The National EUV Accelerator comes to Albany | Hacker News Reader

The National EUV Accelerator comes to Albany

research.ibm.com

What’s the status of IBM these days? They are treated like an obsolete dinosaur in my tech circles, but they’re so big that I wouldn’t be surprised to learn that they are actually cutting edge in some areas (particularly in capital intensive stuff)

I can't speak to most of what they do, but they seem to be at the forefront of quantum computing.

Afaik they're partnered with Samsung on the fab side.

They placed all their bets on enterprise cloud + Quantum.

Azure and AWS?

Well they certainly aren't the biggest but their cloud operation is still pretty large and somewhat nice. Noone is even close to IBM in Quantum cloud offerings.

IBM q2 revenue- 16bb AWS q2 revenue- 26bb Azure q2 revenue 24bb

I really don't understand why people say IBM is dying, its literally one of the most successful and revenue producing companies in the world with plenty of room for growth with their investments in Quantum. The just don't provide as many consumer facing products so everyone assume's they're dead.

I personally don’t know much about quantum or IBM. How relevant are they today? Are they going to become more relevant tomorrow?

Late response, but how are they irrelevant they're 63rd give or take on the Fortune 100 and they're growing just as much as any other company in that size range. They're just not relevant to the average Joe because they don't really make products for us. They make products for other fortune 500 companies.

What is the bull case for a Cloud-based quantum computing offer? Are there really billions of dollars of workloads out there that have any realistic chance of migrating to IBM quantum tech within a decade? Two decades?

All quantum workloads (whatever they will be) are cloud workloads because no one is going to be having their own non-trivial quantum hardware.

Their semiconductor IP division has been cutting edge for decades (particularly before TSMC got copper working).

They kind of act like a bridge between academia and industry, helping to develop academic ideas into something that's viable for high volume manufacturing.

Did the sale of their semi fabs effect the IP side of the business?

I don't know enough about their operations to answer with certainty, but I believe those fabs were generally more mature nodes for mil/gov (in fact I think they were the only semiconductor manufacturers approved for mil/gov use).

Judging by the fact the number of patents they file doesn't seem to have gone down after 2015, I'm going to guess not.

IBM has a track record of pursuing highly visible projects, like chip fabs, PowerPC, Watson etc. and then just seemingly wimping out.

They are low key going wild on AI Agents

Isn’t that true of almost everybody from startup to FAANG? Watson crushed it on jeopardy 15 years ago, but where did that lead go?

The I in FAANG stands for IBM

There was a recent article about IBM:

https://news.ycombinator.com/item?id=41646967

Their goal is to replace staff with AI, although they suck at it:

> A WatsonX chatbot is years behind ChatGPT," Blake said. "Its web interface was horribly broken to the point of being unusable until July 2024, and no one in the entire organization uses it.

IBM isn’t that big. They are way smaller than trillion dollar companies like Google, Microsoft, Apple, Samsung, Amazon etc.

They may have similar revenue to Intel, but they lack focus. Being so spread out they don’t really dominate any segment. What they are great at is marketing compared to say Accenture which is roughly the same size.

Forewarning that this is an incredibly biased viewpoint based on my experience with IBM and their products…

They tend to target slower moving clients, think healthcare or the public sector. They sell them incredibly outdated products which tend to wall them in or limit their choices to other similarly outdated product offerings from IBM.

This begins the support racket, it is not uncommon to see an error from an IBM product instructing you to reach out to your IBM support representative for help in resolving the issue. These support agreements are where quite a lot of their money is made.

It goes deeper, many people who previously worked at IBM end up taking software architect positions within these slow-moving sectors. I am not saying it’s outright, but it seems quite obvious that these people are financially motivated to fold IBM products into any and every opportunity possible.

Basically IBM has developed somewhat of a dark pattern for software sales and they use it to capture their clients. Extricating your organization from the IBM ecosystem is incredibly difficult and often an uphill battle.

Correct

For an organization to remove this, it needs to be treated like a metastasized spread, and every person that has ever touched or supported this needs to be removed from the org

They are working on quantum computing, so they do have cutting edge tech in some sectors, but as mentioned, they make their money on their rather outdated infrastructure.

https://www.nextplatform.com/2024/10/24/ibms-red-hat-acquisi...

> IBM is nearly done getting back the $34 billion that it spent.. to acquire open source software supporter Red Hat.. [which] doubled in size since the acquisition, with an annual run rate of $6.5 billion and delivering a compound annual growth rate “in the mid teens” over those five years.. OpenShift had an annualized run rate of a mere $100 million when IBM took control and is now running at $1.3 billion, a factor of 13X increase over the five years

Let's see if Albany's NanoTech Complex finally takes off this time.

It's had a fairly cursed history of corruption scandals from the Cuomo era of NY state politics.

If Cuomo hadn't mismanaged NY State's business capacity, this would have taken off a decade earlier - Upstate New York always had the right pieces due to Kodak, Nikon, and IBM Microelectronics-turned-GloFo

According to the Oregonian, Intel and Oregon were hoping to land this center.

https://www.oregonlive.com/silicon-forest/2024/10/biden-admi...

Yikes - Intel must not carry a lot of currency with federal planners any more. The fact that Oregon hasn't had a Portland-area research institution for decades in semi tech/hardware engineering probably doesn't help, either. [1] (Or the fact that Chuck Schumer represents NY.)

[1] I know people in the 50s who used to work at the Oregon Graduate Institute, which was very adjacent to Tektronix, Intel, etc.

It was a tragedy for the area when OGI basically vanished. I know a lot of people who earned graduate degrees while working at local companies.

I took a class there while I lived in the area. The guy teaching it worked with me at Intel. It was a good local resource, sad when it was taken over by OHSU.

I suspect that the federal government thinks that Intel will go bankrupt and they don’t want to tie their center to a part of the country without much intellectual capital (no world class universities, etc).

Feds are not going to let America’s best fabrication operator go out of business. But, now that the federal government is paying for it, federal legislators are going to want to see money flowing to various districts so they can tell their constituents how much money they brought back for them.

There is no way that the US is letting Intel fail -- having a strong competitor in advanced chips is a national security matter.

TSMC and Samsung are expanding fabs in the US so the US can make do without Intel if needed.

Yes, but so is Intel, thanks to the CHIPS act.

No, that’s very different from a government perspective, especially in TSMC’s case.

RPI has a Quantum computer (courtesy of IBM) - the only university to have one I believe - and is a highly respected engineering research university, so that likely played a part in Albany getting this.

I find it surprising that Intel didn't help fund the establishment of a world class research university in Portland, or an OSU or OU satellite focused on engineering.

There was something called the Oregon Graduate Institute that the founders of Tektronix helped establish. The issue with Intel is that their management is not in Oregon for the most part so they are not invested in the same way as Phil Knight (Nike’s founder) is. Intel has also historically relied on importing talent from outside Oregon so the lack of a world class research university in the Portland metro area has not hurt Intel too much. I am sure that they regret not having such an institution now though.

> the lack of a world class research university in the Portland metro area has not hurt Intel too much

Hey, hey, OHSU exists y'know! ;)

Absolutely! Great medical school. I was thinking engineering - should have specified that.

Good, the US is going serious about EUV tooling (If I understood well). This would be for the american block.

In the EU block, there is ASML.

Now, the asian block should have its own, but I don't know how if china should not R&D its own and asian democraties should stick to the US and EU blocks.

And RISC-V ISA could make a good interop glue for all of them.

But the more I look at it, the more I think the world is getting divided in roughly 2, so maybe the world would need only 2 centers of EUV tooling (or ultra complex tools for state of the art chips): one for the finance dictatorship block ("democraties",EU,US,etc), and one for the harder classic (including religious) dictatorship block (china,russia,etc)

In this day and time, I'll believe it when it really happens. Depending on what happens next week, the CHIPS act could be repealed early next year.

Ah yes, the classic block the exports of your partners and throw a bunch of subsidies at your own. Thanks for the free market capitalism America!

Some things are more important than free market capitalism, and making sure you’re not dependent on hostile countries for key technology is one of them.

There is no free market in the fab business. Every single player had some level of government support and direct subsidies for decades.

Exactly. ICs and Silicon Valley as we know it wouldn’t exist without the space program.

America never claimed to have a completely free market: https://www.investopedia.com/ask/answers/031815/united-state...

Reply on news.ycombinator.com