Tenstorrent has previewed its optimized AI model, which is powered by the Blackhole servers, offering blazing-fast video generation speeds.
Video Generation In The Blink of an Eye, Tenstorrent Achieves Real-Time Video Processing Speeds With New AI Model Running on Blackhole
AI Video Generation takes a good amount of time; essentially, the faster your hardware, the faster you'll be able to generate videos. While the process has gotten efficient, Tenstorrent is taking things to the next level.
The company previewed its upcoming AI model, Wan2.2-14B, which is trained on 14 billion parameters within servers running several hundred Blackhole accelerators. And the performance is brutally fast.
In a live demonstration given to EETimes, the new AI model was able to generate a 720p video (81 frames, 40 steps) in just five seconds. What's important is that the generated video itself was also 5 seconds long. Tenstorrent also claims that they were able to achieve a record time of 2.4 seconds for a similar 5-second video, which is 10x faster than what competitors manage to achieve.
Tenstorrent uses an optimized version of the Wan2.2-14B AI model, which is created by Prodia, and runs it on four Galaxy servers, featuring 256 Blackhole accelerators. Each Blackhole accelerator features 16 RISC-V cores and 32 GB of GDDR6 memory, offering high-performance AI processing. It's the flagship product within Tenstorrent's AI stack.
Tenstorrent manages to achieve this milestone with its unified architecture approach, which involves zero reliance on proprietary technology and rather uses an open compute platform where the cores, memory, and networking all work together to run massive AI models.
This is just a preview, and Tenstorrent is expected to showcase some more performance demonstrations in the coming week as it is preparing to launch its next-generation cluster-scale systems, so stay tuned.
About the author: A Software Engineer by training and a PC enthusiast by passion, Hassan Mujtaba serves as Wccftech's Senior Editor for hardware section. With years of experience in the industry, he specializes in deep-dive technical analysis of next-generation CPU and GPU architectures, motherboards, and cooling solutions. His work involves not only breaking news on upcoming technologies but also extensive hands-on reviews and benchmarking.
Follow Wccftech on Google to get more of our news coverage in your feeds.
