Rule 11

SEPTEMBER 20, 2024

Why is InfiniBand so popular for building AI networks? Jeff Tantsura joins Tom Ammon and Russ White to discuss networks for AI workloads. Why is InfiniBand so popular for building AI networks? Jeff Tantsura joins Tom Ammon and Russ White to discuss networks for AI workloads. What about Ethernet for AI?

Infiniband 92

Infiniband Ethernet Network Networking 92

Engineering at Meta

AUGUST 5, 2024

AI networks play an important role in interconnecting tens of thousands of GPUs together, forming the foundational infrastructure for training, enabling large models with hundreds of billions of parameters such as LLAMA 3.1 Distributed training, in particular, imposes the most significant strain on data center networking infrastructure.

Network 132

Network Networking Topology Data Centers 132

Engineering at Meta

JUNE 12, 2024

Supporting GenAI at scale has meant rethinking how our software, hardware, and network infrastructure come together. Solving this problem requires a robust and high-speed network infrastructure as well as efficient data transfer protocols and algorithms. requires revisiting trade-offs made for other types of workloads.

Infiniband 141

Infiniband Data Centers Topology Network 141

Packet Pushers

MARCH 25, 2024

Take a Network Break! Nvidia announces new 800G switches, one for Ethernet and one for InfiniBand, for building AI fabrics. Nvidia also announces an “AI supercomputer,” a rack-scale pre-built bundle of Nvidia GPUs and CPUs connected via InfiniBand switches.

Infiniband 52

Infiniband Ethernet NaaS Network 52

Packet Pushers

OCTOBER 25, 2024

Ethernet competes with InfiniBand as a network fabric for AI workloads such as model training. In other words, AI workloads do best with a lossless network. Read more » Ethernet competes with InfiniBand as a network fabric for AI workloads such as model training. Read more »

Ethernet 83

Ethernet Infiniband Bandwidth Network 83

Packet Pushers

NOVEMBER 15, 2021

This week's Network Break examines new 400G switches from Arista, discusses the Wi-Fi Alliance's certification program for the HaLow long-range low-power standard, targets key Nvidia announcements, catches up on the latest in space networking, and more IT news.

Infiniband 52

Infiniband Network Networking 52

Engineering at Meta

MARCH 12, 2024

We are sharing details on the hardware, network, storage, design, performance, and software that help us extract high throughput and reliability for various AI workloads. Network At Meta, we handle hundreds of trillions of AI model executions per day. The other cluster features an NVIDIA Quantum2 InfiniBand fabric.

Infiniband 145

Infiniband Data Centers Server Network 145