A lot of people are ignoring the fact that SoftBank Group is also one of the largest players in the telco market globally (that's how they started) [0] and that played a role in their Arm strategy and Eutelsat OneWeb strategy.
It's useful to look at the Ampere acquisition from the same lens, as Ampere is heavily used for telco usecases [1]
SoftBank Group has heavily invested in O-RAN to act as the base technology for what will become 6G, and they have been making a number of partnerships, investments, and acquisitions specifically around this [2][3].
SoftBank has been working on a 6G/Beyond 5G strategy for several years now [4] - primarily targeting embedded usecases such as autonomous vehicles [5][6]
Can someone knowledgeable explain SoftBank? Is it partly a Japanese sovereign fund...I've become less and less interested in its investments as it seems to simply act like a giant government money channel with questionable investments that don't seem to make sense from a typical VC/PE perspective. Ie is its ultimate goal not merely profit?
SoftBank was founded by Masayoshi Son. It was originally a tech company that resold software (imagine Steam but for desktop apps), but very quickly it morphed into something like a PE firm that would do continual leveraged buyouts of existing firms. They've bought companies that run conferences, tech magazines, one of their more formative M & A deals was to buy the Japanese Vodafone arm. Many of these deals actually went pretty poorly, with Masayoshi overpaying for the firms he bought and then being saddled with very large debts to repay the loans. Masa was wiped out and faced bankruptcy multiple times.
Later on the would move away from leveraged buyouts and start investing in startups more similar to how a VC firm would in the US. He notably invested early in Yahoo and Alibaba.
The company is not a sovereign fund - it's really run at the behest of it's founder. If you want to read more about the story, there was recently a great biography that chronicled the story: Gambling Man (https://www.amazon.com/Gambling-Man-Greatest-Disruptor-Masay...)
> old power in Japan) has been spending the money on recently
Investing in SoftBank.
MUFG, the Custody Bank, and Master Trust all have significant ownership stakes in SoftBank, and most Japanese conglomerates like NTT, NEC, etc tend to work closely with SoftBank on strategic deals
Yes, Ethnically Korean, his grandparents were immigrants. As for cultural differences, I think you could credit his extensive time spent living in the US as much as the immigrant Korean culture.
Japan has a weird split between Japanese nationals, Korean special residents, and actual foreigners (including "normal" Korean immigrants).
It's a whole can of worm I'll only leave here for any else to open, but Son Masayoshi was born in Japan from parents who might have formerly been Japanese (Korea was a Japanese colony) depending on the circumstances.
Many people in similar cases are basically undistinguishable from regular Japanese nationals culturally speaking.
It's an investment company that primarily specializes in Telecom and telecom adjacent industries (eg. Internet companies, Energy, Space Tech, Hardware).
Most people on HN have only heard of SoftBank's Vision Fund, but that's basically an autonomous organization within SoftBank staffed with Deustche Bank alumni who left in droves in 2017-2020.
> Ie is its ultimate goal not merely profit?
Yes. Their goal is profit.
> I've become less and less interested in its investments as it seems to simply act like a giant government money channel with questionable investments that don't seem to make sense from a typical VC/PE perspective
This investment is part of their strategy for 6G and IIoT, which is synergistic with their telecom focus.
There has been work over the past 4-5 years to become the leader for creating the standards for 6G, and SoftBank is trying to position themselves to be that player.
But it's so big (maybe biggest in the world), so its reach and influence is immense. And as a result, it has raised money from nations and other enormous entities.
Run by Masa who many consider an oracle because his placed massive bets years before other people and those bets have paid of big.
> Run by Masa who many consider an oracle because his placed massive bets years before other people and those bets have paid of big.
I'm not sure how many folks these days would consider him an "oracle". He clearly did fabulously well with some early Internet investments, but he also was famous for folly after folly of overpriced investments in the 2010s (the Softbank "Vision Funds").
I'd be curious if there is a simple accounting list of Softbank's major investments ranked from biggest winners to biggest losers. I guess it pretty much highlights the dynamics of the VC business model - you only need a few giant winners to offset the boatload of losers. In Softbank's case, I'm guessing their biggest winners are Yahoo Japan (which was the dominate site in Japan for a long time, and long after the US Yahoo fell into irrelevance) and Alibaba, which saw their early $20 million investment balloon into billions.
But did Softbank have any winners from their 2010s spending spree (along the time where they shoveled good money after bad into WeWork)?
They've made close to $100B on ARM (which they acquired in 2016). This dwarfs the $14B they lost on WeWork. Of course WeWork gets far more attention because it's a such an entertaining story.
The Vision Fund lost on tons more companies than just WeWork though. But your overall point is correct - there were a couple outsized winners (also just found another article that they made a huge gain on DoorDash) that I'm guessing took care of the majority of their losses.
VCs are indeed partly about giant winners offsetting the losers. But according to Taleb, the main part is to convince other people to put money in your fund and get rich from the management fees. So from that perspective, VC is mostly about pretending your fund does well and getting people to buy in. That's probably why VCs love to buy companies in overhyped industries.
Arguably this is classic VC maths, bet on many companies and the few that succeed will succeed enough that the rest don't matter.
YC does the same thing at the very early stage, A16Z and Sequoia do the same at later stages, and arguably Softbank made their money doing the same thing at a later stage again, although all the lines are a little blurred.
So Softbank now competes with Arm's customers and (unless I've missed something on new governance arrangements) also gets access to confidential information on Arm's customers' businesses (plans, volumes etc). Hmmm.
It kinda makes sense though, ARM has been creeping upwards over the years and yet afaik outside of Ampere no ARM server supplier has become serious (outside of inhouse variants such as Amazon Graviton).
So combining ARM and Ampere operations doesn't hurt any current big customers in the mobile field (apart from perhaps Qualcomm but they're in lawsuit mode there anyhow).
They also make sure to make the ARM alternative available to smaller customers (As an option besides AMD and Intel servers) instead of being gobbled up by f.ex. Oracle (making the big server ARM implementations inhouse to 2 of the big platform operators).
Not to mention the silver lining, combining the ops will give it more credence in the enterprise market that seems to have damn good margins (some have attributed AMD's resurgence and Intel's troubles to Intel slipping in server sales).
Largely agree. Arm / Softbank is in a tricky place if there are no credible customers in a given market. Does it just sit on its hands or actually get out there and build some chips! It has helped fund or help some startups that were using Arm designs to get into new markets (eg Annapurna Labs) but sometimes that doesn't work out at all (Nuvia!)
And to be fair I think Rene Haas has been fairly upfront about changing their approach.
They're the only company selling server grade ARM chips but others are rolling their own for internal use, such as Amazons Graviton line. That's Ampere's problem really, their biggest potential customers are so big that they can afford to cut out middlemen like Ampere.
Even if Ampere could afford to undercut the internal development of these hyperscalar ARM chips (which to my knowledge they cannot), there is more value in in-housing ARM chips like Amazon / Microsoft are doing: you can more easily try out wacky ideas in chip / firmware design.
Often times these companies might have an idea for how to save power / tweak cores in a weird way on a CPU, but have a chicken-and-egg problem with not knowing exactly how much it'll reduce COGS. Not knowing that and having to negotiate with a vendor (who will often want to charge more for the feature) means that it's difficult to do unless it's an obvious slam-dunk.
By bringing chip development in-house, adding new features skews more towards a political decision that requires less rigorous financial calculations (e.g. "how much power will this save relative to the vendor_cost++ and our own developer cost?" for 3rd-party chips). It basically allows these cloud providers to ship new features more quickly in their CPUs.
Interesting... that AL73400 part they're using is the first generation Graviton, and AFAICT that's not available on AWS anymore. Maybe Amazon considers it obsolete and is selling their stocks off to third parties like MikroTik?
I use Hetzner ARM machines myself and they're great, and something something i THINK there's less abuse and noisy neighbors on ARM since it's slightly less skiddie friendly (can't run random x86 binaries)
Somewhat but not competing with its customers has been a central part of Arm's proposition for three decades, for good reason. There are at least a lot of competitors for Nvidia's design. Your options if you want to build a server core are a lot more limited.
Ampere is still the only people making a decent purchaseable ARM server. Not loving the idea of there being a bunch of new debt hovering over them now.
The main problem with those is the memory controller starts to degrade a lot when you use all 128 of those cores.
We were doing some testing and > 96 cores at 100% caused a massive degradation in performance. We ended up going with dual 32C/64T Epycs (which cost twice as much) as a result. If they fix it in the Altra One chips they will be back on the table though because they were very good power wise for our workload and quite price competitive in a supermicro chassis.
Nvidia is targeting AI specific or fastest single core performance . While Ampere are targeting general Web CPU usage which is larger number of core per socket. The two are very different. I would have like Nvidia to enter this space as well but they seems to be focusing on everything AI related at the moment.
Or there are some specific workload on web that their CPU benefits and get filtered across ( DB hosting for example ) rather than going into a separate market for general usage.
AFAICT LOM only has three commands "PowerON", "PowerOFF", and "Reset", a far cry from what other BMC provide. It also requires another Mac in the same LAN to be the controller, and you cant use static IPs (which is pretty common for BMC VLANs).
Also it seems unsupported for the Mac Pro (2023) this thread is about?
They also won't provide any guarantees about components continuing to work, such as booting over thunderbolt that some clouds rely on. There's an awful lot of financial risk in providing MacOS as a service.
In addition to the problems other people have pointed out, it just doesn't have comparable specs to something like Ampere's Altra or AMD's Epyc.
24 CPU cores and 192 GB of memory might be enough for a small company's single on-site server, but those will likely want lots of storage and not want to deal with an exotic platform. Anyone who's willing to deal with ARM and macOS on a server will likely want a lot more performance per 1U.
have fun when that cheese grater hangs in the datacenter and you have to physically go in there to push a button, connect a display, a keyboard and a mouse to do troubleshooting, because apple in their infinite wisdom has decided that a rackable mac should not have ipmi.
Proxmox runs fine on M-series Macs. macOS also runs fine with no GUI.
Either way, it wouldn't be a good fit when Ampere exists.
For low power modular compute, the way GitHub and Amazon re-package Mac Minis is not the moest efficient way to get Apple hardware in a rack with remote management (and still no ECC), but it does have a performance per watt and compatibility for macOS CI workers that you can't get any other way.
Ideally, this stops being an issue when the ARM64 ecosystem can build those apps natively. We're pretty close on that one, and it means you can use Ampere and Graviton for that.
Yeah, I know, but given that you leave the user not-logged in and set the items system-wide startup items, the memory pressure on a 64GB system can be seen as negligible with some mental flexibility.
I think the primary use for a rackable server like this is automating building and testing of macOS native, or macOS specific applications you develop, or build them. Or as a shared remote desktop host for specific tasks.
Also we had (recently deprecated) macOS Server package for quite some time. So it's not unsuitable for servers. It's more suitable for specific jobs as a server.
Installing Linux servers are too easy now, so I don't think macOS should evolve a "server-specific" version either.
24 cores isn't that much, and you end up with a huge hulking GPU just sitting there doing nothing, you can't run a well-supported server OS on it so it can't really be used as a server, and it's extremely expensive for what it is. I'm not really sure what makes you think it's worth bringing up.
Bro, come on. Apple hasn't even tried to pretend they're in the server game for at least a decade now. You're paying for dozens of features you'll never use. They don't even have drivers for their own hardware outside of fucking windows, let alone documentation about how their machines are intended to work.
I love my macs but apple will never be a cost or reliability solution to computation unless they do a complete 180. They're a boutique/luxury firm that can choose which markets they want to dominate. If you think there's not a small horde of people at Apple trying to figure out how to pitch bringing the xserve back, I have a bridge to sell you. But unless they can convince CTOs to become sysadmins it'll never be a viable way for a company as large as Apple to drive hardware development.
I think they could be a decent provider for SOHO options... such as the space filled by Synology on the lower end. It's pretty competitive on cost, but Apple could really play it in concert with their wifi/airplay solutions etc. Like a home-base for SOHO.
Then again, the more I think on it, the worse an idea it is. Too locked-in and out for being able to run the likes of Jellyfin, etc. It could be a good option for a "Server" mac.. but would still make more sense for it to run Linux as a base.
I use their servers in a couple of cloud environments. That little things are snappy, for what I have given them as tasks. Wish them best of luck to the future.
Those 1000+ customers at Wiz have 7-8 figure TCV deals (I have personally seen a couple of those contracts), and have pipeline in multiple industry verticals - Tech, Telecom, ONG, Healthcare, Finance, Insurance, etc
An embedded and telco oriented firm like Ampere has weaker margins as they have to manufacture hardware, which drastically reduces the amount of money available.
It's also a smaller and more consolidated market with a weaker TAM.
They are totally separate the Wiz thing is a reverse IDF merger.
The SB purchase is part of Arm completely restructuring their IP philosophy and business model. Ampere was sold for "so little" because Arm could effectively just cut them off as soon as their contract ran out. This was a cleaner transfer.
Arm is bringing everything back in house, ceding the low end to RISC-V and taking the high end back in and producing their own silicon.
This is a wild exaggeration - very few people are a good fit for serving in a military, and while the Israeli army ignores that, and drafts many it shouldn't (resulting in high suicide rates on one hand, and alleged war crimes on the other), it still only drafts ~69% of AMAB and 55% AFAB.
It does gives free technical education and experience for some recruits (not necessarily the top, and that experience isn't always teh best), but of course that affects tech employment - you'd see higher rates of service across tech employees, as that service gave them a jump start on their tech education.
That said, even if the Israeli military decides to draft a person, that person can stand up and refuse. It ain't "free" (may result in imprisonment,) but it's a choice that even if too few make, it is one, and making it is a strong signal about what kind of person they are.
Wouldn’t it be expected that those most capable with technology and software be routed to the intelligence agency, rather than the front lines or something?
Opening up TLDs beyond the originals and CCTLDs completely invalidated assumptions about what domains and ___domain names should mean. A massive disservice to everyone whose intuition was helping them avoid scams and squatting, but hey, at least a lot of money changed hands.
Yeah... I'm pretty burnt on them too... then again, I own about 20 I'm not actively using. Still not fun when you are just trying to come up with a ___domain name for an idea and everything is parked and/or they want thousands for the name.
I think SB might've upped the bid a bit extra with Oracle as the other suitor. With Graviton and Ampere processors owned by hyperscalers it would've been a very unfortunate time for the ARM server market to become stagnant.
Right now ARM is in a golden position to take server market pieces, higher compute densities with simpler i-set and good software support thanks to years of mobile and now Apple support (most language implementators wants things to run on Mac's so bugs and perf have been getting fixed)
If Oracle would've bought them, third parties (in this case including both Microsoft with Azure and Google) would probably have held back on ARM investments (not to mention smaller operators), Qualcomm being in lawsuit mode would also had sown doubt, maybe some other manufacturer would introduce chips in the meanwhile but they would be small and given time Risc-V might even pop up on the horizon as an alternative hurting ARM on server chances even more.
Now they're "independant", with better software support than ever, guaranteeing availability to smaller parties (at a time when many of those are outside the US and wanting alternatives) and a chance to benefit from vertical integration.
Not sure on that... AMD's 128+ core servers are competitive to ARM on power per unit of work and exceed them on total compute. It's still very competitive.
Might be, but that is the thing, while it is tradition to complain about Oracle, in many cases their are the only ones that show up at the door with the suitcase full of money.
So where are the others that supposedly are better than them?
In this case, Softbank will be having ARM and Ampere.
This depends on definition of "better". Maybe your question is better suited as "So, where are the others that supposedly are richer than them?"
In that case, we have the CPU manufacturers, which already do this as the business, Google, which is not that interested, or already make their own chips, Apple, which has an in-house CPU/GPU company, and probably IBM, which also designs and builds their own processors.
So, the only interested shopper in the vicinity is Oracle, and they happen to have the money. The others have the money, but not the interest.
I think the fear is that if Oracle was the purchaser, then that would be good for Oracle, but not necessarily the rest of the ARM ecosystem. Given Softbank’s ARM investment, it is in their interest to try to keep the ecosystem healthy.
I’d love for someone more informed to provide some insider detail, but by appearances Oracle really, really tried to make SPARC a differentiator for them after the Sun acquisition.
They were shipping new systems (M8) based on a consistent in-house evolution of SPARC almost 8 years after the acquisition, a veritable eternity in tech time.
Eventually the market just said, no, we want x64 and a path to Cloud.
One wonders if they’d kept at it for a few more years whether they’d have needed to bid for Altera at all, given the gaps opened up by Intel’s struggles.
Even if Oracle tried, SPARC performance was terrible and their value for money was too. We benchmarked a finance system (Transaction Processing) where we needed 3x the SPARC hardware to keep up with Intel/Linux. Once they bought Sun I couldn't see any future where they made it work with keeping SPARC alive. Best case was keeping Solaris on x86 since that had genuinely better features (Zones and DTrace in particular).
Wrong. It's telco money - SoftBank Group is primarily a telco vendor and provider, and works closely with NEC, DoCoMo, T-Mobile, Vodafone, and Airtel.
SoftBank Vision Fund is a separate subgroup of SoftBank targeted at VC, and was largely operated and managed by ex-Deustche Bank alumni due to Rajeev Misra and Colin Fan poaching their colleagues like Ziyad Al-Ashaikh (the guy who helped land the Saudi PIF deal), Akshay Naheta, Munich Verma, Saleh Romeih, etc
In Vision Fund yes, but Vision Fund is just a part of the larger SoftBank Group, and did NOT partake in this transaction. This was a transaction done by SoftBank Group directly.
And much of Vision Fund's MENA LPs would leave if a couple key Vision Fund partners were to leave - they did the same with Desutche Bank.
What indication is there that x86 is on the decline? Intel might be on the decline, but AMD seems to be doing fine, and Windows on ARM looks to go nowhere, even with Qualcomm’s latest Snapdragon
I don't think ARM cores doing system management stuff that's not exposed to the user as a general purpose CPU really counts. But those SoCs count for sure, I didn't know they made those.
AMD said in 2021 that they were ready to build ARM chips[1], but honestly, I don't see the market for it. AMD could put an ARM frontend on Ryzen (and various interviews said they've prototyped it), but I don't think it would make a significant difference in compute/watt or whatever other metric, and there's not much in the way of ARM only software that people want to run. Maybe macOS, but you can still run that on amd64 afaik.
amd64 is still the default platform for desktop and server software. You can have arguments about the ISA, but for the most part that doesn't actually matter; amd64 is good enough, upstart cpu makers choose ARM not because it's better than amd64, but because it's at least good enough and licenses are available.
I don’t think x86 as an ISA is necessarily declining much and AMD has products with zen4c / zen5c compact cores intended to compete with Arm server products. AMD doesn’t have to pay licensing / royalties for x86 AFAIK and probably would prefer to keep x86 and not start paying arm and creating any adoption barrier form a change of ISA (however small) if they can make competitive x86 cores.
Because people are (mass generalisation) mostly only buying Arm to save money over x86.
For AMD to get into Arm would be to kill their profit margins. Same curse that Intel had.
What ARM core is available for AMD to use that is better than what AMD has been shipping in Turin?
Adapting Zen to a simpler-to-decode architecture would cost more than licensing a core from ARM. The Zen 5 front-end is adapted for working around deficiencies of x86 with dual-decode SMT craziness.
It's useful to look at the Ampere acquisition from the same lens, as Ampere is heavily used for telco usecases [1]
SoftBank Group has heavily invested in O-RAN to act as the base technology for what will become 6G, and they have been making a number of partnerships, investments, and acquisitions specifically around this [2][3].
SoftBank has been working on a 6G/Beyond 5G strategy for several years now [4] - primarily targeting embedded usecases such as autonomous vehicles [5][6]
[0] - https://www.softbank.jp/en/
[1] - https://amperecomputing.com/press/ampere-accelerates-expansi...
[2] - https://www.softbank.jp/en/corp/news/press/sbkk/2024/2024022...
[3] - https://www.o-ran.org/membership
[4] - https://www.suasnews.com/2021/08/softbank-corp-unveils-6g-co...
[5] - https://www.softbank.jp/en/corp/technology/research/news/051...
[6] - https://www.softbank.jp/en/corp/technology/research/news/069...