Sprint EV-DO Rev. A: The 2006 Mobile Broadband Upgrade

In late August 2006, Sprint began talking publicly about a network upgrade called EV-DO Revision A, a faster successor to the EV-DO technology it had been using since 2005. Mobility Today covered the story that same week, in a short post reporting that a Sprint EV-DO Rev. A card was already available and shipping, according to enthusiast forums, even though the faster network itself was not live anywhere yet. This article revisits what Sprint actually announced and rolled out that year, what Mobility Today reported at the time, and how Rev. A's speeds compared, in theory and in practice, to the network it replaced.

What Was Sprint EV-DO Rev. A?

EV-DO, short for Evolution-Data Optimized, was a mobile broadband technology built on the CDMA2000 family used by carriers like Sprint and Verizon Wireless. Revision A was an upgraded version of the original EV-DO standard, generally called Rev. 0, developed to improve upload speeds, reduce latency, increase network capacity, and better support real-time applications such as voice calls carried over data. It was not a brand-new network built from scratch — it was a software and hardware upgrade layered onto the EV-DO infrastructure carriers already had in place.

How EV-DO Fit Into Mobile Broadband Before LTE

Long before LTE became the standard way to get fast data on a phone or laptop, wireless carriers offered mobile broadband through add-on hardware: PC Card and USB modems for laptops, and increasingly data-capable phones. CDMA carriers, including Sprint, Verizon, and Alltel, built their mobile broadband around the EV-DO family. GSM-based carriers, including Cingular, later AT&T, and T-Mobile, took a different technical path built around UMTS and HSDPA. Both approaches were loosely marketed as third-generation, or 3G, service, even though the underlying technologies were not interchangeable or directly compatible with each other.

EV-DO Rev. 0 vs Revision A: What Actually Changed

Rev. 0 was the version of EV-DO Sprint had already deployed starting in 2005. It offered reasonable download performance for the era but comparatively weak upload speeds, higher latency, and limited support for real-time or interactive applications. Revision A kept the same basic downlink-focused design philosophy but meaningfully improved the uplink, reduced latency, added network capacity, and was engineered to better handle applications like voice-over-IP that need more consistent, lower-delay data flow. The table below summarizes the commonly cited peak theoretical rates alongside the typical real-world speeds reported around the time of Sprint's rollout.

These figures are compiled from multiple contemporary industry and press sources from 2006 and should be read as approximate ranges, not guaranteed numbers for any specific device, market, or point in the rollout.

Sprint's Power Vision Network

Sprint marketed its EV-DO-based mobile broadband and multimedia services under the Power Vision brand. Sprint's first commercial EV-DO deployment began in 2005, and by early 2006 the company described the underlying network as reaching more than 150 million people, making it one of the largest mobile broadband footprints in the country at that point. The Rev. A upgrade was layered onto this existing Power Vision footprint rather than representing an entirely new buildout, which is part of why Sprint could talk about turning it on market by market rather than launching a brand-new nationwide network overnight.

The August 2006 Rev. A Announcement

On August 25, 2006, Sprint indicated that Rev. A-capable cards would soon be available, describing them as backward compatible with the existing Rev. 0 network. Four days later, on August 29, 2006, Sprint and Novatel Wireless formally announced commercial availability of the Novatel Wireless S720, described at the time as the first commercially available EV-DO Revision A-capable device in the country. Sprint said it planned to begin turning on the Rev. A network itself during the fourth quarter of 2006, targeting coverage of more than 40 million people by the end of that year, as part of a broader mobile broadband footprint goal of roughly 200 million people across roughly 220 metropolitan areas. Sprint also indicated that once compatible hardware was used in a market with Rev. A actually turned on, customers could expect download speeds averaging somewhere in the 450-850 Kbps range, with substantially faster uploads than Rev. 0 had offered. It is worth being precise here: this was an announcement of a target and of hardware availability, not a claim that the faster network was already live anywhere in August 2006.

Peak Theoretical Speeds vs Typical Real-World Speeds

It is easy, looking back at coverage from this period, to blur two very different kinds of numbers together, so it is worth separating them clearly. EV-DO Revision A's peak theoretical rate, the maximum the specification could support under ideal conditions, was approximately 3.1 Mbps downstream and approximately 1.8 Mbps upstream. Those are engineering ceiling figures, not a description of what an average customer actually experienced while using the network.

The speeds Sprint and contemporary press actually described for typical usage were much lower: roughly 450-850 Kbps downstream and roughly 300-400 Kbps upstream for a customer with compatible hardware in a market where Rev. A had been activated. That real-world range represents a fraction of the theoretical peak, and it still varied depending on distance from a cell site, how many other users were sharing the same tower, local terrain, and the specific device being used. This gap between theoretical peak and typical real-world performance was not unique to Rev. A — it applied to Rev. 0 before it and to competing technologies like HSDPA as well — but it is an important distinction to keep in mind any time a 2006-era 3G technology's speed is quoted without further context.

Sprint's Actual Rollout: San Diego and the First Markets

Announcing a target and turning on a live network are two different things, and the timeline here illustrates that gap well. Sprint's Rev. A network actually went live for the first time in San Diego around late October 2006, roughly two months after the August card and coverage-target announcements. Contemporary coverage from that period described Sprint planning to follow San Diego with roughly twenty additional markets before the end of 2006, with cities named in various contemporary reports including places like Baltimore, Boston, Denver, Kansas City, Las Vegas, Los Angeles, New York, Philadelphia, San Francisco, Seattle, and Washington, D.C. Exact activation dates for each of those individual markets varied across contemporary sources and are not treated here as a single confirmed master schedule — San Diego as the first live Rev. A market is the detail best corroborated across independent 2006 coverage.

Nationwide Expansion Through 2007

Sprint's Rev. A rollout continued well past the initial 40-million-person, end-of-2006 target. Coverage from the following year described Sprint aiming to have the bulk of its network upgraded to Rev. A by the third quarter of 2007, with reporting toward the end of 2007 describing Sprint's Rev. A-upgraded footprint as covering more than 200 million residents. In other words, the 40-million figure from August 2006 described only the initial phase of the rollout, not the eventual scale of the completed upgrade roughly a year later.

Rev. A-Capable Hardware

Several pieces of hardware were associated with Sprint's early Rev. A push. The table below lists the devices most consistently referenced in contemporary 2006 coverage.

None of these specific model names should be assumed to have appeared in Mobility Today's own original coverage, discussed in the next section — the recovered version of that article referred only generically to a new Sprint EV-DO Rev. A card, without naming a manufacturer or model.

What Mobility Today Actually Reported in 2006

The original Mobility Today article was successfully recovered through the Wayback Machine's archived snapshots of mobilitytoday.com. It was a short news post titled Sprint releases EVDO Rev A Card, filed under the site's Wireless News section, written by David Ciccone, and posted on Friday, August 25, 2006 — the same week as Sprint's broader Rev. A announcements. In the post itself, Ciccone credited a tip from a reader called Mick.

The post reported that, according to the crew at EVDO forums, a new Sprint EV-DO Rev. A card was already available and shipping. It explicitly noted that the card was backward compatible with the existing EV-DO Rev. 0 network and that customers would not get faster speeds immediately — only once Sprint actually turned on the Rev. A network in a given area would existing cardholders see faster performance, automatically and without needing to buy new hardware. The post also mentioned that Rev. A would bring much lower latency and would be able to handle VoIP. On pricing, the article listed the card at $99.99 with a two-year Sprint contract, and $174.99 with a one-year contract — figures close to, though not identical in every detail to, the pricing cited by some other contemporary outlets covering the same announcement.

What the recovered article did not do is just as notable. It did not name a specific device manufacturer or model. It did not cite specific download or upload speed figures in Kbps or Mbps. It did not mention San Diego or any other specific launch city. And it did not include any indication that Mobility Today had tested Rev. A hardware or measured speeds itself — the piece reads as a short news blurb summarizing information from EVDO enthusiast forums, not a hands-on review. A reader comment beneath the post asked when Verizon would respond with pricing of its own, reflecting the competitive framing much of the era's coverage used.

Laptops, Data Cards, and Business Use Cases

Before LTE and before smartphone tethering was common, mobile broadband cards and USB modems were one of the main ways to get a laptop online away from Wi-Fi. Road-warrior professionals, field technicians, and traveling salespeople were common users of this kind of hardware. Improved upload speed specifically mattered for tasks like sending larger email attachments, uploading photos or documents from the field, and early forms of mobile video conferencing — all things that Rev. 0's typical uploads of roughly 50-70 Kbps handled poorly, and that Rev. A's substantially faster typical uploads were positioned to improve.

Sprint vs Verizon EV-DO and HSDPA-Based GSM Networks

Verizon Wireless operated on the same underlying CDMA2000 EV-DO technology family as Sprint, but its own Rev. A rollout came later. Verizon commercially launched its Rev. A network in early February 2007, several months behind Sprint's October 2006 San Diego debut, with initial markets reported to include areas around Boston, Richmond, Chicago, Salt Lake City, and parts of Florida.

Meanwhile, GSM-based carriers took a different technical path entirely. Cingular, later rebranded AT&T, commercially launched HSDPA service in December 2005 across roughly sixteen U.S. markets, ahead of Sprint's Rev. A rollout, built on UMTS rather than CDMA2000. HSDPA's theoretical ceiling was considerably higher on paper than EV-DO Rev. A's, though real-world throughput reported during this general period for both technologies tended to land in a broadly comparable few-hundred-Kbps range for typical users. Neither approach clearly dominated the other in everyday use during this era — actual customer experience depended heavily on local network buildout progress and device support at any given time, and this article is not declaring either technology the winner.

What Replaced EV-DO

EV-DO Rev. A was eventually followed by EV-DO Rev. B, a further evolution offering higher throughput in more limited deployments, largely used by a smaller number of operators. More broadly, though, the entire CDMA2000 EV-DO line, along with the competing HSDPA/UMTS path used by GSM carriers, was eventually superseded by LTE, a shared fourth-generation standard that both Sprint and Verizon, along with AT&T and T-Mobile, ultimately adopted. LTE offered dramatically higher real-world throughput and lower latency than any EV-DO revision and gradually made the older CDMA-versus-GSM data-technology split far less relevant at the network level. It bears repeating clearly: EV-DO Revision A is not LTE — it is a distinct, earlier CDMA2000-based 3G technology that predates LTE by several years.

FAQ

What was Sprint EV-DO Rev. A?

An upgraded version of Sprint's EV-DO mobile broadband technology, built on the same CDMA2000 family as the original Rev. 0 network, designed to improve upload speeds, reduce latency, add network capacity, and better support real-time applications like VoIP.

How fast was EV-DO Rev. A?

Its peak theoretical rate was approximately 3.1 Mbps downstream and 1.8 Mbps upstream, but typical real-world speeds reported at the time were much lower — roughly 450-850 Kbps downstream and roughly 300-400 Kbps upstream for a customer with compatible hardware in an upgraded market.

Was Rev. A faster than Rev. 0?

Yes, particularly on the upload side. Rev. 0's typical real-world uploads were roughly 50-70 Kbps, compared with roughly 300-400 Kbps typical for Rev. A, along with meaningfully lower latency.

When did Sprint launch EV-DO Rev. A?

Sprint announced Rev. A-capable hardware and a fourth-quarter 2006 rollout target in late August 2006, then actually turned on its first live Rev. A network in San Diego around late October 2006, with additional markets following into 2007.

Which devices supported Sprint's early Rev. A network?

Contemporary 2006 coverage most consistently pointed to the Novatel Wireless S720 as the first available Rev. A card, followed later that year by the Novatel U720 USB modem, the Pantech PX-500, and the Sierra Wireless AirCard 595.

Did Sprint phones use EV-DO Rev. A?

Rev. A was used across Sprint's broader device lineup over time as compatible hardware became available, though this article focuses specifically on the initial 2006 laptop-card rollout rather than cataloguing every phone that later supported Rev. A.

Was EV-DO considered 3G?

Yes, EV-DO, including both Rev. 0 and Rev. A, was marketed and generally categorized as third-generation, or 3G, mobile technology during this period.

How did EV-DO Rev. A compare to HSDPA?

They were different underlying technologies — EV-DO built on CDMA2000, HSDPA built on UMTS/GSM. HSDPA had a considerably higher theoretical ceiling on paper, but typical real-world speeds for both technologies during this general period were broadly comparable, in a few-hundred-Kbps range for most users.

What replaced EV-DO?

LTE eventually replaced EV-DO, along with HSDPA/UMTS on the GSM side, as the shared fourth-generation standard adopted across Sprint, Verizon, AT&T, and T-Mobile.

Sprint's EV-DO Rev. A upgrade is a useful snapshot of how mobile broadband actually rolled out in the mid-2000s: an announcement in August, a genuine live network weeks later in a single city, and a gradual, market-by-market expansion that took roughly a year to approach nationwide scale. Mobility Today's own contribution to that story, recovered here from the Wayback Machine, was modest but accurate as far as it went — a short, forum-sourced news post about a card shipping ahead of the network it was built for, without speed tests, city names, or hardware specifics of its own.