Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : The *NEW* Frank Coluccio Technology Forum -- Ignore unavailable to you. Want to Upgrade?

To: elmatador who wrote (5242)	3/24/2002 10:10:00 AM
From: Rich Wolf	Respond to of 46821

re capacity utilization, the question is still unanswered as to the appropriate metric. The answer to that depends upon the end-user metric of interest (e.g., if data networks must carry voice, then QoS requiring little latency requires planning for peak usage, like voice networks, or using overlays that prioritize traffic and allow for sustaining high QoS on the data that needs it). Your quote from ML: << Merrill Lynch claimed in two separate reports issued in March and August that network utilization levels remain below historical levels. In its first report it put average optical network utilization at just 2.7% but then revised in the light of industry criticisms that this was a poor measure of network planning. With its revised figure of 6.4% of capacity in use, Merrill Lynch concluded that capacity had increased 60 times in the past 10 years while network utilization had increased just seven times. >> Commentary on the 'revised' ML metric: Message 17228911 << . “Telecom networks, like highways, must be sized to handle demands at peak periods,” said Janice Haber, OFS VP of systems engineering and market development. She pointed out that the "peak-to-average" rate of data networks is much higher than it is in voice-centric ones -- typically ten times higher, in order to accommodate high-traffic bursts. Using the 10:1 peak-to-average rate that Haber said data-driven networks require, Merrill Lynch's estimate that 6.4 percent of capacity was being used at the end of 2001 translates into 64 percent peak usage. This, she said, is dangerously close to the 70 percent threshold beyond which customer service can no longer be guaranteed. >> ====================== It seems the truth must be in between, unfortunately the bound 'looks' tight only on a logarithmic scale, actually quite loose <GGG>. If I'm a network planner, my cost analysis must balance cost of replacing routers etc so I can provide tiered service (to sustain QoS for the part of the peak traffic that needs it, while allowing much larger delays for data traffic) vs treating all data as equal and building to handle peak vs not carrying voice traffic on the data networks vs under-utilizing the bandwidth capacity of the fiber by relying on older switching tech etc etc Thoughts?