Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/nvidia-to-acquire-arm-for-40-billion.13011/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

NVIDIA to Acquire Arm for $40 Billion!

I think it's one of the better potential combinations out there, although NVidia's competitors are going to have some serious concerns about it. However I think those concerns could be managed. NVidia's CPU business is relatively small, so they could probably divest that to avoid being in a position where they would be in direct competition with their licensees.
 
Save ARM Letter:

 
Last edited:
For someone that has been a fanatic for nvidia no matter what until now why are you so against this sale?
 
https://finance.yahoo.com/amphtml/news/nvidia-analyst-says-chipmaker-could-172509465.html



An overlooked "Plan B' scenario would rely on Nvidia leveraging its current R&D assets and new Arm IP license payment to develop Arm-based servers to address an incremental $20 billion-plus market opportunity that's dominated today by Intel Corporation (NASDAQ: INTC) x86 servers, Arya said.

Arm, which dominates low power/low-performance mobile/IoT markets, has a mere 3% share in the server market compared to 90% for Intel, the analyst said.

Nvidia has the ability to create a new category of Arm-based servers that flexibly spread workload across a pre-possessing layer, Arm-based CPU and acceleration layer, which is Nivida's GPU, he said.

The Plan B envisages a $2-billion down payment and expands Nvidia's total addressable market by at least $20 billion, suggesting a more attractive cost-benefit, Arya said.

The option makes it harder for Nvidia to drive Arm's roadmap and eliminates or minimizes its opportunity to cross-sell its graphics and AI technology to Arm's larger mobile and IoT base, the analyst said.


Nvidia wants and intends to alter ARM’s future road map for itself only.
 
4374914-nvidias-arm-acquisition-good-for-nvidia-bad-for-amd


With the purchase of ARM, Nvidia could easily leverage the ARM license deals in its favor. One method would be by uncoupling the ARM CPU and GPU IPs and instead offering an Nvidia GPU in the aforementioned bundle. This would probably occur at higher (and incrementally increasing) licensing fees. This push for Nvidia GPUs over Mali would shake up the entire mobile chip environment.
Bad for AMD
The first effect would be pushing competitors out of the market. For example, unlike Apple’s (AAPL) recent shift to ARM, which I called a blessing in disguise for AMD, Nvidia’s recent ARM-focused action is bad news for AMD. Nvidia could push AMD’s GPUs out of the mobile market. Samsung’s (OTC:SSNLF) upcoming Exynos SoC GPU is being developed by AMD.


Nvidia is likely to put a stop to the AMD ARM GPU. A few possibilities exist. For example, Nvidia could deny AMD the license. It could make the license deal with AMD unreasonably expensive. It could step in and offer Samsung a better deal. But the main reason I believe the Samsung-AMD deal will end is simply the threat of these aforementioned reasons. I discuss this more below; essentially, AMD cannot responsibly continue its reliance on ARM for its GPU deal with Samsung as that would not only make it reliant on Nvidia’s goodwill but also means that it is funneling its money to its direct competitor, which is a clear opportunity cost.
Nvidia’s ARM purchase could also put Qualcomm (QCOM) and PowerVR in danger. Smartphones SoCs aiming for ARM CPUs will see a heightened incentive to get Nvidia GPUs on the cheap. Moreover, as Nvidia GPUs begin eating up the mobile market share, more companies will be driven to code in CUDA, a free software platform released by Nvidia allowing for coding with Nvidia – and only Nvidia – GPUs.
Controlling ARM License Deals
As mentioned above, Nvidia’s ownership of ARM puts it in control of ARM license deals. Nvidia could simply refuse to provide ARM licenses to companies it deems to be in direct competition. In fact, the mere possibility of this is enough of a threat to force Nvidia competitors to leave certain markets. For a company to be reliant on the goodwill of Nvidia to continue providing ARM licenses is a liability, and for the sake of its shareholders, other options will certainly be considered.
Then, there is the licensing fee issue. SoftBank only barely made back its investment on its purchase of ARM via its licensing fees (ARM’s primary revenue stream), but in the hands of Nvidia, these licensing fees are worth more than their tiny ROIs. Nvidia’s ARM-reliant competitors now must “pay a tithe.” Money moving from one competitor to the other – all things being equal – is a change in the market that clearly benefits Nvidia while hurting Nvidia’s rivals.
AMD’s reliance on ARM technology now puts a sizable portion of its licensing spending into a river flowing directly to its direct competitor in the GPU market, essentially partially funding its opposition. This cash flow is clearly bullish for Nvidia, bearish for AMD, and might even result in AMD abandoning ARM for something like RISC-V, an act that would be bearish in at least the short term for AMD due to the investment required.
Moreover, we could see Nvidia produce ARM-based CPUs on the cheap in the near future, effectively entering the ARM CPU market to compete with AMD (and Intel (NASDAQ:INTC)) on two fronts. Nvidia would not need to pay for ARM licensing, giving it an edge. This could allow Nvidia to enter new markets. One obvious market is the datacenter market, in which it can undercut its competitors, possibly pushing bundled deals with its highest-margin products and even offering its main moneymaker, GPUs. This, again, is bearish for AMD, as AMD has been hitting the datacenter market hard as of late.

Nvidia: Opening New Doors and Trade Implications
The purchase of ARM truly does open new doors for Nvidia in ways it did not for SoftBank.
 
If this acquisition is cleared by regulators forget cutting-edge fabs in the US soil. How it will be possible for intel (while its revenue lovers) for example to spend such a lot of money for R&D and to build a cutting-edge fab?
 
Patrick Little, CEO and President of SiFive Inc. in an interview with Reuters claimed that customer interest in their technology is on the rise (due to NVDA/Arm acquisition).

Patrick was a former Qualcomm Inc. (NASDAQ: QCOM) executive and was declared the new CEO and President of SiFive on Thursday.

SiFive operates in the commercial RISC-V processor business, an open-source chip technology. Reportedly, since his appointment, Little was contacted by many of the company’s customers “asking to work more closely on major designs.”

“Already, six of the top ten semiconductor companies are working with SiFive," Little said. "We just got a big basket of funding, and frankly now I know where all that needs to go."

Since June, the company has raised over $125 million in two funding rounds. In August, SiFive raised million from its Series E funding round led by Korean semiconductor supplier SK Hynix Inc. Whereas in June, it raised $65.4 million from a Series D funding led by existing investors and new investor Qualcomm Ventures LLC.

Why Does It Matter: Arm's acquisition by the American company during the United States-China tensions led to concerns that it could directly impact the semiconductor industry worldwide.

The company's co-founder Hermann Houser expressed his views against the acquisition, dubbing it a “disaster” for all of Europe.

According to CCS Insights Analyst Geoff Blaber, the Nvidia takeover could drive chip manufacturers to explore engagement with other open-sourced technology companies.
 
18 months is long enough for Apple to extend its licensing agreement with “current ARM”; and to modify it to include Arm v9 and beyond. In the meantime, Arm v9 architecture may already be introduced by Arm. I predict that after Arm v9, Apple will go ahead on its own; Apple doesn’t need to worry about compatibility after all.
 
I think Apple will not change its roadmap from Arm to RISC-V after all years of efforts and money it already has poured. I’m sure A17, even A18 has already been on the board.
 
Then what is Apple's SoC plan? What about QCOM and the others?
Apple Bet The Farm On Arm. Now NVIDIA's In Charge, Does It Stick Or Tw…


“However, it’s by no means beyond the realms of possibility that Apple will go it alone with its own chip architecture. When Apple announced it was moving Macs to Arm-based processors earlier this summer, it didn’t mention Arm once in the hour-long presentation. It was simply referred to as “Apple silicon.”
With a perpetual license to use the current Arm cores and a likely multi-year wait to see if Nvidia can get its acquisition past regulators, Apple has time on its side to work on its own architecture.”
 
If Nvidia wanted to make ARM-based CPUs, they could just buy a license for a tiny fraction of $40 billion. The only reason I can think of for this step is that Nvidia wants to be in a position of control, and that would be bad news for the ARM ecosystem, which thrived due to ARM's independence.
 
Back
Top