SoftBank considering selling ARM

[MK14]

I'm pleased they are selling it. Because I don't think they have done a brilliant job of running it.
E.g. They don't seem to have made a big push, of arm cores into the server market (but on the other hand, it could be the correct and wisest choice, in business terms). I don't know. (i.e. they were not going to succeed well enough, to make it financially viable, in that sector).

Did you notice how much money went down the drain trying to get more ARMs into the data centre BEFORE Softbank bought ARM?

By and large, yes. That's why, I said it could of been a correct 'business' decision.

But (even if it was the CORRECT business decision), maybe, I'm not 100%% convinced. because there was a potential 'small server and small PC like computer' markets, such as microservers.
Which, seems to have been absorbed by the likes of Raspberry PI's and similar, single board (tiny) computers.

I.e. We already have many mobile phones and tablets, being arm based.

Why not have many microservers and tiny (not PC compatible) PC's, as well ?
(Although there are some/many arm based microservers, such as some NAS boxes).

EDIT:
But I concede. That I DON'T have the market size/profit information, that arm would have available. So, maybe the market place there, is too small, competitive and unprofitable.
Also, arm DON'T actually sell the processors or hardware as such. It is up to other companies, to licence the available arm cpus, put them into production and then sell the microservers, servers and small (non-PC compatible) PC computers, to others.
If that doesn't take place, you can't fully blame arm for that.

AMD, seem to be having a very successful time, with the various new Ryzen cpus, in the desktop and server, market places. So, even Intel are having difficulties, let alone arm in those market areas, who weren't even in those market segments (arm), to any great extent (unless you count the rather small units, such as Raspberry PI's, which have been a success).

[SiliconWizard]

When you say "fact ARM cores are virtually EVERYWHERE".

That was essentially the situation, when SoftBank purchased arm, in 2016, almost exactly 4 years to the day.

I.e. I wasn't criticising the original owners of arm, who did a rather good job.

OK, I probably got confused a bit by that too, as I was mainly talking about the general history of ARM.
But as coppice said, their late attempts at trying to get into the server or desktop computing markets were essentially mistakes and cost them a lot of money.

I don't know how much of the current situation is due to SoftBank. They probably just tried to cut the losses.

I still don't think there's really a market for "microservers" or "small PC computers", at least beyond what already is there. I don't really believe in the microservers thing. What purpose would that serve exactly? Are you thinking about "home" servers or NAS stuff? Many of them are already ARM-based, so that's not a new maket to tap into.

As to small PCs? Traditional desktop computing has already largely declined in favor of mobile devices, in which you already find - mainly - ARM-based processors. Is there really an intermediate market worth investigating at this point? Not so sure. There's a very fine line these days between a small, low-power laptop and a decent tablet. As to any low-power PC that's not a laptop - yeah, this is declining, so targeting a declining market is probably not really smart.

As to what could ARM do to get better results, I don't know. As I mentioned, I suspect they should rethink their licensing schemes to begin with. And yes, address long-term issues such as RISC-V. RISC-V is still not much of a threat in the western world, but still something to keep an eye on, and it's going to be in China soon and probably other eastern countries. I'm wondering whether open-sourcing some ARM cores could be a long-term solution for instance. It would have to be thought through with a complete strategy. Don't know how that would work out.

[MK14]

Are you thinking about "home" servers or NAS stuff? Many of them are already ARM-based, so that's not a new maket to tap into.

As to small PCs?

Those are exactly what I was referring to (but I was also referring to bigger real (arm based) servers) as well, in my earlier posts.

You and coppice, can easily be right.
I am just not geared up to the right, precise business information, to know, if they (various owners of arm), could have done any better in those markets. I.e. take those comments with a large grain of salt (that is, not very correct or accurate, as regards the business merits of that market for arm, tl;dr = Yes I could of been wrong).

I agree, you can already buy, arm based microservers, such as (some versions of) Synlogoy NAS's (which in real terms are microservers, marketed as NAS's). But what I wanted is something which is general purpose, and usable without being tied into someone else's software base, and which doesn't let you install your own OS's and software, in a custom sort of way.
Ie. The Synology NAS's, primarily are limited to Synology software, and their (somewhat limited) packages.

The Raspberry PI (e.g. 4), serves the role of (home use) microserver and/or tiny PC. But, it would be nice to have something ready cased, and more ready to go. Also, having Sata port(s), and one or two other things, would help as well.

Late EDIT: Shortened, turned business section of post into more like a summary.

[bsfeechannel]

A year ago it was China vs the US, and now after the pandemic, it seems the answer is clear -- it is China vs the entire West

On December 18, 2018, Xi Jinping and other Chinese leaders awarded ten foreigners with the China Reform Friendship Medal, China's highest award, at the celebration of the 40th anniversary of China's reform and opening up.

Those were:

Alain Mérieux, French, head of a medical conglomerate, who has contributed since 1978 to China's health care industry and worked closely with China during major public health issues.

Werner Gerich, German, engineer, dubbed Mr. Quality by the Chinese,  who, arrived in 1984, was hired as a consultant at a state-run factory in Wuhan. Contributed to reforming aspects such as product quality, factory hygiene, company organisation structure and factory workers’ livelihoods. Died in 2003. There's a bronze bust of him in Wuhan.

Klaus Schwab, German, engineer and economist, founder and executive chairman of the WEF, who is an openly supporter of the "Belt and Road Initiative".

Konosuke Matsushita, Japanese, founder of Panasonic, who advised Deng Xiaoping in 1978 on how China could work with overseas companies on technical matters and on establishing joint ventures. Panasonic became a household name for Chinese families. Matsushita died in 1989.

Masayoshi Ohira, Japanese foreign minister, who established formal diplomatic relations between China and Japan in 1972, advised Deng, and provided long-term financial assistance and support in the early stage of the opening up. Ohira died in 1980.

Lee Kuan Yew, Singapore's prime minister, who developed a long friendship whith Deng. Deng said, in 1992, that Singapore was an example for China. Although of Chinese descent, Lee's first language was English and he attended the London School of Economics and graduated from Fitzwilliam College, Cambridge, in law. Lee died in 2015.

Juan Antonio Samaranch, Spanish, former president of the International Olympic Commitee, who helped China emerge as a major sporting power and fulfill China's 100 -year dream of hosting the Olympics in 2008. Samaranch died in 2010.

Stephen Perry, English, chairman of the 48 Group, an independent business network commited to promote links with China. He continues the work of his father Jack Perry, who in as early as the 1950s, developed business ties to break China's isolation from the West. He's considered one of Britain's top experts on China.

Maurice Greenberg, American, former chairman and CEO of American International Group. AIG was founded 100  years ago by the American Cornelius V. Starr in Shanghai, China. Greenberg was pivotal in helping China to have US governmental approval for permanent trade relations, for China to join the WTO, and donated millions of dollars to environmental, educational, medical and cultural causes in China.

Robert Lawrence Kuhn, American neuroscientist, investment banker and corporate strategist. Kuhn is a long-time advisor to China's leaders and the Chinese government, to multinational corporations on China strategies and transactions. He's a columnist of China Daily and host and co-producer of "Closer to China" and "China's Challenge" TV series. Heck, he's even an advocate of Xi's eternal presidency and is the author of books with themes dear to the CCP.

Of those ten foreigners, seven are Westerners. Five are alive. And of those, two are American. Although you might find people out there who would say that Japan and Singapore are also part of the West, what is important to see here is that the West, directly or indirectly, is not an enemy of China. The West is a supporter.
.
Maybe what the Covid-19 pandemic made clear is that everybody is a bit disappoint.



And some may be confusing it with enmity.

[olkipukki]

Bllomberg reports that SoftBank may be selling ARM, and perhaps Nvidia would be interested in buying (and Apple isn't).

Let's be positive, neither IBM or Oracle are not interested right now 

Although, why not? ARM-only world became too boring 

[soFPG]

There are some chinese bluetooth SoCs with a proprietary 32-bit RISC CPU claiming they have better power / performance ratio than Cortex M0.

Not sure how valid that statement is but seems like ARM technology is not the way to go for some companies (incl. Espressif). Of course, proprietary CPUs often have worse debug interfaces / software support.

I doubt that with licensed ARM cores something like the ESP8266 or ESP32 could be as cheap as they are.

In case it is not clear, we are not racist. We welcome anyone who is willing to contribute to China and to abide by Chinese laws.

Most engineers are driven by their excitement for the subject. Maybe USA didn't offer to those people what they were searching for, so they went abroad.
This doesn't necessarily mean that they accept the Chinese system.

[soFPG]

W600/W601 is a family of low cost...

Very interesting, learned something new today

[GeorgeOfTheJungle]

Realtek makes those too (ARM+WiFi): Realtek RTL8720DN, dual-band 2.4Ghz / 5Ghz Wi-Fi

E.g.:
https://www.seeedstudio.com/Wio-Terminal-p-4509.html

[soFPG]

And more you can learn is -- don't touch them.

I've already looked at them some time ago and just didn't remember. But what I am remembering now is exactly what you said. Bad software support.
Pretty much the same for chinese bluetooth SoCs. Don't know why they went with their own core (e.g. Telink, which already is the best in terms of software support I came across) if ARM licensing doesn't seem to be a monetary problem.

[SiliconWizard]

The Raspberry PI (e.g. 4), serves the role of (home use) microserver and/or tiny PC. But, it would be nice to have something ready cased, and more ready to go. Also, having Sata port(s), and one or two other things, would help as well.

You can find many different SBCs these days, some with SATA, even M2, PCI express, etc.
Some have cases you can buy too. On most of them, you can either install Android or some Linux distribution.
Look up SBCs - there are now A LOT of them available beyond the RPi.

There are various sources to find them, you can start here for instance: https://www.hackerboards.com/home.php
Example: http://www.banana-pi.org/w2.html

As to ready to go, well, you can also find small boxes, essentially based on some kind of SBC. Most of them are either sold as NAS solutions, or as multimedia boxes. The latter usually provided with Android. Installing another OS is rarely possible though, as they are essentially fully-custom, undocumented boards. But if you want an Android device with an ARM processor that isn't a mobile phone or a tablet, you can certainly buy that.

Else I would stick to SBCs. But still - keep in mind that most issues with them these days (when they are decently designed) is not with the hardware, so ARM could not do much here to promote them, but with the software. Software support is often NOT complete, documentation is sometimes lacking, etc. That is probably due to... the lack of market.

[MK14]

You can find many different SBCs these days, some with SATA, even M2, PCI express, etc.
Some have cases you can buy too. On most of them, you can either install Android or some Linux distribution.
Look up SBCs - there are now A LOT of them available beyond the RPi.

There are various sources to find them, you can start here for instance: https://www.hackerboards.com/home.php
Example: http://www.banana-pi.org/w2.html

As to ready to go, well, you can also find small boxes, essentially based on some kind of SBC. Most of them are either sold as NAS solutions, or as multimedia boxes. The latter usually provided with Android. Installing another OS is rarely possible though, as they are essentially fully-custom, undocumented boards. But if you want an Android device with an ARM processor that isn't a mobile phone or a tablet, you can certainly buy that.

Else I would stick to SBCs. But still - keep in mind that most issues with them these days (when they are decently designed) is not with the hardware, so ARM could not do much here to promote them, but with the software. Software support is often NOT complete, documentation is sometimes lacking, etc. That is probably due to... the lack of market.

Thanks!
There is a bewildering amount of choices, at various price points. Also, some difficult choices to make.
Especially, to go the arm route e.g. Raspberry PI, the X86 direction, or an off the shelf NAS box solution (which sometimes is also an arm processor).

EDIT:
What I've found (or been told, and/or from other sources), is that although there is a huge number of options. Many of the options, are rather poorly supported as regards, OS's, other software, and drivers (especially in the x86 world, where it might mean your favourite operating system, can't install and/or be truly useful, because of a lack of drivers, for the new Soc devices).
Taking that into account, it dramatically reduces, your choices, to only a few (well supported) options.

[wraper]

NVidia is another story. They don't have the same potential issues with competitors, and they are a (albeit fabless) semiconductor company. They sell ICs. They have a few products based on them, but mostly for show-off purposes IMO. Designing and selling ICs is their core business. Apple is NOT a semiconductor company, they probably wouldn't know how to run one and I don't think they ever want to be one.

Apple IS a fabless semiconductor company since quite a while ago. Their CPUs use ARM instruction set but have custom cores since A6 SoC.

[SiliconWizard]

NVidia is another story. They don't have the same potential issues with competitors, and they are a (albeit fabless) semiconductor company. They sell ICs. They have a few products based on them, but mostly for show-off purposes IMO. Designing and selling ICs is their core business. Apple is NOT a semiconductor company, they probably wouldn't know how to run one and I don't think they ever want to be one.

Apple IS a fabless semiconductor company since quite a while ago. Their CPUs use ARM instruction set but have custom cores since A6 SoC.

No they aren't.
They have been designing their own CPUs (at least for their mobile products) based on ARM cores for a while, but AFAIK, that's strictly for their own use. I don't think they sell a single Apple processor to any third-party. Or do they?
So that doesn't make them a semiconductor company. They don't sell ICs. They sell products. So, yeah, they have the internal resources and know-how to design CPUs around ARM cores, but that's not their core business and they still wouldn't want nor probably be able to become a semiconductor company, marketing and selling processors to third parties, handling the tech support for that, even architecturing processors for other uses than their own, and so on. That's a whole other business.

[SiliconWizard]

What I've found (or been told, and/or from other sources), is that although there is a huge number of options. Many of the options, are rather poorly supported as regards, OS's, other software, and drivers (especially in the x86 world, where it might mean your favourite operating system, can't install and/or be truly useful, because of a lack of drivers, for the new Soc devices).
Taking that into account, it dramatically reduces, your choices, to only a few (well supported) options.

Another factor beyond software support is cost. Sure the basic SBCs such as the RPi are cheap, but at this point, the more capable, ARM-based SBCs (with SATA, M2, several GB of RAM, etc) can often be more expensive, in the end, than, say, a typical Mini-ITX motherboard with an entry-level x86 CPU, especially when you factor in all required accessories, additional components, etc.

Yet another factor, hardware-wise, is heat management. Many of them do not come with any heatsinking solution, and the additional heatsinks you can buy are often not enough to get proper heat dissipation unless you use the board with pretty light loads. Finding or making proper heat management may be a bit of a headache, and will add cost.

But software support is certainly a problem, as I already mentioned. It's often pretty poor, and a given SBC is rarely maintained for more than a year, maybe two, after which the manufacturer usually releases a new one and stops supporting the old models. And when some SBC comes out, software support is often not complete.

[GeorgeOfTheJungle]

Armbian works very well in all these chinese ARM SBCs. Install your favourite desktop manager if you need one, apt install whatever else you need, and if there isn't a package simply download the sources and ./configure and make. That's all folks, everything works 99.9% of the times.

https://www.armbian.com/download/

My favourite least expensive SBC is the Orange Pi PC (1GB / 4 cores @1.4GHz / $15), and on the more expensive side the Odroid XU4 (2GB / 8 cores @2GHz / $49).

[GeorgeOfTheJungle]

Apple IS a fabless semiconductor company since quite a while ago. Their CPUs use ARM instruction set but have custom cores since A6 SoC.

No they aren't.

Yes, they are fab-less, as far as I know, they buy Intels and Samsung makes their ARM chips (or is it TSMC now?)

[nctnico]

The Raspberry PI (e.g. 4), serves the role of (home use) microserver and/or tiny PC. But, it would be nice to have something ready cased, and more ready to go. Also, having Sata port(s), and one or two other things, would help as well.

There is a whole flurry of ATX motherboards with an ARM CPU on it. A random pick from Google:
https://www.cnx-software.com/2015/03/27/gigabyte-mp30-ar0-is-an-arm-server-motherboard-powered-by-applied-micro-x-gene-1-soc/

[SiliconWizard]

Apple IS a fabless semiconductor company since quite a while ago. Their CPUs use ARM instruction set but have custom cores since A6 SoC.

No they aren't.

Yes, they are fab-less, as far as I know, they buy Intels and Samsung makes their ARM chips (or is it TSMC now?)

You didn't read what I wrote, nor what the point was.

[MK14]

Another factor beyond software support is cost. Sure the basic SBCs such as the RPi are cheap, but at this point, the more capable, ARM-based SBCs (with SATA, M2, several GB of RAM, etc) can often be more expensive, in the end, than, say, a typical Mini-ITX motherboard with an entry-level x86 CPU, especially when you factor in all required accessories, additional components, etc.

Yet another factor, hardware-wise, is heat management. Many of them do not come with any heatsinking solution, and the additional heatsinks you can buy are often not enough to get proper heat dissipation unless you use the board with pretty light loads. Finding or making proper heat management may be a bit of a headache, and will add cost.

But software support is certainly a problem, as I already mentioned. It's often pretty poor, and a given SBC is rarely maintained for more than a year, maybe two, after which the manufacturer usually releases a new one and stops supporting the old models. And when some SBC comes out, software support is often not complete.

That is exactly what I think as well!
Word perfect!

The Raspberry PI 4, seems to tick most of the (good) boxes, and what limits it has, often have cheap/easy workarounds.
 E.g. There are stick on heatsinks available for it (and even tiny fans, I think, if you really want them).
The software availability for the Raspberry PI 4, is rather good as well (relatively).
But it still has to take a big back seat to a popular X86 solution (i.e. mainstream, such as Ryzen, not a very cheap Intel Atom, SoC, with limited/poor driver support for Linux and other non-windows OS's).

I suppose, part of the blame is arm themselves (NOT arm as a company necessarily, buy the entire industry surrounding the arm cpus).
Because, with the X86 (ignoring some too cheap, incompatible driver, SoC, Intel cheapo processors, i.e. very cut down Atoms and stuff, which mainly only really work well, under Windows OS's), they show huge compatibility, between even some ancient processors in the line up Intel/AMD/Cyrix(gone now)/Via(rare now) etc. Even future cpus, will tend to be compatible as well.

I.e. You can take some software X86, and run it on todays cpis, or old Pentiums, or new cpus, which haven't even been designed yet, but will be downwardly compatible.

Conversely, there is a bewildering range of arm chip architecture versions. Which, although sometimes offer maybe 1 (or 2 ?) generations of compatibility. In general, can be completely incompatible, between different versions, without recompiling software, and potentially other (various levels of difficulty) changes, to make it work//compatible again.
I.e. Early Arm1 software (if there ever was an Arm1, so whatever the first was, maybe Arm2 or 3), won't work on the latest Arm processors (ignoring emulation, recompiling and other workarounds).

But, 386 software (for example), should still work, even on the latest Intel/Amd X86 cpus. Although it might need an updated version of DOS, or virtualisation. In order to actually run it.
But that is more because Windows 10 doesn't support ancient software, rather than the cpu, which actually probably still does.
As far as I know, an ancient 8088 program, would still run on the latest X86 cpus.

Although arm have their reasons for not supporting older generations (except perhaps the last generation). because it makes their chips a bit faster, cleaner and better/cheaper.
It hinders (arguably), the software creation process.
The various (non-standard) SoC features, are another layer of incompatibility.

tl;dr
In theory, a quick recompile, and the old software (arm), can run on the new arm hardware. In practice, things are never, anywhere near, as simple/quick as that in practice. Unfortunately.
So to me, that significantly favours X86 over arm. If OS's are involved.

[GeorgeOfTheJungle]

Also, having Sata port(s), and one or two other things, would help as well.

Most do not have SATA, if you need it use a USB(3?)->SATA and Bob's your uncle. Most of these SBCs can't or can barely keep up with gigabit ethernet (125 MB/s), and the same goes for SATA: you won't get full SATA throughput speeds in any of them.

If the performance were comparable to an Intel CPU, Intel would be bankrupt by now, thus it is NOT, keep that in mind. What you get is the CPU of a mobile phone in a pcb: that's what these SBCs are.

EDIT:
Oh, and by the way, mobile CPUs are not designed to run at full chooch all the time.

[MK14]

The Raspberry PI (e.g. 4), serves the role of (home use) microserver and/or tiny PC. But, it would be nice to have something ready cased, and more ready to go. Also, having Sata port(s), and one or two other things, would help as well.

There is a whole flurry of ATX motherboards with an ARM CPU on it. A random pick from Google:
https://www.cnx-software.com/2015/03/27/gigabyte-mp30-ar0-is-an-arm-server-motherboard-powered-by-applied-micro-x-gene-1-soc/

Over the years. Whenever I've heard about things like that. They have been some combination of:
>>Way, way too expensive (e.g. £1100 for a motherboard including cpu), when much cheaper and more powerful options (X86), are available, for a fraction of that cost.
>>Completely unavailable, or unavailable in my country and/or unavailable except for big business orders
>>Had useless (opinion), ancient, very low (processing power cpus. When even the latest Raspberry PI's, have much more modern/faster arms.
>>There are continual promises, that it will be ready and launched 'NEXTYEAR'. Where nextyear, just never occurs, and the project gets cancelled.

E.g. AMD K12 Arm cpus.  https://www.fudzilla.com/news/pc-hardware/39179-jim-keller-was-not-a-big-fan-of-k12

[GeorgeOfTheJungle]

[ ... ]
As far as I know, an ancient 8088 program, would still run on the latest X86 cpus.
[ ... ]

You seem to think in terms of .exe(s), like a windows user... but Linux doesn't work like that. You apt install things, and if it isn't in the apt repository, you download the sources and ./configure and make it. That's how it works in Linux. And it works very very well, 99.999% of the times.

Edit:
A .exe program has no value in Linux... because the source code is available. You don't need to keep it saved like if it were a treasure, a .exe, because you can always recompile the thing for the latest and greatest CPU you just bought.

[MK14]

Also, having Sata port(s), and one or two other things, would help as well.

Most do not have SATA, if you need it use a USB(3?)->SATA and Bob's your uncle. Most of these SBCs can't or can barely keep up with gigabit ethernet (125 MB/s), and the same goes for SATA: you won't get full SATA throughput speeds in any of them.

If the performance were comparable to an Intel CPU, Intel would be bankrupt by now, thus it is NOT, keep that in mind. What you get is the CPU of a mobile phone in an pcb: that's what these SBCs are.

Thanks for that.
That is exactly what I am planning to do (unless the onboard SD card, is fast enough, big enough and its relative lack of data integrity/reliability/endurance, is acceptable for the intended uses).
Essentially, just buying an external type USB3 interfaced HDD/SSD, or even a flash pen. Depending on the intended applications.
The Lack of real Sata port(s), does kind of hurt, if you are trying to do big things. Such as have one or two 4TB HDDs, to make a big NAS box. USB3 is usable for that, but a bit of a limiting factor.
I suspect heavy duty, fast, totally reliable, uses, would much prefer Sata, over USB3, connections.
If you want RAID mirrors (soft raid, as hard raid, would need proper server hardware, to do it any justice), for improved data-integrity, then Sata becomes even more important.

Disks are changing anyway. As everyone is gradually migrating from HDDs, to SSDs. Which further complicates the issue (since SSDs can be really fast, if you have interfaces that support their full high speed nature).

[wraper]

NVidia is another story. They don't have the same potential issues with competitors, and they are a (albeit fabless) semiconductor company. They sell ICs. They have a few products based on them, but mostly for show-off purposes IMO. Designing and selling ICs is their core business. Apple is NOT a semiconductor company, they probably wouldn't know how to run one and I don't think they ever want to be one.

Apple IS a fabless semiconductor company since quite a while ago. Their CPUs use ARM instruction set but have custom cores since A6 SoC.

No they aren't.
They have been designing their own CPUs (at least for their mobile products) based on ARM cores for a while, but AFAIK, that's strictly for their own use. I don't think they sell a single Apple processor to any third-party. Or do they?
So that doesn't make them a semiconductor company. They don't sell ICs. They sell products. So, yeah, they have the internal resources and know-how to design CPUs around ARM cores, but that's not their core business and they still wouldn't want nor probably be able to become a semiconductor company, marketing and selling processors to third parties, handling the tech support for that, even architecturing processors for other uses than their own, and so on. That's a whole other business.

You have some strange logic. How by not selling ICs to 3rd party you stop being a semiconductor company? Their ICs go into every single phone/tablet/laptop they sell, and that's a huge amount. Way more than many "true semiconductor companies" produce. Not a core business? If production is halted they will have nothing to sell at all. Also those are not "ARM" cores in a meaning that they are cores designed by ARM. They design their own cores. If they wanted to start selling SoC to 3rd parties, they could start doing so in a blink of an eye.

[GeorgeOfTheJungle]

The Lack of real Sata port(s), does kind of hurt, if you are trying to do big things. Such as have one or two 4TB HDDs, to make a big NAS box. USB3 is usable for that, but a bit of a limiting factor.
I suspect heavy duty, fast, totally reliable, uses, would much prefer Sata, over USB3, connections.
If you want RAID mirrors (soft raid, as hard raid, would need proper server hardware, to do it any justice), for improved data-integrity, then Sata becomes even more important.

No problem at all, because 125MB/s (gigabit ethernet speed) is less than the max speed of SATA(3), so the disk(s), even if connected through USB3, won't be the bottleneck.

Software RAID is fully functional in Linux and works very very well, it's called mdadm (apt install mdadm). You can make any raid you want with that, look: https://linux.die.net/man/8/mdadm