hyperscale
description: refers to computing environments that can quickly scale in response to demand, commonly used in large data centres
16 results
The Age of Surveillance Capitalism
by Shoshana Zuboff · 15 Jan 2019 · 918pp · 257,605 words
power would not be operational without the gargantuan material infrastructure that surveillance revenues have bought. Google is the pioneer of “hyperscale,” considered to be “the largest computer network on Earth.”23 Hyperscale operations are found in high-volume information businesses such as telecoms and global payments firms, where data centers require millions
…
computational resources. This pattern, in which a small, highly educated workforce leverages the power of a massive capital-intensive infrastructure, is called “hyperscale.” The historical discontinuity of the hyperscale business operation becomes apparent by comparing seven decades of GM employment levels and market capitalization to recent post-IPO data from Google and
…
is fortified by the successes of the German automobile industry in the twenty-first century, where strong labor institutions formally share decision making authority.13 Hyperscale firms have become emblematic of modern digital capitalism, and as capitalist inventions they present significant social and economic challenges, including their impact on employment and
…
wages, industry concentration, and monopoly.14 In 2017, 24 hyperscale firms operated 320 data centers with anywhere between thousands and millions of servers (Google and Facebook were among the largest).15 Not all
…
are surveillance capitalists, however, and our focus here is restricted to the convergence of these two domains. The surveillance capitalists that operate at hyperscale or outsource to hyperscale operations dramatically diminish any reliance on their societies as sources of employees, and the few for whom they do compete, as we have seen
…
: Houghton Mifflin Harcourt, 2013), 9. 22. Hilbert, “Toward a Synthesis of Cognitive Biases.” 23. Paul Borker, “What Is Hyperscale?” Digital Realty, February 2, 2018, https://www.digitalrealty.com/blog/what-is-hyperscale; Paul McNamara, “What Is Hyperscale and Why Is It so Important to Enterprises?” http://cloudblog.ericsson.com/digital-services/what-is
…
-hyperscale-and-why-is-it-so-important-to-enterprises; James Manyika and Michael Chui, “Digital Era Brings Hyperscale Challenges,” Financial Times, August 13, 2014, http://www.ft.com/intl/cms/s/0/f30051b2-1e36-11e4-bb68-00144feabdc0
…
Centers,” Wired, April 5, 2017, https://www.wired.com/2017/04/building-ai-chip-saved-google-building-dozen-new-data-centers. 24. Smaller firms without hyperscale revenues can leverage some of these capabilities with cloud computing services. 25. Catherine Dong, “The Evolution of Machine Learning,” TechCrunch, August 8, 2017, http://social
…
: Social Science Research Network, May 22, 2017), https://papers.ssrn.com/abstract=2971352. See also Michael Chui and James Manyika, “Competition at the Digital Edge: ‘Hyperscale’ Businesses,” McKinsey Quarterly, March 2015. 15. One hundred more data centers are expected to be online by late 2018. Microsoft invested $20 billion in 2017
…
, and in 2018 Facebook announced plans to invest $20 billion in a new hyperscale data center in Atlanta. According to one industry report, hyperscale firms are also building the world’s networks, especially subsea cables, which means that “a large portion of the global internet
…
traffic is now running through private networks owned or operated by hyperscalers.” In 2016 Facebook and Google teamed up to build a new subsea cable between the US and Hong Kong, described as the highest-capacity transpacific
…
route to date. See João Marges Lima, “Hyperscalers Taking Over the World at an Unprecedented Scale,” Data Economy, April 11, 2017, https://data-economy.com/hyperscalers-taking-world-unprecedented-scale; João Marges Lima, “Facebook, Google Partners in 12,800Km Transpacific Cable Linking
…
-transpacific-cable-linking-us-china; João Marges Lima, “Facebook Could Invest up to $20bn in a Single Hyperscale Data Centre Campus,” Data Economy, January 23, 2018, https://data-economy.com/facebook-invest-20bn-single-hyperscale-data-centre-campus. 16. T. H. Breen, The Marketplace of Revolution: How Consumer Politics Shaped American Independence
…
-profit cities,” 228–232; fortification strategies, 122–127, 341–342; founding of, 67; funding Pentland’s research lab, 417; grants to antigovernment groups, 126; and hyperscale operations (material infrastructure), 188–189, 500; and instrumentarianism, 401–402; and intelligence agencies, 115–120; involvement in electoral politics, 122–124; lack of cooperation with
…
, 290–292; why experience is rendered as behavioral data, 233 human frailty, ideology of, 343 Humanyze (formerly Sociometric Solutions), 424–425 Hwang, Jenq-Neng, 206 hyperscale firms, 188, 500, 501 IBM, 210, 211, 217, 276–277, 417 ICREACH, 117–118 idea flow (Pentland), 431, 434–435, 436, 438 identification with surveillance
…
), 502 Marx, Karl, 99, 222, 406, 598n64 Marxism, 222 Mashable, 235 mass production, 29, 31, 63–64, 85–86, 87–88, 347–348 material infrastructure: hyperscale of, 188–189, 500, 501 Mattel, 266–267 Mayer, Jonathan, 168 McCann Erickson, 288 McClain, Linda, 479 McConnell, Mike, 119–120 McDonald’s, 316 McKinsey
Empire of AI: Dreams and Nightmares in Sam Altman's OpenAI
by Karen Hao · 19 May 2025 · 660pp · 179,531 words
of their computing infrastructure into massive warehouses of servers in rural communities. The data center world became divided: There were the hyperscalers and there was everyone else. The four largest hyperscalers—Google, Microsoft, Amazon, Meta—now spend more money building data centers each year than almost all the others, relatively unknown developers
…
like Equinix and Digital Realty, combined. It’s difficult to imagine what a hyperscale data center looks like if you’ve never seen one. Mél Hogan, an associate professor at Queen’s University in Canada who studies AI, infrastructure
…
to write about them roughly a decade ago. “Now football fields don’t even come close to the imaginaries of the required size,” she says. Hyperscalers call their data centers “campuses”—large tracts of land that rival the largest Ivy League universities, with several massive buildings densely packed with racks on
…
percent in 2022; AI computing globally could use more energy than all of India, the world’s third-largest electricity consumer. This scale—the mega-hyperscale—has created startling environmental consequences. And yet, in the very same moment, corporate obfuscation of that impact has reached new heights. Since Emma Strubell’s
…
their model training. “I barely got any answers,” she says. “People were just not even responding or saying that this is confidential information.” Even as hyperscalers have spoken loudly in public about the sustainability of their computing infrastructure, executives at Microsoft have admitted internally that the intermittent availability of renewable energy
…
expected costs to hit $10 billion using Nvidia’s latest B100s, yet another generation after H100s. That was a staggering amount, considering the most expensive hyperscale data centers then hovered around $1 to $2 billion. He didn’t plan to stop there, casually floating the idea of the $100 billion supercomputer
…
different battle is waging in the heart of Chile, over the government’s embrace of the tech industry’s data centers themselves. The faster the hyperscalers have expanded, outpacing the supply of land and power in their typical regions of operation, the more aggressively they have pushed to lay claim to
…
Uruguayan environmental ministry’s website, which lists major industrial projects, he came across the company’s proposal for the data center. Pena had read about hyperscalers using potable water, even during major droughts, and the activism of communities like MOSACAT that had resisted the projects. But when he downloaded the details
…
America, and plans to begin the permitting process for an air-cooled data center in Cerrillos when needed. But that hasn’t slowed down other hyperscalers from entering Latin America. In 2022, Microsoft finalized the location for its data center in Chile, shortly after its second investment into OpenAI—right back
…
IN TEXT “If we are going to develop”: Author interview with Martín Tironi Rodó, June 2024. GO TO NOTE REFERENCE IN TEXT The four largest hyperscalers: Author interviews with Alan Howard, a cloud and data center analyst at the technology consultancy firm Omdia, August and September 2023. GO TO NOTE REFERENCE
…
TO NOTE REFERENCE IN TEXT Now developers use: Rich Miller, “The Gigawatt Data Center Campus Is Coming,” Data Center Frontier, April 29, 2024, datacenterfrontier.com/hyperscale/article/55021675/the-gigawatt-data-center-campus-is-coming. GO TO NOTE REFERENCE IN TEXT A rack of GPUs: Author interviews with Hogan, August 2023
…
the EPA’s estimate before January 2024 that a smartphone charge consumed 0.012 kWh of energy. GO TO NOTE REFERENCE IN TEXT Even as hyperscalers: Transcript of meeting. GO TO NOTE REFERENCE IN TEXT build their campuses in threes: Interview with Alan Howard, August 2023. GO TO NOTE REFERENCE IN
…
. See intelligence human longevity, 186–87 human rights, 19–20, 197, 294 Hurd, Will, 321 Huyen, Chip, 52 Hydrazine Capital, 35–36, 38, 41, 69 hyperscalers, 274–75, 277, 279–80, 285, 294, 296 I IBM, 100, 161 Watson, 99 Imagen, 242 ImageNet, 47, 59–60, 100, 101, 117–18, 259
…
, Bob, 69, 156, 236–37, 244, 373, 404, 405–6 Mechanical Turk (MTurk), 194–95, 202–3 Meena, 153 megacampuses, 275–76, 283–84 mega-hyperscale, 276 Mejias, Ulises A., 104 meritocracy, 36 Messerschmidt, Neily, 334–35 Meta, 51 AI investments, 105 compute, 305 content moderation, 190, 192, 209 data centers
The Dark Cloud: How the Digital World Is Costing the Earth
by Guillaume Pitron · 14 Jun 2023 · 271pp · 79,355 words
Montreal, one of the foremost conferences of the digital industry is taking place: Digital Infra Network Montreal. The event is abuzz with topics such as ‘hyperscale data centres’, ‘digital infrastructure energy efficiency’, ‘long-haul fibre optic networks’, ‘peripherals’, and ‘cloud computing’. Let’s face it: these are opaque concepts to most
…
data centres more or less the same size as Equinix AM4.7 In the midst of this web of steel and concrete thrive over 500 ‘hyperscale’ data centres with surface areas sometimes as big as a football pitch. (See appendix 4.) Paul Benoit describes this concentration of cables, security equipment, IT
…
fact, that ‘One of the industries making the most progress in terms of energy efficiency is data centres.’17 A greater centralisation of data in hyperscale structures that can accommodate thousands of servers is already optimising storage. Engineers in Switzerland and the Netherlands have come up with technology that cools data
…
data-centre experts believe that hosting data so far away from hubs results in a latency that is barely acceptable in the internet ecosystem. Alongside hyperscale infrastructure, the future therefore lies with edge computing — a network of micro data centres spread out in close proximity to users. These ‘short information circuits
…
the previous expedition. ‘The work is very physical’, said du Plessis. Not to mention counterintuitive: what could possibly be the link between super calculators and hyperscale data centres, and a string of sailors handling a miserable grapnel in the middle of the Atlantic? But these seacombers may very well be seen
The Measure of Progress: Counting What Really Matters
by Diane Coyle · 15 Apr 2025 · 321pp · 112,477 words
in 2006 by opening a previously internal function to external customers). It is followed by Microsoft’s Azure and Google (along with AWS known as hyperscalers), then some much smaller providers including Alibaba, IBM, Salesforce, Oracle, and Tencent. A summary by the UN task force looking at updating the SNA in
…
by a country mile, followed by Microsoft’s Azure and (at a distance) Google, with a competitive fringe of smaller players trailing far behind these hyperscalers. AWS leads everywhere, as the cloud emerged from its decision in 2006 to offer an internal functionality to outside customers. The report found that the
…
UK market is seeing revenues grow by 35–40 per cent a year. Each of the hyperscalers is building its own ecosystem, from the basic networks up to the software layer, and that competition focuses on customer acquisition. For like Hotel California
…
other IT infrastructure provisioning s ervices”). A more detailed breakdown would of course be desirable but would perhaps require different reporting by the hyperscalers, including revenues by country; Amazon’s published results do not contain statistically useful detail, and it is unclear how the company reports to each
…
country. Purchase, production, and consumption may occur in different places. In any case, what should be m easured? Monetary payments between customers and hyperscaler, if the latter would provide the data, of course. But how should the statistics track the volume flow of storage and software services that are
…
Supplier controls the infrastructure; customer controls everything else figure 6.4. The cloud computing stack. Source: Ofcom (2023a, b). exercise is not straightforward as the hyperscalers offer a huge variety of s ervices, frequently introducing new options that improve quality, and do not provide information on the extent of purchases of
…
a privacy imperative. In a 2022 report, McKinsey estimated that three-quarters of countries had such a rule (Parekh et al.). For large economies, the hyperscalers w ill have 170 Chapter Six multiple data centres inside the relevant political boundary, so this kind of rule is largely unproblematic—there will be
…
routers, but some also buy servers or alternatively computation services from cloud providers who themselves buy servers or—more likely in the case of the hyperscalers—the components to build their own servers. Some of t hese w ill be imported, for example, the chips from Taiwan or South K orea
…
, 63; estimating stock of, 227; including health status in, 215 Human Development Index (HDI), 236, 243 human footprint, 205, 206 hybrid work, 120–21, 163 hyperscalers, 89, 167–69, 169–70, 196 ICT (digital information and communication technologies): enabling remote work, 163; factoryless goods production and, 79; firm-level productivity and
Architecting Modern Data Platforms: A Guide to Enterprise Hadoop at Scale
by Jan Kunigk, Ian Buss, Paul Wilkinson and Lars George · 8 Jan 2019 · 1,409pp · 205,237 words
—or at least tightly couple compute and storage—or spend millions on your storage network infrastructure and operations, which typically only makes sense in the hyperscale context of public clouds. Everything Is Java Well, first of all, it’s not. Storage systems in the big data open source ecosystem today include
…
between your quorum services are not detrimental to HDFS performance under stress scenarios. Finally, quorum spanning with three datacenters is a very relevant case in hyperscale public cloud services. As you’ll see in Chapter 16, public cloud providers offer extremely capable network infrastructure between distinct datacenters in a given service
…
I store all my data in the cloud and process it efficiently? The answer is yes. Public cloud providers operate at such scale (often called hyperscale), that Hadoop environments and their high demand for I/O throughput can be accommodated at reasonable prices. In the meantime, Hadoop distributors have also acted
…
you run Hadoop in the public cloud, the very same technological concepts and their imperfections are at work under the hood, but the provider’s hyperscale infrastructure allows you to make yourself less exposed to them. If you are completely familiar with the underlying concepts and would prefer to jump straight
…
strong anti-affinity is often not possible. As we see in Chapter 16, there is limited support for this with public cloud providers. But the hyperscale characteristics in public clouds typically offer acceptable durability characteristics and give us ways to work around this, via backups. If you are implementing a private
…
to understand what is going on when the discussion turns to general aspects of service isolation and security in the cloud provider’s backbone. Although hyperscale public clouds depend on SDN through and through, and its benefits are obvious and enticing for on-premises environments as well, current implementations often prove
…
we don’t exclude ourselves from that group) have a desire to build their own great and elegant solutions. As we will see, the large hyperscale clouds by Amazon, Microsoft, and Google are elegant in many respects, but much of their greatness is achieved simply by economies of scale. Instead of
…
public cloud offering as alternative to cold standby clusters, Restore public cloud solutions, Solutions in the Public Cloud-Summary quorum spanning with three datacenters in hyperscale public clouds, Quorum spanning with three datacenters security in, Security in the Cloud-Summary Cloud Foundry, VMware and Pivotal Cloud Foundry cloud providersauditing capabilities, Auditing
The Thinking Machine: Jensen Huang, Nvidia, and the World's Most Coveted Microchip
by Stephen Witt · 8 Apr 2025 · 260pp · 82,629 words
Loop 9. CUDA 10. Resonance 11. AlexNet Part II 12. O.I.A.L.O. 13. Superintelligence 14. The Good Year 15. The Transformer 16. Hyperscale 17. Money 18. Spaceships 19. Power 20. The Most Important Stock on Earth 21. Jensen 22. The Fear 23. The Thinking Machine Acknowledgments About the
…
Truck was being replicated at other major tech firms, including Amazon, Oracle, and Microsoft. The business plan for these “cloud-service providers” was to build hyperscale data centers running tens or even hundreds of thousands of GPUs, then lease them to corporate customers. Cloud providers sold computing power in the manner
…
-concept, then the transformer was the jet engine. Shazeer and Uszkoreit, working together on a whiteboard, ensured everything about the transformer was built to accommodate hyperscale architecture, with massive amounts of data, massive numbers of parameters, and massive GPU clusters. As the project built momentum, the work grew frenzied; following one
…
there, an integrated, warehouse-sized solution that Huang no longer referred to as a supercomputer or a data center, but as an “AI factory.” Sixteen Hyperscale Huang’s only interest now was scale. He did not see AI as an emergent machine superintelligence and was dismissive of direct analogies to biology
…
charging $20 per month to access GPT-4. By March 2023, the product was approaching two million subscribers. The synthesis of the transformer architecture with hyperscale parallel computing resulted in a Cambrian explosion of AI services. Microsoft built Copilot, an autocomplete tool for computer code that programmers found indispensable. (So successful
…
superpods were wired with thick bundles of cable and fitted with advanced air- and liquid-cooling systems to keep from overheating. The biggest of the hyperscale data centers measured their annual power requirements in gigawatts—more than the output of a nuclear reactor, enough to power Minneapolis. Continuously upgrading Nvidia chips
Gambling Man
by Lionel Barber · 3 Oct 2024 · 424pp · 123,730 words
. Masa knows he has ground to make up, fast. The balance of power in the technology arms race rests not with SoftBank but with the ‘hyperscalers’ like Amazon, Microsoft, Google and Meta, who are deep into AI. These are the new East India Companies, businesses with global reach, vast stores of
…
Masa cannot touch the likes of Amazon, Google and Microsoft, he is nowhere near out of options. He is a customer and supplier to the hyperscalers. His new chip-making venture involves billions of dollars of investment, drawing in Middle East money and partners from all over the world as well
The Docker Book
by James Turnbull · 13 Jul 2014 · 265pp · 60,880 words
the potential exposures of the hypervisor layer itself. Despite these limitations, containers have been deployed in a variety of use cases. They are popular for hyperscale deployments of multi-tenant services, for lightweight sandboxing, and, despite concerns about their security, as process isolation environments. Indeed, one of the more common examples
…
-alone sandbox environments for developing, testing, and teaching technologies, such as the Unix shell or a programming language. Software as a Service applications; Highly performant, hyperscale deployments of hosts. You can see a list of some of the early projects built on and around the Docker ecosystem in the blog post
The Network Imperative: How to Survive and Grow in the Age of Digital Business Models
by Barry Libert and Megan Beck · 6 Jun 2016 · 285pp · 58,517 words
, based on intangibles and networks or risk falling behind. The Network Imperative provides the why and how to survive and thrive in the age of hyperscale digital networks. It defines the ten principles for network organizations and a five-step process for pivoting your organization toward today’s most valuable and
…
help those firms that are not digital start-ups or technology superstars bridge the gap and create unprecedented growth and value in the age of hyperscale digital networks. You may feel that network disruption is a distant concern for your business or irrelevant for your industry, and that you have more
Superminds: The Surprising Power of People and Computers Thinking Together
by Thomas W. Malone · 14 May 2018 · 344pp · 104,077 words
Malone is back to remind us that the real impact of technology will come not only from AI but also from harnessing human minds at hyperscale. In this terrific, well-researched, and highly readable book, he explores provocatively and practically the opportunities and challenges that superminds will help us address in
Building Secure and Reliable Systems: Best Practices for Designing, Implementing, and Maintaining Systems
by Heather Adkins, Betsy Beyer, Paul Blankinship, Ana Oprea, Piotr Lewandowski and Adam Stubblefield · 29 Mar 2020 · 1,380pp · 190,710 words
The Fourth Industrial Revolution
by Klaus Schwab · 11 Jan 2016 · 179pp · 43,441 words
Internet for the People: The Fight for Our Digital Future
by Ben Tarnoff · 13 Jun 2022 · 234pp · 67,589 words
Vassal State
by Angus Hanton · 25 Mar 2024 · 277pp · 81,718 words
A New History of the Future in 100 Objects: A Fiction
by Adrian Hon · 5 Oct 2020 · 340pp · 101,675 words
Fancy Bear Goes Phishing: The Dark History of the Information Age, in Five Extraordinary Hacks
by Scott J. Shapiro · 523pp · 154,042 words