T-Mobile during its morning media event today in New York City made further indirect allusions to possible iPhone plans. Carrier officials when asked about the option in the post-exclusive US environment didn't confirm or deny plans but instead told the media to "ask Apple." The need to adapt to T-Mobile's 1,700MHz 3G band was cited as one obstacle, although T-Mobile made allusions to 3G chipsets having "evolved" to support many more cellular frequencies and thus covering both AT&T and T-Mobile users.

"We're not part of the [iPhone] chipset today," president Philipp Humm said. "But we have chipsets which support five, or up to 10 spectrum bands in the market, so we should expect there will be more degrees of freedom going forward."

Four- or five-band 3G chipsets are still rare in the industry. Apple uses a pentaband radio in the iPhone 4 but uses the fifth band for an NTT DoCoMo frequency that it has yet to officially support, not T-Mobile. Nokia has been the most common adopter of pentaband hardware and has equipped at least the E7 and N8 with 3G that can work on AT&T, T-Mobile USA and most international carriers.

The language, while not a confirmation, is significant and may be a clue as to ongoing talks. Verizon used very similar language to effectively acknowledge that a CDMA iPhone was coming soon without either pre-announcing the device or setting too many expectations.

T-Mobile's best opportunity may be the iPhone 5, which could have a dual-mode chipset with both CDMA and GSM support. If it included both 850MHz and 1,700MHz support on 3G, it could let Apple produce a single phone that supports virtually all carriers.

The fourth-place US carrier has a relatively small user base at less than half what Verizon offers, but Apple may choose to support it both to put pressure on Android and to address other carriers. In Canada, Videotron claimed it would get the iPhone using the same frequency as T-Mobile. Other recent carrier startups such as Wind Mobile, Mobilicity and Public Mobile are using the same technology.