* [Starlink] mems optical switching @ 2023-03-18 22:19 Dave Taht 2023-03-19 12:58 ` Michael Richardson 2023-03-19 14:16 ` Brandon Butterworth 0 siblings, 2 replies; 8+ messages in thread From: Dave Taht @ 2023-03-18 22:19 UTC (permalink / raw) To: Dave Taht via Starlink I think, in general, we have gathered here quite a few folk that are interested in new technology, and the list does veer away from starlink issues more often than not. Today, this about google's mems switching tech hit, and I keep wondering where else it could be applied. https://www.semianalysis.com/p/google-apollo-the-3-billion-game -- Come Heckle Mar 6-9 at: https://www.understandinglatency.com/ Dave Täht CEO, TekLibre, LLC ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-18 22:19 [Starlink] mems optical switching Dave Taht @ 2023-03-19 12:58 ` Michael Richardson 2023-03-19 14:16 ` Brandon Butterworth 1 sibling, 0 replies; 8+ messages in thread From: Michael Richardson @ 2023-03-19 12:58 UTC (permalink / raw) To: Dave Taht via Starlink [-- Attachment #1: Type: text/plain, Size: 690 bytes --] Dave Taht via Starlink <starlink@lists.bufferbloat.net> wrote: > I think, in general, we have gathered here quite a few folk that are > interested in new technology, and the list > does veer away from starlink issues more often than not. Today, this > about google's mems switching tech hit, > and I keep wondering where else it could be applied. > https://www.semianalysis.com/p/google-apollo-the-3-billion-game An interesting read. Starlink content: could they avoid latency and power consumption by using optical switching in the satellites. I suspect it's irrelevant, because the satellites don't have more than about four lasers. (I think I heard four) [-- Attachment #2: signature.asc --] [-- Type: application/pgp-signature, Size: 487 bytes --] ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-18 22:19 [Starlink] mems optical switching Dave Taht 2023-03-19 12:58 ` Michael Richardson @ 2023-03-19 14:16 ` Brandon Butterworth 2023-03-19 14:33 ` Christian von der Ropp 1 sibling, 1 reply; 8+ messages in thread From: Brandon Butterworth @ 2023-03-19 14:16 UTC (permalink / raw) To: Dave Taht; +Cc: Dave Taht via Starlink, brandon On Sat Mar 18, 2023 at 03:19:49PM -0700, Dave Taht via Starlink wrote: > Today, this about google's mems switching tech hit, They've been talking about it since last year, seems to have got a hype bump recently. Who expected circuit switching to make a comeback? > I keep wondering where else it could be applied. They've been used for a long time, eg almost 20 years ago - https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf There is a goal of optical packet switching, until then you're limited to where there are limited flows of long enough duration to make the change from packet to circuit switching viable. So mostly automated testing. I've dabbled with the idea in an archive use case where very few of a large set of storage nodes need to connect to a moderate number of servers. For some cases we could have zero switches. The goal was a mostly dark infrastructure and many 1000s of storage nodes, removing the switches saves a lot of power. Commercial optical switches are expensive so I was looking at making an optical strowger as I wanted a high fan out not large n^2. In the mobile world they are looking at doing flexible bandwidth per node with coherent optics over gpon fibre plant, allocating variable amounts of spectrum to each, which could be adapted to a similar circuit model. It'd be no use to google as they want the full bandwidth between each node but as dwdm coherent optic costs come down you could imagine doing the same with a full channel between each pair, so like a conventional WSS but cheaper. If it wasn't for the optics cost I suspect they'd have done that reducing switching time to a channel change. brandon ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-19 14:16 ` Brandon Butterworth @ 2023-03-19 14:33 ` Christian von der Ropp 2023-03-20 11:32 ` Dave Taht 0 siblings, 1 reply; 8+ messages in thread From: Christian von der Ropp @ 2023-03-19 14:33 UTC (permalink / raw) To: starlink All-optical switching could greatly reduce complexity and power consumption on the satellites at the cost of flexibility. Up to 44 satellites in an orbital plane would use individual wavelengths which would be passed on transparently down the daisy chain and only satellites in range of gateways would convert the optical signals back into electrical ones, and send them down to earth while they pass a gateway. This would result in relatively short duty cycles, hence less power draw per orbit and less heat dissipation issues. Actually I've been suspecting that the SDA targets all-optical switching for the Transport Layer constellation as I don't seen any other immediate reason for the requirement of their OISL standard to require wavelength switching within the ITU channel grid for LCTs (see p. 18 of the OISL 3.0: https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf) As a matter of fact tuneable wavelenghts were already required in the draft version of the OISL standard published in April 2020: https://twitter.com/Megaconstellati/status/1310336728595562499 -Christian Am 19.03.2023 um 16:16 schrieb Brandon Butterworth via Starlink: > On Sat Mar 18, 2023 at 03:19:49PM -0700, Dave Taht via Starlink wrote: >> Today, this about google's mems switching tech hit, > They've been talking about it since last year, seems to have got > a hype bump recently. > > Who expected circuit switching to make a comeback? > >> I keep wondering where else it could be applied. > They've been used for a long time, eg almost 20 years ago - > https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf > > There is a goal of optical packet switching, until then you're > limited to where there are limited flows of long enough duration > to make the change from packet to circuit switching viable. So mostly > automated testing. > > I've dabbled with the idea in an archive use case where very few of > a large set of storage nodes need to connect to a moderate number > of servers. For some cases we could have zero switches. The goal was > a mostly dark infrastructure and many 1000s of storage nodes, > removing the switches saves a lot of power. > > Commercial optical switches are expensive so I was looking at > making an optical strowger as I wanted a high fan out not > large n^2. > > In the mobile world they are looking at doing flexible bandwidth > per node with coherent optics over gpon fibre plant, allocating > variable amounts of spectrum to each, which could be adapted to a > similar circuit model. It'd be no use to google as they want the > full bandwidth between each node but as dwdm coherent optic costs > come down you could imagine doing the same with a full channel > between each pair, so like a conventional WSS but cheaper. If it > wasn't for the optics cost I suspect they'd have done that reducing > switching time to a channel change. > > brandon > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-19 14:33 ` Christian von der Ropp @ 2023-03-20 11:32 ` Dave Taht 2023-03-20 11:42 ` David Lang 2023-03-20 11:46 ` Mike Puchol 0 siblings, 2 replies; 8+ messages in thread From: Dave Taht @ 2023-03-20 11:32 UTC (permalink / raw) To: Christian von der Ropp; +Cc: starlink We haven't heard much about the starlink ISL links lately. Any sign they are working anywhere yet? On Sun, Mar 19, 2023 at 7:33 AM Christian von der Ropp via Starlink <starlink@lists.bufferbloat.net> wrote: > > All-optical switching could greatly reduce complexity and power > consumption on the satellites at the cost of flexibility. Up to 44 > satellites in an orbital plane would use individual wavelengths which > would be passed on transparently down the daisy chain and only > satellites in range of gateways would convert the optical signals back > into electrical ones, and send them down to earth while they pass a > gateway. This would result in relatively short duty cycles, hence less > power draw per orbit and less heat dissipation issues. > > Actually I've been suspecting that the SDA targets all-optical switching > for the Transport Layer constellation as I don't seen any other > immediate reason for the requirement of their OISL standard to require > wavelength switching within the ITU channel grid for LCTs (see p. 18 of > the OISL 3.0: > https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf) > > As a matter of fact tuneable wavelenghts were already required in the > draft version of the OISL standard published in April 2020: > https://twitter.com/Megaconstellati/status/1310336728595562499 > > -Christian > > Am 19.03.2023 um 16:16 schrieb Brandon Butterworth via Starlink: > > On Sat Mar 18, 2023 at 03:19:49PM -0700, Dave Taht via Starlink wrote: > >> Today, this about google's mems switching tech hit, > > They've been talking about it since last year, seems to have got > > a hype bump recently. > > > > Who expected circuit switching to make a comeback? > > > >> I keep wondering where else it could be applied. > > They've been used for a long time, eg almost 20 years ago - > > https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf > > > > There is a goal of optical packet switching, until then you're > > limited to where there are limited flows of long enough duration > > to make the change from packet to circuit switching viable. So mostly > > automated testing. > > > > I've dabbled with the idea in an archive use case where very few of > > a large set of storage nodes need to connect to a moderate number > > of servers. For some cases we could have zero switches. The goal was > > a mostly dark infrastructure and many 1000s of storage nodes, > > removing the switches saves a lot of power. > > > > Commercial optical switches are expensive so I was looking at > > making an optical strowger as I wanted a high fan out not > > large n^2. > > > > In the mobile world they are looking at doing flexible bandwidth > > per node with coherent optics over gpon fibre plant, allocating > > variable amounts of spectrum to each, which could be adapted to a > > similar circuit model. It'd be no use to google as they want the > > full bandwidth between each node but as dwdm coherent optic costs > > come down you could imagine doing the same with a full channel > > between each pair, so like a conventional WSS but cheaper. If it > > wasn't for the optics cost I suspect they'd have done that reducing > > switching time to a channel change. > > > > brandon > > _______________________________________________ > > Starlink mailing list > > Starlink@lists.bufferbloat.net > > https://lists.bufferbloat.net/listinfo/starlink > > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink -- Come Heckle Mar 6-9 at: https://www.understandinglatency.com/ Dave Täht CEO, TekLibre, LLC ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-20 11:32 ` Dave Taht @ 2023-03-20 11:42 ` David Lang 2023-03-20 11:46 ` Mike Puchol 1 sibling, 0 replies; 8+ messages in thread From: David Lang @ 2023-03-20 11:42 UTC (permalink / raw) To: Dave Taht; +Cc: Christian von der Ropp, starlink [-- Attachment #1: Type: text/plain, Size: 4510 bytes --] Their coverage of the poles and oceans is growing outside of the 'bent pipe' model, so I think that's proof that they are working. I am in southern california and commonly go out a gateway in vancouver BC, I think that's more than a simple bent-pipe hop away. David Lang On Mon, 20 Mar 2023, Dave Taht via Starlink wrote: > Date: Mon, 20 Mar 2023 04:32:19 -0700 > From: Dave Taht via Starlink <starlink@lists.bufferbloat.net> > Reply-To: Dave Taht <dave.taht@gmail.com> > To: Christian von der Ropp <cvdr@vdr.net> > Cc: starlink@lists.bufferbloat.net > Subject: Re: [Starlink] mems optical switching > > We haven't heard much about the starlink ISL links lately. Any sign > they are working anywhere yet? > > On Sun, Mar 19, 2023 at 7:33 AM Christian von der Ropp via Starlink > <starlink@lists.bufferbloat.net> wrote: >> >> All-optical switching could greatly reduce complexity and power >> consumption on the satellites at the cost of flexibility. Up to 44 >> satellites in an orbital plane would use individual wavelengths which >> would be passed on transparently down the daisy chain and only >> satellites in range of gateways would convert the optical signals back >> into electrical ones, and send them down to earth while they pass a >> gateway. This would result in relatively short duty cycles, hence less >> power draw per orbit and less heat dissipation issues. >> >> Actually I've been suspecting that the SDA targets all-optical switching >> for the Transport Layer constellation as I don't seen any other >> immediate reason for the requirement of their OISL standard to require >> wavelength switching within the ITU channel grid for LCTs (see p. 18 of >> the OISL 3.0: >> https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf) >> >> As a matter of fact tuneable wavelenghts were already required in the >> draft version of the OISL standard published in April 2020: >> https://twitter.com/Megaconstellati/status/1310336728595562499 >> >> -Christian >> >> Am 19.03.2023 um 16:16 schrieb Brandon Butterworth via Starlink: >> > On Sat Mar 18, 2023 at 03:19:49PM -0700, Dave Taht via Starlink wrote: >> >> Today, this about google's mems switching tech hit, >> > They've been talking about it since last year, seems to have got >> > a hype bump recently. >> > >> > Who expected circuit switching to make a comeback? >> > >> >> I keep wondering where else it could be applied. >> > They've been used for a long time, eg almost 20 years ago - >> > https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf >> > >> > There is a goal of optical packet switching, until then you're >> > limited to where there are limited flows of long enough duration >> > to make the change from packet to circuit switching viable. So mostly >> > automated testing. >> > >> > I've dabbled with the idea in an archive use case where very few of >> > a large set of storage nodes need to connect to a moderate number >> > of servers. For some cases we could have zero switches. The goal was >> > a mostly dark infrastructure and many 1000s of storage nodes, >> > removing the switches saves a lot of power. >> > >> > Commercial optical switches are expensive so I was looking at >> > making an optical strowger as I wanted a high fan out not >> > large n^2. >> > >> > In the mobile world they are looking at doing flexible bandwidth >> > per node with coherent optics over gpon fibre plant, allocating >> > variable amounts of spectrum to each, which could be adapted to a >> > similar circuit model. It'd be no use to google as they want the >> > full bandwidth between each node but as dwdm coherent optic costs >> > come down you could imagine doing the same with a full channel >> > between each pair, so like a conventional WSS but cheaper. If it >> > wasn't for the optics cost I suspect they'd have done that reducing >> > switching time to a channel change. >> > >> > brandon >> > _______________________________________________ >> > Starlink mailing list >> > Starlink@lists.bufferbloat.net >> > https://lists.bufferbloat.net/listinfo/starlink >> >> _______________________________________________ >> Starlink mailing list >> Starlink@lists.bufferbloat.net >> https://lists.bufferbloat.net/listinfo/starlink > > > > -- > Come Heckle Mar 6-9 at: https://www.understandinglatency.com/ > Dave Täht CEO, TekLibre, LLC > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-20 11:32 ` Dave Taht 2023-03-20 11:42 ` David Lang @ 2023-03-20 11:46 ` Mike Puchol 2023-03-20 18:55 ` Ulrich Speidel 1 sibling, 1 reply; 8+ messages in thread From: Mike Puchol @ 2023-03-20 11:46 UTC (permalink / raw) To: Dave Taht, starlink [-- Attachment #1: Type: text/plain, Size: 5145 bytes --] I tested them in the middle of nowhere (literally) and it worked, no gateways anywhere even close. Latencies varied between 20-30ms and up to 400ms (to the POP), and there were occasional outages. What was fascinating was that I could predict when an outage would take place by watching starlink.sx and seeing that the ISL-linked satellites would dissapear from view with no others to take over, and also predicting when I'd have service again by watching the next ISL-linked sat approach the field-of-view. The other test I did was, while watching the tracker and the azimuth to the satellite, place myself at the right position between the antenna and the satellite, and confirm that the link would drop. All of the above was possible because a) there weren't that many ISL satellites in orbit at the time, and b) the density of satellites in the region I was testing was low, so only 1-2 satellites could be serving my location at a time, making tests easier. Best, Mike On Mar 20, 2023 at 12:32 +0100, Dave Taht via Starlink <starlink@lists.bufferbloat.net>, wrote: > We haven't heard much about the starlink ISL links lately. Any sign > they are working anywhere yet? > > On Sun, Mar 19, 2023 at 7:33 AM Christian von der Ropp via Starlink > <starlink@lists.bufferbloat.net> wrote: > > > > All-optical switching could greatly reduce complexity and power > > consumption on the satellites at the cost of flexibility. Up to 44 > > satellites in an orbital plane would use individual wavelengths which > > would be passed on transparently down the daisy chain and only > > satellites in range of gateways would convert the optical signals back > > into electrical ones, and send them down to earth while they pass a > > gateway. This would result in relatively short duty cycles, hence less > > power draw per orbit and less heat dissipation issues. > > > > Actually I've been suspecting that the SDA targets all-optical switching > > for the Transport Layer constellation as I don't seen any other > > immediate reason for the requirement of their OISL standard to require > > wavelength switching within the ITU channel grid for LCTs (see p. 18 of > > the OISL 3.0: > > https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf) > > > > As a matter of fact tuneable wavelenghts were already required in the > > draft version of the OISL standard published in April 2020: > > https://twitter.com/Megaconstellati/status/1310336728595562499 > > > > -Christian > > > > Am 19.03.2023 um 16:16 schrieb Brandon Butterworth via Starlink: > > > On Sat Mar 18, 2023 at 03:19:49PM -0700, Dave Taht via Starlink wrote: > > > > Today, this about google's mems switching tech hit, > > > They've been talking about it since last year, seems to have got > > > a hype bump recently. > > > > > > Who expected circuit switching to make a comeback? > > > > > > > I keep wondering where else it could be applied. > > > They've been used for a long time, eg almost 20 years ago - > > > https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf > > > > > > There is a goal of optical packet switching, until then you're > > > limited to where there are limited flows of long enough duration > > > to make the change from packet to circuit switching viable. So mostly > > > automated testing. > > > > > > I've dabbled with the idea in an archive use case where very few of > > > a large set of storage nodes need to connect to a moderate number > > > of servers. For some cases we could have zero switches. The goal was > > > a mostly dark infrastructure and many 1000s of storage nodes, > > > removing the switches saves a lot of power. > > > > > > Commercial optical switches are expensive so I was looking at > > > making an optical strowger as I wanted a high fan out not > > > large n^2. > > > > > > In the mobile world they are looking at doing flexible bandwidth > > > per node with coherent optics over gpon fibre plant, allocating > > > variable amounts of spectrum to each, which could be adapted to a > > > similar circuit model. It'd be no use to google as they want the > > > full bandwidth between each node but as dwdm coherent optic costs > > > come down you could imagine doing the same with a full channel > > > between each pair, so like a conventional WSS but cheaper. If it > > > wasn't for the optics cost I suspect they'd have done that reducing > > > switching time to a channel change. > > > > > > brandon > > > _______________________________________________ > > > Starlink mailing list > > > Starlink@lists.bufferbloat.net > > > https://lists.bufferbloat.net/listinfo/starlink > > > > _______________________________________________ > > Starlink mailing list > > Starlink@lists.bufferbloat.net > > https://lists.bufferbloat.net/listinfo/starlink > > > > -- > Come Heckle Mar 6-9 at: https://www.understandinglatency.com/ > Dave Täht CEO, TekLibre, LLC > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink [-- Attachment #2: Type: text/html, Size: 5946 bytes --] ^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [Starlink] mems optical switching 2023-03-20 11:46 ` Mike Puchol @ 2023-03-20 18:55 ` Ulrich Speidel 0 siblings, 0 replies; 8+ messages in thread From: Ulrich Speidel @ 2023-03-20 18:55 UTC (permalink / raw) To: starlink [-- Attachment #1: Type: text/plain, Size: 7635 bytes --] Not so fast! A number of issues here, and one of them is that the question itself is unclear. There are several levels of "working": * A satellite forwarding uplink traffic to an adjacent peer within the same orbital plane for downlink. o Dto for cross-plane * A satellite forwarding incoming laser link traffic to another satellite in the same orbital plane. o Dto for cross-plane * Routing to select destinations only. * Routing to everywhere. * Doing so efficiently globally under load. I'm pretty sure that they are managing somewhere around the first 3-4 major bullet points by now, with probably some outages (I note that Starlink now advertises global maritime coverage, at a price around 10 times that of a land-based subscription). But the last point is where it truly gets hard. The other thing worth remembering is that distance isn't necessarily a good indicator - it depends on where that distance accrues. Southern California to southern BC is around 2000 km, and the zone around the US-Canada border is where satellite density is highest - and capacity attracts in Starlink. Note also that gateways are generally in spots where there is little obstruction in terms of elevation, so in the case of Vancouver BC can probably serve birds that are just a few degrees over the horizon (which also helps keeping the beam out of geostationary trouble). Read: a lot stays bent pipe. On 21/03/2023 12:46 am, Mike Puchol via Starlink wrote: > I tested them in the middle of nowhere (literally) and it worked, no > gateways anywhere even close. Latencies varied between 20-30ms and up > to 400ms (to the POP), and there were occasional outages. > > What was fascinating was that I could predict when an outage would > take place by watching starlink.sx and seeing that the ISL-linked > satellites would dissapear from view with no others to take over, and > also predicting when I'd have service again by watching the next > ISL-linked sat approach the field-of-view. > > The other test I did was, while watching the tracker and the azimuth > to the satellite, place myself at the right position between the > antenna and the satellite, and confirm that the link would drop. > > All of the above was possible because a) there weren't that many ISL > satellites in orbit at the time, and b) the density of satellites in > the region I was testing was low, so only 1-2 satellites could be > serving my location at a time, making tests easier. > > Best, > > Mike > On Mar 20, 2023 at 12:32 +0100, Dave Taht via Starlink > <starlink@lists.bufferbloat.net>, wrote: >> We haven't heard much about the starlink ISL links lately. Any sign >> they are working anywhere yet? >> >> On Sun, Mar 19, 2023 at 7:33 AM Christian von der Ropp via Starlink >> <starlink@lists.bufferbloat.net> wrote: >>> >>> All-optical switching could greatly reduce complexity and power >>> consumption on the satellites at the cost of flexibility. Up to 44 >>> satellites in an orbital plane would use individual wavelengths which >>> would be passed on transparently down the daisy chain and only >>> satellites in range of gateways would convert the optical signals back >>> into electrical ones, and send them down to earth while they pass a >>> gateway. This would result in relatively short duty cycles, hence less >>> power draw per orbit and less heat dissipation issues. >>> >>> Actually I've been suspecting that the SDA targets all-optical switching >>> for the Transport Layer constellation as I don't seen any other >>> immediate reason for the requirement of their OISL standard to require >>> wavelength switching within the ITU channel grid for LCTs (see p. 18 of >>> the OISL 3.0: >>> https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf) >>> <https://www.sda.mil/wp-content/uploads/2022/04/SDA-OCT-Standard-v3.0.pdf)> >>> >>> As a matter of fact tuneable wavelenghts were already required in the >>> draft version of the OISL standard published in April 2020: >>> https://twitter.com/Megaconstellati/status/1310336728595562499 >>> <https://twitter.com/Megaconstellati/status/1310336728595562499> >>> >>> -Christian >>> >>> Am 19.03.2023 um 16:16 schrieb Brandon Butterworth via Starlink: >>>> On Sat Mar 18, 2023 at 03:19:49PM -0700, Dave Taht via Starlink wrote: >>>>> Today, this about google's mems switching tech hit, >>>> They've been talking about it since last year, seems to have got >>>> a hype bump recently. >>>> >>>> Who expected circuit switching to make a comeback? >>>> >>>>> I keep wondering where else it could be applied. >>>> They've been used for a long time, eg almost 20 years ago - >>>> https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf >>>> <https://archive.nanog.org/meetings/nanog32/presentations/zwart.pdf> >>>> >>>> There is a goal of optical packet switching, until then you're >>>> limited to where there are limited flows of long enough duration >>>> to make the change from packet to circuit switching viable. So mostly >>>> automated testing. >>>> >>>> I've dabbled with the idea in an archive use case where very few of >>>> a large set of storage nodes need to connect to a moderate number >>>> of servers. For some cases we could have zero switches. The goal was >>>> a mostly dark infrastructure and many 1000s of storage nodes, >>>> removing the switches saves a lot of power. >>>> >>>> Commercial optical switches are expensive so I was looking at >>>> making an optical strowger as I wanted a high fan out not >>>> large n^2. >>>> >>>> In the mobile world they are looking at doing flexible bandwidth >>>> per node with coherent optics over gpon fibre plant, allocating >>>> variable amounts of spectrum to each, which could be adapted to a >>>> similar circuit model. It'd be no use to google as they want the >>>> full bandwidth between each node but as dwdm coherent optic costs >>>> come down you could imagine doing the same with a full channel >>>> between each pair, so like a conventional WSS but cheaper. If it >>>> wasn't for the optics cost I suspect they'd have done that reducing >>>> switching time to a channel change. >>>> >>>> brandon >>>> _______________________________________________ >>>> Starlink mailing list >>>> Starlink@lists.bufferbloat.net >>>> https://lists.bufferbloat.net/listinfo/starlink >>>> <https://lists.bufferbloat.net/listinfo/starlink> >>> >>> _______________________________________________ >>> Starlink mailing list >>> Starlink@lists.bufferbloat.net >>> https://lists.bufferbloat.net/listinfo/starlink >>> <https://lists.bufferbloat.net/listinfo/starlink> >> >> >> >> -- >> Come Heckle Mar 6-9 at: https://www.understandinglatency.com/ >> <https://www.understandinglatency.com/> >> Dave Täht CEO, TekLibre, LLC >> _______________________________________________ >> Starlink mailing list >> Starlink@lists.bufferbloat.net >> https://lists.bufferbloat.net/listinfo/starlink >> <https://lists.bufferbloat.net/listinfo/starlink> > > _______________________________________________ > Starlink mailing list > Starlink@lists.bufferbloat.net > https://lists.bufferbloat.net/listinfo/starlink > -- **************************************************************** Dr. Ulrich Speidel School of Computer Science Room 303S.594 (City Campus) The University of Auckland u.speidel@auckland.ac.nz http://www.cs.auckland.ac.nz/~ulrich/ **************************************************************** [-- Attachment #2: Type: text/html, Size: 11823 bytes --] ^ permalink raw reply [flat|nested] 8+ messages in thread
end of thread, other threads:[~2023-03-20 18:55 UTC | newest] Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed) -- links below jump to the message on this page -- 2023-03-18 22:19 [Starlink] mems optical switching Dave Taht 2023-03-19 12:58 ` Michael Richardson 2023-03-19 14:16 ` Brandon Butterworth 2023-03-19 14:33 ` Christian von der Ropp 2023-03-20 11:32 ` Dave Taht 2023-03-20 11:42 ` David Lang 2023-03-20 11:46 ` Mike Puchol 2023-03-20 18:55 ` Ulrich Speidel
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox