compute.layer.re

Compute

AI compute is scaling through large cloud commitments and increasingly system-level architectures that combine accelerators, memory and high-bandwidth interconnects.

What changed

ecosystem integration

DeepSeek and Huawei deepen Ascend software-stack integration

DeepSeek said it was open-sourcing programming infrastructure for Huawei Ascend chips and jointly advancing supernode software, including TileLang-related work.

Status
verified
Confidence
97%
Canonical ID
evt-deepseek-huawei-stack-20260930
agent infrastructure release

NVIDIA launches Open Agent Safety Platform

NVIDIA released OpenShell and the Sentry reference design to enforce agent boundaries across software and hardware, including policy, identity, tool and data-access controls.

Status
verified
Confidence
99%
Canonical ID
evt-nvidia-agent-safety-20260928
infrastructure commitment

Anthropic signs long-term Akamai cloud capacity deal

Anthropic and Akamai entered a seven-year $11.6 billion cloud services agreement, with Akamai planning major supporting capital expenditure and component purchases.

Status
verified
Confidence
98%
Canonical ID
evt-anthropic-akamai-cloud-20260924
governance update

EU advances data-center energy transparency and performance framework

The European Commission proposed a common data-center rating scheme and opened work toward minimum performance standards covering energy and water use, efficiency and grid integration.

Status
verified
Confidence
99%
Canonical ID
evt-eu-datacenter-framework-20260921
compute platform release

Huawei unveils Atlas 960E SuperPoD

Huawei announced a large-scale AI compute system built around UnifiedBus, NPO interconnects and up to 4,096 NPUs, alongside continued CANN open-source development.

Status
verified
Confidence
98%
Canonical ID
evt-huawei-atlas960e-20260917

Recent signals