Kneron's new hardware can also support RAG retrieval-augmented generation workflows and the company now has multiple capabilities for training and fine-tuning models that run on its hardware. The company's CEO, Albert Liu, highlighted that one of the key differentiators for Kneron's technology is its low power consumption. The new KL830 has a peak power consumption of only 2 watts, allowing Kneron's chips to be integrated into various devices, including PCs, without the need for additional cooling solutions.
Key takeaways:
- Kneron, a silicon vendor, is rolling out its next generation KL830 neural processing unit (NPU) and providing a glimpse into the future KL 1140 which is set to debut in 2025.
- The company is also expanding its AI server portfolio with the KNEO 330 Edge GPT server that enables offline inference capabilities.
- Kneron's technology aims to improve power and efficiency of AI workloads, offering an alternative to GPU technology.
- The company's new KL830 chip has a peak power consumption of only 2 watts, allowing it to be integrated into various devices without the need for additional cooling solutions.