Name: AI Infrastructure Silicon Valley - Executive Series
Start: 2024-08-16T16:00:00.000Z
End: 2024-08-16T21:45:00.000Z

Friday, August 16 (UTC)

ON DEMAND

Craig Walker, Dialpad

The event featured Craig Walker, CEO of Dialpad, discussing the impact of AI technology on the market. With real-time AI integration in their cloud-based platform, Dialpad's user experience has been revolutionized. The acquisition of TalkIQ in 2018 allowed for real-time translation and transcription capabilities. The shift towards GenAI has improved Dialpad's ability to offer advanced AI features for contact center and sales teams. Emphasizing the importance of in-house AI capabilities for security, accuracy, and cost-effectiveness, Dialpad plans to develop industry-specific AI models for various business needs. By providing a unified communication solution, Dialpad aims to offer reliable, scalable AI-driven services. Prioritizing reliability, architecture, and future-proofing for AI technologies enables businesses to gain a competitive edge. Challenges in implementing AI models are discussed, especially for larger companies with high traffic and limitations with third-party models. The growing demand for skilled AI engineers, particularly among younger generations, is highlighted in the conversation.

Craig Walker
Founder & CEO Dialpad
ON DEMAND

Vijay Nagarajan, Broadcom

Vijay Nagarajan
VP, Strategy & Marketing Broadcom
ON DEMAND

Sid Sheth, d-Matrix

D-Matrix focuses on AI inference for generative AI, making it affordable and commercially viable. With a platform tailored for transformer workloads and interconnects for scaling AI models, they target data centers of all sizes. Their platform is designed for inference only, accelerating generative AI with transformer architectures. Planning a product launch later this year, D-Matrix aims to help customers deploy AI solutions and see a return on their investments. In the discussion, the importance of inference in utilizing AI effectively is emphasized, along with the potential for fine-tuning AI models based on new learnings. The target customer base includes big data centers, cloud scale data centers, and specialized AI or GPU clouds. The solution involves plugging in a server with accelerator cards to enhance performance within existing setups. The session concludes participation in the AI Silicon Valley Leadership series.

Sid Sheth
CEO d-Matrix
ON DEMAND

Larry Yang, Phononic

Larry Yang
CPO Phononic
ON DEMAND

Gopi Sirineni, Axiado

The CEO of Axiado, Gopi Sirineni, discusses the boom of semiconductors, with a focus on accelerated computing for security. Their solution, TCUs, aim to provide a last line of defense against cyber attacks, especially ransomware attacks. The secret sauce of their technology includes custom AI engines trained to detect various types of attacks. Axiado's hardware solution, powered by TSMC, is designed to be a more efficient and powerful alternative to current security measures. It offers features such as dynamic thermal management and communication between blades in a rack. The company's focus on hardware and system-level security, leveraging AI capabilities, fills a crucial gap in the market. Gopi's advice for entrepreneurs is to focus on optimizing hardware for AI applications and delve into kernel-level coding for long-term success in the evergreen market. Axiado's unique approach to security, combining AI and hardware, addresses the growing need for more robust and comprehensive security solutions in today's data-driven world.
Viewed.

Gopi Sirineni
Founder, President & CEO Axiado
ON DEMAND

Bruno Geoly, Lumindt

Lumindt's CEO, Bruno Geoly, explained the company's innovative solid state hydrogen storage technology, aiming to offer reliable energy storage in a world with increasing demand. They have deployed systems in California and plan to reduce grid reliance. The technology enables economical large-scale energy storage. Lumindt seeks to become a micro-utility, providing green energy and reliability. With $3.1 million raised, they are looking for funding and hiring engineers to develop the technology further with patents in progress.

Bruno Geoly
CEO Lumindt Labs
ON DEMAND

Alexander Gallego, Redpanda Data

The Silicon Valley AI Infrastructure Leaders Program is hosted by TheCUBE. Alex Gallego, founder of Redpanda Data, discusses the shift from batch processing to real-time data processing for large banks and content delivery networks. Redpanda focuses on streaming data and AI integration to enhance user experiences. They prioritize data sovereignty and governance by sending models to data. Alex highlights trade-offs between model size and latency, mentioning open source models like Llama and Mistral. Businesses in finance and healthcare are cautious about sharing data with external models to differentiate between bots and real users in dynamic systems. The complexity of modern data infrastructure is discussed, with emphasis on Apache Iceberg as a unifying format supported by various platforms. Redpanda provides an immutable log for real-time data processing and benefits from offloading data retention to hyper cloud storage. The evolution from Hadoop to modern data solutions is reflected on, focusing on governance and metadata management in data catalogs. The importance of architecture in analytics and platform engineering is emphasized, especially as platform teams make decisions related to gen AI. The shift towards platformization requires a focus on data catalog, open table formats, and intelligent apps, moving away from traditional documentation methods. Ensuring safety in deploying models involves basic security principles and understanding the decision-making process. The importance of explainability in AI models and the concept of a data supply chain for tracking data origin is highlighted. Businesses are advised to approach architecture thoughtfully and securely in the era of platformization and AI adoption.

Alexander Gallego
Founder & CEO Redpanda
ON DEMAND

Hao Zhong, Scaleflux

Hao Zhong
CEO ScaleFlux
ON DEMAND

Chris Walker, Untether AI

Chris Walker
CEO Untether AI
ON DEMAND

Honghao Deng, butlr.io

Honghao Deng
CEO, Co-founder Butlr
ON DEMAND

Raj Yavatkar, Juniper Networks

The Silicon Valley AI Infrastructure Leaders Program featured Raj Yavatkar, CTO of Juniper Networks. Networking is crucial for building large GPU clusters with high performance and low latency. Companies are shifting towards on-prem infrastructure for cost savings and data privacy. Juniper Networks has focused on AI, developing an AI-native networking platform and 800 gig switches for congestion management. Raj stresses the need for efficient leaf-spine and super-spine networks. Challenges in data centers include feeding accelerators, data security, and automated root cause analysis. Application-aware assurance enables monitoring and diagnostics. Juniper Networks leads in innovative networking solutions for AI.

Raj Yavatkar
Chief Technology Officer Juniper Networks
ON DEMAND

Ami Badani, ARM

Ami Badani
CMO Arm
ON DEMAND

Rodrigo Liang, SambaNova Systems

Rodrigo Lang, CEO of SambaNova, discusses the company's focus on building chips to compete with NVIDIA and running large models quickly. SambaNova offers a full stack solution for deploying models without needing to learn CUDA. By using Dataflow chips, the company achieves superior performance with low power consumption. SambaNova's emphasis on accuracy, large models, and privacy is attracting early adopters in government and service providers. The energy-efficient approach to AI workloads is impacting business budgets positively. The company targets mid to advanced-scale customers, like Global 2000 companies, seeking efficient AI solutions.

Rodrigo Liang
CEO SambaNova Systems
ON DEMAND

Charles Giancarlo, Pure Storage

The discussion focused on the transformation in the hardware and systems industry, driven by advancements in compute, networking, and storage technologies, particularly for AI applications. Speed and efficiency in training AI models were emphasized, as well as the need for cloud-like environments in data centers. The evolving landscape of cloud computing was explored, with a focus on the practicality of a hybrid cloud approach for enterprises. Storage for AI applications highlighted the importance of write speed and data accessibility. Pure Storage's commitment to product enhancement without hardware replacement was noted, along with their collaboration with NVIDIA in AI deployments. The role of AI in managing and optimizing storage infrastructure was discussed, as well as the advantages of being an independent storage company. The concept of software-defined storage and the integration of Portworx and Kubernetes were mentioned as advancements in storage solutions. Virtualization of storage is a growing trend, with predictions of AI integration taking more years to become widely applicable. Pure Storage has shown growth and aims to become a leader in all-flash storage, particularly in replacing spinning disks with more efficient solutions for hyperscalers. CEO Charlie Giancarlo is recognized as a leader in the industry.

Charles Giancarlo
CEO & Chairman Pure Storage
ON DEMAND

Krishna Rangasayee, SiMa.ai

Krishna Rangasayee leads SiMa.ai, a company focused on AI infrastructure for the embedded edge market. They work to bring AI capabilities to industries like robotics, automotive, and medical. SiMa.ai's platform is designed for power efficiency, ease of use, and scalability. They prioritize software optimization for high performance. By enabling a no-code ML environment, they aim to simplify AI application deployment at the edge. The company has partnerships in sectors like medical imaging, factory automation, and national defense. SiMa.ai believes in the bright future of AI and the evolving industry landscape. They stress the importance of learning from failures and successes, being humble, and embracing challenges. Despite the highs and lows of startup life, staying focused on goals is crucial. The impact of AI leaders is significant, creating value and opportunities for growth. The fast pace of the startup world can be rewarding, leading to personal growth and innovation. Resilience, determination, and a positive mindset are key for navigating the challenges of entrepreneurship.

Krishna Rangasayee
Founder & CEO SiMa.ai
ON DEMAND

Dave Lazovsky, Celestial AI

Dave Lazovsky
CEO Celestial AI
ON DEMAND

Shehram Jamal | NVIDIA

Shehram Jamal
Chief Product Officer for Data & Models Humain