dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
121
share rss forum feed

rradina

join:2000-08-08
Chesterfield, MO

Bandwith to where?

I linked to and read the "specific testing approach" page from this thread's article. However, I couldn't figure out what test site they used to derive the numbers.

I have three speed tests on my iPhone: DSLReports, Ookla and SpeedTest. DSLReports and Ookla let me choose the destination server. Due to Internet "weather" and packet routes, I have found that the destination server can have a huge impact on performance. For instance, I'm located in St. Louis and sometimes I achieve significantly higher performance (1.3Mbps vs. 500Kbps) by choosing a Chicago-based server over a St. Louis-based server.

While I have no problem with them using a laptop for the tests, they claim that they use it over a smart phone because of the "high processing requirements" of the test application. That makes me pause because I don't think a bandwidth test should require substantial processing power. For nearly ten years DSLReports has offered a Java benchmark test. As a "virtual" language Java requires more resources than a native application. Still, I've never known the DSL Reports Java benchmark to require vast quantities of memory or processing power... even on PCs from five or six years ago.

Bottom line -- I think we need more information about how the tests were conducted to draw meaningful conclusions. As my graduate-level statistics professor always said, "In descending order of trust, there are liars, damn liars and statisticians."