Techub News News: The NVIDIA Vera Rubin platform, in its first MLPerf benchmark test, achieved leading performance in training, post-training, and agentic inference. The platform swept all benchmarks of MLPerf Training v6.1 and in its first submission to the MLPerf Endpoints test, its agentic AI throughput per watt achieved more than 10 times generational improvement. In DeepSeek-V3 benchmark, Vera Rubin NVL72 from 256 units GPU scaled to 1024 units GPU maintained approximately 97% efficiency, reducing training time from 16 minutes to about 4 minutes. NVIDIA also used the GB300 system to run benchmarks on 10,240 units of GPU , and Microsoft Azure also used the GB200 system to achieve matching results on NVIDIA GPU of the same scale. In the MLPerf Endpoints agentic inference test, the Vera Rubin platform also led. For example, in the provisional results for Kimi K3 , a large open-source model built for agentic workflows, the Vera Rubin NVL72 system delivered more than GB300 NVL72 times the throughput per watt of NVIDIA 10 at the same response speed. (NVIDIA Blog)

如您认为本页面(或本平台)内容存在侵权或违规情形,请您在发现之日起尽快与我们联系,并提供:权利人姓名/名称及联系方式、被侵权作品名称及权属证明、具体侵权链接及情况说明。我们在核实后将及时处理相关内容。
联系方式:creator@techub.news
商务合作:contacting@techub.news
点击链接下载 Techub News App,获取最新讯息:
https://www.techub.news/download
(或扫描下方二维码下载)

All information provided on this website is for reference only.
This website does not guarantee the accuracy, validity, timeliness, or completeness of the information.
Any actions taken based on the information provided by this website are at the user's own risk.
Techub News News: The NVIDIA Vera Rubin platform, in its first MLPerf benchmark test, achieved leading performance in training, post-training, and agentic inference. The platform swept all benchmarks of MLPerf Training v6.1 and in its first submission to the MLPerf Endpoints test, its agentic AI throughput per watt achieved more than 10 times generational improvement. In DeepSeek-V3 benchmark, Vera Rubin NVL72 from 256 units GPU scaled to 1024 units GPU maintained approximately 97% efficiency, reducing training time from 16 minutes to about 4 minutes. NVIDIA also used the GB300 system to run benchmarks on 10,240 units of GPU , and Microsoft Azure also used the GB200 system to achieve matching results on NVIDIA GPU of the same scale. In the MLPerf Endpoints agentic inference test, the Vera Rubin platform also led. For example, in the provisional results for Kimi K3 , a large open-source model built for agentic workflows, the Vera Rubin NVL72 system delivered more than GB300 NVL72 times the throughput per watt of NVIDIA 10 at the same response speed. (NVIDIA Blog)

如您认为本页面(或本平台)内容存在侵权或违规情形,请您在发现之日起尽快与我们联系,并提供:权利人姓名/名称及联系方式、被侵权作品名称及权属证明、具体侵权链接及情况说明。我们在核实后将及时处理相关内容。
联系方式:creator@techub.news
商务合作:contacting@techub.news
点击链接下载 Techub News App,获取最新讯息:
https://www.techub.news/download
(或扫描下方二维码下载)

All information provided on this website is for reference only. The website does not guarantee the accuracy, validity, timeliness, or completeness of the information. Any actions based on the information provided are at the user’s own risk.