⚙️ Inside GTC 2025 ⚙️
How NVIDIA Is building the infrastructure of the Future

From March 18–21, NVIDIA hosted its annual GPU Technology Conference (GTC) — a four-day global event focused on the future of artificial intelligence, accelerated computing, robotics, and advanced technologies. Often referred to as the “Super Bowl of AI,” GTC has become the premier stage for unveiling what's next in tech.

On opening day, CEO Jensen Huang introduced the new Blackwell GPU architecture, engineered to power the most advanced AI models. He also presented NVIDIA’s technology roadmap through 2028, featuring upcoming architectures: Blackwell Ultra, Vera Rubin, Vera Rubin Ultra, and Feynman.

Beyond hardware, GTC 2025 made it clear: NVIDIA is building a full-stack AI ecosystem. With over 900 sessions, live demonstrations, and expert panels, the conference provided deep insights into real-world AI applications across industries.

Key announcements included:

• A new AI supercomputer, the GB200 NVL72, designed for large-scale generative AI workloads

• Tools to help enterprises integrate generative AI efficiently and securely

• Strategic partnerships with:

  • Microsoft, AWS, and Google Cloud to bring Blackwell to the cloud

  • SAP, Snowflake, and ServiceNow to embed generative AI into enterprise platforms

  • Amgen and Recursion to accelerate drug discovery using digital biology

  • Mercedes-Benz for next-gen autonomous driving systems

  • Siemens and Foxconn to deploy AI in manufacturing through digital twin technology

In short: GTC 2025 wasn’t just about what’s new — it was about what’s next.

🤖 Figure AI's BotQ Platform 🤖

While many AI systems remain locked inside computers, Figure AI's new BotQ platform is bringing advanced intelligence to humanoid robots capable of physical work. This marks a crucial step in the journey from digital-only AI to systems that can manipulate the physical world with human-like dexterity and problem-solving. The potential impact spans from manufacturing and construction to healthcare and elder care, where robots can handle labor shortages and dangerous tasks. What makes BotQ particularly noteworthy is its focus on transfer learning between virtual and physical environments, allowing robots to apply knowledge gained in simulation to real-world problems with minimal additional training.

🎙️ OpenAI's Advanced Voice Models 🎙️

OpenAI's new voice technology represents a fundamental shift in how we'll interact with AI systems. These models offer customizable speech patterns that sound remarkably human while improving transcription accuracy to near-perfect levels. The implications extend far beyond just better virtual assistants - these tools will enable new forms of content creation, accessibility solutions for vision-impaired users, and more natural human-computer interaction across all digital platforms. For businesses, this opens possibilities for consistent brand voices across all customer touchpoints, while creative professionals gain tools to produce audio content at scale without traditional recording studios.

INTERESTING NEWS

📈 AI task capability doubling every 7 months 📈

This week's research revealing that AI capabilities are doubling every 7 months is perhaps the most important meta-trend to understand in the field. This exponential improvement rate - faster than Moore's Law - means we should expect dramatic capability jumps in shorter timeframes than previously anticipated. For context, this suggests that in just two years, AI systems might be eight times more capable than today's frontier models. This has profound implications for workforce planning, business strategy, and regulatory approaches. Anyone making long-term decisions should consider that capabilities that seem like science fiction today may become available before their current strategic plan concludes.

📰 World's first AI-generated newspaper edition published 📰

An Italian media company has published the world's first entirely AI-generated newspaper edition, marking a significant milestone in content creation. While AI writing tools have been available for years, this represents a leap to production-grade content at the scale of a full publication. The newspaper reportedly maintained its editorial standardswhile dramatically reducing production time and costs. This development signals that content industries - from journalism to marketing - are reaching an inflection point where AI generation becomes viable for mainstream use. For professionals across media, communications, and marketing, this serves as a clear indicator that skill sets must evolve to focus on direction and curation rather than production alone.

H-FARM EVENTS ON AI

Here are a few events we organize in our Campus in H-FARM 
to help you stay ahead of the curve.

CLUB AI (April 1, 6PM)

Discover how AI is actually reshaping business with insights from Diego Pizzocaro (CEO of H-FARM AI), Alessandro Conca (GTM Enterprise Italy of ElevenLabs), and Giovanni Benassi (Data & Performance Lead in Google).

Cisco's Event:
Redefining Customer & Workspaces in the Age of AI

Hosted by Cisco in partnership with H-FARM - we will delve into the transformative impact of artificial intelligence on customer experiences and work environments. You will be able to try out the latest tech released by Cisco “CISCO Spatial Meetings” on Apple Vision Pro.

Lean AI Transformation (April 3)

Sponsored by Fischer Consulting in partnership with H-FARM AI - Join an exclusive networking event for business leaders for a hands-on workshop on AI Agents that aims to provide participants with applicable AI insights and solutions for your organization, aperitif included.