Details
- GSMA announced Open Telco AI at MWC 2026 in Barcelona, a global initiative to accelerate telco-grade AI models and systems via open collaboration, anchored by open-telco.ai portal for shared resources, datasets, tools, and benchmarks.
- Key participants include GSMA leading, founding supporters AT&T providing open telco models, AMD supplying Instinct GPUs and ROCm software for training/inference, and TensorWave for hosting.
- AMD showcases Enterprise AI Suite for operationalizing AI from experimentation to production with Kubernetes-native tools for model serving, governance, and telco network automation; EPYC 8005 Server CPUs optimize edge vRAN with high density and wide thermal ranges.
- Addresses GSMA-reported gap where only 16% of gen AI deployments reach networks, contrasting with prior experimentation by enabling validated, scalable telco-specific performance.
- AMD partners with Samsung for AI-powered vRAN commercial deployments and Wind River for Open RAN/AI on shared infrastructure, expanding ecosystem reach.
Impact
AMD's full-stack AI solutions position it centrally in telco's shift to open, virtualized networks, enabling operators to scale AI for automation and resiliency amid rising edge demands. This counters limited gen AI adoption, fostering competition with players like Google Cloud while driving efficient, sustainable deployments. Open collaboration accelerates industry-wide intelligent networks, reducing hardware dependencies and boosting operational efficiency.
