Green Computing with Remote GPU over APN (tsuzumi-7B)

In accordance with the PoC Reference for Green Computing with Remote GPU over APN, we conducted a proof of concept using the small‑scale LLM tsuzumi‑7B.For compute‑dominant workloads where the dataset is relatively small, data transmission time is negligible compared to GPU computation time, even in long‑distance remote GPU environments. While replication and cache strategies are effective for large, data‑intensive workloads, their benefit is limited in lightweight language‑model scenarios such as this PoC.

Download Applicable PoC Reference Document:
PoC Report Participants:

NTT, Inc.

Related projects

Events

Meet the IOWN Global Forum

Members of the Forum gather regularly and speak at industry conferences and events. Find out where you can interact with us next.

Newsletter

Stay up to date

The Forum publishes an informational newsletter and shares industry updates periodically with the latest news, events, and technical information. 

MEMBERSHIP

Join the movement

Join best-in-class partners committed
to solving challenges in creating the next-generation communications infrastructure of tomorrow.

Searching...

In accordance with the PoC Reference for Green Computing with Remote GPU over APN, we conducted a proof of concept using the small‑scale LLM tsuzumi‑7B.For compute‑dominant workloads where the dataset is relatively small, data transmission time is negligible compared to GPU computation time, even in long‑distance remote GPU