Cloud Gaming
OnLive
OnLive streamed console-quality games from remote data centers years before home internet could deliver them lag-free. Real-world latency and ruinous infrastructure costs collapsed the company in 2012; it was fire-sold for $4.8M after a valuation near $1.8B.
- Company
- OnLive
- Started
- 2010
- Ended
- 2015
- Valuation at peak vs. fire-sale price
- $1.8B → $4.8M
- Collapse speed
- Rapid
- Preventability
- Medium
- Lesson transfer
- Industry-wide
- Last reviewed
- 2026-07-22
Narrative
The story
The ambition
OnLive promised to make expensive gaming hardware obsolete. Founded by Steve Perlman, it streamed demanding games rendered on servers in its data centers to any screen — a PC, a TV, a tablet, or a tiny microconsole — for a monthly fee. It was, in 2010, a genuinely futuristic idea: your console lived in the cloud.
The rise
The service launched in the United States on June 17, 2010, running on five North American data centers, and drew real excitement as it added top-tier titles. For a moment OnLive looked like the future of gaming arriving early.
The cracks
Physics and economics both pushed back. Over ordinary consumer connections, latency ran around 150 milliseconds — well above the sub-80ms OnLive had promised — and video-compression artifacts marred fast-action games, leaving the experience hovering at the edge of playable. And the model was upside-down: running always-on server capacity cost far more than a small base of subscribers could ever cover.
The collapse
On August 17, 2012 — just over two years after launch — OnLive laid off its staff and sold its assets through an assignment for the benefit of creditors, an insolvency process that wiped out employee equity. The company that had once carried a valuation near $1.8 billion went for a reported $4.8 million.
The aftermath
A successor entity kept the lights on until Sony bought OnLive's patents in April 2015 and shut the service down on April 30, 2015. The technology fed into Sony's own streaming ambitions — and cloud gaming, the idea OnLive was too early for, returned years later on faster networks.
The lessons
Being right too early is indistinguishable from being wrong. OnLive's vision was sound and later vindicated, but it launched into a world whose bandwidth and latency could not deliver the experience, and it carried the cost of always-on infrastructure with too few subscribers to pay for it. Timing and unit economics, not vision, decide whether a futuristic product survives to see its moment.
Causal timeline
Failure Anatomy
- 2010-06
Gaming from the cloud
OnLive launched in the US on June 17, 2010, streaming demanding games from its data centers to PCs, TVs, tablets, and a microconsole. [1]
- 2010
Physics pushes back
Over consumer connections latency ran ~150ms (versus a promised sub-80ms) and compression marred fast games, leaving play at the edge of acceptable. [2]
Bad timing - 2012
Upside-down economics
Running always-on server capacity cost far more than a small subscriber base could cover. [3]
Unsustainable economics - 2012-08
Collapse and fire sale
On August 17, 2012, OnLive laid off its staff and sold its assets via an insolvency process for a reported $4.8M, against an earlier ~$1.8B valuation. [4]
Unsustainable economics - 2015-04
Patents to Sony, service dark
A successor ran the service until Sony bought OnLive's patents in April 2015 and shut it down on April 30, 2015. [5]
Structured analysis
What Went Wrong
Root causes
Ahead of the network. Consumer internet in 2010–2012 could not deliver cloud gaming lag-free; real-world latency (~150ms vs a promised sub-80ms) and compression left the experience at the edge of playable. [2]
Contributing factors
Always-on cost, too few subscribers. Running data-center capacity cost far more than OnLive earned from its small subscriber base, making the business unsustainable. [3]
Immediate trigger
The company collapses. Unable to cover its costs, OnLive laid off its staff and sold its assets through an insolvency process in August 2012. [3] [4]
Visible symptoms
Too few paying subscribers. OnLive's subscriber revenue never approached the cost of the infrastructure required to serve it. [3]
Warning signs
Latency at the edge of playable. Reviewers found real-world latency and compression pushed OnLive to the boundary of acceptable lag, especially in fast-action games. [2]
Affected groups
Evidence
Claims & sources
Every numbered marker in the analysis links to the claim it rests on, and each claim to its sources.
- [1]
OnLive launched its cloud-gaming service in the United States on June 17, 2010, streaming games rendered in remote data centers to PCs, TVs, tablets, and a microconsole for a monthly fee.
- [2]
In practice, consumer-internet latency ran around 150 milliseconds — versus the sub-80ms OnLive had promised — and compression artifacts pushed the experience to the edge of playable, especially in fast-action games.
- [3]
Running always-on data-center capacity cost far more than OnLive earned from its small subscriber base, and the business proved unsustainable.
Moderate Reported explanation OnLive — Wikipedia Source: OnLive Found A Buyer, Cleaned House To Reduce Liability Prior To Acquisition - [4]
On August 17, 2012 — two years after launch — OnLive laid off its staff and sold its assets through an assignment for the benefit of creditors, reportedly for about $4.8 million despite an earlier valuation near $1.8 billion.
- [5]
A successor company kept the service running until Sony bought OnLive's patents in April 2015 and shut the service down on April 30, 2015.
Sources
OnLive — Wikipedia
Wikipedia
Source: OnLive Found A Buyer, Cleaned House To Reduce Liability Prior To Acquisition
TechCrunch · 2012-08-17