Harmonic's cOS turns it into a software business whose margins rise with every cable operator that converts from hardware. cOS virtualizes cable access networks, holds over 95% share of virtualized CCAP and serves every major US cable operator; DOCSIS 4.0 upgrades by Charter, Comcast and Cox against fiber require it. Q1 revenue rose 43% to $121.7M, backlog rose 87% to $582M of mostly software and services, and gross margin is 52%. At 3.5x forward revenue versus peers' 4-6x, it hasn't re-rated. … Show more

Harmonic is a high-risk, high-reward bet on the multi-year DOCSIS 4.0 cable upgrade cycle as it sells its video segment for $145M to become a pure-play broadband company. Q1 2026 broadband revenue rose 43%, backlog hit a record $574M, up 87%, and full-year guidance was raised to $475-495M. It remains GAAP unprofitable, depends heavily on Comcast, has flagged supply chain issues and is early in its shift to recurring revenue. x.com 

Harmonic is the biggest consensus long ever seen, and that crowding means it is likely to be kept down. x.com 

Harmonic is the clearest pure-play winner of a multi-year cable broadband upgrade cycle driven by agentic AI and home compute pulling more bandwidth. Its cOS runs the networks of almost every major cable operator with about 95% share, and replacing it needs 18–24 months of certification, so it owns the market into 2028. Business outside its two largest customers grew 78% and is now most of bookings. Growing 43%, it trades in the low 20s on FY2027 earnings. x.com 

Harmonic is still incredibly early in a global adoption wave: a first smaller European operator has started deploying its solutions, and the giants will have to follow given technical realities. Its cOS virtualized platform dominates the cable core, delivering multi-gigabit DOCSIS 4.0 over existing coax; per @KawzInvests, Comcast, Charter, Cox and Altice run it, with 150 customers and 45.7 million modems. Virtual OLT software opens a $2.5 billion-plus fiber market. x.com  x.com 

Harmonic is turning into a pure-play AI-broadband infrastructure compounder the market still prices as the old company. Selling its video business to MediaKind for $145M leaves pure broadband, with cash for DOCSIS 4.0, fiber and buybacks. Q1 broadband revenue rose 43%, backlog 87% to $582.1M, and FY26 broadband guidance was raised to $475-495M. Its cOS platform first unified FDX and FDD DOCSIS 4.0, so operators needn't switch suppliers, and 33M+ devices create sticky switching costs. x.com  x.com  … Show more

Harmonic's fundamentals kept improving while the market forgot it. It delivered a double beat and raised guidance, and the stock showed strong price action with a clean retest of its 50-week average. x.com 

Harmonic gains pricing power as fiber becomes structurally scarce, because it sits at the platform layer powering fiber and cable networks. The US has 159.6 million fiber miles but needs 372.9 million by 2029, and North American demand grows 22-25% a year while supply expands 12-19%. Fiber prices are up about 70% from the 2021 trough, AI data centers need about 36x more fiber than CPU racks, and data centers may be 30% of global demand by 2027. That makes an asymmetric setup. x.com 

Electrical work is preferable to Dycom's communications-heavy business because it is non-discretionary while carriers delay spending they cannot afford. IES Holdings and Sterling Infrastructure look like better ways to play the data-center scare than neoclouds. Upcoming spectrum auctions may absorb carrier budgets, though Dycom might benefit from the carriers-versus-SpaceX narrative. x.com  x.com  x.com 

Charter is a very asymmetric bet at 2.35x consensus 2027 EPS, letting it buy back shares massively at a P/E of about 2. A modeled bad case with much bigger, continued subscriber losses still yields about $30 of owner earnings. Charter can offset lost subscribers by raising prices, capturing M&A synergies and paying down debt, with many years of deleveraging before EBITDA would fall 20% and debt covenants would bite. x.com  x.com  x.com