Scribd
Upload
66 views4 pages

DDC For AI-ML Architecture

The document discusses how next-generation 800G data center networks can reduce job completion times for AI/ML workloads through increased port speeds, distributed disaggregated chassis, and openness across GPUs, ASICs, software, hardware and network interface cards. It presents the architecture and components of UfiSpace's next-gen 800G distributed data center (DDC) including 800G and 400G spine switches, and 800G, 400G, and 25G leaf switches, as well as the company's network operating systems and hardware platforms. The document encourages contacting UfiSpace for more information on their open disaggregated network solutions for mobile and broadband networks.

Uploaded by

Mohsen Goodarzi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
66 views4 pages

DDC For AI-ML Architecture

The document discusses how next-generation 800G data center networks can reduce job completion times for AI/ML workloads through increased port speeds, distributed disaggregated chassis, and openness across GPUs, ASICs, software, hardware and network interface cards. It presents the architecture and components of UfiSpace's next-gen 800G distributed data center (DDC) including 800G and 400G spine switches, and 800G, 400G, and 25G leaf switches, as well as the company's network operating systems and hardware platforms. The document encourages contacting UfiSpace for more information on their open disaggregated network solutions for mobile and broadband networks.

Uploaded by

Mohsen Goodarzi
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

DDC for AI/ML

Reduced Job Completion Times with Next-Gen Data


Center Networks

2023 Network Port 1.6T


Speeds Development

800G

400G

Distributed
200G
Disaggregated
InfiniBand 100G Chassis

Low Jitter | Flow Agnostic


Next-Gen Capacity
Openness (GPU, ASIC, SW, HW, NIC)
Resilience
Security
DDC for AI/ML Architecture

Next-Gen 800G DDC


Cell-Based
Scheduled 800G Spine | R3 800G Spine | R3
Fabric Switch Fabric Switch
Fabric DDC Fabric
Congestion Free

Low Jitter and 800G Leaf | J3 800G Leaf | J3 800G Leaf | J3


Controlled Latency Packet Switch Packet Switch Packet Switch
Ethernet
No Protocol
Bridging
Up to 32,000 GPUs
at 800G

Perfect Load
Balancing
Versatile DDC Today

Virtual Single Entity


400G Spine | R 400G Spine | R
Router DDC Fabric
Congestion Free

Robust Ethernet
Protocol Security 25G Leaf | J2c 100G Leaf | J2 400G Leaf | J2c+
Ethernet
No Protocol
Bridging
Lossless RoCE
Operation
DDC Software and Hardware Platforms
DDC Network Operating Systems (NOS)

DDC Fabric Switches

S9700-48D | R R3
400G High Density Fabric Switch 800G High Density Fabric Switch
Coming in 2023*

DDC Packet Switches

S9700-23D | J2 S9710-76D | J2c+ J3


400G Packet Switch 400G High Density Packet Switch 800G High Density Packet Switch
Coming in 2023*

S9701-82DC | J2c S9700-53DX | J2


25G High-Density Packet Switch 100G High Density Packet Switch
CONTACT US

Want to know more about UfiSpace field proven open


disaggregated network solutions for mobile and
broadband networks.
Trial Request | Quote Request | Solution Demonstration | Information Inquiry

sales@ufispace.com

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy