Kirthi Devleker

Kirthi K. Devleker is a technology marketing leader at NVIDIA, where he drives the launch and positioning of transformative AI platforms and the GPU architectures that power them. He played a pivotal role in bringing NVIDIA’s groundbreaking Grace Blackwell architecture to market, including the Grace Blackwell and Grace Blackwell Ultra platforms—redefining performance, scalability, and efficiency for generative AI at global scale. Kirthi specializes in crafting compelling messages around NVIDIA’s datacenter GPU technologies, highlighting their performance advantages and ROI for enterprise AI adoption. Previously, at MathWorks, he led the global Medical Devices business unit and spearheaded strategic product management initiatives that guided the Signal Processing group’s roadmap towards AI. His leadership accelerated machine learning integration across medical devices, aerospace and defense and automotive sectors. As a recognized industry voice, Kirthi has delivered keynotes and technical talks at international conferences on AI-driven engineering and simulation. He holds a Master of Science in Electrical Engineering from San Jose State University, with a specialization in signal and image processing.
Avatar photo

Posts by Kirthi Devleker

Rendering of Rubin CPX.
Generative AI

NVIDIA Rubin CPX Accelerates Inference Performance and Efficiency for 1M+ Token Context Workloads

Inference has emerged as the new frontier of complexity in AI. Modern models are evolving into agentic systems capable of multi-step reasoning, persistent... 5 MIN READ
Generative AI

NVFP4 Trains with Precision of 16-Bit and Speed and Efficiency of 4-Bit

In recent years, AI workloads have grown exponentially—not only in the deployment of large language models (LLMs) but also in the demand to process ever more... 9 MIN READ
An image of the NVIDIA Blackwell Ultra system on a black background.
Data Center / Cloud

NVIDIA Blackwell Ultra for the Era of AI Reasoning

For years, advancements in AI have followed a clear trajectory through pretraining scaling: larger models, more data, and greater computational resources lead... 5 MIN READ
Mixture of experts icons for attention kernels.
Generative AI

Automating GPU Kernel Generation with DeepSeek-R1 and Inference Time Scaling

As AI models extend their capabilities to solve more sophisticated challenges, a new scaling law known as test-time scaling or inference-time scaling is... 6 MIN READ
An image of the GB200 NVL72 and NVLink spine.
Data Center / Cloud

NVIDIA GB200 NVL72 Delivers Trillion-Parameter LLM Training and Real-Time Inference

What is the interest in trillion-parameter models? We know many of the use cases today and interest is growing due to the promise of an increased capacity for:... 9 MIN READ
Robotics

Developing AI-Powered Digital Health Applications Using NVIDIA Jetson

Traditional healthcare systems have large amounts of patient data in the form of physiological signals, medical records, provider notes, and comments. The... 17 MIN READ