Connect with us

AI Research

Amazon Web Services builds heat exchanger to cool Nvidia GPUs for AI

Published

on


The letters AI, which stands for “artificial intelligence,” stand at the Amazon Web Services booth at the Hannover Messe industrial trade fair in Hannover, Germany, on March 31, 2025.

Julian Stratenschulte | Picture Alliance | Getty Images

Amazon said Wednesday that its cloud division has developed hardware to cool down next-generation Nvidia graphics processing units that are used for artificial intelligence workloads.

Nvidia’s GPUs, which have powered the generative AI boom, require massive amounts of energy. That means companies using the processors need additional equipment to cool them down.

Amazon considered erecting data centers that could accommodate widespread liquid cooling to make the most of these power-hungry Nvidia GPUs. But that process would have taken too long, and commercially available equipment wouldn’t have worked, Dave Brown, vice president of compute and machine learning services at Amazon Web Services, said in a video posted to YouTube.

“They would take up too much data center floor space or increase water usage substantially,” Brown said. “And while some of these solutions could work for lower volumes at other providers, they simply wouldn’t be enough liquid-cooling capacity to support our scale.”

Rather, Amazon engineers conceived of the In-Row Heat Exchanger, or IRHX, that can be plugged into existing and new data centers. More traditional air cooling was sufficient for previous generations of Nvidia chips.

Customers can now access the AWS service as computing instances that go by the name P6e, Brown wrote in a blog post. The new systems accompany Nvidia’s design for dense computing power. Nvidia’s GB200 NVL72 packs a single rack with 72 Nvidia Blackwell GPUs that are wired together to train and run large AI models.

Computing clusters based on Nvidia’s GB200 NVL72 have previously been available through Microsoft or CoreWeave. AWS is the world’s largest supplier of cloud infrastructure.

Amazon has rolled out its own infrastructure hardware in the past. The company has custom chips for general-purpose computing and for AI, and designed its own storage servers and networking routers. In running homegrown hardware, Amazon depends less on third-party suppliers, which can benefit the company’s bottom line. In the first quarter, AWS delivered the widest operating margin since at least 2014, and the unit is responsible for most of Amazon’s net income.

Microsoft, the second largest cloud provider, has followed Amazon’s lead and made strides in chip development. In 2023, the company designed its own systems called Sidekicks to cool the Maia AI chips it developed.

WATCH: AWS announces latest CPU chip, will deliver record networking speed



Source link

AI Research

Hong Kong start-up IntelliGen AI aims to challenge Google DeepMind in drug discovery

Published

on


IntelliGen AI, an artificial intelligence (AI) start-up founded in Hong Kong, is positioning itself as a competitor to Google DeepMind in the field of drug discovery, as the city increasingly seeks to bolster its AI capabilities.

In an interview with the Post, founder and president Ronald Sun expressed confidence that IntelliGen AI could soon compete globally with Isomorphic Labs, a spin-off of DeepMind, in leveraging AI for drug screening and design.

“For generative science, new breakthroughs and application opportunities are global in nature,” Sun said. “Within 12 to 18 months, we aim to land major, high-value clients on a par with Isomorphic.”

The term “generative science”, although not widely recognised yet, refers to the use of AI to model the natural world and facilitate scientific discovery.

Ronald Sun, founder and president of IntelliGen AI. Photo: Handout

The company’s ambitious plan follows the launch of its IntFold foundational model, which is designed to predict the three-dimensional structures of biomolecules, including proteins. The model’s accuracy levels were comparable to DeepMind’s AlphaFold 3, according to IntelliGen AI.



Source link

Continue Reading

AI Research

Hong Kong start-up IntelliGen AI aims to challenge Google DeepMind in drug discovery

Published

on


IntelliGen AI, an artificial intelligence (AI) start-up founded in Hong Kong, is positioning itself as a competitor to Google DeepMind in the field of drug discovery, as the city increasingly seeks to bolster its AI capabilities.

In an interview with the Post, founder and president Ronald Sun expressed confidence that IntelliGen AI could soon compete globally with Isomorphic Labs, a spin-off of DeepMind, in leveraging AI for drug screening and design.

“For generative science, new breakthroughs and application opportunities are global in nature,” Sun said. “Within 12 to 18 months, we aim to land major, high-value clients on a par with Isomorphic.”

The term “generative science”, although not widely recognised yet, refers to the use of AI to model the natural world and facilitate scientific discovery.

Ronald Sun, founder and president of IntelliGen AI. Photo: Handout

The company’s ambitious plan follows the launch of its IntFold foundational model, which is designed to predict the three-dimensional structures of biomolecules, including proteins. The model’s accuracy levels were comparable to DeepMind’s AlphaFold 3, according to IntelliGen AI.



Source link

Continue Reading

AI Research

Ireckonu’s AI Research Revolutionizes Hospitality with Timely Churn Prevention Strategies

Published

on


Thursday, July 10, 2025

Dr. Rik van Leeuwen, Head of Data Solutions and Customer Success at Ireckonu, has just uncovered the hospitality industry’s first ever methodology for customer churn management.

The historic study, conducted in collaboration with one of the top North American chains, uses artificial intelligence (AI) to determine the very moment the guest is most likely to drift away—and how hotels can intervene to prevent them from doing just that.

The study outcomes confirm that predictive models powered by artificial intelligence are able to effectively calculate the risk of a guest exiting, allowing hotel managers to take action at the point of the moment. The proactive action might involve the issuance of a discount offer, for instance the issuance of the 20% discount email, once the guest has reached 75% churn risk.

These last-moment offers greatly favor the chances of rebooking, hence enhancing the general guest retention figures.

This breakthrough is the result of the culmination of a multi-week research project combining artificial intelligence, machine learning, and advanced data modeling techniques to yield usable insights for hospitality operations. The framework is focused on predicting customer behavior, identifying risk of churn, and providing tailored recommendations for optimizing retention efforts.

The Power of Hospitality Through the Assistance of AI: Evolution from Forecast to Execution

It’s not just about identifying at-risk guests, but more about intervening at the right time. Dr. van Leeuwen identified the importance of not just knowing who is at risk, but knowing the time and how to intervene. “It’s no longer enough to know who’s at risk.

The value is in knowing how and when to react,” added Dr. van Leeuwen. “That’s where hospitality strategy gains the promise of AI.”

The Dr. van Leeuwen system combines the BG/NBD (Beta-Geometric/Negative Binomial Distribution) model for churn probability with reinforcement learning for future engagement.

The BG/NBD model, in general use for subscription and non-subscription companies, anticipates the probability of repeat purchasing by the customer in the future. By including reinforcement learning, the model by Ireckonu doesn’t just anticipate churn by the customer, but determines the best actions to take, allowing for in-the-moment adjustment based on evolving guest behavior.

Unlike traditional “black box” type artificial intelligence systems, in which interpretability is difficult and implementation in routine business environments is complex, Dr. van Leeuwen’s approach emphasizes transparency and flexibility. The model is designed to be a “white-box” system in the sense that managers in the hotels will understand and rely on the system recommendations based on the used data.

Transparency in this context is the driving factor behind adoption in the hospitality industry, where operating decisions have to be efficient and implementable.

From the Lab to the Field: The Practicality of Ireckonu’s Solutions

Ireckonu has already started integrating these learnings into its broader middleware and customer data platform offerings, allowing hotel chains and other hospitality offerings to deploy AI-drived guest retention approaches at the point of operation. The platform integrates seamlessly with existing hotel management systems in operation, allowing businesses to deploy immediate, data-driven action whenever the system recognizes a guest as being at risk.

“We’re not just pushing academic theory” said CEO of Ireckonu, Jan Jaap van Roon. “Rik’s research brings scientific validation to one of the areas where hotels have long underperformed: guest loyalty. That’s not theory—it’s proven, practical insight. And it’s the kind of innovation we promote at Ireckonu.”

The study’s results have universal applicability to the hotel industry and beyond. The company is exploring how to further optimize the model by incorporating additional variables, such as sentiment and dynamically changing the price in response to the customer’s specific churn risk in coming developments of its AI-powered solutions.

Prospects for Future Development and Applications

For the future, Dr. van Leeuwen’s research opens promising opportunities for further refinements to the artificial intelligence model. One potential area in the future where one might realize developments is in the integration of qualitative guest commentary, e.g., customer review sentiment analysis, to the churn forecasting model. By considering not only the quantitative measures but the emotional and experience facets of the guest’s experience, the model would have the potential for even more accuracy in recommendations for retention efforts.

Furthermore, the AI framework developed by the study has the possibility of being extended beyond the hotel industry. Other areas where high frequency, non-contractual customer interactions occur, i.e., retail and services, would be able to utilize corresponding churn prediction models to maximize customer interaction and retention efforts.

Ireckonu’s ongoing investment in research and development is evidence of its dedication to delivering the hospitality business more intelligent, more tailored guest experiences. By utilizing clean, actionable guest data, the company is helping hotels make more effective retention activities and in the end offer the customer service they desire.

Conclusion:

The Future of Hospitality is AI-Driven As the hotel industry becomes increasingly dependent on artificial intelligence and data science to maximize guest retention, Ireckonu’s research sets the standard in churn management. Identifying the exact moment the guest is most likely to disengage, and providing hotels with concrete action steps, Ireckonu is rethinking the way hospitality businesses approach guest loyalty. This breakthrough shines the spotlight not only on the promise of artificial intelligence for the hospitality profession, but also the worth of marrying cutting-edge research with in-the-field application.

The hospitality future will be guided by more informed, data-driven decisions—decisions that maximize the guest experience, enhance guest loyalty, and achieve long-term business success.



Source link

Continue Reading

Trending