Cisco vs Harmonic

Small Language Models: When Smaller Is Better

This MLPills newsletter issue explains small language models (SLMs) — compact AI models designed to run under resource constraints such as limited memory, power, and latency budgets. It clarifies that 'small' is a comparative concept rather than a specific parameter count, and distinguishes SLMs from quantized frontier models and distillation. The article covers four main routes to building SLMs: curated data training, distillation, pruning, and quantization, and highlights examples including Microsoft's Phi-4 family, Google's Gemma 3n and Gemma 4 edge models, Hugging Face's SmolLM3, and Cisco's Antares vulnerability-localization models. It describes ideal use cases like classification, entity extraction, and tool selection, and recommends a layered architecture using deterministic code, small models, large models, and human oversight. The piece also cautions about evaluation, over-pruning, and privacy limitations of local inference.

• A 4B parameter model at 4-bit precision needs roughly 2 GB of memory; the same model at 16-bit needs roughly 8 GB.
• Microsoft's Phi-4 is a 14B parameter model, with a 3.8B Phi-4-Mini sibling and a Phi-4-Multimodal variant.
• Google's Gemma 3n models used per-layer embeddings, KV cache sharing, and activation quantization to reduce memory footprints.

Harmonic Unveils cOS SensAI Intelligence Fabric for Broadband Operations

New Vendor-Agnostic Broadband Intelligence Fabric Reduces Operational Complexity and Improves Service Quality for Broadband Service Providers. SAN JOSE, Calif. — Sept. 8, 2026 — Harmonic (NASDAQ: HLIT) announced the launch of cOS SensAI, a vendor-agnostic intelligence fabric for broadband operations.

Harmonic launches cOS SensAI intelligence infrastructure for broadband

Harmonic, a leading provider of virtualized broadband solutions, has announced the launch of cOS SensAI, a vendor-agnostic, software-driven intelligence infrastructure for broadband operations. SensAI aggregates heterogeneous data from various network sources into a unified information layer, transforming complex operational data into actionable insights. This enables broadband service providers to proactively manage operations, improve customer experiences, and increase Net Promoter Scores. The solution is already being piloted by leading broadband operators like Hotwire Communications and GCI. Analysys Mason estimates the total addressable market for SensAI at approximately $2.2 billion per year for Tier-2 and smaller broadband providers. Harmonic will showcase SensAI at Connected Britain and SCTE TechExpo in September.

• Harmonic introduced cOS SensAI, a vendor-agnostic intelligence infrastructure for broadband networks.
• cOS SensAI integrates data from multiple sources into a unified information layer for actionable insights.
• The solution is being used by Hotwire Communications and GCI to improve operational efficiency.