Great product, but...
...why would Microsoft want to buy the whole company, rather than just buy a few drums of cable?
Unless it wants to block the flow of product to their competitors? Or do some patent trolling?
Major cloud providers and hyperscalers go to considerable lengths to secure technical advantages over their rivals, and Microsoft last week made just such a move by acquiring fiber optic cable maker Lumenisity. Acquiring a cable-maker is less obviously advantageous than AWS creating its own Arm CPUs or Google's tensor …
Because they have an infinite amount of cash in the bank that is losing money.
They can't but a software company without worrying about monopolies, they can't buy a railroad or oil company without worrying about the SEC.
This is close enough to their core business to not make them a hedge fund but far enough not to be anticompetitive
In trading, latency is everything and any comparative advantage on data transmission latencies (even in single digit milliseconds) will be able to drive a real monetary benefit.
Same applied to trade clippers bringing tea in the 19th century - whoever got there first managed to get the best prices.
Gamers also instinctively understand the downside of a laggy connection: You are dead, you just don't know it yet!
For this sort of latency, MS need to be buying boutique fibre ISP's.
From the information provided, MS's primary application would seem to be primarily focused on communications within a datacentre and potentially between Azure datacentres within an area. Having less data "in-flight" would help clustering, off-site backup and failover to a second datacentre.
Latency is only important for high-frequency trading, i.e., some broker trying to interject himself between a buyer and a seller by buying faster and then selling with a mark-up. There is no benefit of high-frequency trading to the wider economy -- it has a negative effect as it drives up the cost of trading. So this shouldn't be a good long term business to invest in, as the business model will vanish once the regulators get their act together and agree on shutting it down.
But there is a benefit of low latency data transmission for almost all aspects of computing. The only question is whether hollow-core fibers are the tool for the job. Trading signal attenuation versus speed means that the benefits of hollow core fibers come with a significant cost. So it's a matter of finding the right niche where this trade-off is worth the cost.
Optoelectronics Research Centre
I get it outside of Commonwealth/British Empire countries use Center and if the place is not on our shores then yeah am fine with that, but at least double check when in Blighty. Annoying people (Like the twit posting this comment) like to nit pick.
@AC "(Since M$ is an American company, El Reg spelled it correctly.)" Yet MS in their blog wrote the company name thus, Optoelectronics Research Centre. Now why the (name of a fashion brand) fcuk* did MS do that?
Because while the noun, adjective, or verb Center (US) is Centre (UK) in this case the word is not being used as a noun, adjective, or verb, it is used as a company name changing the spelling changes the name which no longer refers to the UK entity.
So MS spelled it correctly, El Reg did not. Most likely it was changed by the author's spell checker easy to miss.
*Feel free to use the US spelling.
To the hollow fibre concept but more as a medium to pass a current (idle thoughts on passing current over insulators similar to strip lighting, pretty sure it's highly impractical but still...).
On the speed of hollow fibres though I'd have to wonder on how altitude/humidity would affect said hollow tubes? Or if there would be some way of instead creating a vacuum (obviously not a perfect one, but the point still stands) in them since the maximum speed light can go at is the maximum velocity of sheep in a vacuum and all that.
This very question was discussed briefly at UKNOF. A solid idea if... we/they can overcome the engineering challenges.
Even if you can produce an 80km or 1000km long fibre with a vacuum. What happens if it breaks?! How do you fix that, especially in a deep sea cable setting. Got to get all the water out first?!
Similar way they fix these "air cored" cables if there is a leak underwater, I guess.
I think they are probably concentrating on land based cabling for now, so that they don't have to worry about leaks / pressure and such too much. Although it does bring to mind one other thing - internal condensation due to temperature differences over a length of cable. Lets say it's running in from a warm outside environment to a cool DC environment, or from a winter environment to a warmer environment in a building.
Wonder how it will work under different pressures, especially deep in the water.
I suppose you'd have to purge it first before reapplying the vacuum. Though I would have thought you'd have the same issue even without a vacuum - if water gets in via capillary action I suppose it'd be an issue regardless (especially since water in the open tends to full of stuff that isn't water... Like sand, salt, biological matter etc which I'd assume would lead to "bad things™")
Yes the hollow fibre is quicker, but eight times more lossy. You then need more amplifiers, like EDFA (Europium doped fibre amplifier) which is near zero latency and near noiseless, but there is a limit, and at some point you need to decode and error-correct, which is much slower.
The nice thing about solid fibre is that the "wall" is from a high refractive index (the core) to a low index (the cladding) - which means you get TIR - total internal reflection. The attenuation is just then down to the transparency of the glass, which is amazingly good.
Other common types of fibre, graded-index and single-mode, also rely on a high-index core, and cannot be replicated with an air-core.
"... but eight times more lossy ..."
Where did that information come from? Research (University of Southampton) published in 2020 had HCF comparable to and better than conventional fibre, with further improvement anticipated. (ref https://www.southampton.ac.uk/news/2020/11/hollow-core-fibre.page)
"The nice thing about solid fibre is that the "wall" is from a high refractive index (the core) to a low index (the cladding) - which means you get TIR - total internal reflection. The attenuation is just then down to the transparency of the glass, which is amazingly good."
HCF uses wave-guide principles to keep the beam in the fibre's channel and which has an even lower attenuation.
I seem to recall there has even been work done looking at transmission of power (high energy laser) over HCF.
""... but eight times more lossy ..."
Where did that information come from? "
"The company quotes attenuation rate — signal loss — of 2.5dB per kilometer for 1310nm wavelengths. By comparison Corning's SMF-28, which is commonly used in long-haul networks, manages 0.32dB/km."
You're right! - it only 7.8125 times the attenuation of SMF-28, although I've never seen a long-haul DWDM system that uses 1310nm. That wavelength generally seems to be favoured for short-haul coarse WDM applications.
From the article https://www.southampton.ac.uk/news/2020/11/hollow-core-fibre.page:
"Their first fibres had attenuations of 5 decibels (dB), i.e. only 30% of light transmission, for every metre of fibre. New physical understanding with contributions from the worldwide community, and substantial development in fabrication technology led by the Southampton team, have now led one of the fibres reported in this study to improve this by a factor of 10,000 by achieving an attenuation of only 5 dB every 10 kilometres."
So that is 0,,5 dB /km, getting close to normal SMF already, and they think they can beat SMF in the long run, interesting to see how that evolves.
The one question I have is how are you going to splice such a cable without blocking the core.
I expect that there are a great many interesting aspects to this (it is not an area I am familiar with), but splicing does not look to be a show-stopping problem - I have seen papers from 2013 discussing it.
While not addressing splicing in particular the following 2020 article gives an introduction to HCF:
It would seem that Corning have got solid fibre close to the theoretical limits; HCF offers an approach that might have further potential. Lumenisity seem to have moved the technology from research quantities into commercially useful (for some applications) quantities. ISTR there were some BT trials starting a couple of years back. There is probably a great deal more work to be done, but to build up a commercial business requires investment - Microsoft might have made an astute move.
Of course there can be many a slip twixt cup and lip - but I hope that it turns out well and provides a significant breakthrough of fibre optics research moving into widespread use.
"The one question I have is how are you going to splice such a cable without blocking the core."
Oh here you go, from a search on Prof David Richardson's publications (from his UoS profile):
"Robust low loss splicing of hollow core photonic bandgap fiber to itself"
IIRC, back quite a ways some researchers into HCF for latency pivoted to "energy delivery" in the form of laser surgery using HCF to minimize tissue damage from "waste energy". This was (according to the article I read at the time) in place of a previous system involving mirrors (but hopefully no smoke)
Perhaps MSFT is looking to make the Blue Screen of Death more literal?
Db loss of 2.5 db/km? Soup tins and string are the next item down on that table. Seems to me this is a purely speculative purchase, hinged on the hope that they can somehow make this shit even vaguely comparable to solid fibre on that stat alone. Yes. I remain long on Corning.
There was a radio piece done where they interviewed someone who was running one of the early computerized trading businesses. He told the story of how someone came smugly walking into his office one day and said how they were going to be building a new data center closer to the NYSE than anyone else and they were selling space for some ridiculous eye-watering sum. At first the guy yells at them and tells them to get out of his office, but before they can even make it to the elevator he has someone bring them back and he signs a contract. He realized that he couldn't afford NOT to do this because even a few milliseconds could mean millions of dollars in extra revenue.
Microsoft's own "Project Natick" proposed a solution to this as well: containerised submarine DCs.
I forget the exact quote but a very high percentage of the world's trading centers are close to water, so this offered a way of using the sea for cooling and low latency by being physically close to the trading floor. Using Azure Stack HCI you have pretty much feature parity with standard Azure and a single management plane as well for this.
At the other end of the scale, MS invested in LSEG so there's probably some commitment on the LSEG side to use Azure Cloud, you would imagine: https://www.theregister.com/2022/12/12/london_stock_exchange_ms/?td=rt-3a
It's a spinoff company from Southampton Uni, which is pretty cool, and I think Southampton also worked with MS a couple years ago with Project Silica for data archiving in glass.
If you can shift bits 47% faster between HCI, then that's pretty awesome in an HCI data center, of which MS has hundreds across the globe.
Seems a very smart buy to me!