This Rubin submission uses the new @KVCache_AI Mooncake Store feature of Extended GPU Memory, where KVCache stored on Ve | Hanami
This Rubin submission uses the new @KVCache_AI Mooncake Store feature of Extended GPU Memory, where KVCache stored on Vera's LPDDR5X is pooled across all 18 Rubin compute trays using NVLink instead of the previous method of using RDMA, resulting in greater than 10x better bandwidth. (2/2)
Reply to @SemiAnalysis_: BREAKING NEWS🚨: HBM CHIPS CAN REACH OVER 1000 tok/s/user on giant 700B+ models via agent swarm-written software, according to a new NVIDIA AgentX submission.
Furthermore, on Rubin NVL72, it can achieve 20.9x better perf per TCO than GB300 NVL72 at even "ultrafast" speeds of 325 tok/s/user and up to 9.11x better perf per TCO at 200 tok/s/user. (1/2)🧵(Historical earnings: For fiscal 2026FY (period ended 2026-07-03), Sandisk reported basic EPS of 77.78, diluted EPS of 73.76, and net income of USD 11.433 billion.
Consensus expectations: For the next upcoming period 2027Q2 (event date 2026-10-29), consensus EPS is 46.2722 and consensus revenue is USD 10.677 billion.)