Kirthi Devleker – NVIDIA Technical Blog News and tutorials for developers, data scientists, and IT admins 2025-03-20T22:34:30Z http://www.open-lab.net/blog/feed/ Kirthi Devleker <![CDATA[NVIDIA Blackwell Ultra for the Era of AI Reasoning]]> http://www.open-lab.net/blog/?p=96761 2025-03-20T22:34:30Z 2025-03-19T18:00:15Z For years, advancements in AI have followed a clear trajectory through pretraining scaling: larger models, more data, and greater computational resources lead...]]>

For years, advancements in AI have followed a clear trajectory through pretraining scaling: larger models, more data, and greater computational resources lead to breakthrough capabilities. In the last 5 years, pretraining scaling has increased compute requirements at an incredible rate of 50M times. However, building more intelligent systems is no longer just about pretraining bigger models.

Source

]]>
Kirthi Devleker <![CDATA[Automating GPU Kernel Generation with DeepSeek-R1 and Inference Time Scaling]]> http://www.open-lab.net/blog/?p=95998 2025-02-20T15:56:57Z 2025-02-12T18:00:00Z 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...]]>

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 emerging. Also known as AI reasoning or long-thinking, this technique improves model performance by allocating additional computational resources during inference to evaluate multiple possible outcomes and then selecting the best one…

Source

]]>
2
Kirthi Devleker <![CDATA[NVIDIA GB200 NVL72 Delivers Trillion-Parameter LLM Training and Real-Time Inference]]> http://www.open-lab.net/blog/?p=79550 2024-07-12T14:47:47Z 2024-03-18T23:00:00Z 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:...]]>

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: The benefits are‌ great, but training and deploying large models can be computationally expensive and resource-intensive. Computationally efficient, cost-effective, and energy-efficient systems, architected to deliver real-time…

Source

]]>
13
Kirthi Devleker <![CDATA[Developing AI-Powered Digital Health Applications Using NVIDIA Jetson]]> http://www.open-lab.net/blog/?p=24159 2024-03-14T20:04:36Z 2021-02-25T18:22:00Z Traditional healthcare systems have large amounts of patient data in the form of physiological signals, medical records, provider notes, and comments. The...]]>

Traditional healthcare systems have large amounts of patient data in the form of physiological signals, medical records, provider notes, and comments. The biggest challenges involved in developing digital health applications are analyzing the vast amounts of data available, deriving actionable insights, and developing solutions that can run on embedded devices. Engineers and data scientists…

Source

]]>
0
���˳���97caoporen����