This is cool, but I am more curious about what happens to all the cat 6a cabling in offices, where do we go after 10GBASE-T
Ultra Ethernet Specification v1.0 [pdf]
21–30 of 40 posts
Re: Ultra Ethernet Specification v1.0 [pdf]
#22This is cool, but I am more curious about what happens to all the cat 6a cabling in offices, where do we go after 10GBASE-T
10g is going to remain sufficient for office use for a long time.
Ultra Ethernet will be primarily for cluster/AI/Super computing, trading, server, and backplane scenarios where high speed low-latency throughput turns into money or bang for your buck.
Debbie in accounting will practically never need more than 10g networking to do her job.
Re: Ultra Ethernet Specification v1.0 [pdf]
#23So how long before hardware is available that supports this spec? Would the kit likely be cheaper than infiniband (or even ROCE) equivalents?
Re: Ultra Ethernet Specification v1.0 [pdf]
#24So how long before hardware is available that supports this spec? Would the kit likely be cheaper than infiniband (or even ROCE) equivalents?
It seems there isn't much difference on physical layer with standard 802.3 100Gb/lane. So in principle this is a software addition. Hardware supporting it should be available quickly.
I'm not familiar with infiniband, so can't comment on the pricing.
Re: Ultra Ethernet Specification v1.0 [pdf]
#25This is cool, but I am more curious about what happens to all the cat 6a cabling in offices, where do we go after 10GBASE-T
There's already 25G BASE-T over CAT6 available. I worked on and taped out a 25G PHY in 2019. Cable reach is 30m instead of 100m though. There's work going on 40G for 10m reach.
Re: Ultra Ethernet Specification v1.0 [pdf]
#26This is cool, but I am more curious about what happens to all the cat 6a cabling in offices, where do we go after 10GBASE-T
Desktop computer are barely on 2.5GBASE-T, and there's still 5G in-between. I reckon it'll be 5-10 years until we'll see widespread 10G rollout, and it'll easily last another 10 before we need something better. The bigger issue is wireless access points. High-end ones are already on 10G today, and unlike desktops PoE is quite critical here.
25GBASE-T and 40GBASE-T are dead-on-arrival: the spec exists, but nobody ever bothered to actually ship it. It requires yet another cable upgrade and the maximum distance is too short to be usable. And that's before we even looked at power consumption.
The obvious answer is fiber, but that's orders of magnitude more complicated to roll out. The fiber itself is a massive pain to install when you want to make it to-length on site, the currently-popular connectors require a lot of babysitting, it can't be backwards compatible with the old RJ45 stuff, and it can't do PoE. Pulling a standard single-mode fiber pair with LC connectors to every cubicle and access point? Not exactly an attractive option.
I personally think we're probably going to see some kind of Frankensteined (think GG45-like) RJ45-with-fiber pop up, but that won't be any time soon.
Re: Ultra Ethernet Specification v1.0 [pdf]
#27This is cool, but I am more curious about what happens to all the cat 6a cabling in offices, where do we go after 10GBASE-T
10g is going to remain sufficient for office use for a long time. Ultra Ethernet will be primarily for cluster/AI/Super computing, trading, server, and backplane scenarios where high speed low-latency throughput turns into money or bang for your buck. Debbie in accounting will practically never need more than 10g networking to do her job.
> Debbie in accounting will practically never need more than 10g networking to do her job.
And 640K of RAM ought to be enough for anyone. Who knows, perhaps in 2050 Edward will be mailing Debbie iterations of their 500GB AI fraud model for local inference?
The question isn't if 10G will need to be replaced with something better, but when.
Re: Ultra Ethernet Specification v1.0 [pdf]
#28Earlier quoted context omitted.
10g is going to remain sufficient for office use for a long time. Ultra Ethernet will be primarily for cluster/AI/Super computing, trading, server, and backplane scenarios where high speed low-latency throughput turns into money or bang for your buck. Debbie in accounting will practically never need more than 10g networking to do her job.
> Debbie in accounting will practically never need more than 10g networking to do her job. And 640K of RAM ought to be enough for anyone. Who knows, perhaps in 2050 Edward will be mailing Debbie iterations of their 500GB AI fraud model for local inference? The question isn't if 10G will need to be replaced with something better, but when .
Unless there are sea changes in the global infrastructure of computing in the next 25 years, that 500gb AI fraud model would be hosted on a server and made available to debbie through a web connection.
Further, almost every person on an internal network still uses 1g, which has been going strong for 25 years so far and will likely only cease being the standard when it's no longer economical for manufacturers to produce 1g hardware.
Re: Ultra Ethernet Specification v1.0 [pdf]
#29While currently it seems like the only (?) InfiniBand vendor is Nvida/Mellanox, there used to be more folks selling the gear. For example Intel used to sell switches (had some for an Isilon backend): * https://www.intel.com/content/dam/support/us/en/documents/ne... (PDF) Officially there are a bunch of folks in the IB alliance: * https://www.infinibandta.org/member-listing/
Intel's Omni-Path was very similar but a little different from Infiniband. Anyone remember any details on how; I'm forgetting? Omni-Path was sold off to Cornelis Networks in 2020. Cornelis just released new 400Gbps products last week, their CN5000 line, with 48 and 576 port switches. https://www.crn.com/news/components-peripherals/intel-spins-...
> Intel's Omni-Path was very similar but a little different from Infiniband.
While Intel may have had the 'IB-adjacent' Omni-Path, they also had plain-old IB as well. I ran these/similar switches for an Isilon back-end network (starting when Isilon was still an independent company, pre-EMC buyout, pre-Dell buyout, in the OneFS 6.x days):
* https://www.dell.com/support/kbdoc/en-us/000019215/install-a...
Re: Ultra Ethernet Specification v1.0 [pdf]
#30This is cool, but I am more curious about what happens to all the cat 6a cabling in offices, where do we go after 10GBASE-T
Desktop computer are barely on 2.5GBASE-T, and there's still 5G in-between. I reckon it'll be 5-10 years until we'll see widespread 10G rollout, and it'll easily last another 10 before we need something better. The bigger issue is wireless access points. High-end ones are already on 10G today, and unlike desktops PoE is quite critical here. 25GBASE-T and 40GBASE-T are dead-on-arrival: the spec exists, but nobody ever…
The Realtek RTL8127 chip for 10GbE will cost $10, draw 1.95 W and is designed for motherboards. So I think you're a bit pessimistic on timelines...
https://www.tomshardware.com/networking/realteks-usd10-tiny-...