podcast
details
.com
Share
科技热点新闻
不加一块GPU,多榨出15%算力:大模型圈开始对网络“动刀”
Listen
Description
1.这场从“堆硬件”向“挖效率”的转向,正在对上游供应链产生实质影响;
2.ZCube架构所需的交换机和光模块数量比原有方案少了三分之一。
podcast
details
Print
Share
Want to check another podcast?
Enter the RSS feed of a podcast, and see all of their public statistics.
Check it
Made by
Alex Barredo
. Send your feedback to
alex@barredo.es
.