Qualcomm introduces HBC near-memory AI architecture and AI250, AI350 accelerators

Qualcomm has introduced its High Bandwidth Compute (HBC) near-memory AI architecture, along with AI250 and AI350 accelerators, aiming to improve AI workload scalability. The new technology promises 6x higher bandwidth-per-watt compared to HBM and 200x capacity compared to on-chip SRAM.

The company also unveiled a comprehensive data center roadmap, including the Dragonfly C1000 CPU, HBC, and Dragonfly AI300 accelerator, designed to support growing demand for AI and data center workloads.

In other AI developments, a project called 'Plant Talk' uses AI to interpret plant needs through sensors, allowing plants to 'talk' to humans. This system monitors environmental conditions and translates data into human-comprehensible language.

The AI economy presents opportunities for work done in new ways and efficiency gains. Companies can modernize systems, write software, and run operations with AI and agents alongside people.

AI is being tested in various applications, such as the Casper Police Department's use of a platform called TRULEO to review body-worn camera footage and help officers write police reports. Additionally, discussions at the VivaTech conference revolved around AI risks, cybersecurity threats, and the need for AI sovereignty.

An AI model called Claude has made predictions about America's future, including potential sea-level rise and city transformations. Meanwhile, enterprises are generating and shipping AI-generated code but struggling to control and secure it.

The AI market is experiencing volatility, with investors reevaluating their expectations for AI-related gains. There is also a growing backlash against AI, with concerns about its impact on employment, inequality, and transparency.

As AI becomes more prevalent, leaders must establish clear accountability and governance before delegating tasks to AI. This includes deciding who owns the outcome and ensuring responsible AI deployment.

Key Takeaways

• Qualcomm introduces HBC near-memory AI architecture and AI250, AI350 accelerators to improve AI workload scalability. • Qualcomm unveils data center roadmap with Dragonfly C1000 CPU, HBC, and Dragonfly AI300 accelerator. • 'Plant Talk' project uses AI to interpret plant needs through sensors. • AI economy presents opportunities for new work methods and efficiency gains. • Casper Police Department tests AI for reviewing body-worn camera footage. • VivaTech discussions focus on AI risks, cybersecurity, and AI sovereignty. • Claude AI model makes predictions about America's future. • Enterprises struggle to control and secure AI-generated code. • AI market experiences volatility due to reevaluation of expectations. • Growing AI backlash raises concerns about employment, inequality, and transparency.

Qualcomm unveils fast AI architecture

Qualcomm introduces its HBC near-memory AI architecture, AI250 and AI350 accelerators, promising 6x higher bandwidth-per-watt compared to HBM and 200x capacity compared to on-chip SRAM. The technology aims to break the memory wall and enable AI workloads to scale linearly.

Qualcomm data center roadmap

Qualcomm unveils a comprehensive data center roadmap for the agentic AI era, including the Dragonfly C1000 CPU, High Bandwidth Compute (HBC), and Dragonfly AI300 accelerator. These innovations support growing demand for AI and data center workloads.

Plants get voice with AI

A project called 'Plant Talk' uses AI to interpret plant needs through sensors, allowing plants to 'talk' to humans. The system monitors environmental conditions and translates data into human-comprehensible language.

AI economy opportunities

The AI economy presents two great opportunities: work done in entirely new ways and efficiency gains. Companies can modernize systems, write software, and run operations with AI and agents alongside people.

Casper police test AI

The Casper Police Department tests artificial intelligence to review body-worn camera footage and help officers write police reports. The platform, TRULEO, converts audio from body camera recordings into text.

VivaTech AI discussions

VivaTech conference discussions revolve around AI risks, cybersecurity threats, and the need for AI sovereignty. Experts emphasize the importance of responsible AI development and deployment.

AI predictions for 2276

An AI model, Claude, makes predictions about America's future, including a potential sea-level rise and the transformation of cities. The predictions offer a glimpse into a possible future.

AI-generated code concerns

Enterprises are generating and shipping AI-generated code, but struggling to control and secure it. The use of AI coding tools has created new bottlenecks and governance challenges.

AI market volatility

The AI market experiences volatility, with investors reevaluating their expectations for AI-related gains. The sell-off highlights the fragility of risk sentiment in the AI sector.

AI backlash growing

The AI backlash is growing, with concerns about AI's impact on employment, inequality, and transparency. Organizations must develop and deploy AI responsibly.

AI delegation and accountability

Before delegating tasks to AI, leaders must decide who owns the outcome. Establishing clear accountability and governance is crucial for responsible AI deployment.