FRAMINGHAM 7 MARCH 2011 - The iPhone 4 on AT&T's network downloaded data twice as fast, on average, as the iPhone on Verizon Wireless, according to thousands of recent field tests in five U.S. cities conducted by Metrico Wireless, an independent mobile device performance evaluation company.
Metrico's tests looked at several other variables, however, and found the average Web page load time was nearly the same on iPhone 4 onboth networks, a Metrico official said.
Also, when the iPhone 4 was in a moving vehicle, the AT&T model successfully finished about 10% more download session than Verizon's. But when the iPhones were stationary, the Verizon iPhone had a 10% better success rate in uploading data than the AT&T iPhone, Metrico said.
Metrico did not reveal actual time measurements for any of its results, including the data downloads and uploads or Web page loading times, prior to publication of its full study.
Some of Metrico's findings are at odds with several smaller spot reports done last month that found Verizon's iPhone performed better on several criteria. But a Metrico official noted that those quick studies were based on only a handful of test samples, often in a single city, including San Francisco, where AT&T's coverage for the iPhone has been consistently criticized.
AT&T admitted more than a year ago that it had network problems in downtown San Francisco and Manhattan and had begun upgrades.
"Metrico's results are based on both stationary and mobile [while the user is moving] scenarios at [multiple] locations over five markets and with sample counts in the thousands," said Rich McNally, vice president of information products for Metrico.
Metrico did not conduct the study for either AT&T and Verizon, but has compiled the data in a $4,950 report that any carrier can purchase, McNally said. Despite widespread concerns about AT&T's network in downtown San Francisco, Metrico did not include that city in its study, although it did include New York City. The other four locations were Seattle, Baltimore-Washington, Chicago and Dallas.
The earlier spot reports may have shown different results because they may have been based on relatively fewer users on a single Verizon cell tower, analysts said. That's because fewer customers would have subscribed in the early days after Verizon began selling the iPhone on Feb. 10. Fewer customers on a single cell tower usually means better performance.
For the iPhone study, Metrico ran more than 10,000 Web page downloads in the five cities, as well as more than 2,000 data download and upload tests and nearly 4,000 voice calls. Voice calls on the AT&T iPhone have often been the biggest point of criticism with AT&T's network, but Metrico would not reveal which carrier had the better overall voice performance.
Sign up for Computerworld eNewsletters.