T-Mobile ends up dead last in task success while AT&T’s Nokia decision is a winner

50% of AT&T’s wireless traffic runs over Open RAN 

Unlike Closed RAN, which uses hardware and software from a single vendor, Open RAN can mix hardware and software from different vendors. For example, an Open RAN can include a radio unit from Company A and software from Company B. 50% of AT&T‘s traffic currently runs over an Open RAN system and AT&T has a goal to get that to 70% by the end of this year.

Why use Open RAN? Well, it allows carriers to lower costs over time because the vendors try to get the business from the wireless provider by lowering their costs. It also allows the carriers to use general-purpose servers instead of proprietary machines. It also allows for the integration of third-party AI tools to help with traffic optimization, predictive maintenance, energy efficiency, and spectrum management.

Why using Open RAN has made AT&T a hypocrite

So where does the hypocrisy come in? AT&T has used its embrace of Open RAN to slowly drop Nokia gear from its network, replacing the latter’s equipment with gear built by Ericsson. From the first half of 2025 to the second half, Ericsson equipment made up 73.84% of AT&T‘s Macro sites (the “backbone” of a wireless network) increasing to 84.21% in the second half of the year, with Nokia equipment making up the balance.

Despite moving to open up its Macro sites by using Open RAN, the shift has actually resulted in AT&T increasing its dependence on one particular networking vendor, Ericsson. The latter has become AT&T‘s “sole main supplier.”

RootMetrics’ testing shows AT&T at the top and T-Mobile at the bottom

1. Ericsson’s AT&T markets
2. Ericsson’s Verizon markets
3. Samsung’s Verizon markets
4. Nokia’s AT&T markets
5. Nokia’s T-Mobile markets
6. Ericsson’s T-Mobile markets

Keep in mind that these rankings are based on the performance of equipment sourced from individual vendors. T-Mobile finished at the bottom in reliability and task success. The carrier still leads in Median Download speeds (the only carrier to hit over 100Mbps in all 125 major markets), Mbps, and 5G availability (95.2% vs. 81% for AT&T).

AT&T Store. | Image by AT&T

While overall scores were not revealed, Mike Dano, lead industry analyst at Ookla, told Light Reading about the test results, “It’s a composite of all the different things that you can see about the performance of that particular vendor’s equipment, and, to be really clear, it’s looking at how is the phone performing.” According to Dano, the results were very close, adding, “It’s not like one is way out ahead.”

Ookla’s lead analyst says the age of the gear used might have hurt T-Mobile’s results

It appears that the age of the equipment used was a factor in the rankings. T-Mobile‘s reputation for network quality makes it surprising that the carrier’s Nokia-powered RAN sites and its RAN sites using Ericsson gear finished fifth and sixth out of six ranking positions. 

Ookla’s Dano pointed out something that we mention often: T-Mobile‘s use of the 2.5GHz mid-band spectrum it acquired in the Sprint transaction gave it an early lead in 5G. However, it is possible that when it comes to RAN performance, the older equipment used by T-Mobile was responsible for its last-place rankings.

This is a possibility and a reason why it might be true comes from a comment made by Per Narvinger, the head of Ericsson’s mobile networks business group. Narvinger pointed out that Ericsson is now using semiconductors built on more advanced process nodes, such as 3nm, for its 5G products.