SoftBank's Arm registers for blockbuster U.S. IPO

[erek]

As Intel’s fab business continues to falter: https://www.digitimes.com/news/a20230428VL201/intel-foundry-chips+components.html

“Earlier this year, Arm rebuffed a campaign from the British government to list its shares in London and said it would pursue a flotation on a U.S. exchange.

Arm's IPO preparations are being led by Goldman Sachs Group Inc (GS.N), JPMorgan Chase & Co (JPM.N), Barclays (BARC.L)“

Source: https://www.reuters.com/markets/deals/softbanks-arm-registers-blockbuster-us-ipo-sources-2023-04-29/

 

[sharknice]

I thought Softbank was buying Blockbuster for a second after reading that headline.

 

[Red Falcon]

...and everyone was worried that ARM would be taken over by NVIDIA, because this is so much better.

 

[Twisted Kidney]

...and everyone was worried that ARM would be taken over by NVIDIA, because this is so much better.

THis is definitely better. A monolithic, monopolistic, market crushing corporate entity (nVidia) is slightly worse than a company annihilating its own founding principals in the pursuit of shareholder value.

Slightly.

 

[atp1916]

Suprised WeWork didn't teach these people a lesson (about the US market)

 

[toast0]

THis is definitely better. A monolithic, monopolistic, market crushing corporate entity (nVidia) is slightly worse than a company annihilating its own founding principals in the pursuit of shareholder value.

Slightly.

I'm pretty sure Arm was publicly traded for years before Softbank bought them, not sure if it's better to be owned by Softbank or the public market.

 

[Lakados]

Well SoftBank is playing hardball and doing all the things everybody feared NVidia would do and more. This IPO is a cash grab to pay for the SoC’s they need to develop and deliver since 3’rd parties will no longer be allowed to modify the chips or design their own packages.

 

[emphy]

Well SoftBank is playing hardball and doing all the things everybody feared NVidia would do and more. This IPO is a cash grab to pay for the SoC’s they need to develop and deliver since 3’rd parties will no longer be allowed to modify the chips or design their own packages.

The ipo is a cash grab to pay for softbank's miserable big-tech gambles. Arm itself won't be seeing a single cent.

 

[Lakados]

The ipo is a cash grab to pay for softbank's miserable big-tech gambles. Arm itself won't be seeing a single cent.

Well ARM is tanking a few lawsuits right now and SoftBank’s name is more apt than ever as they seem quite flacid.

I’m sure legally they have to funnel it through ARM, but how much of it stays there is anybody’s guess.

 

[david123]

Softbank must be in big trouble. They sold all alibaba stock. Now want to get cash from ARM. Not sure how long they can last, considering the world economy going downward spiral.

 

[erek]

Suprised WeWork didn't teach these people a lesson (about the US market)

""The We Company" was described in its IPO filing as an umbrella with three major arms: WeWork, WeLive, and WeGrow.

The antics that led to Neumann stepping down from his CEO role included reports of him smoking weed on a private jet, serving employees tequila shots after discussing layoffs, and trademarking the term "We" and then forcing WeWork to buy it for $5.9 million."

In addition to the concerns about WeWork's governance, business model, and financials, the S-1 filing also contained some odd statements that raised eyebrows. Some of these included:

1. "We are a community company committed to maximum global impact. Our mission is to elevate the world's consciousness."
2. "Our global potential is limitless."
3. "We have redefined the way people work and elevated the world's consciousness."

These statements, along with other grandiose and vague language, led some observers to question WeWork's business model and management team, as well as the overall hype surrounding the company. The statements also raised concerns about whether the company was more focused on its image and brand than on its actual operations and financials.

 

[Verge]

Source: https://www.reuters.com/markets/deals/softbanks-arm-registers-blockbuster-us-ipo-sources-2023-04-29/

What is so "blockbuster" about ARM being traded on a us exchange.


This has already been a thing. https://www.barchart.com/stocks/quotes/ARMH

 

[Lakados]

What is so "blockbuster" about ARM being traded on a us exchange.


This has already been a thing. https://www.barchart.com/stocks/quotes/ARMH

What’s so blockbuster is the opening price…
They expect to value ARM around $50B this time around…

 

[erek]

What’s so blockbuster is the opening price…
They expect to value ARM around $50B this time around…

Arm Ltd Files for IPO on Nasdaq, Aiming to Raise $8-10 Billion​

by AleksandarK Today, 08:29 Discuss (0 Comments)
According to the latest report from Reuters, Arm Ltd has filled the documents for its initial public offering (IPO) efforts in hopes of getting publically traded later this year. The stock exchange of choice is Nasdaq, where Softbank plans to list Arm's shares publicly. Seeking to raise anywhere between 8-10 billion US Dollars, the company's market capitalization has yet to be determined. If any factor is to go by, NVIDIA tried to acquire Arm Ltd for 40 billion US Dollars, which ultimately failed due to regulators rejecting the deal.

As a reminder, Arm is changing its licensing model to boost royalties, which we reported about here. Goldman Sachs, JPMorgan Chase & Co, Barclays, and Mizuho Financial Group guide the IPO efforts.

 

[Lakados]

Arm Ltd Files for IPO on Nasdaq, Aiming to Raise $8-10 Billion​

by AleksandarK Today, 08:29 Discuss (0 Comments)
According to the latest report from Reuters, Arm Ltd has filled the documents for its initial public offering (IPO) efforts in hopes of getting publically traded later this year. The stock exchange of choice is Nasdaq, where Softbank plans to list Arm's shares publicly. Seeking to raise anywhere between 8-10 billion US Dollars, the company's market capitalization has yet to be determined. If any factor is to go by, NVIDIA tried to acquire Arm Ltd for 40 billion US Dollars, which ultimately failed due to regulators rejecting the deal.

As a reminder, Arm is changing its licensing model to boost royalties, which we reported about here. Goldman Sachs, JPMorgan Chase & Co, Barclays, and Mizuho Financial Group guide the IPO efforts.

Well hot fuck!
Not even a month ago all the big firms were all $30-70B and the market settled at 8-10…
Jesus that has got to be a feels bad moment for somebody there.

 

[staknhalo]

They missed the 2020/2021 IPO boat

Should have been sold to Nvidia, would have literally been the best thing for them, themselves, from just the company alone standpoint

 

[Lakados]

They missed the 2020/2021 IPO boat

Should have been sold to Nvidia, would have literally been the best thing for them, themselves, from just the company alone standpoint

But no regulators were afraid they would raise prices, change the licensing model, and remove the ability for vendors to customize the designs…

So instead ARM has decided to change from a per chip fee to one based on the selling price of the final product ( how Qualcomm licenses its modems). They are removing the ability to customize the chips and instead licensing full SoC’s, and they are instead issuing those licenses to the device manufacturers so they can potentially cut Broadcom, Qualcomm, Mediatek, and the likes out completely and contract to fabs directly…

 

[Verge]

What’s so blockbuster is the opening price…
They expect to value ARM around $50B this time around…

18b in growth over 7 years?

Seems pretty normal. In the same period AMD went from 4b to 140b


This is pretty normal stuff.

 

[Lakados]

18b in growth over 7 years?

Seems pretty normal. In the same period AMD went from 4b to 140b


This is pretty normal stuff.

Well “the experts” were expecting up to $80B… Market settled between 8-10… so feels bad to be bad I suppose. So more of. $36B to $10B thing…

 

[toast0]

Well hot fuck!
Not even a month ago all the big firms were all $30-70B and the market settled at 8-10…
Jesus that has got to be a feels bad moment for somebody there.

Selling the whole shebang for $30-70B is different than IPO to raise $10B --- that $30-70B is what goes to Softbank, and this $10B is what stays with the newly independent Arm. Softbank would still sell its shares for something (probably not all the shares at the IPO).

 

[erek]

Well hot fuck!
Not even a month ago all the big firms were all $30-70B and the market settled at 8-10…
Jesus that has got to be a feels bad moment for somebody there.

:feels_bad: for NVIDIA?

(Nvidia) has terminated its proposed acquisition of Arm Ltd. (Arm) from SoftBank Group Corp., FTC Bureau of Competition Director Holly Vedova issued this statement: “More than two months into its litigation with the FTC, Nvidia this week announced that it would abandon its acquisition of Arm.Feb 14, 2022
https://www.ftc.gov/news-events/new...on-nvidia-corps-attempted-acquisition-arm-ltd

 

[staknhalo]

:feels_bad: for NVIDIA?

(Nvidia) has terminated its proposed acquisition of Arm Ltd. (Arm) from SoftBank Group Corp., FTC Bureau of Competition Director Holly Vedova issued this statement: “More than two months into its litigation with the FTC, Nvidia this week announced that it would abandon its acquisition of Arm.Feb 14, 2022
https://www.ftc.gov/news-events/new...on-nvidia-corps-attempted-acquisition-arm-ltd

Yes in that Nvidia even realizes ARM would be valued more now if they were a Nvidia company (it would just be rolled into the price of Nvidia)

Sad indeed

 

[david123]

But no regulators were afraid they would raise prices, change the licensing model, and remove the ability for vendors to customize the designs…

So instead ARM has decided to change from a per chip fee to one based on the selling price of the final product ( how Qualcomm licenses its modems). They are removing the ability to customize the chips and instead licensing full SoC’s, and they are instead issuing those licenses to the device manufacturers so they can potentially cut Broadcom, Qualcomm, Mediatek, and the likes out completely and contract to fabs directly…

ARM is kind of committing suicide. It is a huge risk to shift from a mature and successful business model. But anyway probably it is desperate as Softbank.

 

[Lakados]

ARM is kind of committing suicide. It is a huge risk to shift from a mature and successful business model. But anyway probably it is desperate as Softbank.

What successful business model? they lose money every quarter because their licensing revenue doesn't cover R&D...
Broadcom, Qualcomm, MediaTek, and the rest make more money selling ARM chips than ARM makes licensing chips to them they also spend more in R&D each than ARM makes in revinue. Their business model 100% needs to change and that is why they all fought against Nvidia's purchase because they knew it would have to and it would cost them money.

 

[Red Falcon]

NVIDIA was offering $40 billion USD for ARM, and even when it failed they still had to pay a mandatory $2 billion USD fee to them.
Now the the IPO is going for $10 billion, way to go SoftBank.

 

[Lakados]

NVIDIA was offering $40 billion USD for ARM, and even when it failed they still had to pay a mandatory $2 billion USD fee to them.
Now the the IPO is going for $10 billion, way to go SoftBank.

But they did get a 20-year license from ARM for that $2B so they are going to be exempt from the impending license changes longer than anybody else, Nvidia is safe until 2042 but the rest are all up for renegotiation between 2024 and 2028. So assuming ARM is successful in being spun off and doesn't end up being sold for scrap to the highest bidder then Nvidia and Apple will be the only players able to customize their own designs.

 

[Red Falcon]

But they did get a 20-year license from ARM for that $2B so they are going to be exempt from the impending license changes longer than anybody else, Nvidia is safe until 2042 but the rest are all up for renegotiation between 2024 and 2028. So assuming ARM is successful in being spun off and doesn't end up being sold for scrap to the highest bidder then Nvidia and Apple will be the only players able to customize their own designs.

This is true, at least there was a bit of a silver-lining, and NVIDIA at least made the investment cost worthwhile.

 

[erek]

NVIDIA was offering $40 billion USD for ARM, and even when it failed they still had to pay a mandatory $2 billion USD fee to them.
Now the the IPO is going for $10 billion, way to go SoftBank.

Arm co-founder partly blames ‘Brexit idiocy’ for US flotation​

https://www.theguardian.com/business/2023/may/02/arm-brexit-float-us-hermann-hauser-uk-london

Also remember SoftBanks loss of the WeWork Tech company failed IPO

“How much did SoftBank lose in WeWork?


8, SoftBank reported quarterly earnings, revealing that it had lost roughly $23 billion in just three months. Some of this loss, the largest in company history”

 

[david123]

What successful business model? they lose money every quarter because their licensing revenue doesn't cover R&D...
Broadcom, Qualcomm, MediaTek, and the rest make more money selling ARM chips than ARM makes licensing chips to them they also spend more in R&D each than ARM makes in revinue. Their business model 100% needs to change and that is why they all fought against Nvidia's purchase because they knew it would have to and it would cost them money.

There are many other ways to improve their business model, cut their own cost, or incrementally change things. This abrupt huge change is dangerous for any mature company. For example, the cost for an ARM core in customer's SOC increase by more 10 times now. Such aggressive pricing uprising certainly will drive a lot of legacy customers away. And ARM is also having fight with Qualcomm on legal now. These kind of move makes it more than money alone.

As far as I know, many companies (Apple, NVidia, AMD, Qualcomm) are really actively migrating their embedded core from ARM to RISC-V. This migration has a huge cost for these companies. If ARM does the change incrementally, these companies probably would not do the migration, or at least do it slower to avoid the migration pain. It will give ARM more breath room to face the RISC-V challenge. The embedded core migration will speed up the main CPU facing user application, and the data center/server CPU to migrate. That's how ARM is speeding up their downward spiral.

 

[Lakados]

There are many other ways to improve their business model, cut their own cost, or incrementally change things. This abrupt huge change is dangerous for any mature company. For example, the cost for an ARM core in customer's SOC increase by more 10 times now. Such aggressive pricing uprising certainly will drive a lot of legacy customers away. And ARM is also having fight with Qualcomm on legal now. These kind of move makes it more than money alone.

As far as I know, many companies (Apple, NVidia, AMD, Qualcomm) are really actively migrating their embedded core from ARM to RISC-V. This migration has a huge cost for these companies. If ARM does the change incrementally, these companies probably would not do the migration, or at least do it slower to avoid the migration pain. It will give ARM more breath room to face the RISC-V challenge. The embedded core migration will speed up the main CPU facing user application, and the data center/server CPU to migrate. That's how ARM is speeding up their downward spiral.

The cost for an ARM SoC may have gone up by 10x but ARM signs license deals for 20 years. So arm still only gets something like $0.02 per chip sold regardless of how much it sells for. Because Qualcomm and the likes signed those deals in the early 2000’s and mid to late 90’s.
RISC-V has its place but it is still a decade away from being viable for a mainstay product. Most of the money they are spending on RISC is to replace ARM’s IO controllers for vastly better memory and data speeds. It is also the mainstay for many dedicated encryption chips for RSA and the likes RISC mostly gets incorporated into other architectures as a fixed function bit of silicon not as an actual CPU, the ISA for RISC doesn’t really go general processing very well yet.

Broadcom, Qualcomm, MediaTek, something like 90% of their revenue comes from selling ARM silicon, and each of them makes vastly more money at it than ARM does. SoftBank has tried a number of times to renegotiate and when that failed they said fuck it were selling ARM we all know how that went so this is where things are now.

The companies mostly selling ARM stuff now don’t care what happens to SoftBank ARM because if that goes tits up they can fall back on ARM China and continue doing what they are doing now while lawyers fight it out for a decade or more in court.

 

[erek]

The cost for an ARM SoC may have gone up by 10x but ARM signs license deals for 20 years. So arm still only gets something like $0.02 per chip sold regardless of how much it sells for. Because Qualcomm and the likes signed those deals in the early 2000’s and mid to late 90’s.
RISC-V has its place but it is still a decade away from being viable for a mainstay product. Most of the money they are spending on RISC is to replace ARM’s IO controllers for vastly better memory and data speeds. It is also the mainstay for many dedicated encryption chips for RSA and the likes RISC mostly gets incorporated into other architectures as a fixed function bit of silicon not as an actual CPU, the ISA for RISC doesn’t really go general processing very well yet.

Broadcom, Qualcomm, MediaTek, something like 90% of their revenue comes from selling ARM silicon, and each of them makes vastly more money at it than ARM does. SoftBank has tried a number of times to renegotiate and when that failed they said fuck it were selling ARM we all know how that went so this is where things are now.

The companies mostly selling ARM stuff now don’t care what happens to SoftBank ARM because if that goes tits up they can fall back on ARM China and continue doing what they are doing now while lawyers fight it out for a decade or more in court.

Set to be pitched

https://www.bloomberg.com/news/arti...-ipo-of-chipmaker-arm?leadSource=uverify wall