BlockBeats News, January 6 — NVIDIA (NVDA.O) announced that its highly anticipated new “Rubin” data center product will be launched later this year. Customers will be able to test the technology, helping to accelerate the development of artificial intelligence. NVIDIA stated that all six Rubin chips have been delivered from manufacturing partners and have passed several key tests, indicating that customer deployment is proceeding as planned. Its latest Rubin accelerator offers a 3.5x improvement in training performance over the previous Blackwell generation and a 5x increase in speed for running AI software. The company emphasized that systems based on Rubin will have lower operating costs than those based on Blackwell, as they require fewer components to achieve the same performance. It is reported that Microsoft and other major remote computing service providers will be among the first to adopt this new hardware in the second half of the year. (Jin10)
View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
NVIDIA: The new Rubin chip is progressing smoothly, helping to accelerate the operation speed of artificial intelligence
BlockBeats News, January 6 — NVIDIA (NVDA.O) announced that its highly anticipated new “Rubin” data center product will be launched later this year. Customers will be able to test the technology, helping to accelerate the development of artificial intelligence. NVIDIA stated that all six Rubin chips have been delivered from manufacturing partners and have passed several key tests, indicating that customer deployment is proceeding as planned. Its latest Rubin accelerator offers a 3.5x improvement in training performance over the previous Blackwell generation and a 5x increase in speed for running AI software. The company emphasized that systems based on Rubin will have lower operating costs than those based on Blackwell, as they require fewer components to achieve the same performance. It is reported that Microsoft and other major remote computing service providers will be among the first to adopt this new hardware in the second half of the year. (Jin10)