データセンター

AIネットワーク意思決定フレームワーク:スピード、ROI、戦略的自由度を最適化するGPUファブリック

The decision to invest in AI infrastructure centers on GPUs. But GPU compute efficiency is ultimately gated by the network connecting them. In large-scale AI training clusters, the network fabric is not infrastructure support; it is a primary determinant of how much of your GPU investment actually delivers productive compute cycles.

A suboptimal network can reduce GPU utilization by up to 50% during the compute-exchange-update phase of distributed training. When GPU clusters cost thousands of dollars per hour to operate, network inefficiency is a direct revenue loss.

Why the Network is the AI Bottleneck

Modern AI training uses distributed data parallelism: multiple GPU nodes process data simultaneously and periodically synchronize gradients. The synchronization step (the All-Reduce operation) requires all GPUs to exchange gradient data through the network simultaneously. This creates extreme traffic bursts that expose every weakness in the fabric:

  • Packet loss: RoCEv2 (RDMA over Converged Ethernet) is loss-sensitive. A single dropped packet triggers retransmission that stalls the entire GPU collective operation until the packet is recovered.
  • Latency spikes: jitter during All-Reduce operations extends Job Completion Time (JCT) proportionally to the number of GPUs in the cluster.
  • Congestion spreading: without proper flow control, congestion on one link spreads to others through head-of-line blocking, degrading the entire fabric.

The Three AI Fabric Design Decisions

Three-pillar AI fabric decision framework: Performance (lossless transport), Scale (switching capacity and topology), and Freedom (open hardware and NOS choice), converging on OcNOS as the open platform
The three AI fabric design pillars: Performance (lossless RoCEv2 transport), Scale (400G/800G Clos topology), and Freedom (open hardware, no vendor lock-in). OcNOS delivers all three on open, disaggregated platforms.

OcNOS AI Fabric Capabilities

OcNOS 7.0 delivers a complete lossless AI fabric stack on Broadcom Tomahawk 5-based open hardware:

  • Priority-Based Flow Control (PFC): per-priority pause frames prevent RoCEv2 packet drops at the switch level
  • Enhanced Transmission Selection (ETS): allocates guaranteed bandwidth to RoCEv2 traffic classes (typically 70-80% of fabric bandwidth)
  • DCBX: automatically negotiates PFC and ETS parameters with GPU servers, eliminating manual server-side configuration
  • Explicit Congestion Notification(ECN): signals congestion to senders before queues fill, enabling proactive rate reduction
  • EVPN-VXLAN multi-tenancy: isolates multiple AI workloads or tenants on the same physical fabric

Open vs. Proprietary AI Fabric: TCO at Scale

評価項目 Proprietary (Arista, Cisco) OcNOS on Open Hardware
Spine switch (51.2T, 64x800G) $150K-$250K per switch $60K-$100K per switch
NOS licensing Per-feature or bundle; complex All-inclusive per platform
Transceiver lock-in Often required for warranty Third-party optics supported
Hardware vendor flexibility 単一ベンダー UfiSpace, Edgecore, Celestica, others
Automation interfaces Vendor-specific APIs gNMI, NETCONF, OpenConfig, Ansible

For a 100-node GPU cluster requiring 10 spine switches, the hardware cost difference alone is $500K-$1.5M over a 3-year lifecycle.


IP Infusion Engineering Team

さらに詳しく。資料をお持ち帰りください。

製品データシートと、このページよりもさらに踏み込んだ簡潔で技術的なダウンロード資料です。

データシート

OcNOS-DC データシート

OcNOS-DC の完全な仕様。EVPN-VXLAN と Ethernet for AI の機能セット、ソフトウェア SKU、サポート対象ハードウェアプラットフォーム、ソリューション発注ガイドを収録しています。

データシートを入手
ソリューションブリーフ

OcNOS 800G ロスレス AI Fabric

Tomahawk 5 スパイン上のロスレス RoCEv2 ファブリック:SKU 階層、認定済みプラットフォーム、導入アーキテクチャ。

概要資料を入手
ソリューションブリーフ

EVPN-VXLAN データセンターファブリック

EVPN-VXLAN によるリーフスパイン型データセンターファブリック:テナントごとの VRF と VNI、EVPN マルチホーミング、BGP unnumbered、ルートターゲットフィルタリング。

概要資料を入手
共有