Semiwiki on LinkedIn Semiwiki on Facebook Semiwiki on Twitter Register / Log in 
Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/threads/u-s-backed-chip-startup-xlight-aims-to-raise-350-million-for-an-euv-alternative.25366/
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/EmailDomainReplace] => 1000010
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2031070
            [XFI] => 1060170
        )

    [wordpress] => /var/www/html
)

U.S.-backed chip startup xLight aims to raise $350 million for an EUV alternative

Daniel Nenni

Daniel Nenni

Founder
Staff member
1782336992212.png


The U.S.-backed chip startup xLight is negotiating to raise $350 million, with Boardman Bay Capital Management and Bain Capital expected to lead the round. The U.S. government has invested $150 million in xLight through the CHIPS and Science Act, bringing xLight’s total funding to approximately $200 million.

Former Intel CEO Pat Gelsinger serves as executive chairman of xLight, and his firm, Playground Global, invested in xLight in 2025. xLight plans to lobby ASML, TSMC, Intel, and Micron Technology to participate in this financing round, aiming to develop alternative EUV lithography technologies.

ASML CEO Christophe Fouquet revealed that ASML is collaborating with xLight on technology demonstrations. xLight intends to build its first prototype site in Albany, New York, and has signed a non-binding project financing agreement worth $4.2 billion with lenders. Substrate is also developing X-ray lithography technology, directly competing with ASML.

Company Overview & Goal
xLight aims to invent an alternative light source for semiconductor manufacturing that lowers the cost and time required to produce advanced AI server chips.
  • The Problem: Current EUV lithography machines (primarily manufactured by ASML) rely on laser-produced plasma, which xLight cites as a major bottleneck.
  • The Solution: The startup is building free-electron lasers to beam ultraviolet light for printing microscopic patterns on silicon wafers.

Recent Developments
  • Executive Leadership: Former Intel CEO Pat Gelsinger took over as executive chairman in March.
  • Funding: The company is currently in talks with firms like Bain Capital and Boardman Bay Capital Management to lead the new $350 million funding round.
  • U.S. Stake: The $150 million CHIPS Act award makes the U.S. government one of the startup's largest shareholders.
  • Prototype: xLight has signed nonbinding agreements securing up to $4.2 billion in project financing to build its facilities, with the first prototype being developed in Albany, New York. The first working light source is targeted for 2028.

www.theinformation.com

U.S.-Backed Chipmaking Startup Chaired by Former Intel CEO Is Raising $350 Million

XLight, a startup developing advanced lasers for semiconductor manufacturing, is in talks to raise $350 million in funding from a pair of investment firms just weeks after receiving substantial investment from the U.S. Department of Commerce, according to the company’s verbal and written pitch ...
www.theinformation.com www.theinformation.com
 
Daniel Nenni said:
Recent Developments
  • Executive Leadership: Former Intel CEO Pat Gelsinger took over as executive chairman in March.
  • Funding: The company is currently in talks with firms like Bain Capital and Boardman Bay Capital Management to lead the new $350 million funding round.
  • U.S. Stake: The $150 million CHIPS Act award makes the U.S. government one of the startup's largest shareholders.
  • Prototype: xLight has signed nonbinding agreements securing up to $4.2 billion in project financing to build its facilities, with the first prototype being developed in Albany, New York. The first working light source is targeted for 2028.

www.theinformation.com

U.S.-Backed Chipmaking Startup Chaired by Former Intel CEO Is Raising $350 Million

XLight, a startup developing advanced lasers for semiconductor manufacturing, is in talks to raise $350 million in funding from a pair of investment firms just weeks after receiving substantial investment from the U.S. Department of Commerce, according to the company’s verbal and written pitch ...
www.theinformation.com www.theinformation.com
Click to expand...

I recently talked at an event with the former, now retired, ASML CEO about this xLight initiative. My impression is that ASML does not feel threathened by this FEL stuff at all. ASML also looked into FELs for many many years.

His as was my conclusion (already before) are the same: you need redundancy, so a second FEL for each xLight setup (that delivers EUV light to say 10 EUV steppers), in case the first FEL has major down time. This is a very expensive redundancy!!

And presently, ASML ships every EUV machine with a (soon) 800-1000 Watt EUV source. They ship about 60-100 EUV machines EVERY YEAR. So, they bring some 60-100 kWatt EUV source power on the Fab market EVERY year.

A couple of 10 kWatt FELS from xLight is not going to move the needle very much.....and who is going to integrate them with the scanner? Is xLight going to build EUV-scanners/steppers as well? ASML will probably not spend too much time with xLight on adapting their tools for xLight FEL EUV sources would be my guess.

My personal note, PG is now steering the strategy on this burning through 350 M$? A lot of enthusiasm, but perhaps making a similar mistake as with INTEL 2.0, where is the market for your capacity?

Oh well, it keeps PG off the street and brings in some bucks for his chimney to burn at home, and perhaps, perhaps he has a direct line to upstairs (devine) information......


PS when the Street and small investors give EM 80 B$ to play, PG must think 350 M$ is noise, I should be able to find some suckers....
 
Last edited:
user nl said:
I recently talked at an event with the former, now retired, ASML CEO about this xLight initiative. My impression is that ASML does not feel threathened by this FEL stuff at all. ASML also looked into FELs for many many years.

His as was my conclusion (already before) are the same: you need redundancy, so a second FEL for each xLight setup (that delivers EUV light to say 10 EUV steppers), in case the first FEL has major down time. This is a very expensive redundancy!!

And presently, ASML ships every EUV machine with a (soon) 800-1000 Watt EUV source. They ship about 60-100 EUV machines EVERY YEAR. So, they bring some 60-100 kWatt EUV source power on the Fab market EVERY year.

A couple of 10 kWatt FELS from xLight is not going to move the needle very much.....and who is going to integrate them with the scanner? Is xLight going to build EUV-scanners/steppers as well? ASML will probably not spend too much time with xLight on adapting their tools for xLight FEL EUV sources would be my guess.
Click to expand...
What would be the extent of ASML's mentioned collaboration with xLight, I wonder.
 
Fred Chen said:
What would be the extent of ASML's mentioned collaboration with xLight, I wonder.
Click to expand...
ASML will always "collaborate" just to know what these guys are doing and why it won't work. In case it would get interesting, they would buy them.......And getting the knowledgable people, and IP, in their own ASML-empire. Knowledgeable people are worth their weight in GOLD!
 
Last edited:
Scientifically, an accelerator-based EUV source is a solvable problem given enough time and money. It will take way more than 3-4 years. There will be many engineering obstacles, some of which are not even known yet. Integrating it into a real litho tool? What's the incentive? Almost zero. It's a fun project that sounds great in a presentation! But you don't know unless you try, right? Keeps the crazier folks off the street.
 
user nl said:
His as was my conclusion (already before) are the same: you need redundancy, so a second FEL for each xLight setup (that delivers EUV light to say 10 EUV steppers), in case the first FEL has major down time. This is a very expensive redundancy!!

And presently, ASML ships every EUV machine with a (soon) 800-1000 Watt EUV source. They ship about 60-100 EUV machines EVERY YEAR. So, they bring some 60-100 kWatt EUV source power on the Fab market EVERY year.
Click to expand...

Does ASML make it's own light sources for EUV - or are they the only source in the world for such light sources?

As expensive as xLight's options may be, from a (foundry) business continuity planning perspective - being sole sourced on suppliers is a risk of it's own that carries a value.
 
Xebec said:
Does ASML make it's own light sources for EUV - or are they the only source in the world for such light sources?

As expensive as xLight's options may be, from a (foundry) business continuity planning perspective - being sole sourced on suppliers is a risk of it's own that carries a value.
Click to expand...

If you want to know the latest on EUV sources, scan the program and book of abstracts here from the June 2026 EUV source workshop:
https://euvlitho.com/

The people from (German company) Trumpf, that build the huge CO2 laser systems for the present EUV sources of ASML, are working with the Jena-solid state laser people, that they acquired in 2022, to develope 2 um solid state lasers for a much more efficient wall plug efficiency of the future ASML EUV tools. And also a much smaller footprint of the laser-floor needed below the EUV scanner, than is currently needed to place the huge CO2 laser that blasts the Ti-droplets in the EUV source chamber, at 70 kHz reprate or so currently.


See page 11 in the book of abstracts:
2µm Fiber Lasers as Efficient Next Generation EUV Drive Lasers

Our results demonstrate the potential of Tm-doped fiber amplifiers and pave the way towards efficient, low footprint drive laser technology with high wall plug efficiencies (>20%) and potential EUV-power far beyond 1kW.



There were also xLight people talking about the FEL stuff: Dinh Nguyen xLight FEL for EUV Lithography. Here a pdf linke to a talk on the web, this talk was from 2024 looks like:
https://euvlitho.com/2024/P43.pdf
 
Xebec said:
Does ASML make it's own light sources for EUV - or are they the only source in the world for such light sources?
Click to expand...

Way back in the early 00's, ASML started formally working with Cymer, a San Diego-based provider of excimer lasers, for DUV lithography. They formally partnered starting in 2007 on a laser produced plasma EUV source. ASML bought Cymer in 2012.
 
user nl said:
If you want to know the latest on EUV sources, scan the program and book of abstracts here from the June 2026 EUV source workshop:
https://euvlitho.com/

The people from (German company) Trumpf, that build the huge CO2 laser systems for the present EUV sources of ASML, are working with the Jena-solid state laser people, that they acquired in 2022, to develope 2 um solid state lasers for a much more efficient wall plug efficiency of the future ASML EUV tools. And also a much smaller footprint of the laser-floor needed below the EUV scanner, than is currently needed to place the huge CO2 laser that blasts the Ti-droplets in the EUV source chamber, at 70 kHz reprate or so currently.


See page 11 in the book of abstracts:
2µm Fiber Lasers as Efficient Next Generation EUV Drive Lasers

Our results demonstrate the potential of Tm-doped fiber amplifiers and pave the way towards efficient, low footprint drive laser technology with high wall plug efficiencies (>20%) and potential EUV-power far beyond 1kW.



There were also xLight people talking about the FEL stuff: Dinh Nguyen xLight FEL for EUV Lithography. Here a pdf linke to a talk on the web, this talk was from 2024 looks like:
https://euvlitho.com/2024/P43.pdf
Click to expand...

Thank you for the links.

I never realized how physically large the EUV light sources were. I had an idea they weren't small, but. Impressive.

xLight's FEL source looks monsterous compared to Trumpf's / ASML's CO2 solution.

..

Is it fair to take away that the FEL solution is more complex in terms of shaping and delivering the laser light (due to size/distance/number of mirrors), but less complex in terms of not needing to deal with plasma/tin "droplets"? (And would the lack of tin droplets also mean the pellicels / masks would last longer?)
 
Xebec said:
And would the lack of tin droplets also mean the pellicels / masks would last longer?
Click to expand...
It looks like the Sn source plasma is far enough removed from the reticle, the mask would be more likely to get Sn from the MOR if that is being used.
1783349316471.png

The pellicle is mainly being limited by the power heating, in addition to hydrogen exposure (for the rest of the system). The mask (w/ or w/o pellicle) is limited by cleans. The contamination is likely from the resist on the wafer as well as other parts of the chamber.
 
user nl said:
I recently talked at an event with the former, now retired, ASML CEO about this xLight initiative. My impression is that ASML does not feel threathened by this FEL stuff at all. ASML also looked into FELs for many many years.
Click to expand...

The tin plasma EUV source was once a "lower cost" and "lower complexity" option vs. "building the fab around a particle accelerator"

Of course, over they years, it became anything, but simple, and low cost, and you have to build the fab around an infrared laser instead.
 
Paul2 said:
The tin plasma EUV source was once a "lower cost" and "lower complexity" option vs. "building the fab around a particle accelerator"

Of course, over they years, it became anything, but simple, and low cost, and you have to build the fab around an infrared laser instead.
Click to expand...

The Fab is not built around the CO2 laser. Each and every EUV tool has its individual CO2 laser, beneath the EUV scanner

My feeling with the xLight FEL, the way TSMC and others built their Fabs is totally out of sync with the way the xLight tool would be built and would supply their light to say 10-20 EUV tools.

And the redundancy needed is just a massive cost, you need a complete second FEL, in case there is major down time with the first FEL

My feeling is that the new nanosecond (50-80ns) duration 2 um wavelength solid state lasers that are being developed now by Trumpf and Jena AFS, will reduce the needed wall Plug Power quite a bit, and thus increase the energy efficiency of the Laser produced plasma EUV tremendously.

As said every year 60-100 EU tools, with each tool 800-1000 Watt EUV, there is no way the xLight FEL is going to be economically competitive enough for TSMC and others to switch gears.

Let's see if Trump or some other of his Wall Street buddies are going to give 350 M$ (the price of a single high NA EUV tool more or less) to have PatG burn through this pocket money the coming 3-5 years for his FEL development of xLight. My feeling is this FEL project for Fabs will die the coming 3 years......Simply not workable....
 
user nl said:
Let's see if Trump or some other of his Wall Street buddies are going to give 350 M$ (the price of a single high NA EUV tool more or less) to have PatG burn through this pocket money the coming 3-5 years for his FEL development of xLight. My feeling is this FEL project for Fabs will die the coming 3 years......Simply not workable....
Click to expand...

I wish our (USA) government was more fiscally responsible, but $350M is nothing compared to the $1,000,000M+ spent on US military. All someone has to do is pitch this as "strategic" and the money flows.. for better or (usually) worse.
 
You must log in or register to reply here.
Back
Top