/
Navigation
Chronicles
Browse all articles
Explore
Semantic exploration
Research
Entity momentum
Nexus
Correlations & relationships
Story Arc
Topic evolution
Drift Map
Semantic trajectory animation
Posts
Analysis & commentary
Pulse API
Tech news intelligence API
Browse
Entities
Companies, people, products, technologies
Domains
Browse by publication source
Handles
Browse by social media handle
Detection
Concept Search
Semantic similarity search
High Impact Stories
Top coverage by position
Sentiment Analysis
Positive/negative coverage
Anomaly Detection
Unusual coverage patterns
Analysis
Rivalry Report
Compare two entities head-to-head
Semantic Pivots
Narrative discontinuities
Crisis Response
Event recovery patterns
Connected
Search: /
Command: ⌘K
Embeddings: large
TEXXR

Chronicles

The story behind the story

← → days · ↑ ↓ browse · Enter similar · o open

Skype - Is 10,000 Page Faults per Second normal?

I am writing this as a follow-up to yesterday's Tech Talk about Skype.  I mentioned this issue then, and I have received a number of requests for additional information about it.  There have been some new developments today, as well.

ZDNet J.A. Watson

Context & Ripple Effects

Two years after Paul Kedrosky's Fear of a Skype Planet framed Skype's peer-to-peer design as an internet-scale phenomenon, the service has spent late 2007 pushing outward fast: the [[a:none|3 Skypephone]] handset launch, a MySpaceIM voice-chat partnership, a Facebook app launcher for 18 games — alongside customer backlash over stripped London 0207 numbers. Growth, not stability, has been the story.

ZDNet's Tech Talk now turns attention inward, flagging that the Windows client generates roughly 10,000 page faults per second, with readers requesting more detail and new developments reported today. It matters because Skype's P2P model treats ordinary users' machines as part of its infrastructure, so client-side overhead is not a private matter — it is distributed across every endpoint.

First-order effects

  • Windows users running Skype carry a measurable memory-paging load on their own machines, since the P2P design uses participant endpoints rather than Skype-operated servers to carry the network.

Second-order effects

  • Partners who just embedded Skype in constrained environments — the 3 Skypephone handset and MySpaceIM — inherit whatever resource-footprint questions this scrutiny surfaces, and mobile hardware leaves far less headroom for paging churn than desktops.

Third-order effects

  • If endpoint-based infrastructure keeps scaling while client devices grow lighter and more mobile, the economics of offloading network work onto users' machines come into question — an architectural pressure point that would eventually force a choice between P2P purity and server-side control.

The trend: Consumer VoIP's peer-to-peer model, which shifts infrastructure costs onto end-user devices, faces growing technical scrutiny precisely as services like Skype extend onto phones and third-party clients.