RDMA
T-Head Panmai 920: Building Smarter 10,000-GPU Clusters
·2242 words·11 mins
T-Head
Panmai 920
SmartNIC
AI Infrastructure
RDMA
RoCE
Data Center
Networking
GPU Clusters
Presto: Running TCP on Programmable Switches at Terabit Speed
·3502 words·17 mins
Presto
TCP
Programmable Switches
RMT
P4
Data Plane
Network Offload
RDMA
Terabit Networking
REPS Explained: Lightweight Load Balancing for AI Clusters at Scale
·1629 words·8 mins
Networking
AI Infrastructure
Ultra Ethernet
Data Centers
Load Balancing
SmartNIC
RDMA
High-Performance Computing
AI Super-Cluster Interconnects: NVIDIA, Google, and China's Networking Strategies
·1312 words·7 mins
AI Infrastructure
Networking
NVIDIA
Google TPU
InfiniBand
RDMA
High-Performance Computing
Data Centers
RFC 9800 Compressed SRv6: The Hidden Engine Behind AI Superclusters
·1781 words·9 mins
RFC 9800
SRv6
AI Datacenter
Networking
MRC
OpenAI
Microsoft
RDMA
Ethernet
High-Performance Computing
OpenAI MRC Protocol Powers 100,000-GPU AI Superclusters
·1211 words·6 mins
OpenAI
MRC
AI Networking
RDMA
RoCE
NVIDIA
AMD
Broadcom
Microsoft
Supercomputing
6 ns Hardware Timer: IEEE TC Breakthrough for RDMA and DPU
·782 words·4 mins
Hardware Timer
IEEE TC
RDMA
DPU
Programmable Data Plane
FPGA
ASIC
Networking
RDMA Explained: The Backbone of High-Performance Networking
·595 words·3 mins
RDMA
Networking
HPC
AI Infrastructure
Data Center
Distributed Systems
ByteDance veRoCE Makes RDMA Work on Lossy Networks
·653 words·4 mins
RDMA
RoCE
Data Center Networking
AI Infrastructure
ByteDance
RDMA Explained: How Remote Direct Memory Access Works
·666 words·4 mins
RDMA
HPC
RoCEv2
Networking