SoftBank’s proposed sale of Arm to NVIDIA should be blocked. While there are major UK national interest concerns about this transaction, the threat and likelihood of severe anticompetitive effects are global and more fundamental. NVIDIA’s control over Arm will throttle the advance of other Arm licensees in cloud datacentres. NVIDIA will extend its ascendancy there into edge computing—as is essential in emerging 5G networks and applications like Augmented Reality (AR), Virtual Reality (VR) and the Industrial Internet of Things (IIoT)—and stifle the potential for heterogeneous computing—including Central Processing Unit (CPU), Graphics Processing Unit (GPU) and Digital Signal Processor (DSP)1—in the wider ecosystem including smartphones and IoT devices.
A series of revolutions and hegemonies
Various tech firms have risen to dominance, ingraining themselves and taking the vast majority of their market’s profits for many years, while at the same time stifling their rivals’ abilities to innovate and compete. They invariably attempt to retain such positions indefinitely. Some have only been dislodged after decades, if and when major technological changes and shifts in market demand make disruption possible. In other cases, as with Microsoft and Intel a couple of decades ago, and with Facebook and Google right now, antitrust interventions after the fact have tried to limit the harms stemming from this abusive dominance and entrenchment. However, by the time the antitrust interventions have started, it is often too late to undo the damage.
In computing, IBM’s longstanding dominance with centralised mainframes was eroded by the introduction of minicomputers, first from Digital Equipment Corporation in the 1970s, including lower cost technology and with more widespread applications and usage. Introduction of highly decentralised personal computing based on low-cost microprocessors in the 1980s was initially to IBM’s benefit, but its strategic missteps soon enabled the “Wintel” vertical duopoly of Microsoft Windows software and Intel CPU chips to prevail there. In contrast, many Arm licensees have significantly shared in supply of the CPU chips for mobile phones, consumer electronic products and industrial devices. Smartphones have become the predominant personal computing devices in the last five years2. Mass adoption of broadband Internet—on PCs since 2000 and on Apple and Android smartphones with 4G LTE since 2012—has cultivated demand for this decentralised computing. While around 95% of smartphone CPUs are based on Arm’s instruction set architecture and processor designs, Apple has taken the majority of mobile device profits throughout the last decade and with more than 60 percent of these in Q3 2020. This shift to decentralised computing and broadband connectivity has also significantly fostered the rise of cloud computing in centralised datacentres as devices are continuously online and increasingly seek interaction with large troves of continuously updated data including web content and video. Intel was the main beneficiary in chips with its x86 CPUs dominating in the servers there, as well as in PCs for 40 years, but it is increasingly being contested as workloads change to include the graphics and Artificial Intelligence (AI) for which parallel processing by GPUs is best suited.
Consequently, NVIDIA’s fortunes have improved substantially over the last few years. While it initially flourished with its historic pre-eminence in GPU-based 3D graphics acceleration in the large and fast-growing market for high-performance gaming on PCs, it has significantly grown its revenues by adapting its GPU technology architecture for general-purpose parallel processing and scaling it up for use in datacentre servers. By introducing CUDA, NVIDIA’s proprietary programming environment for its GPUs, NVIDIA has grown its developer network in support of its GPU dominance. As more developers use the CUDA platform, graphics and “visualisation” capabilities are being harnessed by NVIDIA’s GPUs in numerous applications: in media and entertainment, engineering and architectural construction, product design and manufacturing, surgeon training, and vaccine development. GPUs are also particularly well suited to AI processing, as used in intelligent agents including Alexa and Siri, which are fuelling rapid growth in cloud datacentre demand.
Recent demand growth for datacentre computing is substantially for NVIDIA’s GPUs—from public cloud services providers such as Amazon Web Services—while demand for CPU market leader Intel’s processors is moderating there and has “collapsed” in enterprise and government datacentres.
The next shift in computing technology supply is underway as the cloud expands from centralised datacentres to the edge, with increasingly parallel processing soon to connect tens of billions of personal and IoT devices. While definitions for “edge” may differ and even encompass terminal devices, my definition—in the context of 5G—is close to the air interface on the network side. Powerful computing with low latency is essential for 5G signal processing and network operation (e.g. with massive MIMO radios and AI-based automated network configuration) and for real-time 5G applications including those in AR, VR and in IoT, such as with AI in Cellular Vehicle to Everything (C-V2x)-based autonomous driving and in factories (i.e. Industry 4.0 and IIoT). According to Intel’s incoming CEO Pat Gelsinger: “5G is going to represent a platform that is redefining edge computing; it will open up smart cities, smart factories, it will displace Wi-Fi. This is a powerful technology. It will also be deployed in private 5G environments as well”.
With rapid growth for GPUs in the datacentre, NVIDIA is also positioned for growth in edge computing as workloads there will also increasingly demand GPU processing. NVIDIA’s competitive challenge is that its GPUs currently need to interoperate with Intel’s x86 CPUs, which still dominate in datacentres, the edge and PCs. Intel has competing offerings and ambitions in GPUs.
Dominant opportunities
NVIDIA is now jockeying for position to dominate the ecosystem for CPUs as well as GPUs though its control of Arm. It is being encouraged by those favouring NVIDIA’s ecosystem dominance and the exceptional profits this will generate for NVIDIA.
With control of Arm’s intellectual property—as a licensing company, that is Arm’s entire being— NVIDIA will make itself the overwhelming beneficiary from Intel’s fall from dominance in datacentres due to the emergence of Arm CPUs in datacentres. While establishing itself as the leading processor supplier to the datacentre—with its Arm-based CPUs, as well as GPUs where NVIDIA already reigns supreme—NVIDIA will also be able to expand this position to the emerging cloud edge, where substantial growth is anticipated. In establishing technological and market dominance across the entire extended cloud, NVIDIA will garner the market power also to subordinate the rest of the broader ecosystem of devices including smartphones and emerging IoT devices. That is where numerous Arm licensees currently compete vigorously, without significant market share concentration and with customers benefiting greatly in many ways including constant innovation, wide choice, and low prices. It is also where the new paradigm of heterogeneous computing— with CPU, GPU and DSP for image, video, graphics and radio wave processing at low power—was developed during the revolutionary growth of smartphones since the introduction of the iPhone in 2007 and Android in 2008. This is where there is now also great potential for Machine Learning (ML) and AI processing, despite this mostly being cloud-based so far. This vibrant ecosystem will be marginalised because NVIDIA will have the incentive and means to dumb-down the workloads allocated to devices, in subservience to where NVIDIA will obtain architectural control and dominance in the enlarged cloud.
NVIDIA will surely also exploit the opportunity to extend its strengthening position into PCs, where it is already the leading supplier of GPUs for gaming. Arm-based CPUs are displacing x86 there with support from Microsoft. Similarly, Apple has switched from Intel to Arm-based CPUs for its Mac computers.
Ain’t nothing goin’ on but the rent
NVIDIA and associated stakeholders are very incentivised for the firm to seek hegemony by extracting economic rents at the expense of ARM licensees. These customers will no longer be treated even-handedly by a hitherto neutral Arm, but relegated by NVIDIA in its pursuit of significant long-term competitive advantage and dominance.
While Arm’s executives seem unperturbed by the prospect of Arm’s “Switzerland” business model being broken, they are highly motivated to complete the acquisition, including $1.5 billion in NVIDIA stock for Arm employees, equivalent to $230,000 each.
The large and rapid rise in NVIDIA’s stock price and market capitalisation around the time the proposed transaction was revealed indicates the stock market’s expectation of upcoming economic rents. With the prospect of these monopoly profits baked into the stock price, it became irresistible for NVIDIA’s management to do whatever is necessary to preserve those gains. These and other NVIDIA staff immediately found themselves with deep-in-the-money stock options, including some who may have to wait several years for those to become vested and thus exercisable to cash in.
While it is generally a target’s stock price that rises on an acquisition announcement—rather than that of the acquirer—NVIDIA’s market capitalisation rose by around $74 billion3, versus only $8.6 billion for Arm from the price SoftBank paid for it in 2016 to what NVIDIA has agreed to pay now4. The increase for NVIDIA would be even bigger, but for the discount in NVIDIA’s stock price reflecting the significant probability and rising expectations that the acquisition will be opposed by competition authorities. For example, Morningstar Equity Research based its November 2020 Fair Value estimate for NVIDIA on a 50 percent probability the deal will not be approved due to “regulatory risk”. Morningstar valued NVIDIA stock conservatively at $340 versus a market price of $537.61 on that date with the possibility of this acquisition at $52 billion in NVIDIA’s Fair Value, and at two times that (i.e. $104 billion) if the deal closes. A lot of economic rents for many years are required to generate that all that uplift in present value — far more than can be envisaged beyond organic growth already envisaged and also reflected in NVIDIA’s lofty stock price and market valuation.
Why now
SoftBank needed to sell off its acquisition of Arm due to financial pressures including poor performance of other investments.
When SoftBank acquired Arm in 2016, SoftBank vowed to let Arm continue operating independently. This ensured Arm’s neutral licensing model was preserved, which is also what its licensees want and expect when they sign up and renew agreements with Arm. This allowed competition and innovation to flourish among Arm licensees who competed with each other, but not with Arm. As a financial investor, SoftBank was good to its word. While ongoing independent operation for Arm with the possibility of an IPO has been touted since 2018, SoftBank could wait no longer to cash in its investment.
When SoftBank came under financial pressure with its first quarterly loss in 14 years—resulting from a $8.9 billion hit at its Vision Fund—which was reported in November 2019, it became eager to liquidate profitable investments to support its poorly performing investment in WeWork. SoftBank peddled Arm around for a trade sale, including to major Arm licensee Apple. By September 2020, NVIDIA had agreed to purchase Arm for $40 billion.
NVIDIA urgently seeks ownership of Arm because Arm would give NVIDIA full control over the only significant industry standard CPU instruction set architecture other than x86. Industry standards—proprietary (e.g. x86) or otherwise (e.g. licensable like Arm ISA)— are vital and valuable because they cultivate large software development ecosystems. The next phase of processor-chip competition requires heterogeneous computing capabilities integrated in the same hardware—right down to the chip level. This will at least include CPU and GPU. Dominance and tying between GPU and CPU will foreclose competition.
NVIDIA aims to dominate cloud computing including central datacentres and extending to the edge with its own combination of these processors. While NVIDIA’s core historic strength and most of its business are in the GPU accelerators used with PCs and datacentre servers, Arm’s leadership is in the CPUs—with high performance and low power consumption—that prevail in mobile devices, and that are in increasing use in datacentres.
Arm’s standalone value is in the intellectual property ownership of its proprietary architecture and instruction set. By contrast, Arm’s value to NVIDIA is in providing an ability to exclude competitors. Arm CPUs attract large ecosystems of developers, including those using iOS, Android and increasingly Windows. Even though Arm is often incorrectly called a chip company or even a chip maker, it has neither a silicon foundry nor does it implement commercial chip designs itself. Instead, it licenses its intellectual property to several hundred firms that implement Arm’s instruction set and reference designs in various commercially manufactured chips.
Preserving the benefits of competition
A key objective of antitrust policy is to preserve the benefits that arise from competition, including better quality through innovation, increasing customer choice of products and producers, and ensuring economically efficient prices. While economic efficiencies may accrue from horizontal or vertical consolidation and increasing scale, it can be anti-competitive if it destroys competition and its benefits are largely hoarded by dominant suppliers rather than passed on to consumers.
In computing, vertically and horizontally integrated suppliers and those with dominant scales have tended to amass the financial benefits of integration for themselves. Economic efficiencies they pass on to customers are greatly diminished by the monopoly rents they extract for a decade or more and by the sclerosis they inflict on ecosystems and competitors until the next major disruption might become possible. For many years, IBM flourished with high profit margins from its highly vertically-integrated supply in mainframes, including chips, finished hardware products and operating software. Customers only began to benefit from new computing paradigms and more suppliers with lower prices when this near monopolist in centralised enterprise data processing was undermined by the advance of departmental and then personal computing. While Intel and Microsoft obtained economic efficiencies for many years, the dominance of their Wintel alliance in PCs and servers limited customer choice and downward pricing pressure, with much of the economic rewards retained rather than passed on to consumers. Intel also obtained economic efficiencies for many years with its vertical integration between chip design and manufacture. Recently, however, the foundry has become something of a competitive impediment as other, fabless, semiconductor companies including NVIDIA and other Arm licensees have exploited TSMC’s leadership with 7nm and even smaller process nodes.
The only way that innovation, choice and efficient pricing can be assured is when there is sufficient horizontal and vertical separation with different technologies and many players significantly in competition, rather than in cahoots. The Wintel alliance is in demise. Microsoft is now very supportive of ending the x86 monopoly and supporting ARM-based CPUs both in the datacentre and in PCs. The dominance of Intel is, at last, also threatened by the rise of parallel computing including that required for AI with GPU processing, and with the rise of edge computing. The latter will be significantly driven by emerging 5G including OpenRAN, whose implementers seeking off-the-shelf “industry standard” computing platforms are amenable to solutions that are not Wintel-based. Arm-based implementations are increasingly attractive in the data centre and edge due to competitive performance, low power usage, and because this alternative industry standard typically costs less than x86. Lower capital costs and power consumption (i.e. also lower operating costs) are well proven in devices where Arm-based chip designs prevail and most of the 150 billion Arm-based chips shipped have been employed.
Why not
The acquisition of Arm by NVIDIA threatens to destroy the currently auspicious state of affairs with cloud computing expanding to the edge, a vibrant market in device technologies and Intel’s processor-chip dominance in decline. The enlarged cloud with heterogeneous computing enables workloads to be distributed to where they can be processed most effectively and efficiently. While NVIDIA argues that acquisition of Arm will help it compete in the datacentre where Intel dominates, NVIDIA is already able to do that with its GPU-based Accelerated Computing Platform, Mellanox Networking acquisition and CPU licensing with Arm. The preferential treatment NVIDIA will obtain through acquisition will impede the ability of other Arm licensees also to be significant competitive constraints against Intel, and against NVIDIA as its muscle in datacentres, the edge, PCs and in the wider device ecosystem increases. Technical and commercial ties between NVIDIA’s GPUs and its privileged CPUs with ownership of Arm will also be against customers’ desire for best of breed capabilities, choice and full competition. As Nvidia’s dominance throughout the datacentre increases, its need to innovate will similarly decrease as it side-lines its competitors.
Just as— even separately-owned—Intel and Microsoft found exclusivity most advantageous, NVIDIA will also find an arrangement like that with Arm most compelling. It will provide competitive strength against x86 dominance. But once entrenched and with other Arm licensees hobbled, NVIDIA will milk the market with excess profits in a horizontal oligopoly with Intel and AMD, and then seek to marginalise or even ultimately exclude those others.
With large up-front and ongoing investments in R&D and developer programs it will be an anathema for NVIDIA to foster market entry and market share fragmentation with other Arm licensees when Arm’s licensing fees are mere titbits in comparison to prospective revenues and profits for NVIDIA in selling its hardware and software. Arm receives licensing royalties of around 2 percent of chip selling prices. That means that NVIDIA will make 50 times more revenue if it sells its own silicon instead.
With such a big difference, if NVIDIA raises Arm’s licensing fees only to mitigate—let alone offset— those relatively large revenues that NVIDIA loses when Arm licensees win product sales in competition with NVIDIA, that could cause antitrust concerns about monopoly pricing. While Arm is effectively a monopoly in device CPUs for smartphones—with its neutral business model—it has never caused such concerns.
Other licensing examples suggest that NVIDIA will forge significant advantages over other Arm licensees through its control of Arm. While Intel has licensed the x86 architecture to AMD, Intel has continued to dominate for decades with around 80 percent market share for the few years until declining into the sixties of percent since 2019. Android is an opensource operating system, but parent Alphabet’s control of the surrounding ecosystem including the Google Play app store has ensured increasing dominance in search, maps and more. While Android app developers are largely complementary to Alphabet, in competition for ecosystem profits, Alphabet wins hands down with an overwhelming share of these.
The proposed acquisition seemingly enables integration of Arm technology into NVIDIA’s designs on a closer and more customized basis. However, over many years, Arm licensees including Apple still, have designed proprietary CPUs using Arm’s instruction set architecture, rather than—among many others— implementing Arm’s processor reference designs. Without ownership of Arm, NVIDIA competes on equal terms with other licensees. It is what NVIDIA will restrict and delay others from accessing with its ownership and control of Arm that should be of concern to antitrust authorities.
Examples in dominant open standards illustrate the importance of preserving neutrality for these and that antitrust scrutiny of firms with major sway over standards development and licensing is warranted. Alphabet’s dominance of the Android ecosystem is an anathema to the OS’s purportedly open source status. While Android has fostered a large, diverse and competitive device market, Google has been fined for using its control of Android to abusively dominate the mobile software ecosystem. Standards setting organizations, including 3GPP, which sets all 4G LTE and 5G cellular standards, are governed and run to ensure no company is able to dominate technology selection and to ensure that standardised technologies are available for licensing to implementers on Fair Reasonable and Non-Discriminatory terms. Nevertheless, since 2014, there have been several investigations, judgments and reversals on alleged anticompetitive licensing practices for 3GPP standards5. There are also concerns that crucial 5G standardisation could be manipulated to favour the interests of particular companies and nations.
While the Arm architecture is proprietary and not legally-bound to FRAND licensing, independence from its licensees and reliance on developing for their needs has preserved it like an open standard available to all on FRAND-like terms. That will be lost with this acquisition. NVIDIA could so easily and subtly degrade or delay others’ access to Arm technology, or NVIDIA could raise costs for these competitors. Even seemingly small privileges in NVIDIA’s access to ARM technologies could result in harmfully-extreme results. That is the nature of competitive advantage in technology markets with network effects and strong “winner takes all” tendencies.
Counter-counter factual
NVIDIA may argue that absent the acquisition of Arm, Intel will continue to dominate in CPUs in the datacentre, edge and in PCs. On the contrary, Intel is already under attack and losing market share to Arm CPU licensees in the datacentre. In competition with NVIDIA and other Arm licensees, public cloud computing “hyperscaler” Amazon Web Services has developed its own Arm-based processors for its datacentres. The parent of hyperscaler Microsoft Azure is also touted to be developing its own Arm-based processors for use in the datacentre and in its Surface tablets to better compete with Apple’s tablets, also including Arm-based processors.
The hyperscalers are also focused on rapid expansion in a market share grab versus their direct competitors, also including Google Cloud and Tencent Cloud, as they all expand into the edge cloud.
Edge computing environments are those outside traditional centralised datacentres, whether those belong to a hyperscaler providing computer services to an organisation, or that organisation owning and operating the facilities to provide those services for itself. Edge computing environments are diverse in their requirements, depending on industry vertical including, for example, smart cities and Industry 4.0 manufacturing, as well as 5G networks. Solutions and suppliers will vary accordingly.
Hyperscalers are cost focused and partner strategically. While some are vertically integrated within the abovementioned self-supply of some hardware and software (e.g. tools for developers) in the datacentre, they will typically partner at the edge. For example, AWS works with network operator Verizon, that has network capillarity, local facilities, engineering staff and an extensive enterprise salesforce.
There are also other contenders, including many with industry vertical strength, who could exploit emerging edge-computing opportunities with Arm-based, rather than x86 computers there. While NVIDIA will need to focus its efforts (i.e. at the datacentre), hundreds of Arm licensees with various different strengths and market targets could exploit numerous market entry opportunities, each with different focussed efforts. For example, while Ericsson already has a tie up with Intel in mobile edge computing, Arm provides an attractive alternative platform that can differentiate the offerings of other OEMs contending in that market. For example, with Arm-based solutions tending to have rather lower power consumption than those based on x86. These other opportunities for Arm through licensees will not be priorities with NVIDIA in control of Arm R&D, as NVIDIA focuses resources on the most lucrative opportunities for itself.
Competition review, industrial policy and national security
While antitrust authorities including the Federal Trade Commission, European Union’s DG Comp and the UK’s Competition and Markets Authority will now assess the harms this transaction will bring, the UK government will also consider national interest issues. Arm’s co-founder Herman Hauser said the sale would be a “disaster”. Concerns also include loss of national control over one of the nation’s most successful technology companies. It has been suggested the US might override the UK’s agreement to export Arm technology to certain nations, when Arm comes under US ownership and as NVIDIA’s US-developed technologies are mixed with Arm’s offerings.
And, of course, the acquisition would be particularly disadvantageous for graphics and AI chip companies like the UK’s Imagination Technologies and Graphcore, as NVIDIA tries to force its products on customers through practices such as tying, while excluding other GPU and CPU suppliers. As Graphcore’s co-founder Nigel Toon said- this transaction is “bad for competition”, “bad for the market overall” and “bad for Britain”.
End notes
1. The spaghetti soup of acronyms in computing from data center to device is more extensive and also includes Network Processor (NPU), Field-Programable Gate Array (FPGA) and Application-Specific Integrated Circuit (ASIC).
2. Among a base or stock of around 18 billion devices used in 2018, 17 percent of these were smartphones and 4 percent of them were tablets, but only 7 percent of them were PCs. Other (non-personal and non-computing) devices included machine-to-machine (23 percent) and non-smartphones (13 percent). Source: Cisco Annual Internet Report 2018-2023, updated March 2020.
3. News reports that NVIDIA was a likely purchaser started emerging in mid-July 2020, when its market capitalization was $254 billion. The corresponding figure on December 16, 2020 was $328 billion. Intel’s market capitalization fell $38 billion in that same period.
4. SoftBank acquired Arm in 2016 for $31.4 billion.
5. The European Commission found that year that both Motorola and Samsung had abused their positions in SEP licensing, even versus mighty Apple. Among many other cases, a California District Court’s ruling on antitrust abuse by Qualcomm in its licensing practices was completely overturned on appeal at the Federal Circuit. However, recent legal developments indicate that how and to whom leading standards technologies are licensed continues to be a contentious matter in various jurisdictions.
