1 of 8

Open Systems for AI

OCP Marcom Plan 2024

Community-driven hyperscale innovation for all

2 of 8

Same Mission, New Use Case

Community

Hyperscalers
Co-location providers
Solution vendors
IT service providers
Telecoms
Enterprise IT

Open Innovation

Hyperscale-proven designs
Trusted-IP model
Specs, best practices, …
Rapid technology adoption
Multi-vendor ecosystem
Future technologies

Data centers Edge

Silicon and systems Physical Facilities

Scope

Connect. Collaborate. Accelerate.

Community-driven hyperscale innovation for all

3 of 8

Next DC Technology Shift

Infrastructure Challenges:

Keeping specialized processors and systems modular with open design and interfaces while in the AI “rush”
Adapting system designs with accelerated computing and new interconnects for xPUs, memory, multiple network tiers
Keeping cool while massively powering up requires new rack and physical facilities standards and sustainability goals
Scale requires more automated systems and facilities management and with AI and robotics

Hyperscale DC Required

Scale and efficiency of IT and physical DC through openness
Flexible disaggregation and SD-x systems management value on open hardware

At-scale HPC and AI

Connect. Collaborate. Accelerate.

Community-driven hyperscale innovation for all

AI and HPC workloads are also distributed and diverse, much more so than other virtualized and cloud-native applications. In the pursuit of greater performance, infrastructure specialization for some workload categories is burgeoning, but risks fracturing the supply chain into silos, raising costs and reducing efficiencies. This is exacerbated by the accelerated drive for innovation to meet new requirements in the hardware ecosystem and rush for all players in the market to provide solutions.

�To prevent these downsides, the OCP facilitates open-sourcing hardware specifications and standardizing building blocks that are forged for the most demanding hyperscale solutions. The OCP’s trusted IP model allows multiple vendors to safely collaborate and specialize their products with greater interoperability and composability which benefits those vendors as well as the customers, moving the entire market forward. Spanning the OCP Projects, the OCP “Open Systems for AI” Strategic Initiative brings together the work of the Community to meet AI's scale and workload diversity.

�In summary, as open-source and standardized systems improve economies and efficiencies for hyperscale data centers and in turn data centers more generally, they do so even more so for AI clusters’ requirements of even more massive resource scale and varying infrastructure needs.

4 of 8

Next DC Technology Shift

Infrastructure Challenges:

Hyperscale DC Required

Scale and efficiency of IT and physical DC through openness
Flexible disaggregation and SD-x systems management value on open hardware

At-scale HPC and AI

Specialization

Scale

Speed

Connect. Collaborate. Accelerate.

Community-driven hyperscale innovation for all

AI and HPC workloads are also distributed and diverse, much more so than other virtualized and cloud-native applications. In the pursuit of greater performance, infrastructure specialization for some workload categories is burgeoning, but risks fracturing the supply chain into silos, raising costs and reducing efficiencies. This is exacerbated by the accelerated drive for innovation to meet new requirements in the hardware ecosystem and rush for all players in the market to provide solutions.

�To prevent these downsides, the OCP facilitates open-sourcing hardware specifications and standardizing building blocks that are forged for the most demanding hyperscale solutions. The OCP’s trusted IP model allows multiple vendors to safely collaborate and specialize their products with greater interoperability and composability which benefits those vendors as well as the customers, moving the entire market forward. Spanning the OCP Projects, the OCP “Open Systems for AI” Strategic Initiative brings together the work of the Community to meet AI's scale and workload diversity.

�In summary, as open-source and standardized systems improve economies and efficiencies for hyperscale data centers and in turn data centers more generally, they do so even more so for AI clusters’ requirements of even more massive resource scale and varying infrastructure needs.

5 of 8

Open Systems for AI - Initiative Work Areas

SYSTEMS MANAGEMENT

DC IT INFRASTRUCTURE

DC PHYSICAL INFRASTRUCTURE

DC Facilities

Server

Test and Validation

Rack and Power

Cooling Environments

AI Automation

Storage

Networking

Hardware Management

Open Platform Firmware

Memory Systems

Open Chiplet Economy

OCP

Future

Technologies Initiative

Sustainability

Security

Link together work areas to create open systems

Connect. Collaborate. Accelerate.

Community-driven hyperscale innovation for all

6 of 8

Assembling Open Systems of Systems

Project level

Cross-stack

Cluster level

Progressively Getting to Holistic AI Clusters Design with Open Source

Connect. Collaborate. Accelerate.

Community-driven hyperscale innovation for all

7 of 8

Marketing and Comms Strategy

CLARIFY

UNIFY

AMPLIFY

Paper and slide deck by Summit
Curate great AI tracks
Ride the wave at other events
Embrace the nature of our eng-led community with long-form technical, educational content

Repeat “Open Systems for AI” in Summit signage, digital media…
Upcycle and derivative content
Promote the boundary spanners
Expand OCP marketplace with OCP Inspired AI contributions, e.g. many HGX-based products

Invite partners, academics, analysts, guest to our calls
Work with other open bodies like UEC, UALink, MLCommons, AI Alliance…
Cross-promote with upstack foundations and benchmarks like LF, MLPerf, DCPerf

Other plan details:

market and messaging background, calendar, KPIs, etc.

Connect. Collaborate. Accelerate.

Community-driven hyperscale innovation for all

8 of 8

Community-driven hyperscale innovation for all

Thank you

opencompute.org/ai