• Cloud-hosting services suffering from buyers’ remorse.
• Many in the UK return to on-prem.
• Limited use-cases for cloud?

The technology space runs on hype cycles. In 2023, talk of AI superseded the previous year’s metaverse speculation. Before that, we saw blockchain and distributed ledger technologies becoming the industry’s darling, fueling a boom-bust in cryptocurrency speculation that made – and lost – billions of dollars. A decade earlier, cloud computing dominated the technology industry’s press pages.

READ NEXT

It’s another record another year for the cloud infrastructure market

But it’s worth noting that the subjects of the hype du jour are, more often than not, created by vendors keen to massively over-promote their wares. As is the case in almost every commercial offering, the impetus to find something new to sell creates answers to questions that have yet to be asked.

But after the hype has died down, the gray dawn of reality begins to shed light on the decisions made just a few years ago. And in the case of cloud computing (hindsight has reduced the phrase to lower case), the aftertaste of cloud adoption is dominated by the flavor of overspending, with acid notes of unused capability.

UK turns away from cloud-hosting services

According to a survey by Citrix of UK-based IT leaders as reported by InfoWorld, 43% of respondents said moving applications and associated data to the cloud from on-premise was more expensive than they’d thought. Nearly a quarter (24%) said cloud solutions were failing to meet expectations.

The three big promises of cloud computing were agility/scalability, lower cost, and access to cutting-edge innovation. If we translate the ‘lower cost’ into ‘OPEX, not CAPEX’ (shifting figures around a spreadsheet), what’s interesting about the remaining two benefits is that both refer to infrastructure, not what happens on that infrastructure.

Cloud hosting services allow applications to run on systems that can scale according to changing levels of demand. But without significant re-engineering or even rewriting, many applications simply can’t take advantage of the rapidity of scale-up/down on offer. The exception is data storage, which can easily be expanded, given deep enough pockets.

But most applications that have been in reliable production for more than the blink of an eye gain little from being cloud-hosted. The obvious exceptions are applications written using microservices: containers that can be quickly replicated and torn down. ‘Cloud-native application’ has become a differentiating term denoting a code base divided into replicable elements that can be controlled independently.

Access to cutting-edge technologies, cloud’s third great hope, is in no way unique to cloud vendors. There’s a degree of automation in creating new infrastructure, which is certainly made simple. But organizations demand systems without vendor lock-in so that they can, theoretically at least, migrate from cloud to cloud to hunt down the best value for money. Attempts by cloud vendors to make open source technologies unique (MongoDB on AWS is a good case in point) are heading for failure.

READ NEXT

AWS, Google, Microsoft growth slows as cloud spending reduces

And therein is the core lesson about cloud computing. What’s offered is most simply described as ‘someone else’s computer .’ There are layers of usability placed on top, and users can choose from a menu of pre-vetted platforms and technologies that ‘just work.’ And that facility may be useful if, for example, users create new applications based on microservices that operate on data already embedded into cloud-based workflows. But as nearly half of UK IT decision-makers have found, doing so can be more expensive than was predicted.

Why containers, after all?

It’s also worth noting that the industry standard for fleets of containers, Kubernetes, dedicates a great number of lines of code to what should happen when containers crash: keeping systems going when component parts fail. Less fashionable but more mature technologies, like VMs or FreeBSD jails, may not be as glamorous in boardroom discussions but offer a more solid basis on which to build out new features on relatively reliable existing applications.

Companies hoping to sidestep the IT staff shortage may also look at the big cloud vendors for solutions. But the slew of AWS/GCP certifications now available indicate that trained and therefore expensive personnel are a pre-requisite, regardless of where systems are located.

Cloud-hosting services don’t offer their facilities primarily for the good of users. They aim to have as many users signed up as possible, whether or not their clients’ needs are best met by what’s on offer. It’s a good match in some cases, but more companies who rode with the ‘stampede to the cloud’ are regretting being caught up in the melee.

“Outrunning the Wall Cloud” by Wesley Fryer is licensed under CC BY-SA 2.0.

The post Clouds darken in UK as on-prem makes a comeback appeared first on TechHQ.

• Dynatrace can now deploy causal AI to deliver certainty of results.
• This fits a particular niche of need for enterprises that GenAI can’t deliver.
• It’s also delivering a carbon calculator that goes beyond standard, vague models.

READ NEXT

Dynatrace Perform: here’s what you missed

From causal AI to harsh deletion; after a run of exciting announcements at Perform 2024, we spoke to Dynatrace’s CTO and co-founder, Bernd Greifeneder, to get some insight on the technology behind the observability platform.

As the “tech guy,” how do you approach the marketing side of things? How do you get across the importance of Dynatrace to those who don’t “get” the tech?

Right now we are on that journey – actually, this Perform is the first one explicitly messaging to executives. It’s worked out great, I’m getting fantastic feedback. We also ran breakout sessions with Q&A’s on this three by three matrix to drive innovation by topics like business analytics, cloud modernization and user experience.

Then, we have the cost optimization because every executive needs to fund something. I can explain ten ways to reduce tool sprawl alone with Dynatrace. Cloud cost coupled with carbon is obviously a big topic, and the third layer is risk mitigation.

No one can afford an outage, no one can afford a security breach – we help with both.

Bernd Greifeneder presented Dynatrace’s new products on the mainstage at Perform 2024.

Executives have always asked me how to get to the next level of use cases. I think that’s another opportunity; in the past we were mostly focused on middle management. If we first give executives the value proposition, they can go down to the next level of scale, implementing the use cases they wanted.

The other aspect is extending to the left. It’s more than bridging development with middle management, because you can’t leave it just to developers. You still need DevOps and platform engineering to maintain consistency and think about the bigger picture. Otherwise it’s a disaster!

How has changing governance around data sovereignty affected Dynatrace clients – if at all?

[At Perform 2024, Bernd announced Dynatrace OpenPipeline, a single pipeline for petabyte-scale ingestion of data into the Dynatrace platform, fuelling secure and cost-effective analytics, AI, and automation – THQ.]

Well, we have lots of engagements on the data side – governance and privacy. For instance, with OpenPipeline it’s all about privacy because when customers just collect data it’s hard to avoid it being transported.

It’s best not to capture or store it, but in a production environment you have to. We can qualify out the data at our agent level and maintain interest in it throughout the pipeline. We have detection of what is sensitive data to ensure it isn’t stored – when it is, say if analytics require it to be, you have custom account names on the platform.

That means you can inform specific customers when an issue was found and fixed, but still have proper access control.

We also allow harsh deletion; the competition offers soft deletion only. The difference is that although soft deletion marks something as deleted, it’s still actually there.

Dynatrace’s hard deletion enables the highest standard of compliance in data privacy. Obviously, in the bigger scheme of Grail in the platform, we have lots of certifications from HIPAA and others on data governance and data privacy.

[Dynatrace has used AI on its platform for years; this year it’s adding a genAI assistant to the stack and introducing an AI observability platform for their customers – THQ.]

What makes your use of AI different from what’s already out there? How are you working to dispel mistrust?

READ NEXT

What happened when Dynatrace CEO chose to disrupt rather than be disrupted

Would you want to get into an autonomous car run by ChatGPT? Of course not, we don’t trust it. You never know what’s coming – and that’s exactly the issue. That’s why Dynatrace’s Davis hypermodal AI is a combination of predictive, causal and generative AI.

Generative AI is the latest addition to Davis, intended as an assistant for humans, not to drive automation. The issue is the indeterminism of GenAI: you never know what you’ll get, and you can’t repeat the same thing with it over and over. That’s why you can’t automate with it, or at least automate in the sense of driving a car.

What does it mean then for running operations? For a company, isn’t this like driving a car? It can’t go down, it can’t be insecure, it can’t be too risky. This is where causal AI is the exact opposite of nondeterministic, meaning Dynatrace’s Davis causal AI produces the same results over and over, if given the same prompts.

It’s based on actual facts. It’s about causation not correlation, really inferring. In realtime, a graph is created so you can clearly see dependencies.

For example, you can identify the database that had a leak and caused a password to be compromised and know for certain that a problem arose from this – that’s the precision only causal AI provides.

Generative AI might be able to identify a high probability that the database leak caused the issue, but it would also think maybe it came from that other source.

This is also why all the automation that Dynatrace does is based on such high-quality data. The key differentiator is the contextual analytics. We feed this high-quality, contextual data into Davis and causal AI helps drive standard automation so customers can run their production environments in a way that lets them sleep well.

Observability is another way of building that trust – your AI observability platform lets customers see where it’s implemented and where it isn’t working.

Yeah, customers are starting to implement in the hope that generative AI will solve problems for them. With a lot of it, no one really knows how helpful it is. We know from ChatGPT that there is some value there, but you need to observe it because you never know what it’s doing.

Because of its nondeterministic nature, you never know what it’s doing performance wise and cost wise, output wise.

What about the partnership with Lloyds? Where do you see that going?

Especially for Dynatrace, the topic of sustainability and FinOps go hand in hand and continue to rapidly grow. We’ve also implemented sophisticated algorithms to precisely calculate carbon, which is really challenging.

Here’s a story that demonstrates how challenging it is: enterprise companies need to fulfil stewardship requirements. To do so, they might hire another company that’s known in the market to help with carbon calculation.

But the way they do it is to apply a factor to the amount the enterprise spends with AWS or Google Cloud, say, and provide a lump sum of carbon emissions – how can you optimize that?

The result is totally inaccurate, too, because some companies negotiate better deals with hyperscalers; the money spent doesn’t exactly correlate to usage. You need deep observability to know where the key carbon consumption is, whether those areas truly need to be run the way they are.

We apply that to this detailed, very granular information of millions of monitored entities. With Lloyds, for example, optimization allowed a cut of 75 grams of carbon per user transaction, which ultimately adds up to more and more.

Our full coverage of Dynatrace Perform is here, and in the next part of this article, you can read a conversation with Dynatrace VP of marketing Stefan Greifender.

The post A conversation with Dynatrace’s CTO appeared first on TechHQ.

• Raimondo warns against unwanted access for China to US cloud technology to build AI.
• The Secretary of Commerce is acting to block use of US tech for AI by China due to “security concerns.”
• The move, impacting players like Amazon and Microsoft, is anticipated to escalate tech tensions with China.

The long-standing rivalry between the US and China has evolved into many facades over the last decade. The intensifying competition underscores economic supremacy and national security concerns, shaping the dynamics of a burgeoning tech war. Last year, the battleground extended into the development of AI, but this year, the US has indicated the desire to control and dominate local cloud computing services. 

READ NEXT

Will US cloud computing restrictions on China have any impact?

Recent proposals suggest stringent measures to curb China’s access to US cloud computing firms, fueled by concerns over the potential exploitation of American technology for AI advancement. In a recent interview, US Secretary of Commerce Gina Raimondo emphasized the need to prevent non-state actors and China from utilizing American cloud infrastructure to train their AI models.

“We’re beginning the process of requiring US cloud companies to tell us every time a non-US entity uses their cloud to train a large language model,” Raimondo said at an event on January 27. Raimondo, however, did not name any countries or firms about which she was particularly concerned. Still, the maneuver is anticipated to intensify the technological trade war between the US and China, and signify a notable step toward the politicization of cloud provision.

The focal point of this battle lies in recognizing that controlling access to cloud computing is equivalent to safeguarding national interests. Raimondo parallels the control exerted through export restrictions on chips, which are integral to American cloud data centers. As the US strives to maintain technological supremacy, closing avenues for potential malicious activity becomes imperative.

Therefore, the proposal mandates explicitly firms like Amazon and Google to gather, store, and scrutinize customer data, resembling the weight of stringent “know-your-customer” regulations akin to those shaping the financial sector. Conversely, China has been aggressively pursuing AI development, seeking to establish itself as a global leader in the field. 

The US concerns stem from the dual-use nature of AI technologies, which can have both civilian and military applications. The fear is that China’s advancements in AI could potentially be leveraged for strategic military purposes, posing a direct challenge to US national security.

Of AI, cloud computing, and the US-China tech war

China’s Premier Li Qiang (R) speaks with US Commerce Secretary Gina Raimondo during their meeting at the Great Hall of the People in Beijing on August 29, 2023. (Photo by Andy Wong/POOL/AFP).

Although the US broadened chip controls in October, focusing on Chinese firms in 40+ nations, a gap remains. That is why it is paramount for the US to address how Chinese companies can still leverage chip capabilities through the cloud. Cloud technology has become the backbone of modern businesses and governments, making it a critical asset in the ongoing tech war. 

From start to finish, cloud computing is inherently political, Trey Herr, director of cyber statecraft at the Atlantic Council, told Raconteur. He said that its reliance on extensive physical infrastructure tied to specific jurisdictions makes it susceptible to local politics, adding that conversations about cloud security inevitably take on political dimensions.

READ NEXT

Huawei ends DC lobbying: What’s next for US-China relations?

In October 2023, Biden mandated the US Department of Commerce mandate disclosures, aiming to uncover foreign actors deploying AI for cyber-mischief. Now, the Commerce Department, building on stringent semiconductor restrictions for China, is exploring the idea of regulating the cloud through export controls. Raimondo said the concern is that Chinese firms could gain computing power via cloud giants like Amazon, Microsoft, and Google.

“We want to make sure we shut down every avenue that the Chinese could have to get access to our models or to train their models,” she said in an interview with Bloomberg last month. In short, China’s strides in AI and cutting-edge technologies are a paramount worry for the administration. After all, despite Washington’s efforts to curtail China’s progress through chip export restrictions and sanctions on Chinese firms, the nation’s tech giants resiliently achieve substantial breakthroughs, challenging the effectiveness of US constraints.

Nevertheless, regulating such activities in the US is still being determined because cloud services, which do not involve physical goods transfer, fall outside export control domains. Thea Kendler, assistant secretary for export administration, mentioned the potential need for additional authority in this space during discussions with lawmakers last month.

Addressing further loopholes, the Commerce Department also plans to conduct surveys on companies developing large language models for their safety tests, as mentioned by Raimondo on Friday. However, specific details about the survey requests were not disclosed.

What are cloud players saying?

As with previous export controls, US cloud providers fear that limitations on their interactions with international customers, lacking reciprocal measures from allied nations, may put American firms at a disadvantage. However, Raimondo said that comments on the proposed rule are welcome until April 29 as the US seeks input before finalizing the regulation.

What is certain is that the cloud will persist as an arena for trade war extensions and geopolitical maneuvers. Nevertheless, this tech war has broader implications for the global tech ecosystem. It prompts questions about data sovereignty, privacy, and the geopolitical alignment of technological alliances. As the US seeks to tighten its grip on the flow of technology, China is compelled to find alternative routes to sustain its AI ambitions.

The outcome will shape the future trajectory of technological innovation, with ramifications extending far beyond cloud computing and AI development. 

US Commerces Secretary Gina Raimondo confirms the full on economic war on China (and US businesses).

The goal of the US is to slow down China's tech advances. US gonna double down on its sanctions on China pic.twitter.com/7MTyq5IpFO

— Carl Zha (@CarlZha) December 7, 2023

The post Biden weighs blocking China’s access to US cloud tech, fearing AI advancement appeared first on TechHQ.

• The data center will be the first to be operated by Google in the UK.
• Google’s 2022 deal with ENGIE adds 100MW wind energy.
• The aim is for 90% carbon-free UK operations by 2025.

READ NEXT

AI cybersecurity tool in collaboration with Google Cloud

In the ever-evolving landscape of cloud computing, Google Cloud is a formidable player, shaping the global data center market with its leading solutions and heavyweight presence. Google Cloud’s commitment to expanding its global footprint is exemplified by its recent announcement of a US$1 billion investment in a new data center in Waltham Cross, Hertfordshire, UK. 

The move not only underscores the company’s dedication to meeting the needs of its European customer base, but also aligns with the UK government’s vision of fostering technological leadership on the global stage. As it is, one of the critical pillars of Google Cloud’s presence in the UK is its substantial investment in cutting-edge data infrastructure. That said, the upcoming data center would be Google’s first in the country.

Illustration of Google’s new UK data Centre in Waltham Cross, Hertfordshire. Source: Google.

“As more individuals embrace the opportunities of the digital economy and AI-driven technologies enhance productivity, creativity, health, and scientific advancements, investing in the necessary technical infrastructure becomes crucial,” Debbie Weinstein, VP of Google and managing director of Google UK & Ireland, said in a statement last week.

In short, this investment will provide vital computing capacity, supporting AI innovation and ensuring dependable digital services for Google Cloud customers and users in the UK and beyond.

Google already operates data centers in various European locations, including the Netherlands, Denmark, Finland, Belgium, and Ireland, where its European headquarters are situated. The company already has a workforce of over 7,000 people in Britain.

Google Cloud’s impact extends far beyond physical infrastructure, though. The company’s cloud services have become integral to businesses across various sectors in the UK. From startups to enterprises, organizations are using Google Cloud’s scalable and flexible solutions to drive efficiency, enhance collaboration, and accelerate innovation. 

The comprehensive nature of Google Cloud’s offerings, including infrastructure as a service (IaaS), platform as a service (PaaS), and software as a service (SaaS), ensures that it caters to the diverse needs of the UK’s business landscape.

That said, the investment in Google’s Waltham Cross data center is part of the company’s ongoing commitment to the UK. It follows other significant assets, such as the US$1 billion acquisition of a Central Saint Giles office in 2022, a development in King’s Cross, and the launch of the Accessibility Discovery Centre, fostering accessible tech across the UK.

“Looking beyond our office spaces, we’re connecting nations through projects like the Grace Hopper subsea cable, linking the UK with the United States and Spain,” Weinstein noted.

“In 2021, we expanded the Google Digital Garage training program with a new AI-focused curriculum, ensuring more Brits can harness the opportunities presented by this transformative technology,” Weinstein concluded. 

Google is investing US$1 billion in a new UK data center to meet rising service demand, supporting Prime Minister Rishi Sunak’s tech leadership ambitions. Source: Google.

24/7 Carbon-free energy by 2030

Google Cloud’s commitment to sustainability also aligns seamlessly with the UK’s environmental goals. The company has been at the forefront of implementing green practices in its data centers, emphasizing energy efficiency and carbon neutrality. “As a pioneer in computing infrastructure, Google’s data centers are some of the most efficient in the world. We’ve set out our ambitious goal to run all of our data centers and campuses on carbon-free energy (CFE), every hour of every day by 2030,” it said.

READ NEXT

AMD to use Google Cloud for enhancing chip design workloads

This aligns with the UK’s ambitious targets to reduce carbon emissions, creating a synergy beyond technological innovation. Google forged a partnership with ENGIE for offshore wind energy from the Moray West wind farm in Scotland, adding 100 MW to the grid and propelling its UK operations towards 90% carbon-free energy by 2025. 

Beyond that, the tech giant said it is delving into groundbreaking solutions, exploring the potential of harnessing data center heat for off-site recovery and benefiting local communities by sharing warmth with nearby homes and businesses.

The post Google’s first data center in the UK: a billion-dollar tech investment appeared first on TechHQ.

Light has revolutionized global telecommunications, and those fast-propagating signals could benefit AI operations too. Progress in areas such as silicon photonics, which enables mathematical operations to be performed using light, could help to lower the energy consumption of querying large language models (LLMs). In fact, given their capacity to perform matrix vector multiplication – applying model weights to inputs – more efficiently than conventional GPUs, optical neural networks could turn out to be the future of inference.

AI inference at light-speed

Nicholas Harris – CEO of Lightmatter, a firm taking optical computing to market – claimed (in an interview with Tech Youtuber John Koetsier) that its light-based general-purpose AI accelerator can handle around eight times the throughput of a server blade from NVIDIA, but using a fifth of the power.

READ NEXT

Considering the future of computing – chemical AI and beyond

Lightmatter’s design runs on a single color of light, but Harris points out that – in principle – it’s possible to use multiple colors and feed the chips with multiple inputs at the same time, each one encoded with a different portion of the spectrum. In his view, the upper limit could be as high as 64 colors, which would dramatically increase the throughput and energy efficiency achievable using optical neural networks.

Lightmatter is pitching its range of products – the first of which could be based around adding high-speed, low-power optical interconnects to silicon chip designs – at suppliers of data center infrastructure. However, optical neural networks capable of carrying out low-energy inference could benefit portable devices too, where having a tiny power consumption is good news for battery life.

Researchers in China have built an integrated photonic convolution acceleration core for wearable devices, publishing their work in the journal Opto-Electronic Science, to highlight the feasibility of using light rather than electrons to drive AI algorithms. Their prototype chip measures just 2.6 x 2.0 mm and is capable of recognizing 10 different types of hand gestures, which could be used to control wearable electronics remotely.

Output from a prototype photonic chip created by Baiheng Zhao, Junwei Cheng, Bo Wu, Dingshan Gao, Hailong Zhou, and Jianji Dong. Image: Opto-Electronic Science.

Because optical neural networks are fast, they can generate predictions at high speed – for example, to recognize signs or other roadside features. Being able to perform AI model inference using light could benefit advanced driver assistance systems (ADAS), which have to react quickly to avoid hazards. A fast reaction time helps to extend the available vehicle stopping distance.

Things become more interesting still when you picture how optical neural networks work and integrate model weights – for example, those capable of differentiating between a dog and a cat.

Optical neural network could lead to intelligent cameras https://t.co/IHiuICpHnv #optics #Imaging #AI #neuralnetwork pic.twitter.com/1NONOSzRlH

— IS&T (@ImagingOrg) August 27, 2019

Optical neural networks have been considered for some time and the energy demands of LLMs could provide the impetus to bring designs to market.

To build a deep neural network, it’s necessary to feed the input layer with training data and propagate those values forward through the so-called hidden layers. Initially, the model weights are randomized and so the likelihood that the output layer prediction will match the training label is slim.

Backpropagation enables the model weights to be changed based on error functions so that, after a few cycles of running the model forwards and backwards, predictions and labels start to match more frequently.

READ NEXT

Demystifying language models – an explainer on how LLMs work

This kind of adjustment is difficult to perform on an optical circuit and developers get around this by simulating their networks digitally using conventional silicon processors. However, once those weights have been established, and don’t need to be changed, they can be written into the optical chip.

Speaking at a TEDx event – Instagram Filters for Robots & Optical Neural Networks – Julie Chang, then a PhD student at Stanford’s Computational Imaging Lab and now an Imaging and Computer Vision Research Scientist at Apple – points out that each layer within a deep neural network can be likened to a filter.

“Information travels through these connected layers and filters are applied on top of each other until finally the network reaches a decision at a high level,” she told the audience gathered in Boston, US.

During the training process that we touched on above, the many-layered neural network is trying to figure out – by readjusting its weights – which is the optimum filter to create to identify all of the features of interest correctly. Doing so means that irrelevant information will be rejected, while useful signals become amplified relative to the noise.

Conventionally, running this inference process – using a trained model to generate a prediction from an input – involves performing a series of matrix multiplications. And due to the number (in some cases billions) of parameters involved, the energy required to do this can be significant.

However, now imagine being able to do this with light – as if you were placing a filter over a camera lens. You point the camera at the scene and the filter, which has been optically encoded with the series of model weights, recognizes objects within its view. It’s a simplification, but the analogy points to how fast and efficient things can be when you swap electrons for photons.

Could optical neural networks be the future of inference? This author thinks so.

The post Could optical neural networks be the future of inference? appeared first on TechHQ.

• During CEO Jensen Huang’s visit to Malaysia, Nvidia announced a partnership with a local conglomerate to build AI infrastructure that will bring the fastest supercomputers to the SEA nation by the middle of 2024
• The one-hour-long briefing saw Huang share his perspectives on generative AI, AGI, Malaysia’s data center sector, and US-China geopolitical tensions.
• Nvidia plans to set up a base in Vietnam to boost its semiconductor industry, recognizing its significance in the Vietnamese market.

When Jensen Huang established Nvidia three decades ago, his vision was to turn his passion into a thriving industry. He shared a foresight with his team, and they began shaping the future of computing, emphasizing the significance of accelerated, graphics-focused processing. Grounding their argument in the success of video game consoles thirty years ago, Nvidia has since transformed into the leading producer of microprocessors driving the AI revolution today.

READ NEXT

AMD gears up for an epic showdown against Nvidia in AI chips

In 2013, Huang, as CEO and co-founder of Nvidia, made a strategic bet on AI, influenced by promising research from the academic computer science community, charting a course that would define the company’s future. The pivotal moment arrived in mid-2023, when it was disclosed that ChatGPT, the AI chatbot dominating headlines since it went live in late 2022, had been trained on an Nvidia supercomputer. 

Subsequently, on May 25, 2023, with the Nasdaq opening, Nvidia’s market value skyrocketed by around two hundred billion dollars, propelling the chipmaker to breach a market cap of US$1 trillion a few days later. Described as the sole “arms dealer” in the ongoing AI war by a Wall Street analyst, today, Nvidia has become synonymous with AI dominance.

 “We now have AI that can understand data of many kinds: text, images, chemicals, proteins, and even the laws of physics. And it can generate content of many kinds,” Huang told reporters during a media roundtable session in Kuala Lumpur last week.

Huang spent an hour with local reporters on his maiden trip to Malaysia. “I have the opportunity to meet your Prime Minister. And as you know, your Prime Minister is the only prime minister in the world with AI as his name, and I’m in the AI business. How can I refuse to meet the AI Prime Minister?” he said. (The Prime Minister’s name is Anwar Ibrahim – AI if abbreviated).

When asked about the purpose of his trip to Malaysia as part of his Asian tour, Huang said he had taken the opportunity to visit many developers, startups, researchers, and young companies. 

“Second, as you know, Malaysia is an excellent infrastructure hub for Southeast Asia, with really terrific companies like YTL,” he added. As some unconfirmed reports suggested, Huang was asked if Nvidia and YTL are working together. But it wasn’t until after the media briefing that the announcement came that confirmed YTL will advance AI development in Malaysia in collaboration with Nvidia.

Nvidia & YTL

On November 8, 2023, Malaysian infrastructure conglomerate, YTL Power International Berhad (YTL), revealed its partnership with Nvidia to develop AI infrastructure, aiming to introduce the fastest supercomputers to Malaysia by mid-2024. The announcement was made during a meeting on Friday, attended by Huang and Prime Minister, Anwar Ibrahim, along with other key figures. 

YTL is a company with a distinct legacy. “[Since computing infrastructure] requires access to land, facilities, and power, I think YTL could play a great role in that,” Huang told reporters before YTL announced later that day. 

(L-R) Tengku Zafrul Abdul Aziz, Minister of Investment, Trade and Industry (Miti), Jensen
Huang Jen-hsun, NVIDIA Corp CEO and Founder, Anwar Ibrahim, Malaysia PM,
Yeoh Seok Hong, MD, YTL Power International Berhad.

Unlike Singapore, the collaboration will give birth to Malaysia’s first supercomputer by Nvidia YTL, which will deploy Nvidia H100 Tensor Core GPUs, which power today’s most advanced AI data centers, and use Nvidia AI Enterprise software to streamline production AI. “Nvidia AI Enterprise includes Nvidia NeMo, an end-to-end, cloud-native framework for building, customizing, and deploying generative AI models from anywhere,” Nvidia explained.

YTL’s AI infrastructure, situated in the YTL Green Data Center Park in Kulai, Johor, a 500 MW facility powered by on-site solar energy, will be owned and managed by YTL Communications Sdn Bhd. The initial phase is expected to be operational by mid-2024. As a pioneer in mobile networks, YTL Communications, under its “Yes” brand, introduced 4G and 5G services in Malaysia. 

According to YTL, the AI infrastructure will support the development of scientific research applications and propel Malaysia toward becoming an AI-driven nation. “YTL not only offers energy-efficient AI infrastructure but also develops AI applications and services, leveraging Nvidia NeMo to create a customized Malay language foundation model reflecting Malaysia’s diverse heritage,” the company said.

Nvidia CEO sees Malaysia as a data center hub

Clad in his signature look – a black leather jacket, often paired with a black T-shirt and black jeans, which he wears every time he is in the public eye – Huang spent an hour with local reporters during his maiden trip to Malaysia. Photo: Tech Wire Asia/TechHQ

Throughout the roundtable, Huang iterated several times the possibility of Southeast Asia becoming a vital technology hub.

“[The region] is very good at many aspects of the technology supply chain and quite excellent in packaging and assembly as well battery manufacturing in particular,” Huang told reporters.

In Malaysia particularly, the data center infrastructure, “a layer of computing which is one of the most important parts of AI and cloud, is very successful here,” Huang said. “I think we’re going to see Southeast Asia participate across the entire technology stack, and Malaysia play a role in cloud infrastructure,” he added.

Huang also explained how the AI revolution has given birth to a new type of data center: the AI data center.

“This type of data center is unlike the old one; it’s designed to produce while the previous kind was designed to hold data. That’s why it was called the data center. It was designed to hold all of our files and run applications,” he explained. The AI data centers, according to the tech titan, are “really factories.” 

READ NEXT

Nvidia’s Q3 revenue soars thanks to Singapore

“It takes raw data and transforms it into real, valuable data. That’s called artificial intelligence, a transformation that is basically like manufacturing,” Huang emphasized. Therefore, he believes Malaysia’s excellent manufacturing and Nvidia’s data center technology expertise could turn the country into a hub for AI manufacturing.

“China is crucial to us,” Nvidia CEO reiterates

The topic of China and its geopolitical tensions with the US frequently arises wherever Huang is, given that the Eastern powerhouse accounts for approximately 20% of the US chip giant’s global revenue. At the Malaysian roundtable, Huang was asked about the unfortunate dynamics between the nations. “There’s no guarantee of success,” he said, “so we’ll do our best to succeed.”

For context, the AI chip giant has commanded more than a 90% share of China’s US$7 billion AI chip market. The initial impact of the October 2022 US restrictions targeted the company’s H100 and A100 AI chips, seeking to limit sales to China. In a September 2022 filing, Nvidia indicated that the US government had permitted the continued development of the H100 in China. 

But when Chinese companies shifted to using Nvidia’s H800 and A800 chips, made specifically for the Chinese market, they faced renewed restrictions on those sales imposed by the US in October this year. 

Nvidia CEO predicts Vietnam expansion and AGI.

“The US has determined to regulate our technology, the highest end of our technology, and limit its access to China. The regulations specify the maximum performance we can ship to China, and we will follow the regulations precisely, and very explicitly do so,” Huang told reporters, adding that Nvidia is currently in the process of coming up with new chips for China.

What is AGI, and when will it be a reality?

Reporters asked Huang about AGI–artificial general intelligence–which he defined as a piece of software or a computer that can complete tests that reflect basic intelligence that’s “fairly competitive” with that of an average human. “I believe we can achieve this. A general version perhaps in the next five to eight years,” he predicted.

He explained that while machine learning excels in tasks like recognition and perception, it currently lacks the capability for multistep reasoning, a crucial focus for companies and researchers to achieve AGI, according to Huang. 

Nvidia in Vietnam

The CEO of Nvidia concluded his Southeast Asia visit with a stop in Vietnam, where he plans to create a hub to boost the country’s semiconductor industry, marking it as a crucial market. Although it was another of Huang’s inaugural visits, he expressed the company’s commitment to establishing a center in Vietnam, considering it as Nvidia’s home.

“The base will be for attracting talent from around the world to contribute to the development of Vietnam’s semiconductor ecosystem and digitalization,” said the Vietnamese government, citing Huang after he met with Prime Minister Pham Minh Chinh.

Having previously invested US$250 million in Vietnam, Nvidia is scheduled to explore collaboration agreements on semiconductors with Vietnamese tech companies and authorities during a meeting today, as reported by Reuters last week.

The post Nvidia has its eyes set on Southeast Asia, especially Singapore, Malaysia and Vietnam appeared first on TechHQ.

• Singapore contributed around 15%, or US$2.7 billion, to the quarterly revenue of Nvidia.
• Expenditure on Nvidia chips per capita in the quarter for Singapore was US$600, significantly higher than US$60 in the US and approximately US$3 in China.
• Analysts suggest this is because of the city-state’s significant density of data centers and cloud service providers.

In Singapore, the data center sector has served as a cornerstone of technological advancement, propelling the digital ambitions of businesses spanning diverse industries in the last two decades. Renowned for its strategic location, resilient infrastructure, government backing, and commitment to sustainability, Singapore has continued to assert its prominence in the global data center landscape. This significance recently came to the forefront as Nvidia Corp revealed that the city-state alone contributes a substantial 15% to the quarterly revenue of the US chip giant, underscoring Singapore’s pivotal role in the technology sector’s global narrative.

READ NEXT

Nvidia CEO: “decades” before US reaches chip supply chain independence

As per a filing with the US Securities and Exchange Commission, Singapore emerged as a pivotal contributor to the recent financial triumph of the US chip giant. For the quarter ending October, Singapore accounted for a notable US$2.7 billion out of the total US$18 billion revenue, reflecting an extraordinary surge of 404.1% compared to the US$562 million reported in the corresponding quarter of the previous year. This impressive performance outpaced Nvidia’s overall revenue growth, which stood at 205.5% from a year ago.

The growth puts Singapore ahead of every country except the US (35%), Taiwan (24%), and China, including Hong Kong (22%), based on CNBC’s observation. In the third quarter, 80% of Nvidia’s sales, as disclosed in the SEC filing, came from the data center segment. The remaining portion was attributed to gaming, professional visualization, automotive, and other sectors.

“Cloud service providers drove roughly half of data center revenue, while consumer internet companies and enterprises comprised approximately the other half,” said Nvidia in the filing. That said, Singapore had its advantagea, considering it is a global data center hub, hosting significant players such as Amazon Web Services, Microsoft Azure, IBM Softlayer, and Google Cloud. 

What’s more, due to a robust network supported by 24 submarine cables, the country is also the landing site for a dense network of undersea cables, connecting it to other parts of Asia, Europe, Africa, Australia, and the US. A quick check on the Speedtest Global Index by Ookla shows Singapore has the world’s highest median fixed broadband speed.

Even Citi analysts acknowledged in a November 27 report that “Singapore is also a growing area of specialized CSPs standing up data centers in the region. The contrast becomes more pronounced when accounting for Singapore’s size. On a per capita basis, Singapore spent US$600 on Nvidia chips in the quarter, whereas the US spent only US$60 and China spent approximately US$3 per capita.

“That’s the billing location of the customer and not necessarily the point of consumption,” said Srikanth Chandrashekhar on LinkedIn, responding to a post by former Temasek director Sang Shin. Sang Shin had suggested the chips might be bound for data centers in Singapore, which seems reasonable since most Nvidia chips are headed for data centers, and Singapore has many such facilities.

Singing the praises of Singapore. Source: LinkedIn

What’s next for Singapore’s data center sector?

According to an article by ASEAN Briefing, 7% of total electricity consumption in Singapore goes to data centers, and it is projected to reach 12% by 2030. The city-state will likely attract more players in the market, especially after lifting a moratorium on data centers in January 2022. Initially enacted in 2019, this moratorium responded to the considerable energy consumption associated with data centers.

Singapore has rapidly emerged as a prime destination for this pivotal industry due to its technological prowess, regulatory strength, and enticing incentives. Firstly, the Pioneer Certificate Incentive (PC) program encourages companies, including those in the data center sector, to enhance their capabilities and undertake new or expanded activities in Singapore. 

READ NEXT

Nvidia unveils H200, and three more chips tailored for China

The incentive is aimed at companies involved in global or regional headquarters (HQ) activities, managing, coordinating, and controlling business operations for a group of companies. Designed to drive substantial investment contributions and foster advancements in leading industries, the PC aligns with the characteristics and potential of the data center sector. 

The incentive is a win-win situation for both companies and the city-state as to qualify, businesses must introduce advanced technology, skillsets, or know-how, surpassing prevailing standards in Singapore. Additionally, they should engage in pioneering activities that substantially contribute to the economy.

Another allure of incentives includes GST waivers on importing data center equipment and covering servers, networking gear, and cooling systems. Then there’s Singapore’s dedication to sustainability that stands out through initiatives such as the SS 564 Green Data Centers Standard and the Data Center Carbon Footprint Assessment (DC-CFA) program. 

The nation’s commitment to data security and privacy is also reflected in its regulatory framework, notably the Personal Data Protection Act (PDPA) and the Cybersecurity Act, fostering a trustworthy environment for data center operations.

Singapore – making data centers work better.

The post Nvidia’s Q3 revenue soars thanks to Singapore appeared first on TechHQ.

• AWS unveils new features & products at re:Invent 2023.
  
• Includes Trainium2 and Graviton4 AI chips, Amazon Q Chatbot, Titan Image Generator preview.
• Plus, Guardrails for Amazon Bedrock.

In a year marked by infinite discussions on generative AI in tech conferences, AWS re:Invent 2023 seamlessly continues the trend. The conference kicked off on November 27 and has already showcased integrating AI tools and services.

READ NEXT

Generative AI is getting more expensive

Amazon has been actively working to dispel the notion that it lags behind in the AI race. Over the past year, following the release of ChatGPT by OpenAI, major players like Google, Microsoft, and others have raised their stakes, introducing chatbots and making substantial investments in AI development. Selipsky’s keynote this year highlighted AWS’s comprehensive investment in the AI lifecycle, emphasizing high end infrastructure, advanced virtualization, petabyte-scale networking, hyperscale clustering, and tailored tools for model development.

In a keynote lasting nearly 2.5 hours, Selipsky conveyed AWS’s focus on meeting the diverse needs of organizations engaged in building AI models.

Adam Selipsky, AWS CEO, during his keynote address, sharing about innovations in data, infrastructure, & AI/ML. Source: AWS’s livestream.

Trainium2 & Graviton4: AI model training chips

Amazon took the stage at the re:Invent conference to introduce its latest chip generation for model training and inferencing, combatting the growing demand for generative AI on GPUs, with Nvidia’s offerings in short supply.

AWS’s first chip, AWS Trainium2, aims to provide up to four times improved performance and two times better energy efficiency than its predecessor, Trainium, introduced in December 2020. Amazon plans to make it accessible in EC Trn2 instances, organized in clusters of 16 chips. Trainium2 is scalable and can reach deployments of up to 100,000 chips in AWS’ EC2 UltraCluster product.

The company said this provides supercomputer-class performance with up to 65 exaflops of compute power. (“Exaflops” and “teraflops” measure how many floating-point operations a chip can perform per second.) This power will enable AWS customers to train large language models with 300 billion parameters in weeks rather than months.

AWS Graviton4 and AWS Trainium2 processors. Source: AWS

“Trainium2 chips are designed for high-performance training of models with trillions of parameters. This scale enables customers to train large language models with 300 billion parameters in weeks rather than months. The cost-effective Trn2 instances aim to accelerate advances in generative AI by delivering high-scale ML training performance,” Amazon said in a press release. “With each successive generation of chip, AWS delivers better price performance and energy efficiency, giving customers even more options—in addition to chip/instance combinations featuring the latest chips from third parties like AMD, Intel, and NVIDIA—to run virtually any application or workload on Amazon Elastic Compute Cloud (Amazon EC2).”

Amazon did not specify the release date for Trainium2 instances to AWS customers, except for indicating they will be available “sometime next year.” The second chip introduced was Graviton4.

Graviton4 marks the fourth generation delivered in five years and is “the most powerful and energy-efficient chip we have ever built for a broad range of workloads,” David Brown, vice president of Compute and Networking at AWS, said. According to the press release, Graviton4 provides up to 30% better compute performance, 50% more cores, and 75% more memory bandwidth than current generation Graviton3 processors.

AWS & Nvidia deepening ties at re:Invent 2023

AWS CEO Adam Selipsky and Nvidia’s CEO Jensen Huang at the re:Invent 2023.

AWS’s strategy doesn’t solely rely on selling affordable Amazon-branded products; similar to its online retail marketplace, Amazon’s cloud platform will showcase premium products from other vendors including Nvidia.

After Microsoft unveiled its Nvidia H200 GPUs for the Azure cloud, Amazon made parallel announcements at the Reinvent conference. AWS disclosed plans to offer access to Nvidia’s latest H200 AI graphics processing units alongside its own chips, including the new models.

YOU MIGHT LIKE

GENERATIVE AI

How generative AI in Google is powering change in every sector

“AWS and Nvidia have collaborated for over 13 years, beginning with the world’s first GPU cloud instance. Today, we offer the widest range of Nvidia GPU solutions for workloads including graphics, gaming, high-performance computing, machine learning, and now, generative AI,” Selipsky said. “We continue to innovate with Nvidia to make AWS the best place to run GPUs, combining next-gen Nvidia Grace Hopper Superchips with AWS’s EFA powerful networking, EC2 UltraClusters’ hyper-scale clustering, and Nitro’s advanced virtualization capabilities.”

A detailed explanation of AWS and Nvidia’s collaboration can be found in this standalone article.

Amazon Q: a chatbot for businesses

In the competitive landscape of AI assistants, Amazon has entered the fray with its offering, Amazon Q. Developed by the company’s cloud computing division; this workplace-focused chatbot is

distinctively tailored for corporate use, steering clear of consumer applications. “We think Q has the potential to become a work companion for millions and millions of people in their work life,” Selipsky told The New York Times.

Amazon Q is designed to assist employees with their daily tasks, from summarizing strategy documents to handling internal support tickets and addressing queries related to company policies. Positioned in the corporate chatbot arena, it will contend with counterparts like Microsoft’s Copilot, Google’s Duet AI, and OpenAI’s ChatGPT Enterprise.

Titan Image Generator and Guardrails for Bedrock

Aligning itself with the multitude of tech giants and startups that have ventured into this domain, Amazon is introducing an image generator. Unveiled at AWS re:Invent 2023, Amazon said the Titan Image Generator is now in preview on Bedrock for AWS users. As part of the Titan generative AI models, it can generate new images based on text or customize existing ones.

#AWS is putting the “art” back into “artificial intelligence” with the new Amazon Titan Image Generator. #MachineLearning #AI

Learn more. https://t.co/pOzi2TaPPk pic.twitter.com/NvRdCtLuiu

— Amazon Web Services (@awscloud) November 29, 2023

“[You] can use the model to easily swap out an existing [image] background to a background of a rainforest [for example],” Swami Sivasubramanian, VP for data and machine learning services at AWS, said onstage. “[And you] can use the model to seamlessly swap out backgrounds to generate lifestyle images, all while retaining the image’s main subject and creating a few more options.”

AWS also unveiled Guardrails for Amazon Bedrock, enabling consistent implementation of safeguards ensuring user experiences align with company policies. “These guardrails facilitate the definition of denied topics and content filters, removing undesirable content from interactions,” AWS noted in a blog post.

AWS customers can now develop a tailored generative model, designed specifically to perform in their unique domain, with fine-tuning of Command on Amazon Bedrock. Source: AWS

Applied to all large language models in Amazon Bedrock, including fine-tuned models and Agents, guardrails deploy preferences across applications, promoting “safe innovation while managing user experiences.”

The post GenAI unveilings at AWS re:Invent 2023: AI chips to chatbot appeared first on TechHQ.

• While APAC is flourishing, Cloudera sensed more challenges from North Asia, particularly with the geopolitical situation between China and the US.
• TechHQ sat down with Remus Lim, Cloudera’s VP of Asia Pacific and Japan, to discuss AI adoption, hybrid cloud, and the region’s prospects.
• Lim underscored the challenges companies faced in China, emphasizing how geopolitical tensions with the US have led to a governmental shift away from American products.

Since it was founded in 2008, Cloudera has been a dominant force in the Big Data field. Managing 25 exabytes of data today, the company effectively addresses data management challenges in public and private clouds, empowering customers to handle and derive value from their data efficiently. 

READ NEXT

Net Zero – hybridization and hyperscalers

Patrick Moorhead aptly wrote in an article for Forbes that the evolution of Cloudera is evident across its versions, from Cloudera 1.0, dedicated to constructing an open source enterprise data platform, to Cloudera 2.0, which brought together Hortonworks and Cloudera to expedite the journey to a hybrid cloud. Finally, Cloudera 3.0 marked a significant milestone by establishing the first genuine hybrid, multi-cloud data platform.

Today, the platforms provided by Cloudera are more timely than ever, considering how data has been multiplying at an astonishing rate. The surge urgently demands efficient data management across organizations and industries, and Cloudera lets customers manage their proprietary data across private and public cloud environments.

After all, in a post-ChatGPT era, generative AI and large language models (LLMs) are only as good as the data they’ve been trained on. On average today, nine out of ten of the largest global companies in various industries “trust Cloudera with their data assets,” according to CEO Charles Sansbury. That figure resulted in 25 million terabytes of data being managed with Cloudera solutions. 

25 exabytes – that’s playing with th ebig boys. Source: Tech Wire Asia on X

Although the company does not specialize in generative AI today, its Cloudera Data Platform (CDP) and related tools have been crucial in supporting the data infrastructure needed for generative AI applications in enterprises. 

During Cloudera’s recently concluded Evolve conference held in New York, TechHQ  sat downwith Remus Lim, VP of Asia Pacific and Japan, discussing AI adoption among enterprises, the prominence of hybrid cloud, and the region’s prospects in the near future. 

The interview has been edited for length and clarity.

THQ: What’s your role, and how has it been since you assumed it?

I’ve been managing this role for almost two years now. ASEAN is the largest territory across the Asia Pacific (APAC), and there have been various challenges and different contexts from a regional perspective. For context, APAC in the Cloudera context comprises ASEAN, India, China, Korea, Japan, and ANZ (Australia and New Zealand). We separated Indonesia from ASEAN. We have a reasonably big team in Indonesia, as we do quite a bit of business in the country, with the largest telco being our biggest customer. In fact, all the telcos and banks in Indonesia are our clients.

In terms of the last year, there were more challenges from North Asia, particularly with the geopolitical situation between China and the US. The Chinese market, in particular, has been facing a lot of uncertainty because of the geopolitical and political situation and the economic conundrum. On the geopolitical front, the Chinese government is encouraging local enterprises not to use US products. It’s not a law, but it’s recommended. 

Remus Lim, VP Asia Pacific and Japan, Cloudera.

So, we are seeing a few headwinds over there. As for the rest of the region, Japan and Korea are substantial manufacturing countries. We know the sector is in a slump right now, with companies being more conservative and pulling back on spending. On top of that, the strong US currency across our region needs to be helping. In a nutshell, geopolitical and economic situations, as well as currency, are the key issues. 

But having said that, our key customers are still investing and expanding fast despite delayed expansion plans. Hopefully, next year, things will be better.

THQ: Is China a big market for Cloudera? If yes, how are you circumventing the downturn?

China is a big market. If I ranked them, it would be second. We’re very pragmatic. We continue to sell our values and solutions, and at the end of the day, we will convince customers that we are providing them with a platform. We are not building an IP; we are not taking your data. It is a platform that merely helps you to manage the data. 

So, it is challenging. Specific organizations will be okay, but some will not. We have some of the largest organizations in our portfolio, like Shanghai Pudong Development Bank (SPDB), a state-owned bank. Though they are under pressure, they are still investing in us. On the flip side, with telcos like China Mobile, that has been tougher. We do have partnerships with local cloud providers like Alibaba though. 

THQ: Your chief revenue officer, Frank O’Dowd, mentioned that the recently concluded third quarter was APAC’s biggest. Could you elaborate on that?

Source: LinkCXO Global Pvt. Ltd. on X

The success was due to three key regions, out of which two saw a record-breaking quarter: ANZ. India specifically grew by over 400% because we closed a huge deal. But it has not just been for this quarter. This year, India has been doing exceptionally well, and in the last quarter, we closed a deal with PhonePe, an Indian Fintech company. 

One of the critical things PhonePe wanted out of us was the assurance that Cloudera should not go 100% on the cloud, because they also believed in on-premise strategy and wanted to ensure we did, too. That resonated well with us, so it worked well in our favor.

THQ: Does that mean that the momentum for Cloudera still lies within the telco and the financial industry?

As you can imagine, our most significant sector is customer, which has a lot of complex data. That is typically the financial, telco, government, and manufacturing sectors. These days, there are more vested interests in smart manufacturing and autonomous driving, and we are seeing that in China as well. Take SAIC-Volkswagen Automobile, the Sino-German joint venture operated by SAIC and Volkswagen Group, one of China’s oldest automotive joint ventures.

Cloudera is aiding SAIC Volkswagen in constructing a vehicle data monitoring platform to meet regulatory requirements for smart vehicle data in its Internet of Vehicles (IoV) business. Deploying CDP has resulted in substantial performance enhancements, reducing costs and optimizing data operations efficiency. The implementation has led to a 67% reduction in storage space needs, enhancing vehicle monitoring efficiency and optimizing vehicle data management for SVW.

Remus Lim in conversation with THQ. Photo: TechHQ.

THQ: Enterprises are still wary when it comes to deploying generative AI. What are the narratives like on that in the APAC?

OCBC, one of the largest financial institution in Southeast Asia based in Singapore, is one of the early adopters of AI and machine learning. It started five years back, setting up an AI Lab and recruiting the right people. Today, it has around 200 data scientists. The challenge, however, is that not all banks invest that way. So when we take the story other banks in Japan, Korea, and other countries, they’re all trying to do something. 

But the challenge is that they don’t have the skills, so they hit a brick wall before starting. You need to have a vision and build a foundation. So, for customers who are not at that stage, it is difficult for them to get up there. But AI, in general, has been around for a while, including machine learning. The typical use case for banks is fraud detection, so, there’s a maturity curve.

Generative AI, however, is something new, and most organizations will need to figure out how to use it.

THQ: But have enterprises grasped how data should be collected, cleaned, and managed overall for them to move towards utilizing it for generative AI?

Absolutely. I have been in the business intelligence or data analytics area for over ten years, and it was the same conversation back then. It is just now getting bigger because of the size of the data and the volume we hold today. So yeah, you’re right; people are still talking about silos, inconsistency, and the readiness to merge. That challenge is still there, and the ongoing conversation will not change because technology keeps evolving.

That is perhaps why we adopt an open ecosystem to provide solutions – because there is no such thing as “one size fits all .” It is essential for us also to commit ourselves to the open source environment so that all these ecosystems can work together.

THQ: What do you see as the biggest priority for the region, and what sort of customers is Cloudera targeting going forward over the next year? 

For us, it’s about moving up the value chain. When we started, we were on Hadoop, and when Cloudera and Hortonworks merged, we came up with a new product called CDP to re-architect the whole thing, modernizing the entire stack. So it became a single architecture, on-prem or on the cloud, which is more of a cloud-native architecture, providing us the agility and the scalability using separation of compute and storage. So we started that. 

READ NEXT

Generative AI is getting more expensive

In the past three years, we’ve been trying to make sure that our customers move to CDP from legacy, and we have crossed that hurdle. The next thing is, how do I then get our customers to tap onto what we promise, because there is a lot more potential that they can push, like how OCBC had moved from CDH (Cloudera’s 100% open source platform distribution, including Apache Hadoop, built specifically to meet enterprise demands) to CDP. 

Then, we found out that they had a big plan for AI. When they started moving to CDP, we introduced them to data services on CML (CDP Machine Learning), and that’s when the whole thing flew.

They began to see the value of CML, which is data services, and because of the whole ecosystem, they could plug in Hugging Face and LLM models and create the entire ecosystem into the CML as the base. So that makes a lot of value for them. You’d need to have trusted data to have trusted AI. The foundation of that is CDP, which manages the trusted database.

THQ: It is a very competitive landscape today. How does Cloudera sets itself apart from the rest of the lot, including hyperscalers

We are in a competitive environment; we cooperate and compete. There are undoubtedly many rivals, and once we realize that, we will be able to understand how to operate and go into the market. Our key differentiator is the fact that Cloudera is genuinely hybrid. 

If you look at all our competitors, everyone is eager to jump on top, and their strategy is 100% cloud, whereas we are hybrid. So we are in a unique position where we provide all rights: on-premise and cloud. Not just cloud; we even support multi-clouds like AWS, Azure, and others. So we don’t get customers locked in. From that perspective, we are in a very unique position.

THQ: What are Cloudera’s plans for the region?

We provide the capability to our customers to build because we’re still a platform. We are also building those capabilities into our platform. Our challenge was trying to move our legacy into the new platform. Now that has happened, how can we propel? So there is that.

We are building a natural language LLM where developers can use natural language to submit queries. So that is in the works and should be out soon. That will enhance our product with generative AI capabilities, and that helps to improve the operational efficiency for customers.

The post Cloudera: APAC shines, and it’s not all gloom and doom in China appeared first on TechHQ.

• An immutable desktop operating system can help businesses scale without fear of overwriting core data.
• The same immutable desktop principle can even be useful in data center scenarios.
• Transactional updates and declarative descriptions help ensure that an immutable desktop operating system can keep running smoothly.

Deploying and managing computers at scale is a challenging task faced by several IT functions in organizations.

In addition to the physically apparent machines on every desktop, the need for duplicate machines (real or virtual) is increasingly common as demand for computing power grows. Clusters of less powerful computers, linked by software or hardware layers, often take high-burden roles, a method that’s both scalable (simply add or subtract more resources as required), more economical (it’s cheaper to buy 1000 standard servers than a supercomputer), and more resilient (dying machines can be swapped out from a cluster without service interruption).

The immutable desktop and parallel changes.

READ NEXT

Behind Red Hat Enterprise Linux’s dropping of LibreOffice

A constant factor in all these situations is one of reproduce-ability: hardware and software on a computing instance should be identical to its peers when it’s commissioned and brought online, and all systems have to be upgraded, optimized, and changed in parallel.

The obvious analogy is a room full of desktop computers running, say, Windows 10. If 20 of the total 100 are upgraded by their daily users to Windows 11, the local helpdesk now has to support two operating systems, each with its own foibles, bugs, and issues. Throw in the need for users to use software specific to their role or use their own software, and it’s apparent that in no time at all, every deployed machine would become its own entity, at variance with the others.

So. Many. Linked. Computers. They have to work perfectly in parallel, or what you have is a chaos engine. (Image generated by AI).

That situation is burdensome for users (who can’t hot-desk from machine to machine, for example) and support and maintenance staff, who are at that point essentially maintaining 100 discrete systems. That’s why you need an immutable desktop.

In data centers or clouds, the same problems exist, often at a larger scale, although, thankfully, rarely involving much to do with Microsoft or Windows.

Fleets of servers running specific software need to be identical if they share the burden of running, for instance, a mission-critical application. For that reason, workloads are often run on virtualized machines (VMs) or, increasingly, containers, where individual resources can be created and destroyed by software as needed.

Replication in virtualized or containerized environments is relatively simple, but other issues exist around cybersecurity, skill sets required by systems administrators, resourcing, and vendor lock-in.

There are several technologies designed specifically for infrastructure deployment. Puppet, Chef, Ansible, and Salt are declarative tools with which specialists can create and attenuate the infrastructure that runs applications and services.

Need an immutable desktop?

The complexities of most applications and their required infrastructure mean that these tools can be as complex as many programming languages, with elements like loops, conditions, branches, and subroutines acting as an everyday part of a script or playbook. Accordingly, the personnel skilled in these tools are well-paid: $100k-$200k salaries for a middleweight Ansible engineer, for example, are commonplace.

In general, to reduce the resources required to build, deploy, and maintain software in production, it’s important for there to be as little disparity as possible between every aspect of the systems through which a project moves during its course from the developer’s desktop to the data center.

The immutable desktop – iteration benefits.

Teams of developers working on a project, therefore, should be building their code against identical versions of all the software that’s part of the project.

As the project iterates, moves through to testing and to eventual deployment, version control and homogeneity are hugely important. That continuity significantly lowers the time taken to develop a finished project, but any updates, improvements, or security fixes to an application in production need to be mass-deployed (after appropriate testing).

Iteration – made easier and less likely to go crunch 30 seconds before the final output.

To aid the processes described above, developers, infrastructure engineers and administrators will often talk about immutable operating systems or software, transactional updates, and declarative methods of deployment.

An immutable system is one where software, from the operating system upwards, cannot be changed easily by its users and, therefore, by bad actors. Security is a significant advantage of immutable systems, but with immutability come the difficulties of making necessary changes after the software is instantiated.

READ NEXT

Linux might not be as safe as we thought

Immutable systems have been around for decades and can commonly be found in IoT devices, from machinery controls and environmental sensors to networking equipment. Manufacturers of such systems will often allow an area of their code to be overwritten at a specific level, termed firmware. Upgrading these types of devices, therefore, is often termed flashing the firmware, which refers to electronically rewriting code that’s then held semi-permanently in a device’s long-term storage.

A way of describing less common changes or updates to running systems (and flashing firmware is one such instance) is transactional updates. These occur at specific times or on particular events and, critically, should be designed not to affect running systems. An ideal transactional software upgrade, for example, may occur, and end-users see no noticeable changes.

Transactional software updates are also said to be atomic (encapsulated or discrete) and so can be rolled back if the new version does not work as required.

Immutability can also apply to a computer’s software (as opposed to firmware), be it a server or desktop machine. That means the software can be altered by users but only up to a certain point, one that’s set so as not to affect the smooth running of other software on the machine or other users. On a desktop computer, for example, an immutable operating system grants its users the ability to run their own software, make their own settings, and change many aspects of their environment. Any changes can be rolled back atomically, and no alterations can be made to core systems, so the platform remains viable for others.

To achieve immutability, it’s usually necessary for software to be installed alongside all its dependencies, such as libraries, graphical front end, and so forth. That setup is ideal for software development teams, for example, where homogeneity of the base on which software is built is essential. For example, every developer will build against the same version of, for instance, Java.

The same facility is also highly valuable on servers that host multiple applications. As an example, application A can run using library B version 1.1, while application X can also run using library B, but version 1.2. In a ‘normal’ system, library B could only have a single version – version 1.2 would replace 1.1. Anything depending on the specifics of version 1.1 would error out.

The need for granularity.

To establish this granularity, operating systems and software in general are often created declaratively, using roughly the same methods as the those systems that create infrastructure (Ansible, Puppet, etc.). With a declarative software system, a text file or files contain details of exactly how a system should be configured, who or what its users are, what software (and dependencies of those applications) is to run, and so on. On the invocation of a command or transactional event (after a reboot, for example), the declarative statements are read and acted on, the operating system configures itself, pulls and installs required elements, creates users, deploys applications, and so on.

For mass deployments of software, having a simple, easily-propagated text-based description of the full setup is invaluable. Not only are the files themselves tiny compared to the often huge applications to which they refer (text files are highly portable), but any changes to be reproduced at scale are invoked by simply distributing a new version of the human-readable text file.

READ NEXT

Linux Kernel 5.19s New Features Overshadowed By Rumors of 6.0

Mass deployment technologies have existed for decades but were often comprised of several parts: to take a historical example, a base operating system was created by hand (a so-called gold image) that was built on using applications and dependencies pulled from source and pushed slowly to targets by third-party software. Updates and patches were pushed by the same third-party application, and an agent often had to be pre-installed on each target.

With immutable systems configured declaratively and changed transactionally, the maintainer of systems gets granular control. Individual users and separate applications can run concurrently independently of one another, and the only configuration tool required is a simple text editor.

The post The power of an immutable desktop appeared first on TechHQ.