In the ever-evolving landscape of technology, artificial intelligence (AI) stands out as a transformative force, mimicking human cognitive functions to solve complex problems efficiently. Now, Nvidia has introduced the Blackwell platform, which will help organisations use AI more efficiently. With this platform, they can run large language models much faster and cheaper than before, saving energy along the way.
“Powering a new era of computing, Nvidia today announced that the Nvidia Blackwell platform has arrived — enabling organisations everywhere to build and run real-time generative AI on trillion-parameter large language models at up to 25x less cost and energy consumption than its predecessor,” the company announced in a press release on March 18th.
Also read: OpenAI, Meta and Tesla desperately want this Nvidia GPU, but why?
The Blackwell GPU architecture includes six transformative technologies that speed up computing. These advances will drive progress in data processing, engineering simulation, electronic design automation, computer-aided drug design, quantum computing and generative AI.
These are all growing fields where NVIDIA sees big opportunities.
Also read: NVIDIA unveils DLSS 3.5 with new Ray Reconstruction feature, here’s how it works
Big companies like Amazon Web Services, Dell Technologies, Google, Meta, Microsoft, OpenAI, Oracle, Tesla, and xAI are likely to start using Blackwell.
Nvidia claims that the new Blackwell platform is the “world’s most powerful chip”. It is packed with 208 billion transistors, and is manufactured using a custom-built 4NP TSMC process.
Also, Blackwell is said to support double the compute and model sizes with new 4-bit floating point AI inference capabilities.
The Blackwell architecture includes features that can use AI for preventive maintenance. This means it can check for problems and predict when things might break down. This helps keep systems running smoothly, even for big AI projects that need to run continuously for weeks or months.
The Blackwell platform offers advanced security features to safeguard AI models and customer data without slowing down performance. It includes support for encryption protocols that are essential for industries like healthcare and finance, where privacy is crucial.
Additionally, Blackwell includes a specialised engine for decompression, which speeds up database queries. This means faster performance for data analytics and data science.
Nvidia revealed that Blackwell-based products will be available from partners starting later this year.