It's like they're determined to intentionally come up with the most obtuse and confusing naming conventions possible. Which is how you end up with a situation where USB 3.0 == USB 3.1 Gen 1 == USB 3.2 Gen 1x1 == SuperSpeed USB 5Gbps, because having four different names for exactly the same thing is obviously the right approach.
This is no accident. The USB consortium is run by PC vendors who want to be able to call their crappy old widget from last year USB 3.2 without any hardware change, and hide the fact that it is old and slow in the fine print.
Read the last page of the article. The USB maintainers have decided to make USB4 even more of a garbage fire of confusion than previous versions. USB4-20gb and USB-40gb ports are only required to support 10gb data rates for USB drives (and can count bandwidth to alternate data stream devices, ie parallel video out) toward the total.
IMO it's past time to disband the current USB group, and create a new C(ompetant)SB organization to maintain future standards that bans anyone involved in the USB 3.x/4.x nomenclature from membership.
Well, we can always ditch USB3.1 / USB4 as consumers. Instead ask for USB-A (3.0) for legacy support, and then opt for USB-C (TB4) instead. Just say "hell no" to the USB-IF consortium and their broken standards.
Yet, that might be okay for now/per individual, but it's not gonna solve anything in the long run. And even worse, we can't actually do that because there is a lack of options in the market (ie You don't get to choose which ports your device has, you are stuck with whatever option they deem good for you). Tough times.
Stop spreading FUD and read the specs, they're freely available for download. USB4 operates at 20Gbps (Gen 2x2) or 40Gbps (Gen 3x3) but it's a tunneling protocol, i.e. a *totally different* protocol than USB3. Backwards compatibility is mandatory and at least USB3 10Gbps (Gen 2x1) as well USB 2.0 signaling is required. USB 2.0, 3.2 and 4 involve different signaling and different protocols.
USB4 supports tunneling of PCIe and DisplayPort packets in addition to native USB packets so that the total bandwidth can be flexibly and efficiently shared. Not bothering to understand this and instead spewing internet outrage helps no one.
@repoman27's technical explanation is A-OK :) The move to tunnelling is definitely a big step for USB.
However, @DanNeely's comment is also correct from a *consumer* viewpoint ; Would a regular non-tech savvy consumer care about tunnelling? If he sees USB4 20Gbps, would it be unfair for the person to expect his USB SuperSpeed 20Gbps device to work to its full potential in it?
All these problems could have been avoided if USB 3.2 Gen 2x2 capabilities were integrated into USB4 as mandatory.
Oh, totally. But on the other hand, if Intel had just included USB 3.2 dual-lane operation in Tiger Lake / Thunderbolt 4, nobody would have realized it was optional, because it would have been supported everywhere one might expect it. I’m afraid that Intel intends to completely forego USB 3.2 dual-lane operation in favor of Thunderbolt.
It mainly boils down to USB 3.2 @ 20 Gbit not being the same implementation as USB 4 @ 20 GBit, a very year 2020 problem to have.
There is still interoperability between the two but that knocks the speeds down to 10 Gbit. That is noticeable for things like storage doing transfers on fast SSDs etc. From a consumer stand point though, things will still work which I would rank as more important.
I will say that these issues lay with the USB consortium as they've created a mess of specifications that makes it difficult for things to work *as advertised*. Historically they also have needed to crackdown to lazy implementations and bad cables. Give that group a bit of teeth to enforce their spec and things would be far better for consumers today.
Well, at the risk of beating a dead horse, nothing prevents a USB4 host/device from also supporting USB 3.2 20Gbps. It’s just that it’s optional, Intel didn’t include it in Tiger Lake, and I’m not aware of anything on their roadmap that will in the near future.
You can choose to blame the USB-IF, Intel, or the fact that Intel pretty much runs the USB-IF. But at this point Intel hasn’t shown us any products containing USB 3.2 20Gbps IP, and without Intel on board, I’m not sure what kind of future that standard has.
When I read it, it sounded like the issue was with USB 3.X devices operating on a USB4.0 chipset. My assumption, and I don't think it was explicitly addressed, is that USB4.0 would be "full speed" per it's own specs. Of course, it wasn't explicitly addressed (from what I took away from this), so I phrased my comment as a question; "So wait for USB4 devices then?"
You’re reading is pretty much on the money, but the answer to your question is a bit trickier.
The first USB4 hosts to hit the market will (probably) be Intel Tiger Lake based products which have integrated Thunderbolt 4 and support USB4 40Gbps. The USB4 spec requires backwards compatibility with USB 3.2, including both the Gen 1 (5Gbps) and Gen 2 (10Gbps) PHYs. It does not, however, require USB3 dual-lane operation (Gen 2x2, 20Gbps), and Intel has not included this optional feature in the controller integrated into Tiger Lake CPUs.
If Intel doesn’t have any plans for integrated USB 3.2 20Gbps, I fail to see how it becomes widespread, unless Apple and AMD both embrace it in their future chipsets. On the other hand, USB 3.2 devices are probably always going to be cheaper than Thunderbolt or USB4 gear. Paying the premium for a USB 3.2 20Gbps device today is somewhat questionable, unless you have a capable host or the performance difference compared to other products when connected to a 10Gbps port is worth it to you.
Why will USB3.2 20Gbps devices always be cheaper than USB43 gear? It seems likely USB3.2 20 Gbps will be a niche product and without being produced in high volume, street price won't come down. USB4 might be initially expensive, but volume production and competition may bring street prices down to what we see today for USB3.2 Gen 1 5Gbps devices.
Because USB4 is essentially Thunderbolt, but even more complicated. It will always take way more silicon and way more power than USB3 on the same node. Thus it’s never going to be as cheap. Economies of scale can’t solve everything.
Sounds about right. I don't have a sudden need for USB 3.2 dual lane. It seems like the best case for more parties to forgo support for it altogether and push USB 4.0. On a related note, USB 3.X has always been very buggy and unstable/unreliable for me, so hopefully 4.X fixes some of that.
The ASMedia ASM2364 is the PCIe NVMe to USB3 20Gbps bridge chip used by the devices being tested for this review.
The ASMedia ASM3242 is the USB3 20Gbps host controller that was used for testing these USB3 20Gbps drives.
The Intel JHL6540 is a Thunderbolt controller which includes an integrated USB3 10Gbps host controller for interoperability with USB3 devices. This was used to test these USB3 drives while connected to a USB3 10Gbps host. The ASM3242 is winning because it supports twice the USB3 signaling rate as the JHL6540.
When connected to a Thunderbolt 3 device, the JHL6540 supports up to 40 Gbit/s. In some of the tests, the ASM3242 card was actually plugged into a Thunderbolt 3 enclosure and connected to the host PC via the JHL6540.
This was tried and rejected by the world already, under the name SATA Express. A whole generation of motherboards shipped with SATA Express ports but nobody made any drives of any sort to use with the port. (IIRC, WD had a demo at Computex one year.) The closest I ever came to using a SATA Express port for anything was the clever ASrock adapter that let you repurpose the port to create a pair of USB 3.1 ports, Type A and C with 10Gb/s support, that went into a front drive bay.
Performance is one thing, and I understand it's the primary concern in some cases. In most cases though, compatibility not performance is the main issue, and we run into 2 problems: 1- it's hard to know what *should* work. A USB-C port doesn't mean anything at all by itself, there's not even a color code as a quick hint. Any consumer tech that requires to RTFM is failing at a very basic level. 2- even stuff that should work sometimes doesn't. Apparently USB-PD charging on MacBooks works much better on one side than on the other. I've seen a lot of issues with video, even simple storage/LAN stuff. The goal of USB is laudable. The way they're going at it is laughable.
Even with speeds it varies so wildly by device its silly. The real only advantage i found with a SSD for portable drive is the size and weight is better. Performance is Meh, because most people aren't using it for own devices so much as plugging it into someone elses. I've seen top selling drives that will barely get usb 2.0, and even then the read/write to usb speed is insanely different between devices.
I forgot to mention that even cables mater so much, i'm not talking about scam monster cable stuff, i'm talking just even same brand to brand, can get a bulk 20 pack of usb-c cables, and each one could be different in speed.
Nope. USB 1.0 defined both low-speed 1.5 Mbit/s and full-speed 12 Mbit/s signaling. It just sucked, which is why 1.1 was released to fix a bunch of issues that were encountered in real-world implementations.
Why do boot, office, PCMark10 runs. DAS is only about copying data from internal to external move the DAS to other system and do the opposite (copy data from external to internal). Users don't have ram drives or use robocopy (you already have the synthetic benchmarks).
Just use fast internal drives (that people actually buy) and start dragging files/folders with File Explorer like humans do :)
I boot from USB external media all the time. It's a pretty common part of many workflows. So is running VM's that reside on external media. Or working with large media files that live on external media dedicated to a particular project / client.
I was bike commuting from home to work with my SSD (Samsung X5 Thunderbolt 3) in my pocket and booting Linux off it at either end. Easier than trying to sync data, plus I only had to maintain the, like, 1000 development tools I need in once place.
I have been reading Anandtech for 16 years now and this is the most outright confusing article that I have read for the reasons listed 1. I disagree with the testbed choice-- IT IS NOT A SYSTEM,IT IS A HACK . I do not think that it has realworld applicability. Could you try a couple of Ryzen system. Failing that just stop giving 16 threads q32 results. I cannot think of a case where a 1 TB, bus powered DAS would be used in a realworld use requiring 16*32 random reads. Can you? 2. I agree with @danneely @spunjji @stormyparis, USB 3.2 is a mess. Even rtfm fails. I would request your reviewers to add the supported speeds to each system, board and flagship mobile review. With 8k video (and limited storage) it is not unthinkable to use one of these devices to transfer files.
The article is a description of our attempts to make use of USB 3.2 Gen 2x2 as a regular consumer. Rarely do people go out on a limb and make a new PC for something that could be achieved with an expansion card. That is the reason for the piece being described as 'ecosystem review', rather than a review of just the WD_BLACK P50 and the SanDisk Extreme PRO v2.
The testbed choice was an attempt to use the three 'direct-attached storage testbeds' we have used since starting the reviews set back in 2014. It did end up as a 'hack', but that serves the purpose of this particular review well. If you see many more Gen 2x2 reviews using the same Haswell testbed, then your complaint is justified.
AnandTech's editors do not operate out of a central location. Most (including me) are freelancers spread out throughout the world. As far as testing out a couple of Ryzen systems goes - all the Ryzen PCs with me are SFF machines without a PCIe expansion slot. I am loath to building a new testbed at *this point in time* because USB4 is just around the corner. It is better to build a testbed that can serve the purpose well for at least 2 to 3 years.
As for bus-powered DAS and the 4KQ32T16 workload - if you don't think it is suitable, feel free to disregard the numbers. The limited applicability of the workload is exactly the reason we felt it was OK to present the results from the other workloads on the same machine.
Thank you for addressing my concerns. In light of this I have 2 fresh ones 1. As you also agree the host system is a hack-- it should not become the regular testbed. I do not bother with T16q32 results in any review, because I do not have any workload that can generate such a workload. 2. Please again I would like like that all devices with USB 3xyz are labelled with speeds and power output supported.
The issues they run into here make me think of the practical problems I have with USB.
I have a few recent Windows laptops that have USB 3 Type A and Type C ports.
In the old days you were supposed to be able to plug a hub into a host then plug a hub into a hub and do it again and have it work. The spec said you could do it and you really could.
In the USB3 spec I don't see anything promised as to what kind of configurations are supported and I find I can't take it for granted that I plug my PC into a 4-way hub and can then plug my monitor into the hub and then plug my keyboard and mouse into the monitor.
Some configurations work but then I plug in the RealSense camera and my mouse stops working; or maybe the SD card reader connects and disconnects all the time and I am always hearing the notification tone for that.
The hard way I learned "at most four way hubs" and "never plug a hub into a hub" and I've finally settled in on something that works but plug in a USB hard drive and I pray that the filesystem doesn't get corrupted.
I know this will read old-fashioned, but I really wish that one of these USB flavors would allow true "serial" use, i.e. the ability to daisy-chain USB devices directly in the plug. With most "ultraportables" now down to 1-2 ports, USB charging is often made absurd by the then-lost connectivity. If one could simply plug another device into the back of the male USB plug, that issue would be moot. Is there such a solution? I will gladly stay on 3.1 or 3.2 if that feature would be enabled.
In theory, they all do. You just need to embed a USB hub to create that topology. It costs money, adds complexity, and consumes power, but plenty of dongles and some chargers already do this.
Seems to me this whole fiasco is just an inevitable merge of USB (dumb/cheap interface) to Thunderbolt (smart-ish/expensive interface) to produce a middle-of-the-road option. Just how far can we go with the only choice of cheap & dumb vs smart & expensive? When the dust settles in a couple years we'll all be more-or-less happy and the next, greatest thing will be all over the tech news!!
Thunderbolt is essentially proprietary to Intel, and without Intel supporting USB 3.2 2x2, that makes 2x2 essentially proprietary to AMD. Did anyone ever release a JHL7440 based Thunderbolt 3 SSD that's actually compatible with regular USB-C ports? Cuz at least a USB 3.2 2x2 drive will *work* when plugged in to an Intel (Thunderbolt 3/USB4) system port, whereas most Thunderbolt drives won't even connect to an AMD (USB 3.2 2x2) system port.
Intel has already “launched” Tiger Lake CPUs with integrated Thunderbolt 4 / USB4 as well as the Goshen Ridge 4-port hub/device and a couple retimer chips. Additionally, they announced the Maple Ridge host controllers which will be available “later this year”. I’d reckon we’ll also see Apple Silicon Macs with Thunderbolt 4 / USB4 by the end of the year.
Hmmm, really thinking about the thired party controller chips that are likely to be integrated into motherboards and drives on non-Tigerlake platforms.
Much as the Asmedia chip is used today to add 2x2 (after being announced on AT about 18 months ago).
I’d have to imagine ASMedia and AMD have something in the works, but to my knowledge they haven’t said anything publicly. Which means Intel and Apple are probably going to have a significant head start with Thunderbolt 4, and Intel will be the only source for discrete USB4 silicon for a while.
Apologies for the emoji, but I don’t see how this USB 3.2 Gen 2x2 standard will ever take off given the travails and issues described in the article. Best to kill it and bury it somewhere deep.
Honestly, what on earth was the thinking behind the ridiculous naming (and renaming) of the USB versions?
It used to be simple and clear, but now, between the crazy naming and mess that is USB-C (is it USB or Thunderbolt? Is the cable USB 2, 3, 3.1 Gen 2? Does it support 100W power?), how is anyone supposed to make sense on it.
On a related note, I would love to see you guys do some kind of investigation into why we're five years into this standard and one still cannot buy an actual USB-C hub (i.e. not a port replicator).
A 3.2 hub with gen 2 ports and a 2x2 uplink would be cool!
I wanted a 10gbps / gen 2 hub and got the StarTech HB31C3A1CS, which at least has a USB-C gen 2 uplink and a single USB-C gen 2 port (plus type A ports). Don't know if you can do any better than that right now.
Although it's still not exactly what you're looking for, I've tried (unsuccessfully) to get people to understand what a unicorn the IOGEAR GUH3C22P is. Link: https://www.iogear.com/product/GUH3C22P
It's a 5-port USB3 10Gbps hub with a tethered USB Type-C cable on the UFP (which supports up to 85W USB PD source), two (2!) downstream facing USB Type-C ports (one of which supports up to 100W USB PD sink), and two USB Type-A ports (one of which supports up to 7.5W USB BC).
No alt mode support like for DisplayPort. I haven't found a portable type-C hub that supports DisplayPort alt mode over downstream type-C ports although some desktop docks support it.
I am building hyperconvergent clusters for fun and for work, the home-lab one out of silent/passive 32GB RAM, 1TB SATA-SSD J5005 Atoms, the next iteration most likely from 15Watt-TDP-NUCs, an i7-10700U with 64GB RAM, 1TB NVMe SSD in testing.
Clusters need short-latency, high-bandwidth interconnects, Infiniband is a classic in data centers, but NUCs offer 1Gbit Ethernet pretty much exclusively, Intel struggling to do 2.5Gbit there while Thunderbolt and USB3/4 could do much better. Only they aren’t peer-to-peer and a TB 10Gbase-T adapter sets you back further than the NUC itself, while adding lots of latency and TCP/IP, while I want RDMA.
So could we please pause for a moment and think on how we can build fabrics out of USB-X? Thunderbolt/USB4 is already about PCIe lanes, but most likely with multi-root excluded to maintain market segmentation and reduce validation effort.
I hate how the industry keeps going to 90% of something really useful and then concentrating on 200% speed instead of creating real value.
Uh, Thunderbolt and USB4 are explicitly designed to support host-to-host communications already. OS / software support can be a limiting factor, but the hardware is built for it.
Let me tell you: You just made my day! Or more likely one or two week-ends!
Not being a Mac guy, I had completely ignored Thunderbolt for a long time and never learned that it supported networking natively. From the Intel docs it looks a bit similar to Mellanox VPI and host-chaining: I can use 100Gbit links there without any switch to link three machines in a kind of “token ring” manner for Ethernet (these are hybrid adapters that would also support Infiniband, but drivers support is only support Ethernet for host-chaining). Unfortunately the effective bandwidth is only around 35GByte/s for direct hops and slows to 16GByte/s once it has to pass through another host: Not as much of an upgrade over 10Gbase-T as you’d hope for: I never really got into testing latencies, which is where the Infiniband personality of those adapters should shine.
And that’s where with TB I am hoping for significant improvements over Ethernet apart from native 40Gbit/s speed: Just right for Gluster storage!
I also used to try to get Ethernet over fiber-channel working years ago, when they were throwing out 4Gbit adapters in the data center, but even if it was specified as a standard, Ethernet over fiber never got driver support and at the higher speeds the trend went the other direction.
So I’ll definitely try to make the direct connection over TB3 work: CentOS8 should have kernel support for TB networking and the best news is that it doesn’t have to wait for TB4, but should work with TB3, too.
I’ve just seen what seemed like an incredibly cheap 4-way TB switch recommended by Anton Shilov on the TomsHardware side of this enterprise, which unfortunately is only on pre-order for now (https://eshop.macsales.com/shop/owc-thunderbolt-hu... but supposed to support TB networking. Since the NUCs are single port TB3 only, that should still do the trick and be upgradable to TB4 for just around $150… The 5Gbit USB3 Aquantia NIC wasn’t much cheaper and even the 2.5Gbit USB3 NIC are still around $40.
Exciting, exciting all that: Thank you very much for those links!
...except... I don't think that "switch" will be supporting multiple masters, same as USB.
If it did, Intel would have shot themselves in the foot: 40Gbit networking on NUCs and laptops with little more than passive ables, that's almost as bad as adding a system management mode on 80486L and finding that it can be abused to implement a hypervisor (Mendel and Diane started VMware with that trick).
Yet that's exactly what consumers really thirst for.
According to the USB4 spec, host-to-host tunneling is possible via inter-domain links even when using a hub, but of course YMMV.
Thanks for pointing out that Thunderbolt 4 hub, that's awesome! And way cheaper than I was expecting. It looks like a Goshen Ridge hub based on Intel's Thunderbolt 4 Compact Dock reference design to me. However, I'm a little worried that OWC's wording implies that it is not backwards compatible with Thunderbolt 3 hosts. That would seem insane, and it may just be an OS support issue at this juncture, but it sure sounds like it's for Thunderbolt 4 PCs only. Aargh!
It *might* work, and cost may not be prohibitive & worth a shot! I don't see networking mentioned at all on the product page. Also, aren't thunderbolt 3 cables required to be active to do 40GB speeds?
IP over Thunderbolt is a thing, but you'll never see it advertised. Much like IP over IEEE 1394 (a.k.a. FireWire) was also a thing... that nobody knew about. Where we're going, we don't need Ethernet!
It is a Thunderbolt 4 / USB4 hub, almost certainly based on the recently launched Intel JHL8440 chip. Per the USB4 spec, it contains a router which "includes a flat point-to-point, configurable switch necessary to create the internal paths between adapters", in addition to a PCIe switch as well as USB 3.2 and USB 2.0 hubs. It is also required to support Thunderbolt 3 interoperability and DisplayPort Alternate mode on all downstream facing ports. It's the real deal.
Oh, also, the max length for passive cables at 40 Gbit/s is 0.8 m, and FWIW Apple sells theirs for $39. Full featured active cables are available in lengths up to 2.0 m, but cost $129 (from Apple).
And throughput for host-to-host communications will be limited by the available bandwidth of the PCIe link between the host CPU and Thunderbolt controller. For Thunderbolt 3 hosts, that tends to be a PCIe Gen3 x4 link which results in real-world throughput of about 22 Gbit/s after accounting for protocol overhead.
So Thunderbolt isn't exactly going to be a panacea, but if you can live with those constraints, it is still a very fast interface.
i want one (or 2 - wonder if they can be cascaded?)! Should be interesting to see follow ons from this or other vendors for cost or features. I've been eyeballing the 2.5baseT switch recently released by QNAP but this might be a better option for me.
Hi Anandtech, can you do a performance comparison of 2 PCs networked via Thunderbolt 3 to Thunderbolt 3 standard cables (since it allows running 10Gbps Ethernet) vs networking using USB 20 Gbps (Gen 2x2) to RJ-45 10 Gbps Ethernet adapters between 2 PCs? Which is faster? I bet the Thunderbolt 3 connection will be faster due to the protocol overhead of USB.
With the stupidity (or anti-consumerist behavior) so glaringly obvious, It's almost as if USB-IF want people to stick with Lightning for as long as possible.
I've got a question. Let's say I purchase a gen 2x2 External SSD like Firecuda with R/W speed around 2000mb/s, will it be compatible with M1 Mac Mini? It has a thunderbolt port/USB 4.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
Facebook
Twitter
RSS
81 Comments
Back to Article
vol.2 - Monday, October 5, 2020 - link
So wait for USB4 devices then?
boredsysadmin - Monday, October 5, 2020 - link
No, USB 3.3 Rev 2 Gen 5 10.13 Finalvol.2 - Monday, October 5, 2020 - link
So USB4 will be gimped to 10gb rates for USB4 devices? Regardless, what makes you say 3.3, etc?magreen - Monday, October 5, 2020 - link
He's joking. Poking fun at the ridiculous (re)naming conventions of the USB-IFGuspaz - Monday, October 5, 2020 - link
It's like they're determined to intentionally come up with the most obtuse and confusing naming conventions possible. Which is how you end up with a situation where USB 3.0 == USB 3.1 Gen 1 == USB 3.2 Gen 1x1 == SuperSpeed USB 5Gbps, because having four different names for exactly the same thing is obviously the right approach.BikeDude - Tuesday, October 6, 2020 - link
Screw you guys.I'm sticking with USB _Full_ Speed. That is the fastest of them all. Full speed ahead!
Oh...wait...
Alistair - Tuesday, March 23, 2021 - link
it's a year later, and I'm still laughing so hard at this thread :)throAU - Sunday, October 11, 2020 - link
This is no accident. The USB consortium is run by PC vendors who want to be able to call their crappy old widget from last year USB 3.2 without any hardware change, and hide the fact that it is old and slow in the fine print.throAU - Sunday, October 11, 2020 - link
USB 4.1 Gen 1 will probably still be 5 gigabitHuzzam - Wednesday, October 14, 2020 - link
no way man, i'm sticking with USB 3.3. Rev 2 Gen 5 10.12 beta... so called "final" causes covid and gayness, one of which is really badDanNeely - Monday, October 5, 2020 - link
Read the last page of the article. The USB maintainers have decided to make USB4 even more of a garbage fire of confusion than previous versions. USB4-20gb and USB-40gb ports are only required to support 10gb data rates for USB drives (and can count bandwidth to alternate data stream devices, ie parallel video out) toward the total.IMO it's past time to disband the current USB group, and create a new C(ompetant)SB organization to maintain future standards that bans anyone involved in the USB 3.x/4.x nomenclature from membership.
Spunjji - Monday, October 5, 2020 - link
I'm actually inclined to agree. It's gone way, way past being a joke at this stage.Kangal - Tuesday, October 6, 2020 - link
Well, we can always ditch USB3.1 / USB4 as consumers.Instead ask for USB-A (3.0) for legacy support, and then opt for USB-C (TB4) instead. Just say "hell no" to the USB-IF consortium and their broken standards.
Yet, that might be okay for now/per individual, but it's not gonna solve anything in the long run. And even worse, we can't actually do that because there is a lack of options in the market (ie You don't get to choose which ports your device has, you are stuck with whatever option they deem good for you). Tough times.
repoman27 - Monday, October 5, 2020 - link
Stop spreading FUD and read the specs, they're freely available for download. USB4 operates at 20Gbps (Gen 2x2) or 40Gbps (Gen 3x3) but it's a tunneling protocol, i.e. a *totally different* protocol than USB3. Backwards compatibility is mandatory and at least USB3 10Gbps (Gen 2x1) as well USB 2.0 signaling is required. USB 2.0, 3.2 and 4 involve different signaling and different protocols.USB4 supports tunneling of PCIe and DisplayPort packets in addition to native USB packets so that the total bandwidth can be flexibly and efficiently shared. Not bothering to understand this and instead spewing internet outrage helps no one.
repoman27 - Monday, October 5, 2020 - link
D'oh, that should have been "40Gbps (Gen 3x2)".ganeshts - Monday, October 5, 2020 - link
@repoman27's technical explanation is A-OK :) The move to tunnelling is definitely a big step for USB.However, @DanNeely's comment is also correct from a *consumer* viewpoint ; Would a regular non-tech savvy consumer care about tunnelling? If he sees USB4 20Gbps, would it be unfair for the person to expect his USB SuperSpeed 20Gbps device to work to its full potential in it?
All these problems could have been avoided if USB 3.2 Gen 2x2 capabilities were integrated into USB4 as mandatory.
repoman27 - Monday, October 5, 2020 - link
Oh, totally. But on the other hand, if Intel had just included USB 3.2 dual-lane operation in Tiger Lake / Thunderbolt 4, nobody would have realized it was optional, because it would have been supported everywhere one might expect it. I’m afraid that Intel intends to completely forego USB 3.2 dual-lane operation in favor of Thunderbolt.Kevin G - Monday, October 5, 2020 - link
It mainly boils down to USB 3.2 @ 20 Gbit not being the same implementation as USB 4 @ 20 GBit, a very year 2020 problem to have.There is still interoperability between the two but that knocks the speeds down to 10 Gbit. That is noticeable for things like storage doing transfers on fast SSDs etc. From a consumer stand point though, things will still work which I would rank as more important.
I will say that these issues lay with the USB consortium as they've created a mess of specifications that makes it difficult for things to work *as advertised*. Historically they also have needed to crackdown to lazy implementations and bad cables. Give that group a bit of teeth to enforce their spec and things would be far better for consumers today.
repoman27 - Monday, October 5, 2020 - link
Well, at the risk of beating a dead horse, nothing prevents a USB4 host/device from also supporting USB 3.2 20Gbps. It’s just that it’s optional, Intel didn’t include it in Tiger Lake, and I’m not aware of anything on their roadmap that will in the near future.You can choose to blame the USB-IF, Intel, or the fact that Intel pretty much runs the USB-IF. But at this point Intel hasn’t shown us any products containing USB 3.2 20Gbps IP, and without Intel on board, I’m not sure what kind of future that standard has.
Spunjji - Wednesday, October 14, 2020 - link
Makes sense. They've been artificially holding the standard back ever since they decided they'd rather push Thunderbolt over USB 3.six_tymes - Monday, October 5, 2020 - link
spot on. thank you for posting truths.vol.2 - Monday, October 5, 2020 - link
When I read it, it sounded like the issue was with USB 3.X devices operating on a USB4.0 chipset. My assumption, and I don't think it was explicitly addressed, is that USB4.0 would be "full speed" per it's own specs. Of course, it wasn't explicitly addressed (from what I took away from this), so I phrased my comment as a question; "So wait for USB4 devices then?"repoman27 - Monday, October 5, 2020 - link
You’re reading is pretty much on the money, but the answer to your question is a bit trickier.The first USB4 hosts to hit the market will (probably) be Intel Tiger Lake based products which have integrated Thunderbolt 4 and support USB4 40Gbps. The USB4 spec requires backwards compatibility with USB 3.2, including both the Gen 1 (5Gbps) and Gen 2 (10Gbps) PHYs. It does not, however, require USB3 dual-lane operation (Gen 2x2, 20Gbps), and Intel has not included this optional feature in the controller integrated into Tiger Lake CPUs.
If Intel doesn’t have any plans for integrated USB 3.2 20Gbps, I fail to see how it becomes widespread, unless Apple and AMD both embrace it in their future chipsets. On the other hand, USB 3.2 devices are probably always going to be cheaper than Thunderbolt or USB4 gear. Paying the premium for a USB 3.2 20Gbps device today is somewhat questionable, unless you have a capable host or the performance difference compared to other products when connected to a 10Gbps port is worth it to you.
magreen - Monday, October 5, 2020 - link
Why will USB3.2 20Gbps devices always be cheaper than USB43 gear? It seems likely USB3.2 20 Gbps will be a niche product and without being produced in high volume, street price won't come down. USB4 might be initially expensive, but volume production and competition may bring street prices down to what we see today for USB3.2 Gen 1 5Gbps devices.repoman27 - Monday, October 5, 2020 - link
Because USB4 is essentially Thunderbolt, but even more complicated. It will always take way more silicon and way more power than USB3 on the same node. Thus it’s never going to be as cheap. Economies of scale can’t solve everything.vol.2 - Thursday, October 8, 2020 - link
Sounds about right. I don't have a sudden need for USB 3.2 dual lane. It seems like the best case for more parties to forgo support for it altogether and push USB 4.0. On a related note, USB 3.X has always been very buggy and unstable/unreliable for me, so hopefully 4.X fixes some of that.Meteor2 - Monday, October 26, 2020 - link
Not if you plug a USB4-20Gbps or a USB4-40Gbps SSD into them.YB1064 - Tuesday, October 6, 2020 - link
Why is the ASM2364 winning in every scenario compared to Thunderbolt, if TB offers higher bandwidth? Am I reading this incorrectly?WTF is this godawful nomenclature dumpster fire??? The idiots on the USB standards committee need to be flogged with extreme prejudice.
repoman27 - Tuesday, October 6, 2020 - link
The ASMedia ASM2364 is the PCIe NVMe to USB3 20Gbps bridge chip used by the devices being tested for this review.The ASMedia ASM3242 is the USB3 20Gbps host controller that was used for testing these USB3 20Gbps drives.
The Intel JHL6540 is a Thunderbolt controller which includes an integrated USB3 10Gbps host controller for interoperability with USB3 devices. This was used to test these USB3 drives while connected to a USB3 10Gbps host. The ASM3242 is winning because it supports twice the USB3 signaling rate as the JHL6540.
When connected to a Thunderbolt 3 device, the JHL6540 supports up to 40 Gbit/s. In some of the tests, the ASM3242 card was actually plugged into a Thunderbolt 3 enclosure and connected to the host PC via the JHL6540.
Googer - Tuesday, October 6, 2020 - link
It's kind of sad to think that USB is now faster than SATA. Will there ever be a SATA 4 for SSD and future bulk storage technologies?epobirs - Wednesday, October 7, 2020 - link
This was tried and rejected by the world already, under the name SATA Express. A whole generation of motherboards shipped with SATA Express ports but nobody made any drives of any sort to use with the port. (IIRC, WD had a demo at Computex one year.) The closest I ever came to using a SATA Express port for anything was the clever ASrock adapter that let you repurpose the port to create a pair of USB 3.1 ports, Type A and C with 10Gb/s support, that went into a front drive bay.Once NVMe caught on it just didn't make much sense to pursue a direct successor to SATA.
https://en.wikipedia.org/wiki/SATA_Express
Hrunga_Zmuda - Tuesday, October 13, 2020 - link
Exactly.StormyParis - Monday, October 5, 2020 - link
Performance is one thing, and I understand it's the primary concern in some cases.In most cases though, compatibility not performance is the main issue, and we run into 2 problems:
1- it's hard to know what *should* work. A USB-C port doesn't mean anything at all by itself, there's not even a color code as a quick hint. Any consumer tech that requires to RTFM is failing at a very basic level.
2- even stuff that should work sometimes doesn't. Apparently USB-PD charging on MacBooks works much better on one side than on the other. I've seen a lot of issues with video, even simple storage/LAN stuff.
The goal of USB is laudable. The way they're going at it is laughable.
drexnx - Monday, October 5, 2020 - link
yeah, USB used to mean it just worked, didn't have to think about it or read anything. Literally plug and play.now? no clue, unless the mfg puts iconography to let you know what each port can do.
imaheadcase - Monday, October 5, 2020 - link
Even with speeds it varies so wildly by device its silly. The real only advantage i found with a SSD for portable drive is the size and weight is better. Performance is Meh, because most people aren't using it for own devices so much as plugging it into someone elses. I've seen top selling drives that will barely get usb 2.0, and even then the read/write to usb speed is insanely different between devices.imaheadcase - Monday, October 5, 2020 - link
I forgot to mention that even cables mater so much, i'm not talking about scam monster cable stuff, i'm talking just even same brand to brand, can get a bulk 20 pack of usb-c cables, and each one could be different in speed.BeethovensCat - Monday, October 5, 2020 - link
Agree!! A complete joke! Have Patriot and a Sandisk SSD and they don't work with the same USB C cable. How can USB have come to this?drexnx - Monday, October 5, 2020 - link
this is really getting back to the dark ages of RS232 where there were different baud rates for different peripherals, now that I think about itdontlistentome - Monday, October 5, 2020 - link
I'll bite. Bit of a correction on diagram, more complexity needed.USB 1.0 was 1.5Mbps, USB 1.1 was 12Mbps
repoman27 - Monday, October 5, 2020 - link
Nope. USB 1.0 defined both low-speed 1.5 Mbit/s and full-speed 12 Mbit/s signaling. It just sucked, which is why 1.1 was released to fix a bunch of issues that were encountered in real-world implementations.henkhilti - Monday, October 5, 2020 - link
Why do boot, office, PCMark10 runs.DAS is only about copying data from internal to external move the DAS to other system and do the opposite (copy data from external to internal).
Users don't have ram drives or use robocopy (you already have the synthetic benchmarks).
Just use fast internal drives (that people actually buy) and start dragging files/folders with File Explorer like humans do :)
repoman27 - Monday, October 5, 2020 - link
I boot from USB external media all the time. It's a pretty common part of many workflows. So is running VM's that reside on external media. Or working with large media files that live on external media dedicated to a particular project / client.hubick - Tuesday, October 6, 2020 - link
I was bike commuting from home to work with my SSD (Samsung X5 Thunderbolt 3) in my pocket and booting Linux off it at either end. Easier than trying to sync data, plus I only had to maintain the, like, 1000 development tools I need in once place.drajitshnew - Monday, October 5, 2020 - link
I have been reading Anandtech for 16 years now and this is the most outright confusing article that I have read for the reasons listed1. I disagree with the testbed choice-- IT IS NOT A SYSTEM,IT IS A HACK . I do not think that it has realworld applicability. Could you try a couple of Ryzen system. Failing that just stop giving 16 threads q32 results. I cannot think of a case where a 1 TB, bus powered DAS would be used in a realworld use requiring 16*32 random reads. Can you?
2. I agree with @danneely @spunjji @stormyparis, USB 3.2 is a mess. Even rtfm fails. I would request your reviewers to add the supported speeds to each system, board and flagship mobile review. With 8k video (and limited storage) it is not unthinkable to use one of these devices to transfer files.
ganeshts - Monday, October 5, 2020 - link
The article is a description of our attempts to make use of USB 3.2 Gen 2x2 as a regular consumer. Rarely do people go out on a limb and make a new PC for something that could be achieved with an expansion card. That is the reason for the piece being described as 'ecosystem review', rather than a review of just the WD_BLACK P50 and the SanDisk Extreme PRO v2.The testbed choice was an attempt to use the three 'direct-attached storage testbeds' we have used since starting the reviews set back in 2014. It did end up as a 'hack', but that serves the purpose of this particular review well. If you see many more Gen 2x2 reviews using the same Haswell testbed, then your complaint is justified.
AnandTech's editors do not operate out of a central location. Most (including me) are freelancers spread out throughout the world. As far as testing out a couple of Ryzen systems goes - all the Ryzen PCs with me are SFF machines without a PCIe expansion slot. I am loath to building a new testbed at *this point in time* because USB4 is just around the corner. It is better to build a testbed that can serve the purpose well for at least 2 to 3 years.
As for bus-powered DAS and the 4KQ32T16 workload - if you don't think it is suitable, feel free to disregard the numbers. The limited applicability of the workload is exactly the reason we felt it was OK to present the results from the other workloads on the same machine.
supdawgwtfd - Monday, October 5, 2020 - link
What editors?drajitshnew - Tuesday, October 6, 2020 - link
Thank you for addressing my concerns. In light of this I have 2 fresh ones1. As you also agree the host system is a hack-- it should not become the regular testbed. I do not bother with T16q32 results in any review, because I do not have any workload that can generate such a workload.
2. Please again I would like like that all devices with USB 3xyz are labelled with speeds and power output supported.
PaulHoule - Monday, October 5, 2020 - link
The issues they run into here make me think of the practical problems I have with USB.I have a few recent Windows laptops that have USB 3 Type A and Type C ports.
In the old days you were supposed to be able to plug a hub into a host then plug a hub into a hub and do it again and have it work. The spec said you could do it and you really could.
In the USB3 spec I don't see anything promised as to what kind of configurations are supported and I find I can't take it for granted that I plug my PC into a 4-way hub and can then plug my monitor into the hub and then plug my keyboard and mouse into the monitor.
Some configurations work but then I plug in the RealSense camera and my mouse stops working; or maybe the SD card reader connects and disconnects all the time and I am always hearing the notification tone for that.
The hard way I learned "at most four way hubs" and "never plug a hub into a hub" and I've finally settled in on something that works but plug in a USB hard drive and I pray that the filesystem doesn't get corrupted.
Let's see an expose on that!
eastcoast_pete - Monday, October 5, 2020 - link
I know this will read old-fashioned, but I really wish that one of these USB flavors would allow true "serial" use, i.e. the ability to daisy-chain USB devices directly in the plug. With most "ultraportables" now down to 1-2 ports, USB charging is often made absurd by the then-lost connectivity. If one could simply plug another device into the back of the male USB plug, that issue would be moot. Is there such a solution? I will gladly stay on 3.1 or 3.2 if that feature would be enabled.repoman27 - Monday, October 5, 2020 - link
In theory, they all do. You just need to embed a USB hub to create that topology. It costs money, adds complexity, and consumes power, but plenty of dongles and some chargers already do this.CaptainChaos - Monday, October 5, 2020 - link
Seems to me this whole fiasco is just an inevitable merge of USB (dumb/cheap interface) to Thunderbolt (smart-ish/expensive interface) to produce a middle-of-the-road option. Just how far can we go with the only choice of cheap & dumb vs smart & expensive? When the dust settles in a couple years we'll all be more-or-less happy and the next, greatest thing will be all over the tech news!!hubick - Tuesday, October 6, 2020 - link
Thunderbolt is essentially proprietary to Intel, and without Intel supporting USB 3.2 2x2, that makes 2x2 essentially proprietary to AMD. Did anyone ever release a JHL7440 based Thunderbolt 3 SSD that's actually compatible with regular USB-C ports? Cuz at least a USB 3.2 2x2 drive will *work* when plugged in to an Intel (Thunderbolt 3/USB4) system port, whereas most Thunderbolt drives won't even connect to an AMD (USB 3.2 2x2) system port.hubick - Sunday, November 29, 2020 - link
The LaCie "Rugged SSD Pro" (STHZ1000800 & STHZ2000800) appear to be the first JHL7440 based Thunderbolt SSD with USB fallback support.R3MF - Tuesday, October 6, 2020 - link
Have we had any indication of launch dates for USB4 controller chips?And any indication of the lag between launch and products utilising them arriving on shelves?
repoman27 - Tuesday, October 6, 2020 - link
Intel has already “launched” Tiger Lake CPUs with integrated Thunderbolt 4 / USB4 as well as the Goshen Ridge 4-port hub/device and a couple retimer chips. Additionally, they announced the Maple Ridge host controllers which will be available “later this year”. I’d reckon we’ll also see Apple Silicon Macs with Thunderbolt 4 / USB4 by the end of the year.R3MF - Tuesday, October 6, 2020 - link
Hmmm, really thinking about the thired party controller chips that are likely to be integrated into motherboards and drives on non-Tigerlake platforms.Much as the Asmedia chip is used today to add 2x2 (after being announced on AT about 18 months ago).
repoman27 - Tuesday, October 6, 2020 - link
That’d be the aforementioned Maple Ridge chips.I’d have to imagine ASMedia and AMD have something in the works, but to my knowledge they haven’t said anything publicly. Which means Intel and Apple are probably going to have a significant head start with Thunderbolt 4, and Intel will be the only source for discrete USB4 silicon for a while.
Tomatotech - Tuesday, October 6, 2020 - link
Dancing_coffin_pallbearers.gif🕺🏻🕺🏻 ⚰️ 🕺🏻 🕺🏻
Apologies for the emoji, but I don’t see how this USB 3.2 Gen 2x2 standard will ever take off given the travails and issues described in the article. Best to kill it and bury it somewhere deep.
PeachNCream - Tuesday, October 6, 2020 - link
"USB has emerged as the mainstream interface of choice for data transfer from computing platforms to external storage devices."The opening line made me laugh a little since USB has been the dominant PC interface for a solid 15 years now.
Mitch89 - Tuesday, October 6, 2020 - link
Honestly, what on earth was the thinking behind the ridiculous naming (and renaming) of the USB versions?It used to be simple and clear, but now, between the crazy naming and mess that is USB-C (is it USB or Thunderbolt? Is the cable USB 2, 3, 3.1 Gen 2? Does it support 100W power?), how is anyone supposed to make sense on it.
Eric_WVGG - Tuesday, October 6, 2020 - link
On a related note, I would love to see you guys do some kind of investigation into why we're five years into this standard and one still cannot buy an actual USB-C hub (i.e. not a port replicator).hubick - Tuesday, October 6, 2020 - link
A 3.2 hub with gen 2 ports and a 2x2 uplink would be cool!I wanted a 10gbps / gen 2 hub and got the StarTech HB31C3A1CS, which at least has a USB-C gen 2 uplink and a single USB-C gen 2 port (plus type A ports). Don't know if you can do any better than that right now.
repoman27 - Tuesday, October 6, 2020 - link
Although it's still not exactly what you're looking for, I've tried (unsuccessfully) to get people to understand what a unicorn the IOGEAR GUH3C22P is. Link: https://www.iogear.com/product/GUH3C22PIt's a 5-port USB3 10Gbps hub with a tethered USB Type-C cable on the UFP (which supports up to 85W USB PD source), two (2!) downstream facing USB Type-C ports (one of which supports up to 100W USB PD sink), and two USB Type-A ports (one of which supports up to 7.5W USB BC).
serendip - Thursday, October 8, 2020 - link
No alt mode support like for DisplayPort. I haven't found a portable type-C hub that supports DisplayPort alt mode over downstream type-C ports although some desktop docks support it.stephenbrooks - Tuesday, October 6, 2020 - link
What if I want a USB 20Gbps port on the front of my computer?Can I get a USB-C front panel and somehow connect the cable internally to the PCI-E USB card?
abufrejoval - Tuesday, October 6, 2020 - link
I am building hyperconvergent clusters for fun and for work, the home-lab one out of silent/passive 32GB RAM, 1TB SATA-SSD J5005 Atoms, the next iteration most likely from 15Watt-TDP-NUCs, an i7-10700U with 64GB RAM, 1TB NVMe SSD in testing.Clusters need short-latency, high-bandwidth interconnects, Infiniband is a classic in data centers, but NUCs offer 1Gbit Ethernet pretty much exclusively, Intel struggling to do 2.5Gbit there while Thunderbolt and USB3/4 could do much better. Only they aren’t peer-to-peer and a TB 10Gbase-T adapter sets you back further than the NUC itself, while adding lots of latency and TCP/IP, while I want RDMA.
So could we please pause for a moment and think on how we can build fabrics out of USB-X? Thunderbolt/USB4 is already about PCIe lanes, but most likely with multi-root excluded to maintain market segmentation and reduce validation effort.
I hate how the industry keeps going to 90% of something really useful and then concentrating on 200% speed instead of creating real value.
repoman27 - Wednesday, October 7, 2020 - link
Uh, Thunderbolt and USB4 are explicitly designed to support host-to-host communications already. OS / software support can be a limiting factor, but the hardware is built for it.Existing off-the-shelf solutions:
https://www.dataonstorage.com/products-solutions/k...
https://www.gosymply.com/symplyworkspace
https://www.areca.com.tw/products/thunderbolt-8050...
http://www.accusys.com.tw/T-Share/
IP over Thunderbolt is also available:
https://thunderbolttechnology.net/sites/default/fi...™%20Networking%20Bridging%20and%20Routing%20Instructional%20White%20Paper.pdf
https://support.apple.com/guide/mac-help/ip-thunde...
repoman27 - Wednesday, October 7, 2020 - link
Stupid Intel URL with ™ symbol. Let's try that again:https://thunderbolttechnology.net/sites/default/fi...
abufrejoval - Wednesday, October 7, 2020 - link
Let me tell you: You just made my day! Or more likely one or two week-ends!Not being a Mac guy, I had completely ignored Thunderbolt for a long time and never learned that it supported networking natively. From the Intel docs it looks a bit similar to Mellanox VPI and host-chaining: I can use 100Gbit links there without any switch to link three machines in a kind of “token ring” manner for Ethernet (these are hybrid adapters that would also support Infiniband, but drivers support is only support Ethernet for host-chaining). Unfortunately the effective bandwidth is only around 35GByte/s for direct hops and slows to 16GByte/s once it has to pass through another host: Not as much of an upgrade over 10Gbase-T as you’d hope for: I never really got into testing latencies, which is where the Infiniband personality of those adapters should shine.
And that’s where with TB I am hoping for significant improvements over Ethernet apart from native 40Gbit/s speed: Just right for Gluster storage!
I also used to try to get Ethernet over fiber-channel working years ago, when they were throwing out 4Gbit adapters in the data center, but even if it was specified as a standard, Ethernet over fiber never got driver support and at the higher speeds the trend went the other direction.
So I’ll definitely try to make the direct connection over TB3 work: CentOS8 should have kernel support for TB networking and the best news is that it doesn’t have to wait for TB4, but should work with TB3, too.
I’ve just seen what seemed like an incredibly cheap 4-way TB switch recommended by Anton Shilov on the TomsHardware side of this enterprise, which unfortunately is only on pre-order for now (https://eshop.macsales.com/shop/owc-thunderbolt-hu... but supposed to support TB networking. Since the NUCs are single port TB3 only, that should still do the trick and be upgradable to TB4 for just around $150… The 5Gbit USB3 Aquantia NIC wasn’t much cheaper and even the 2.5Gbit USB3 NIC are still around $40.
Exciting, exciting all that: Thank you very much for those links!
abufrejoval - Wednesday, October 7, 2020 - link
...except... I don't think that "switch" will be supporting multiple masters, same as USB.If it did, Intel would have shot themselves in the foot: 40Gbit networking on NUCs and laptops with little more than passive ables, that's almost as bad as adding a system management mode on 80486L and finding that it can be abused to implement a hypervisor (Mendel and Diane started VMware with that trick).
Yet that's exactly what consumers really thirst for.
repoman27 - Wednesday, October 7, 2020 - link
According to the USB4 spec, host-to-host tunneling is possible via inter-domain links even when using a hub, but of course YMMV.Thanks for pointing out that Thunderbolt 4 hub, that's awesome! And way cheaper than I was expecting. It looks like a Goshen Ridge hub based on Intel's Thunderbolt 4 Compact Dock reference design to me. However, I'm a little worried that OWC's wording implies that it is not backwards compatible with Thunderbolt 3 hosts. That would seem insane, and it may just be an OS support issue at this juncture, but it sure sounds like it's for Thunderbolt 4 PCs only. Aargh!
CaptainChaos - Wednesday, October 7, 2020 - link
It *might* work, and cost may not be prohibitive & worth a shot! I don't see networking mentioned at all on the product page. Also, aren't thunderbolt 3 cables required to be active to do 40GB speeds?CaptainChaos - Wednesday, October 7, 2020 - link
... plus it's described as a hub and not a switch so expectations would need to be tempered accordingly!repoman27 - Wednesday, October 7, 2020 - link
IP over Thunderbolt is a thing, but you'll never see it advertised. Much like IP over IEEE 1394 (a.k.a. FireWire) was also a thing... that nobody knew about. Where we're going, we don't need Ethernet!It is a Thunderbolt 4 / USB4 hub, almost certainly based on the recently launched Intel JHL8440 chip. Per the USB4 spec, it contains a router which "includes a flat point-to-point, configurable switch necessary to create the internal paths between adapters", in addition to a PCIe switch as well as USB 3.2 and USB 2.0 hubs. It is also required to support Thunderbolt 3 interoperability and DisplayPort Alternate mode on all downstream facing ports. It's the real deal.
repoman27 - Wednesday, October 7, 2020 - link
Oh, also, the max length for passive cables at 40 Gbit/s is 0.8 m, and FWIW Apple sells theirs for $39. Full featured active cables are available in lengths up to 2.0 m, but cost $129 (from Apple).And throughput for host-to-host communications will be limited by the available bandwidth of the PCIe link between the host CPU and Thunderbolt controller. For Thunderbolt 3 hosts, that tends to be a PCIe Gen3 x4 link which results in real-world throughput of about 22 Gbit/s after accounting for protocol overhead.
So Thunderbolt isn't exactly going to be a panacea, but if you can live with those constraints, it is still a very fast interface.
CaptainChaos - Wednesday, October 7, 2020 - link
i want one (or 2 - wonder if they can be cascaded?)! Should be interesting to see follow ons from this or other vendors for cost or features. I've been eyeballing the 2.5baseT switch recently released by QNAP but this might be a better option for me.CaptainChaos - Wednesday, October 7, 2020 - link
... except that I guess I'd need to use a PC to bridge TB subnet to a 10 GB ethernet network :-(Deprectuod - Saturday, October 10, 2020 - link
It looks good!xpclient - Tuesday, October 13, 2020 - link
Hi Anandtech, can you do a performance comparison of 2 PCs networked via Thunderbolt 3 to Thunderbolt 3 standard cables (since it allows running 10Gbps Ethernet) vs networking using USB 20 Gbps (Gen 2x2) to RJ-45 10 Gbps Ethernet adapters between 2 PCs? Which is faster? I bet the Thunderbolt 3 connection will be faster due to the protocol overhead of USB.Howard - Sunday, October 18, 2020 - link
With the stupidity (or anti-consumerist behavior) so glaringly obvious, It's almost as if USB-IF want people to stick with Lightning for as long as possible.rishaby - Wednesday, November 25, 2020 - link
Hi,I've got a question. Let's say I purchase a gen 2x2 External SSD like Firecuda with R/W speed around 2000mb/s, will it be compatible with M1 Mac Mini? It has a thunderbolt port/USB 4.
https://scrn.li/ZUg87oBQPvBq7w
So it should work, right?