Ph.D student demonstrates a single, working laser on silicon | Hacker News Reader

Ph.D student demonstrates a single, working laser on silicon

brighterworld.mcmaster.ca

213 pointscroes59 comments

This is neither low cost nor commercially relevant. It’s an exotic, low-temperature (so incompatible with subsequent steps of most wafer processes) deposition of rare earth material. The lasing wavelength is all wrong (1.9 um). It requires a pump laser in the 1.6 um range which is the usual range of interest. The efficiency is not very good. Couldn’t find the linewidth while scanning rapidly. It’s a run of the mill academic paper.

There exist commercial electrically pumped lasers on silicon that are being continuously improved through different approaches. Intel is one of the leaders on that.

Is there not a single dimension in which the paper is novel? If not, it seems like a mistake to publish this.

Not my area of expertise, but something can be academically novel and interesting without having any commercial or industrial applicability.

Isn't this what usually happens when the PR department of a university runs away with a story?

It's novel, it's a huge iteration in the space. But, it's also being misrepresented as more significant than it is due to the summary being friendly to laypeople.

Not really. It's a different approach with different tradeoffs that are just as bad as the previously known ones.

That's probably why it wasn't picked up by a better journal and that nobody outside of the university where it happened is talking about it.

And on top of that, you can make almost any material lase. There have been entertaining experiments where folks used yogurt, apple juice, hair dye, and other fun liquids in pumped lasers, exploiting some molecular band transition that was accidentally good enough.

Thanks for the review. Unfortunately knowing universities this read a little too much like a puff piece. Of the same article was on a news site I'd then go looking for their dept website. When they publish this themselves my reaction is a little more sceptical and glib unfortunately.

Hmm, couldn't you do the doping as the last processing step? Can you switch the laser with on-chip electronics, even if you can't pump it electrically? 1.9 um sounds like it would be just fine for communications.

(I haven't read the paper, just the PR fluff.)

> This is neither low cost nor commercially relevant.

That's the case for a lot of research. It paves the way to be iterated on.

Silicon photonics (microchip lasers and particle accelerators) is an active area of research that has had some promising improvements lately.

Research like this is likely to result in a laser miniaturization breakthrough like that of MEMs sensors, which will have big impacts if the money and energy cost of a putting a laser on a device drops to near zero.

Great talk on the state of the art: https://www.youtube.com/watch?v=t5EPeuLxaE0

The real test will be if they can mount it on a jet and do a fly-by to fill an entire house with popcorn.

Deep reference! This comment made my morning. Thank you!

real genius

> Encouraged by her supervisor to expand her academic horizons, she considered studying in the U.S. but was unable to obtain a visa.

> Her father suggested she think of Canada.

Paper: https://onlinelibrary.wiley.com/doi/10.1002/lpor.202100348

This is important work. Communications bandwidth limits the efficiency of applications on multi-core computers. On-chip lasers can be used for core-core and core-RAM optical interconnects. Such optical interconnects have advantages over electrical interconnects:

- Optical lines can run tangent to the chip surface. This frees up precious chip space and layers. It also changes processor geometry from 2D to 3D.

- Longer electrical interconnects use more power than short ones. This means that electrical interconnect power consumption is more than linear in the number of cores. Optical lines use nearly the same power regardless of length. This means that optical interconnect power consumption is linear in the number of cores.

- Light travels twice as fast as electrical signals in wires. Therefore, optical lines have half the latency of electrical ones. Latency is especially important with large processors because the distance between any two cores is larger.

Taken together, these mean that we should eventually be able to use optical interconnects to create processors with 1,000 times more cores (than processors with electrical interconnects) and fixed power consumption per core regardless of processor size.

The state-of-the-art in massively-multi-core processors:

- Cerebras WSE-2 chip [0] has 850,000 cores and uses 15,000 W of power [1] to achieve 220 PB/s of inter-chip bandwidth.

- Esperanto ET-SoC-1 chip has 1,088 cores and uses 20 W of power [0]. I could not find out its interconnect bandwidth.

[0] https://cerebras.net/chip/

[1] https://www.tomshardware.com/news/cerebras-wafer-scale-engin...

[2] https://www.esperanto.ai

[3] https://www.esperanto.ai/wp-content/uploads/2021/08/Esperant...

Very interesting points. Thanks. Maybe in the future we will have high bandwidth chip to chip standards (like PCIe, and network chip to chip standards like XFI) with optical communication for better latency, especially in large switches with latency constraints.

There are several misconceptions (as pointed by others) relating the trade-off of electrical vs optical for interconnects. Optical interconnect are promising, and in fact, at some point they will become almost unavoidable: bandwidth vs path length. If you want to dive into the fundamental reasons, the following paper might help: https://www.researchgate.net/publication/224504016_Miller_D_...

Thank you for sharing that article.

Miller p.4, says:

> Typical results have energies per bit of 2 – 30 pJ/bit in recent demonstrations. (1 pJ/bit is the same as 1 mW/(Gb/s); the latter is a more common way of stating the unit in the electrical interconnect literature, though the former relates more obviously to the physics of the interconnect devices.) The best current results for transceivers are ~ 2.8 – 6.5 pJ/bit for board or backplane interconnects [17], and ~ 2 pJ/bit [16] for moderate length chip-to-chip interconnects with a relatively ideal electrical channel.

Miller p.12:

> Hence, from the above discussion of systems energy targets, we have optical device energy targets of 10 – 20 fJ/bit for off-chip interconnects, and ~ 2 – 10 fJ/bit for on-chip interconnects.

Kiani's on-chip laser begins operating at 16mW of power. If the laser is part of a chip running at 5GHz, switching the laser at the same clock yields 5Gbit/s. This is 3.2 mW/(Gb/s) = 3.2 pJ/bit. This is still 3 orders of magnitude more than the 10 fJ/bit target.

Light does not, in fact, travel faster in glass than electrical signals in wires. It is a bit slower, instead. Even in vacuum or air, light is only half-again as fast, not twice as fast.

The appeal of optical interconnect is not speed, which is less, but isolation and noise resistance.

Good point about isolation and noise.

I got the propagation numbers from Wikipedia:

> the refractive index of glass is typically around 1.5, meaning that light in glass travels at c / 1.5 ≈ 200000 km/s [0]

> the refractive index of air for visible light is about 1.0003 [0]

> Wires have an approximate propagation delay of 1 ns for every 6 inches (15 cm) of length. [1]

The speed of light in air is c/1.0003, or 299,700 km/s. In glass, it is c/1.5, or 200,000 km/s. If Wikipedia is correct, wire propagation delay is about 15 cm/ns, or 150,000 km/s, which is still slower than in glass. If that number is off, would you please update the Wikipedia article with the correct number? It would be good to include a better reference than the university textbook it currently cites.

Also, I don't think we should assume that a future processor optical mesh will carry signals in glass. I expect it will use micro-randomized mirrors and protect the optics with inert gas. The randomized mirrors will form random optical lines between cores, on the same chip or between chips. The lines will spontaneously reconfigure due to mechanical shock or changes in temperature. The OS will detect such changes and automatically reconfigure the software to optimally utilize the current connections. This should allow hot-swapping CPU cores, and upgrading or downgrading processors on live computers.

There is a scene in the film 2001 Space Odyssey where (MOVIE SPOILER) the protagonist removes pieces of a computer and the AI running on it becomes slower and slower. Such a thing may actually happen.

[0] https://en.wikipedia.org/wiki/Speed_of_light

[1] https://en.wikipedia.org/wiki/Signal_propagation_delay#Elect...

Propagation delay of electrical signals depends on properties of the insulation. In coaxial cable, the insulation is foam, mostly air, so faster than light in glass. On a chip, the wire often has glass and silicon on one side, organic passivation goo on the other. So, propagation there will be slower than in coaxial cable, and could even be a little slower than in fiber.

Optical interconnect via straight-line paths through inert gas, steered by mirrors, could indeed be strictly faster than through fiber. At the same time, propagation across a wafer cannot take less than a nanosecond, at best, so 1.0ns, 1.5ns. The total delay will be a product of message marshalling and routing, typically dozens or more times that.

Real-time adaptation may be necessary for managing flaky quantum bits, but the more of it that happens, the less total speed you get.

This laser requires a second laser to pump it in order for it to work. It's kind of defeating the point of things.

A lot of lasers require a pump laser, it seems their argument is that this can be useful for low cost production in something like the telcom world.

Not really. Pumped lasers are the norm for quite a few materials that are hard to use as primary light source, but as a secondary lightsource they are plenty useful.

Note that lots of other lasers are pumped as well, for instance, some lasers were originally pumped by flashing Xenon light from spiral wound flash tubes into Ruby crystal rods grown at great expense to get the gas to emit coherent light. The energy efficiency is important so that you don't end up overheating your secondary laser but the principle has plenty of uses and isn't even restricted to using light as the starting point.

One possible way to use this is to use a somewhat higher power laser to illuminate a substrate at right angles to the desired outputs, which would allow for a hybrid where only the primary is made with exotic materials and the remainder can be made in Silicon. Right now every laser output needs to be on a physically different die than the Silicon based circuitry that drives it, this technique could do away with that intermediary step substantially reducing the costs of certain telecommunications devices.

Finally, once something has been pushed to lase it usually doesn't take very long before improvements are made that allow it to do so cheaper, faster or independent of the original light source, it is that first step that is the really hard one.

One joke I heard from someone who is deep inside laser research is that when all is said and done there are probably not a whole lot of substances that can't be made to emit coherent light, the trick is to find out how to do it in the first place, and then to do it repeatedly and affordably.

Really cool, but the article could show a picture of the actual laser light (maybe it's not visible light?)

It appears to be emitting in the IR part of the spectrum (1.9 um).

I'm skeptical of university press releases. Is this the "holy grail" as they claim it to be? :D

I'll also add that "holy grail" type discoveries tend to be published in higher tier journals than this one.

Ehhhh. Not at all true. Or at least that's maybe 50% true. Many high impact papers in my field, including "holy grail" type papers, have been published in "low tier" journals. Some scientists would much rather publish quality research in a timely manner than go through the Nature review process and waste a lot of their time. It's extremely dependent on the disposition of the lead author.

For the record I can't at all speak to the quality of this research paper, I'm only trying to address this comment specifically.

So true

For a so-so discovery, you need the cred of a high-tier journal far more than they need your paper.

For an actual holy grail, ~every journal in the field needs the cred of your paper far more than you need them.

If you have a holy-grail, do journals like Nature do things to try to lure you into publishing with them instead of the one this author chose?

This is exactly what should stop, so if this is the case of someone publishing an actual breakthrough in a lower tier journal then that should be encouraged, rather than punished by outright dismissing the contribution based on where it is published.

Otherwise we'll never get rid of those cartels.

That's like asking smart programmers should stop working for FAANG and work for minimum wage at a non-profit, to help save the world. The academic perks at multiple levels are deeply tied to publishing in top tier journals. It's not going to change unless there is a wholesale reorganization of incentives, which is unlikely to happen. Open sourcing the research publications for taxpayer funded research is probably what we can strive to achieve.

No, it really isn't like that. I'm not asking someone to do anything, I'm just observing that if someone already does something that we do not berate them for it but instead take their contribution at face value rather than on account of the venue.

It might very well be a big deal.

This has been done with Indium Phosphide (InP). In fact, lasers, modulators and supporting circuitry have have been integrated together on InP in what is usually referred to as "Photonic Integrated Circuits". Infinera has been doing this for ~20 years. The drawback is you need a very specialized InP fab (which ultimately means high-cost and low volume).

If similar stuff could be done on ordinary silicon, that changes the game and opens up a huge range of possibilities.

Does it need to be ordinary silicon, or ordinary silicon processes (including temperature?). One seems much easier than the other, the later seems much more likely to be cheap (I can build a big laser on ordinary silicon today). Much harder to do with ordinary silicon in a somewhat standard process flow.

To be fair, it's a "holy Grail" of the very niche field of photonics. On the other hand, the lack of easily-integrated lasers has been a major hurdle to photonics becoming anything other than a niche field.

After all, the first transistor wasn't remotely competitive with existing vacuum diodes in use at the time.

Breakthrough using extremely rare earth https://en.wikipedia.org/wiki/Thulium which comes at a high price and only provides a low power laser?

Wierd that the article says its low cost. Certainly not a holy grail.

Thulium is only required in small quantities. Here is a back-of-the-envelope calculation: molar mass of thulium: 169g/mol, density: 9.32g/cm3, i.e. 0.055 mol/cm^3 [0]. That's ~ 3.3*10^22 atoms per cubic centimeter.

From the article: A thulium ion dopant concentration of 4.0 × 10^20 cm−3 was measured using Rutherford backscattering spectrometry. [1], i.e. a concentration of roughly 1/100 compared to pure material.

Taking the first search result at face value, a cubic centimeter of 99.9% pure Thulium costs $85 [2]. Since the disks seem to have a diameter of 40μm, only a minuscule fraction of a cubic centimeter of material is needed for each.

I would conclude that the material costs for these disks are not as large as initially assumed.

[0] https://en.wikipedia.org/wiki/Thulium

[1] https://onlinelibrary.wiley.com/doi/10.1002/lpor.202100348#p... (page 8, left column halfway down the page)

[2] https://luciteria.com/metal-cubes/thulium-metal-cubes

The term rare earth is a bit misleading. It's not rare as in uncommon, it's rare as in dilute.

Rare earths are elements that don't form concentrated ores, and instead are spread diffusely in common rocks. Most soils contain between 0.4 and 0.8 ppm Thulium. Luckily Earth has a lot of rock, so in absolute terms there are still lots of rare earth material.

The problem is that rare earths are mined by processing prodigious amounts of earth, typically the waste from some other mining activity. Like most large scale mining operations, it is labor intensive and environmentally destructive. Naturally this leads to such operations being located where labor is cheap and environmental regulations are lax. When people express concern about access to Rare Earths, they really mean access to materials from countries (often one specific country) with cheap labor and poor regulation.

Thulium's expense is due to limited production, as there are few industrial uses for it and those applications that do utilize it tend to need only microscopic amounts, but with high purity.

Axel did it first

Any press release that evokes a response of "oh yeah, my master he already has one and it's very nice-ahhh" is probably heavy on hype and light on substance.

"Please don't post shallow dismissals, especially of other people's work. A good critical comment teaches us something."

https://news.ycombinator.com/newsguidelines.html

What do you mean master?

I believe the part in quotes (that include master) is a reference to a scene in Monty Python and the Holy Grail[0]. In this case, master refers to Guy de Lombard[1].

While the OP may not contribute to deeper discourse on the topic I believe it does point out an issue with press releases vs the research proper. Most often I prefer the research paper to the PR. Then again, in this case, the paper link is right there at the beginning of the second sentence.

If there were a takeaway, it might be something like "submit the research link rather than the press release.

[0] https://www.youtube.com/watch?v=GYcopzJ-T9w

[1] http://www.mit.edu/afs.new/sipb/user/ayshames/Python/FRENCH....

This is correct.

I'm extremely skeptical, because I've heard claims of silicon lasers for a long time but it never materialized.

Also the article seems to plug unrelated scientists with vague affiliations to the University and a weird political spin to the whole thing regarding the author's home country's relationship with the US... despite it not being a US university.

Weird.

> Also the article seems to plug unrelated scientists with vague affiliations to the University.

The only other scientists I noticed were her supervisor and the coauthors of the paper on her research. Did I miss someone?

> a weird political spin to the whole thing regarding the author's home country's relationship with the US... despite it not being a US university

That was in the short biographical section at the end, describing where she got her undergraduate degree and how she ended up McMasters. It is hardly a "weird political spin" to note that she considered US universities but could not get a visa.

I doubt GP would have even commented if she had a more typical bio.

> The only other scientists I noticed were her supervisor and the coauthors of the paper on her research. Did I miss someone?

"Miarabbas Kiani's curiosity-based research is reflected in the story of another McMaster student. Donna Strickland received her BEng in Engineering Physics from McMaster in 1981 and went on to doctoral studies at the University of Rochester. Strickland's work on pulsed lasers with her PhD supervisor Gérard Mourou would lead to their Nobel Prize in Physics in 2018"

I guess they wanted to have the "Nobel" keyword in there so they found the closest person to the institution (some undergrad from 40 years ago)?

> That was in the short biographical section at the end, describing where she got her undergraduate degree and how she ended up McMasters. It is hardly a "weird political spin" to note that she considered US universities but could not get a visa.

I just don't know if my alma mater would brag that "hey, since this guy couldn't get into somewhere else so he picked us!". I get that it's an undergrad's joke to laugh at "The Other Institute of Technology" but for a press release? The whole thing about the visa thing feels forced.

I was a grad student in Canada more than a decade ago. At the time, I noticed a disproportionate number of Iranian students among my peers, many of them very talented and who hailed from selective universities like Sharif. I befriended many of them and most of them told me they ended up in Canada because of how onerous it was to obtain US student visas (even though they had applied and gotten into good schools).

This seeming non sequitur in the press release seems to be referring to that context — seems like it’s adding a human element to what is a dry technical press release.

The narrative doesn’t point to any specific to any executive in power at the time but is more of a pointed reference to US visa policies since 1979 turning talent away.

> who hailed from selective universities like Sharif

Selective for the country I'm sure. Not the world.

> they ended up in Canada because of how onerous it was to obtain US student visas (even though they had applied and gotten into good schools).

I'm sure they were admitted to great US schools. Or at least claimed to. The visa process is long because of potential technological transfers. Also, the fact that countries like Canada aren't as strict is one of the reasons businesses like SpaceX that are subject to ITAR can't hire international applicants, even from countries like Canada.

The "weird political spin":

>Miarabbas Kiani received her master’s in electrical engineering from Shiraz University in Iran, where she specialized in photonic and optoelectronic devices. Encouraged by her supervisor to expand her academic horizons, she considered studying in the U.S. but was unable to obtain a visa. Her father suggested she think of Canada.

Presumably, this is some sort of bizarre prompted response and it's a stretch to think Kiani was simply asked "why did you decide to study at McMaster University" and this was her actual experience.

I noticed this quote as well. It is pointed.

You have to read further to see this happened in 2018. Presumably, this was directed at the policies of the executive at the time.

Indeed. I wonder if they would have mentioned it had her visa denial had happened in 2014 rather than 2018, during a certain administration...

What's puzzling is that it's not a US University. I can't recall any US University press release about innovation in physics mentioning some foreign country's policy in a weird political ending. Maybe it's cultural?

Do you think it's possible it only appears pointed to you because of your own bias?

Reply on news.ycombinator.com