at one of the smaller booths in GITEX Europe last month, so Relativity Space no longer stands out for 3D printing rockets.
I have no idea where any of the AI companies are going. It looks like the sector is in a stage where there's a lot of low-hanging fruit, and therefore rapid improvements are relatively easy to stumble upon. But the same dynamics mean that's a red-ocean environment, whereas (IMO) the best innovation is a blue-ocean environment: https://www.blueoceanstrategy.com/tools/red-ocean-vs-blue-oc...
Hadn't heard the red- vs blue-ocean distinction before. Thanks. The bullet points are similar to Thiel's zero to one thesis. (Right?)
FWIW, I'm currently most excited by efforts addressing climate crisis. Like advanced geothermal, better heat pumps, improving the grid, etc.
If I understand, most would be considered blue-ocean.
Most definitely face the chicken & egg challenge. As in the merit and benefits are apparent. But someone has to underwrite lofting a new market(s) from scratch.
I've said as much in other threads. Given mediocre sales and the high existing competition for the futuristic vision for their unreleased future products, I think the correct price for TSLA is USD 10-20/share, and no that's not missing a digit.
For Mars… well, I like the idea of settling other worlds and expanding and yada yada yada, but Musk has repeatedly demonstrated he's not got the personality to actually do that right and not lock people out of the base for disagreeing with him. Plus, if you Muntz* a life support system or a food supply or the walls on a Mars base, everyone can die before replacements arrive.
There are several good examples in adjacent-comments, each company suffering from the same common deathknell of becoming largest in its field of similarly over-cumbersome organizations. Failure to steer the ship as it becomes too top-heavy (e.g. kill R&D immediately; outsource all the things).
To echo [1] the Xerox example, the C-suite became entirely made-up of former salesmen who'd become addicted to the pavlovian `click`ing sound their massive copiers made in bill-per-page leases... which blinded them from realizing the digitizing world wasn't going to favor their no-risk-taking approach.
`click` - `click` - `click` – late 60's ASMR erotica for Xerox VPs.
Tech today sells dopamine to end users. Those companies were huge industrial type giants at the time, especially GE, and sold to enterprises large and small.
Companies can do financial tricks but end users can’t.
The linked article doesn’t even try to complete the argument… there’s not a single reason offered why TI could plausibly have increased total revenues more… instead of just spinning their wheels from a higher cost base.
Or to put it another way, why wouldn’t that revenue growth have gone to their competitors?, which did in fact happen.
As many here have said it seems most likely this is hollow PR. It is also possible, though, that the government has decided building chips in the US is actually a national security concern and TI is one of the companies that could make that happen with enough printed money.
Any company large/influential enough to get in front of trump right now is gonna be in the news saying this stuff. The entire admin runs on "the last person to convince trump".
Trump pardoned a bunch of criminals, drug dealers, and other trash people because the right person talked to him at the right time.
This is exactly what people said would happen. It's open season for bribery
Watching from this side of the Atlantic, and being from a country that was under dictorship, I think anyone hoping they only need to hold on to some future elections, haven't paid that much attention during history classes.
The hopeful side is that he is old, and most people voted him in because they did not like inflation and he promised he would bring prices down on day 1.
Trump isn't running again and he has a notoriously poor track record of picking political allies so his "chosen successor" is DOA. After 2028 it will be normal uniparty American politics as usual, with the MAGA/MIGAs going the way of the Tea Party, absorbed and homogenized into the political machine.
The hyperbole about 2024 being the last election in American history will seem silly in retrospect to most, but those who really believe it now will for the most part never admit they got swept up in hysteria and will instead pat themselves on the backs and credit themselves for America narrowly avoiding disaster. A generation or two later, young people will just roll their eyes when their parents talk about any of it, one way or the other, like millennials did when their parents rambled on about Reagan or Nixon.
The hyperbole is based on history with similar events.
Being from 70's generation, with family that had to their encounters with PIDE/DGS, and took part on the liberation forces, I see the supposed hyperbole with different eyes.
The MAGAs are creating a lot of institutional rot at state and local level offices where they are out of the limelight and can continue doing damage after the felon is gone. It is impossible for a rational Republican to get on the ballot because they have to face primaries against the shameless narcissists who've been shown how to con their gullible members.
I really don't think it's all that different from Republicans in the past (particularly the two I named), both in the way they were perceived and received, and what they actually did. Felon and showman, wrecking government institutions, subverting democracy, etc etc.
When in doubt, bet on the status quo. When the status quo falters, bet on a return to the status quo.
Nah, the real problem was the old guy that promised to step down, then refused to do so until the last minute leaving us without the time to choose a real candidate.
Not to mention the fact that we refused to hold the person who tried to overthrow our government accountable for doing so.
yeah, nepotism has always being around, it is just more obvious now, before one would need laundry money through some what seemly acceptable foundations.
I see this argument where I live too. I think it minimises what’s going on. Saying ‘they did it too, but more it was more subtle’ is a defence, and a ‘whataboutism’.
Especially when the reality of building and permitting will mean someone else, even if it’s from the current administration, will be in power long after the check is cashed and ground broken if it even happens. Because of the time scale there’s plenty of time for the “aw shucks it didn’t work out” if that’s the better path. Hitting a few milestones but not enough to actually break ground can be exceptionally profitable.
It's simply not cost effective to manufacture advanced semiconductors at scale in the USA. No company, not Intel, not TI, not GlobalFoundries not even TSMC is clever enough to overcome this.
Now, if there are other reasons to do this like for defense or supply chain security, that's a different story. But it will cost you.
Knowing these facts changes your perspective on these semis-in-the-USA news headlines.
This seems to be a political term, not one the electronics industry uses.[1] "Foundational chips (also called “legacy,” “lagging edge,” and “mature node” semiconductors) are often defined as chips made with a 22nm manufacturing process or above."
Is there actually a lack of 22nm and larger fab capacity in the US? Or is it just that they're not being used much.
FD-SOI fabs are lacking. It's a planar transistor technology used frequently in very low power/thermal devices with small packages and with a high SEU tolerance. Lattice semiconductor use them in their FPGAs, which are the lowest power draw FPGAs on the market at the moment, with a really flat power curve all the way up to 125C. They're used everywhere in A&D and industrial, as root-of-trust devices on servers, and are the largest FPGA manufacturer by volume.
Currently it's fabbed in France (STMicro), Germany (GF), and South Korea (Samsung). No plans to onshore that in the US.
What PA and Fairchild did. We haven’t had a new-entrant opening for fundamental semiconductor design for decades, a fact laid bare in their top engineers’ comp. (No clue if this is what they are actually doing.)
Yes, there is a lack of that tech. Power devices and any mid-complexity mixed signal devices use higher nodes like 22, 55, or even larger geometries. Not a lot of that capacity in the states and it seems to be mostly held by players that own their own fabs. Fabless semiconductor companies don’t have many options besides overseas.
Also already building a new fab up in Sherman, Texas.
Fabs have pretty decent electricity demand, I actually sited a battery against that one after we did a bit of modeling with the new load vs the solar flows in the area. Siting just means make a land deal, figure out easements, and start working its way through the queue. Might not ever get built though, lot of money and hurdles in the way ha.
Every time I see an industrial announcement like this power is the first thing I think of - where it’s gonna come from and what impact it will have on the existing grid.
Most of these projects invest heavily in renewable energy, becuase the scale makes it cost efficient. During the Biden era, there were a bunch of additional incentives to push for renewable investments in these projects, on top of Texas's very generous renewable tech incentives - most ONG companies began investing heavily in renewables all the way back in the 2000s because they're ENERGY conglomerates first and foremost.
If you have a beating pulse and a strong renewable tech IP, you will raise a seed successfully from the Saudi PIF or Cheveron Ventures even in this market. Sadly, this is HN, not Bookface, so most ideas are bad ideas. Sucks too because it was a very fruitful demo days - one of the more impactful over the past several years.
This is what immediately popped in my mind. What’s that definition of insanity again…
I’d love it if things were streamlined when building. Just as a homeowner it’s amazing how hard it can be to get a permit or an inspection. The lame duct tape fix seems to me some weird CMS and ArcGIS, as if that somehow fixes the human and policy problems.
For those out of the loop but who enjoy podcasts, there is a Reply All episode about this [0] though somehow I can't find a direct link as spotify seems to have swallowed all of gimlet and killed the links. I couldn't even find a transcript.
They’ve steadily expanded over the years, I wouldn’t be surprised if this was real. It’s not just calculators, TI chips show up in a lot of military hardware in particular.
Which is why they are not planning to make cutting edge semiconductors in the new plant but "foundational semiconductors" aka old nodes that now cost pennies
If you want to build a fleet of infantry drones and you want to source components from the US there’s a lot of these and many others to buy. Not sure if there’s $60B of them there but they’ll figure out how to waste this much I’m sure.
Well if you look at how Feds has being pumping money into Musk's bank account and what brought down the SVB, then $60B may not be that much. At this point as the country that can prints/(adjust db ledger) $60B is nothing.
With MESI cache coherence, maybe you could migrate the whole workspace for your subroutine into a line of your core's L1D cache, and make it perform like hardware registers while retaining the pleasantly parsimonious TMS 990 architectural semantics?
They’re making a political advertisement to get favors for things they already did. The meat of the announcement is that the plants started under the Biden administration are continuing as planned – which is an accomplishment of sorts given how complex fabs are – and they’d really like to continue getting the CHIPS act funding which Trump periodically talks about cancelling because it was a Biden-era law even though it had strong support from Republicans who favor investing in important domestic industries, especially in their districts:
Number of new jobs for one. They just laid off ~300 people working on an older node in Lehi, Utah, while now claiming they’re going to bring new jobs to Lehi, Utah. Thus “new jobs” don’t really track the actual change in the number of their employees even just at the local level across even moderate timescales.
Similarly hyping up total R&D spending without a tight timeframe is meaningless, they will eventually spend X$ assuming the company survives indefinitely.
More charitably there’s meaningful differences between nodes, but it’s likely a significant portion of the workforce will be rehires.
I feel like this is a fairly small investment on the scale of the semiconductor industry, although it would be large for a single company if we were talking about a single year. See, for example, https://wccftech.com/chinese-chip-giant-smic-shaken-by-tarif... US$7.5 billion in capex at SMIC this year. But SMIC is just the largest of dozens of Mainland China semiconductor firms, and a lot of that investment is going into cutting-edge strategic nodes, not 180nm stuff from last millennium.
Don't get me wrong, you need that stuff from last millennium—not just 180nm but also the 6 μm I suspect "foundational" actually refers to. You need power regulation, precision measurement, and RF frontends in analog, and analog can't scale down the way digital can. And TI actually retains a world-leading position in those chips, with better and cheaper ICs than anything out of Mainland China today. But it's potentially a game of chasing China's taillights.
SiGe might be an exception; even at 130nm, IHP's SiGe BiCMOS process can hit 450GHz oscillation frequency and 350GHz ft, so even at fairly coarse process node sizes, SiGe could be a strategic enabling technology for submillimeter communication such as Starlink, which is reportedly critical for weapons such as Ukraine's naval drones.
CHIPS invested heavily in OSAT and Packaging capacity - especially in TX. Samsung, Micron, OmSemi, and TI took full advantage of that under the Biden admin while Intel and TSMC were fighting on the airwaves to undermine each other.
Much of the rest has been reinvested in SK and India (TI is part of the SCL Mohali modernization RFP) as part of the QUAD+ initiative.
A lot of us in the private and public sector working in this space aren't idiots.
The adjective "foundational" for semiconductors can be ignored here, it is simply that TI will make more of their existing chips in the US with nothing being different in their current IC lineup of mostly MCUs, ADCs, etc simple ICs.
Really? You don't think it means specifically "old"? Because TI also makes new chips like their MSPM0 microcontroller lineup. Are those "foundational" too?
I hate doing a first article, seeing nothing wrong on the IC side of the LED board, then apply power and have the IC explode and rocket off the board, taking traces and pads with it (or if I'm lucky, it just craters its packaging.)
And it is BAD. I'm talking 1/20 kind of bad, off multiple reels (I've gone though tens of thousands of them.) Only TI stuff.
Worth noting that TI primarily makes logic semiconductors, and those don't require the advanced 2nm, 3nm or even 5nm node processes that TSMC's and Samsung's most advanced fabs support. However logic semiconductors may have special operating constraints, like extremes of temperature, that the semiconductors produced by foundries don't require.
The article mentions SiGe at the Sherman site, which is an RF node. These are usually at 22-100+ nm scale, so "mature" or "foundational" compared to 2nm stuff, but the really tight nodes aren't always inherently better for RF.
the TSMC founder was once a Ti employee. Yet due to cism he got denied the opportunity to steer the org. Now it's time to play to trump's whims and earn a quick payout.
I was wondering where the money comes from. This article [1] says it's government money provided through the CHIPS act:
> In December, the [prior] administration finalized a $1.61 billion government subsidy for Texas Instruments to support construction of three new facilities after the company announced plans to invest at least $18 billion under the $52.7 billion CHIPS and Science bill.
Title shortened by me from "Texas Instruments plans to invest more than $60 billion to manufacture billions of foundational semiconductors in the U.S."
How are they going to afford an investment thats ~1/3rd the value of the company? Seems like one of those announcements that no one follows up on to keep them honest?