Google streams games from data centers to browsers

What happened

Technically the service delivered much of what it promised. Games ran in data centers on custom hardware, and the resulting video arrived quickly enough that many titles felt responsive on a good connection, including some that demanded precise timing. A browser tab or an inexpensive television dongle could replace a console entirely. Players on slower or less consistent connections, and anyone with a monthly data cap, had a very different experience, and the whole proposition depended on infrastructure that a game company cannot control and a household often cannot improve.

The commercial design created most of the trouble. Games had to be purchased individually at full price, while a separate subscription covered higher resolution and a rotating handful of titles, which left buyers unsure what they were actually getting. The catalogue was thin at launch and grew slowly, several announced features arrived late, and the company's reputation for closing products made people reluctant to invest in a library they could not move. An internal studio set up to make original games was shut down after roughly a year, which did little to reassure anyone.

The world at the time

Video streaming had reshaped film and television, and the assumption that games would follow was widespread. Earlier attempts had failed on bandwidth and cost, but data centers were now close enough to most users for the latency budget to work. Console makers and publishers were all experimenting with streaming, mostly as an addition to something else.

What it changed

The service closed a little over three years after opening, with purchases refunded, and became the standard cautionary example about digital libraries held on someone else's terms. Streaming itself continued as a feature inside subscriptions rather than a product of its own, which proved the more durable arrangement. The underlying technology remains in use for trials, portable play and low-end devices.