I'm using the sonnet at work. Work paid for it. Yes it's more expensive. I have gotten over 25 gbps bidirectional, but not by much. Maybe 27. A big downside is that it only supports 15W upstream power which I didn't think to check, so since that laptop only has 2 usb-c ports for power input, it's a bit limiting.
The upside, though, was that I plugged it in, got my throughput, and never had to worry ever again. Remember that engineering time is mucho $$. So troubleshooting overheating, divising a solution, manufacturing it, etc, that all eats into the savings.
The $400 dock would be cool though. It seems smaller than the sonnet and fixes my charging problem. Weirdly I didn't think I'd care about not including transceivers but they did actually come in handy, so that is a downside. Might try to get one of these if I need another one in the future.
ConnectX-4 and newer are natively supported. I use a ConnectX-4 LX in a $50 Razer CoreX off of craigslist. Works fine, hit similar limits to what Jeff did.
Using an OCP is the neat trick here. MUCH smaller setup.
As far as the other posters comment about Tx's... that is an entirely different topic. I've never had an issue with FS.com optics personally.
Absolutely not. It's another layer of throughput/compatability bingo on top of the already often quite annoying SFP bingo.
If you are unfamiliar, the reality of the situation is quite bad. Slight hardware revisions of adapters/switches, which if you are lucky can be differentiated by manufacture date or serial number sequence, will not recognize known good modules, and if they do you then have to actually test if they can establish link, then let them burn for 24+ hours to see if they can maintain it without dropping (often, because of forward error correction, you will not start to see flapping until enough correctable errors have accumulated and the buffer overflows). Then there are firmware updates where this process must then be repeated.
You would think with how expensive all of this shit is you would simply be able to buy the officially listed as compatible components but my experience with them has been that a shotgun approach is best for your sanity and deadlines, buying sample kits of each module and trying them all, and then once you find the ones that play nice, buying as many of them as is reasonable and testing all of them.
If I had a dollar every time two modules from the same brand, that came in the same box, from the same production run, did not work with one another, I would be able to test your theory for free. Network Engineering is Fun
For us <25Gbps plebes out there: For god's sake do not buy a USB-C RealTek RTL8156-equipped multi-gig ethernet dongle to use with your Mac. I went through 3, and they're all trash:
Plugable 2.5G USB-C to ethernet adapter (bought November 2025), SABRENT USB C to 2.5Gbps ethernet adapter (bought October 2024), UGREEN USB C 2.5G ethernet adapter (also October 2024). In my experience they immediately fall on their face in a speedtest and drop to a couple hundred Mbps, experience packet loss, etc. This is testing across a few modern Apple Silicon Macs.
I now have a WavLink WL-NWU341G 5Gbps dongle with the RTL-8157 that seems to work well, but I've only run a couple of tests-- I am away from my home setup and don't have wired ethernet run to all the rooms where I'm living.
If you do some Googling, there's some good information out there about it, but when I bought the bad ones, I couldn't find any information corroborating my experience.
As mentioned in the blog post's footnote, the bottleneck may just be on the NAS side:
> this real-world speed limitation could also be related to my use of my lower-power Arm NAS (running on Ampere Altra, with 32 fairly-slow CPU cores). I was testing writing to an array of fast enterprise NVMe SSDs, so they shouldn't be the bottleneck here.
He was only getting 1 Gigabyte of throughput out of the built in 10 Gigabit Ethernet connection and moving to the 25 Gigabit connection only made a marginal impact on throughput, after all.
One of the YouTube comments pointed out the issue is most likely the lack of support for SMB Direct (RDMA) in macOS. Would be nice to repeat the test on a Windows/Linux laptop.
That is just RDMA itself and seems to be aimed directly at distributed inference (no surprise there!). Now they need to update the SMB driver to make use of it.
Yeah, raw RDMA is one thing, making sure other protocols can use it is another... hopefully Apple gets there, but right now SMB with RDMA is probably a tiny, tiny niche.
I do wonder what Apple's custom-built servers are running for networking. I wish they still made Mac Pro-level chassis specifically for better IO. Thunderbolt 5 is nice, but far inferior to a PCIe x16 slot (which can hit 400 Gbps networking) nowadays.
I'd suspect that (from Apple's perspective) any form of file sharing (other than iCloud) is now considered 'extremely niche.'
If they were focused on the needs of the kind of person who moves enough terabytes of files around to need this kind of speed, we'd see them still making either computers with slots, or first-party 25Gbps networking.
I'm not really trying to criticize them for it -- Tim Cook's bean counters clearly did the math and they're certainly free to optimize their Mac R&D spend for things that'll sell college students MacBook Neos and Airs, rather than worrying about the MKBHDs, Linus Sebastians and Jeff Geerlings of the world (this is just a list of people I've seen using extremely fast networking).
Is RDMA something worth looking into for connecting two random Linux computers? I've been extremely underwhelmed by the stability and performance of my NFS shares.
If your storage devices are native PCIe and your network is actually fast enough and your nics and switches have the proper hardware and driver support and your switches are properly configured and your nfs client and server configurations are tweaked just so, and you hold your face in just the right position
I do have stable diffusion running on a non-Nvidia GPU, so I'm familiar with the manner of rituals required before contemplating installing system updates or even suspending the system.
Now, I'm off to determine if an eMMC M.2 module is a native PCIe device…
This is a ConnextX4 on a thunderbolt bridge. This is well trodden territory. Unless some hero of open source has come to bat recently, MacOS doesn't support RDMA with this card or over Ethernet at all for that matter. It's only supported over Thunderbolt.
His iperf3 results show that it's able to go full beans on a single 25g port. A cross platform SMB file copy is kind of a meaningless benchmark in this context.
IMO a pretty cool product would be a network switch providing native thunderbolt interfaces to multiple upstream hosts. That wouldn't magically fix this RDMA issue, though. This is just a separate thought. Seems like the kind of thing Mikrotik should make.
Commented on another top comment but yeah, I've been running a similar setup for about $200 all in. Using the PCIe LX version takes care of much of the heat issue, and an eGPU dock (Razer Core in my case) takes care of the fan. Craigslist and eBay were the sources.
Speeds aren't any better, but at least it was cheap and fiber.
I have a 25GbE setup using Thunderbolt, kinda similar to the author's. Except that my fan is held together by a piece of string, not a 3D-printed enclosure.
I found that SMB on macOS is just bad. If I try to heavily load it (e.g. by doing mass conversions in my Audible library), it can cause stalls and disconnections.
I managed to find a set of parameters (in /etc/nsmbd.conf) that makes it work more-or-less reliably by trial-and-error: dir_cache_off=yes, file_ids_off=yes, notify_off=yes, mc_on=no.
It's still terrible, though. It's so bad that Linux in a Parallels VM works _better_.
It's always been like this, too! SMB or otherwise. Any company/person that has been using macs long enough has been burned by silently corrupted volumes or etc and has a thorough system of multiple on and offsite backups.
It's not just software companies or home enthusiasts either. There are a lot of ie medical practices or accounting/benefits agencies out there who have sworn off Apple for life after having to spend extreme amounts of money on data recovery to hopefully get their years of records back, usually in a do or die situation like an IRS audit which causes them to discover that their basement was flooding and all of the paper records are destroyed.
Dang, here I was feeling slick with 10Gb between my MBP and Mac mini and NAS. That’s the thing, it could always be bigger / faster / more. It’s enough for my workflow, but it’s fun to see people push further.
My default assumption (with no knowledge about this particular case) is that all gear in YouTube videos is sponsored. Too easy to get jealous of all the fancy setups. But I don't want to become a YouTuber so that settles that.
The Mac Studio in most of my videos is my own workstation, purchased a while back.
The 4 Mac Studio cluster I tested with Exo and Thunderbolt RDMA was a loaner cluster from Apple. There's no universe in which I could justify buying that.
It’s crazy that the Ethernet chips consume more power (presumably - because of the heat) than the processor in the computer. How is driving 25Gb/s such a hot task compared to driving multiple 40Gb/s Thunderbolt ports?
As I understand it, it takes a bit of anger to get electromagnetism to wiggle that fast over copper. Imperfect transistors using their full operational envelope to control a sort of chaotic thing? Fields and stuff.
can confirm. anger causes loud fast communication that can be heard over background noise. I've heard anger using light-based signalling like sign language might be more efficient.
There are no shortcuts in networking. It's quite easy to spend a 2-3x+ premium on a switch or adapter capable of non blocking line rate on every port simultaneously only to then have it fall over to a fraction of that in the form of lots of tiny packets. This is why we have things like QoS, flow control, and ECMP
They’re not that bad, but even a couple watts of power with no way to dissipate is going to get hot enough to burn you. Forced air cooling over a heatsink is really effective. 15W under load is much lower than a high end desktop CPU that can pull 150-200W.
See also NVMe drives? They draw enough idle power to drain a laptop battery overnight, and will easily hit 60+C in a bad enclosure. That’s only a 10Gbps module. The USB4 enclosures are all heatsink.
For a 25G NIC, I’d be perfectly unsurprised if it demanded an x16 slot. I’d be more surprised if it didn’t have a power connector like a video card, though.
Maybe the NIC does a lot? In past experience, a whole Xeon CPU couldn't keep up with a 10gig NIC if it was sending/receiving min-size UDP packets. Even with Intel DPDK iirc.
Ethernet is isolated through transformers on the physical layer and they're always sinking power. Thunderbolt is just pushing current directly through.
With ethernet it's for lots of reasons but a big one is not creating ground loops between distant equipment. If you didn't isolate the network cable would be sinking potentially amps of current between distant devices and that just isn't sustainable. You don't have this problem attaching USB between two devices.
It's distance. 25GbaseT is good for 30m+. Thunderbolt is good for 1m without active cables. Making the much more rugged signal is a much more intense job.
And your computer already speaks the PCIe and thunderbolt. It has that probably on CPU. Here we have to receive USB, convert it to Ethernet. On much less up to date process nodes than your CPU.
Third, until very recently for even 10Gbit it meant using a usb4<->pcie bridge chip which itself was hot, to provide a generic pcie connection, which then talked to a conventional network chip, so it was a double hop: USB <-> pcie <-> nic. I don't know if 25Gbit to USB direct convert is available yet but we've seen a wave of vastly cheaper 10Gbit adapters come out that are single chip USB<->nic, probably not pcie at all, that are much much cooler. 25Gbit is probably still not here yet, maybe?
We refurbish laptops and hence distribute .iso / ipsw images to various refurbishing stations. (or also, think of PXE/network boot situations). More images x more laptops refurbished at once = more bandwidth.
Obviously it's possible to move this to a NAS, not use a mac, etc. etc. - but just wanted to chime in with my little use case of moving multi-gig-files around. :-)
I was just looking for a thunderbolt-sfp solution. Super timely!
> Author talks about 4k video, but AI tells me that for "4K production mastering (ProRes 422 HQ, 4:2:2, 10-bit)" the rate is under 2Gbps.
That'd be the rate for 1 stream at 1x playback. Multiple streams because of overlays/crossfades/etc... along with playback at >1x (such as during rendering in particular) will change that significantly.
NAS stuff can benefit greatly and straightforwardly.
A typical NVMe SSD can easily do sequential reads at 7GB/s, which outpaces a 25Gbps connection by more than 2x. NAS is more typically backed by SATA HDDs, but there's some relatively easy way to put an NVMe disk in front of them as a cache.
Do you ever _need_ it? Perhaps not, but there's plenty of common activities just stemming from NAS usage which see a linear speedup, and would continue to get faster up to 100Gbps even.
Normal 4K editing workflows are fine with even 2.5 Gbps wired networking (wired to make latency like when scrubbing not-horrid)... 10 Gbps more necessary if editing ProRes RAW and doing multicam (4-8 streams simultaneously).
I'm in a never-ending pursuit of cutting time when transferring large files though — AI models is one thing, but I have 100-200 GB projects I transfer across for backup or import (using an external TB5 drive capable of 3-5 GB/sec), and if I can do that in 1 or 2 minutes per project instead of 3 or 4, that's a noticeable speedup.
Definitely overkill for 98% of my workflow though. 10 Gbps is fine.
Somewhat off-topic: has anyone implemented USB C based networking?
So for instance you have two or more machines that have USB 4 ports at 20 to 80 Gbps and you plug in the appropriate USB C cables to connect them.
What do you need to make the connections appear as a (point to point) part of a regular TCP/IP network? Seems much cheaper than Ethernet if you already have the ports.
Yup, totally doable. Of course, it’s a point to point link so if you want to build a whole cluster that way then you’ll need a lot of usb-c ports or you’ll need to make more of a mesh topology.
I believe MacOS supports it via the usual network control panel. I’ve used it for transferring data from an older laptop to a newer one. The cable is more expensive than fiber, last time I checked.
I have also used a USB 4 cable to network two MacBook Pros to one another and can confirm that it was very simple to connect and it was fast as heck. The USB 4 cable I used was rated for up to 40 Gbps.
I used to have a thunderbolt ring network for a ceph backend on my k8s cluster. On Linux load the thunderbolt_net kernel module and they show up the same as any other Ethernet interfaces. Just directly connect them and configure the network settings on each end using your normal tools.
Latency was an order of magnitude higher than twisted pair but the throughput was greater than 10Gbps and I didn’t have to spend on 10GbE capable servers and network equipment. I vaguely remember that they had to be proper Thunderbolt not just USB-C but that could have changed it’s been quite a while.
With USB 4, very strictly speaking: It's iffy. USB 4 doesn't require hosts to be able to switch roles and talk to other hosts. It can support this, but support is optional.
But for all certified Thunderbolt 4 and 5 ports, this kind of networking support is a requirement. (It is optional with Thunderbolt 3.)
Yes, there are a few standard(ish) mechanisms for doing this, like [RNDIS][1] and NCM. If you have two Thunderbolt-equipped Macs and connect their ports together, you can use Bonjour/zeroconf for network discovery between them.
You're limited in your network topology because there is no USB Switch. This HN post used two USB4 connections to create a 3-node cluster: https://news.ycombinator.com/item?id=39003469
Your so-called appropriate USB C cable is just two USB-C to Ethernet adapters plus an Ethernet cable. It’s very common on laptops without native Ethernet ports. But in my experience USB isn’t a very good protocol; it generally uses way more CPU than plain PCIe Ethernet. This is especially the case if the chipset is using ECM rather than NCM.
> I can edit 4K video straight off my NAS over the network,
Nothing against Jeff Geerling and I love his YouTube channel and I'm glad he's one of us HNer but this crop of circular "I make YouTube vids about big tx speeds and big NAS because I make vids hence I need big tx speeds and big NAS so I can speak about big tx speeds and big NAS because I make YouTube vids about big tx speeds and big NAS" videos is a bit tiring IMO.
I'm not into cryptocurrencies but, I don't know, host a full Ethereum node, host a Plex/Jellyfin server serving thousands of movies, host LLMs, etc.
I mean: talk about something else than editing vids because you make vids about vids.
The upside, though, was that I plugged it in, got my throughput, and never had to worry ever again. Remember that engineering time is mucho $$. So troubleshooting overheating, divising a solution, manufacturing it, etc, that all eats into the savings.
The $400 dock would be cool though. It seems smaller than the sonnet and fixes my charging problem. Weirdly I didn't think I'd care about not including transceivers but they did actually come in handy, so that is a downside. Might try to get one of these if I need another one in the future.
I initially read this as SONET[1] and was seriously questioning your technical choices.
[1] https://en.wikipedia.org/wiki/Synchronous_optical_networking
what a man does in his own homelab, even if it does involve ATM, is none of our business.
ConnectX-4 and newer are natively supported. I use a ConnectX-4 LX in a $50 Razer CoreX off of craigslist. Works fine, hit similar limits to what Jeff did.
Using an OCP is the neat trick here. MUCH smaller setup.
As far as the other posters comment about Tx's... that is an entirely different topic. I've never had an issue with FS.com optics personally.
If you are unfamiliar, the reality of the situation is quite bad. Slight hardware revisions of adapters/switches, which if you are lucky can be differentiated by manufacture date or serial number sequence, will not recognize known good modules, and if they do you then have to actually test if they can establish link, then let them burn for 24+ hours to see if they can maintain it without dropping (often, because of forward error correction, you will not start to see flapping until enough correctable errors have accumulated and the buffer overflows). Then there are firmware updates where this process must then be repeated.
You would think with how expensive all of this shit is you would simply be able to buy the officially listed as compatible components but my experience with them has been that a shotgun approach is best for your sanity and deadlines, buying sample kits of each module and trying them all, and then once you find the ones that play nice, buying as many of them as is reasonable and testing all of them.
If I had a dollar every time two modules from the same brand, that came in the same box, from the same production run, did not work with one another, I would be able to test your theory for free. Network Engineering is Fun
Plugable 2.5G USB-C to ethernet adapter (bought November 2025), SABRENT USB C to 2.5Gbps ethernet adapter (bought October 2024), UGREEN USB C 2.5G ethernet adapter (also October 2024). In my experience they immediately fall on their face in a speedtest and drop to a couple hundred Mbps, experience packet loss, etc. This is testing across a few modern Apple Silicon Macs.
I now have a WavLink WL-NWU341G 5Gbps dongle with the RTL-8157 that seems to work well, but I've only run a couple of tests-- I am away from my home setup and don't have wired ethernet run to all the rooms where I'm living.
If you do some Googling, there's some good information out there about it, but when I bought the bad ones, I couldn't find any information corroborating my experience.
> this real-world speed limitation could also be related to my use of my lower-power Arm NAS (running on Ampere Altra, with 32 fairly-slow CPU cores). I was testing writing to an array of fast enterprise NVMe SSDs, so they shouldn't be the bottleneck here.
He was only getting 1 Gigabyte of throughput out of the built in 10 Gigabit Ethernet connection and moving to the 25 Gigabit connection only made a marginal impact on throughput, after all.
MacOS has supported RDMA over thunderbolt since March 19th.
* https://learn.microsoft.com/en-us/windows-server/storage/fil...
* https://docs.kernel.org/filesystems/smb/smbdirect.html
I do wonder what Apple's custom-built servers are running for networking. I wish they still made Mac Pro-level chassis specifically for better IO. Thunderbolt 5 is nice, but far inferior to a PCIe x16 slot (which can hit 400 Gbps networking) nowadays.
If they were focused on the needs of the kind of person who moves enough terabytes of files around to need this kind of speed, we'd see them still making either computers with slots, or first-party 25Gbps networking.
I'm not really trying to criticize them for it -- Tim Cook's bean counters clearly did the math and they're certainly free to optimize their Mac R&D spend for things that'll sell college students MacBook Neos and Airs, rather than worrying about the MKBHDs, Linus Sebastians and Jeff Geerlings of the world (this is just a list of people I've seen using extremely fast networking).
The perf is amazing
Now, I'm off to determine if an eMMC M.2 module is a native PCIe device…
His iperf3 results show that it's able to go full beans on a single 25g port. A cross platform SMB file copy is kind of a meaningless benchmark in this context.
IMO a pretty cool product would be a network switch providing native thunderbolt interfaces to multiple upstream hosts. That wouldn't magically fix this RDMA issue, though. This is just a separate thought. Seems like the kind of thing Mikrotik should make.
Speeds aren't any better, but at least it was cheap and fiber.
I found that SMB on macOS is just bad. If I try to heavily load it (e.g. by doing mass conversions in my Audible library), it can cause stalls and disconnections.
I managed to find a set of parameters (in /etc/nsmbd.conf) that makes it work more-or-less reliably by trial-and-error: dir_cache_off=yes, file_ids_off=yes, notify_off=yes, mc_on=no.
It's still terrible, though. It's so bad that Linux in a Parallels VM works _better_.
It's not just software companies or home enthusiasts either. There are a lot of ie medical practices or accounting/benefits agencies out there who have sworn off Apple for life after having to spend extreme amounts of money on data recovery to hopefully get their years of records back, usually in a do or die situation like an IRS audit which causes them to discover that their basement was flooding and all of the paper records are destroyed.
The 4 Mac Studio cluster I tested with Exo and Thunderbolt RDMA was a loaner cluster from Apple. There's no universe in which I could justify buying that.
I always mention any equipment that's provided / loaned / etc. and the nature of my relationships with vendors—I try to also keep this document updated for transparency's sake: https://github.com/geerlingguy/youtube/blob/master/Sponsors....
See also NVMe drives? They draw enough idle power to drain a laptop battery overnight, and will easily hit 60+C in a bad enclosure. That’s only a 10Gbps module. The USB4 enclosures are all heatsink.
The NIC in this setup was released 12 years ago. It's built on a very old process.
you can look up the rated power use of a connectx4 card, they do a lot of offloaded computation, too
With ethernet it's for lots of reasons but a big one is not creating ground loops between distant equipment. If you didn't isolate the network cable would be sinking potentially amps of current between distant devices and that just isn't sustainable. You don't have this problem attaching USB between two devices.
And your computer already speaks the PCIe and thunderbolt. It has that probably on CPU. Here we have to receive USB, convert it to Ethernet. On much less up to date process nodes than your CPU.
Third, until very recently for even 10Gbit it meant using a usb4<->pcie bridge chip which itself was hot, to provide a generic pcie connection, which then talked to a conventional network chip, so it was a double hop: USB <-> pcie <-> nic. I don't know if 25Gbit to USB direct convert is available yet but we've seen a wave of vastly cheaper 10Gbit adapters come out that are single chip USB<->nic, probably not pcie at all, that are much much cooler. 25Gbit is probably still not here yet, maybe?
Author talks about 4k video, but AI tells me that for "4K production mastering (ProRes 422 HQ, 4:2:2, 10-bit)" the rate is under 2Gbps.
I'm not saying it's not needed, - just curious.
Edit: https://www.jeffgeerling.com/blog/2025/15-tb-vram-on-mac-stu...
Obviously it's possible to move this to a NAS, not use a mac, etc. etc. - but just wanted to chime in with my little use case of moving multi-gig-files around. :-)
I was just looking for a thunderbolt-sfp solution. Super timely!
That'd be the rate for 1 stream at 1x playback. Multiple streams because of overlays/crossfades/etc... along with playback at >1x (such as during rendering in particular) will change that significantly.
A typical NVMe SSD can easily do sequential reads at 7GB/s, which outpaces a 25Gbps connection by more than 2x. NAS is more typically backed by SATA HDDs, but there's some relatively easy way to put an NVMe disk in front of them as a cache.
Do you ever _need_ it? Perhaps not, but there's plenty of common activities just stemming from NAS usage which see a linear speedup, and would continue to get faster up to 100Gbps even.
I'm in a never-ending pursuit of cutting time when transferring large files though — AI models is one thing, but I have 100-200 GB projects I transfer across for backup or import (using an external TB5 drive capable of 3-5 GB/sec), and if I can do that in 1 or 2 minutes per project instead of 3 or 4, that's a noticeable speedup.
Definitely overkill for 98% of my workflow though. 10 Gbps is fine.
So for instance you have two or more machines that have USB 4 ports at 20 to 80 Gbps and you plug in the appropriate USB C cables to connect them.
What do you need to make the connections appear as a (point to point) part of a regular TCP/IP network? Seems much cheaper than Ethernet if you already have the ports.
I believe MacOS supports it via the usual network control panel. I’ve used it for transferring data from an older laptop to a newer one. The cable is more expensive than fiber, last time I checked.
Latency was an order of magnitude higher than twisted pair but the throughput was greater than 10Gbps and I didn’t have to spend on 10GbE capable servers and network equipment. I vaguely remember that they had to be proper Thunderbolt not just USB-C but that could have changed it’s been quite a while.
Thunderbolt 4 is profile with USB4 tunneling, 40Gbps speed, and 100W power. Thunderbolt 5 is 80Gbps and 240W power.
But for all certified Thunderbolt 4 and 5 ports, this kind of networking support is a requirement. (It is optional with Thunderbolt 3.)
The rest is up to the operating system.
[1]: https://en.wikipedia.org/wiki/RNDIS
Nothing against Jeff Geerling and I love his YouTube channel and I'm glad he's one of us HNer but this crop of circular "I make YouTube vids about big tx speeds and big NAS because I make vids hence I need big tx speeds and big NAS so I can speak about big tx speeds and big NAS because I make YouTube vids about big tx speeds and big NAS" videos is a bit tiring IMO.
I'm not into cryptocurrencies but, I don't know, host a full Ethereum node, host a Plex/Jellyfin server serving thousands of movies, host LLMs, etc.
I mean: talk about something else than editing vids because you make vids about vids.