Ford is overhauling how it designs electric vehicles, shrinking battery packs and dramatically reducing wiring in an effort to bring EV prices closer to those of traditional gasoline-powered cars. The strategy marks a major shift in how the automaker approaches electrification, focusing less on maximizing battery size and more on improving overall efficiency and cost structure.

A New EV Platform Focused on Efficiency

For years, automakers have chased longer driving range by installing larger batteries. But battery packs can account for up to 40% of an EV’s cost and a significant portion of its weight. Ford believes simply adding more battery capacity is not the answer to making electric vehicles affordable for mainstream buyers.

Instead, the company is developing a new Universal Electric Vehicle (UEV) platform that prioritizes efficiency at every level. Engineers are cutting thousands of feet of wiring, reducing parts count, and simplifying electrical architecture. In some cases, Ford has reduced wiring length by roughly 4,000 feet, trimming weight and material costs in the process.

The company is also moving toward zonal electrical systems and 48-volt architectures, consolidating components and improving energy management. By improving aerodynamics and reducing overall vehicle weight, Ford says it can maintain competitive range even with smaller battery packs.

Ford’s move comes at a critical time

While EV adoption is growing, high upfront costs remain one of the biggest barriers to mass-market acceptance. Many consumers are hesitant to pay a premium over gasoline vehicles, especially when interest rates and economic uncertainty weigh on purchasing decisions.

By shrinking batteries and reducing manufacturing complexity, Ford aims to lower sticker prices without sacrificing performance or usability. The company has publicly discussed targeting a mid-size electric pickup starting around $30,000 – a price point that would put it in direct competition with traditional gas-powered trucks.

This strategy could help close the cost gap between EVs and internal combustion vehicles, accelerating adoption without relying heavily on government incentives. Smaller batteries also mean lighter vehicles, which can improve efficiency and handling while reducing strain on supply chains for critical battery materials.

For consumers, Ford’s approach could translate into more affordable electric vehicles that still deliver practical range and everyday usability. Instead of chasing 400-mile battery capacities that many drivers rarely need, Ford is focusing on optimizing efficiency so that smaller packs go further.

That means drivers could see EVs priced closer to comparable gas vehicles while still benefiting from lower fuel and maintenance costs over time. Reduced complexity may also improve long-term reliability and simplify repairs.

In short, the shift is about delivering value – not just range numbers

Ford plans to introduce its first vehicles built on the new UEV platform in 2027, beginning with a mid-size electric pickup. Additional models across different segments are expected to follow.

If successful, this engineering rethink could reshape how the industry approaches EV design. Rather than competing on ever-larger batteries, automakers may pivot toward smarter architecture, weight reduction, and system simplification as the path to affordability.

For Ford, the message is clear: the future of electric vehicles may not be bigger – it may be leaner, lighter, and significantly less expensive.

from the no-such-thing-as-a-free-surveillance-tech dept

Surveillance technology vendors, federal agencies, and wealthy private donors have long helped provide local law enforcement “free” access to surveillance equipment that bypasses local oversight. The result is predictable: serious accountability gaps and data pipelines to other entities, including Immigration and Customs Enforcement (ICE), that expose millions of people to harm.

The cost of “free” surveillance tools — like automated license plate readers (ALPRs), networked cameras, face recognition, drones, and data aggregation and analysis platforms — is measured not in tax dollars, but in the erosion of civil liberties. 

The collection and sharing of our data quietly generates detailed records of people’s movements and associations that can be exposed, hacked, or repurposed without their knowledge or consent. Those records weaken sanctuary and First Amendment protections while facilitating the targeting of vulnerable people.   

Cities can and should use their power to reject federal grants, vendor trials, donations from wealthy individuals, or participation in partnerships that facilitate surveillance and experimentation with spy tech. 

If these projects are greenlit, oversight is imperative. Mechanisms like public hearings, competitive bidding, public records transparency, and city council supervision aid to ensure these acquisitions include basic safeguards — like use policies, audits, and consequences for misuse — to protect the public from abuse and from creeping contracts that grow into whole suites of products. 

Clear policies and oversight mechanisms must be in place before using any surveillance tools, free or not, and communities and their elected officials must be at the center of every decision about whether to bring these tools in at all.

Here are some of the most common methods “free” surveillance tech makes its way into communities.

Trials and Pilots

Police departments are regularly offered free access to surveillance tools and software through trials and pilot programs that often aren’t accompanied by appropriate use policies. In many jurisdictions, trials do not trigger the same requirements to go before decision-makers outside the police department. This means the public may have no idea that a pilot program for surveillance technology is happening in their city. 

In Denver, Colorado, the police department is running trials of possible unmanned aerial vehicles (UAVs) for a drone-as-first-responder (DFR) program from two competing drone vendors: Flock Safety Aerodome drones (through August 2026) and drones from the company Skydio, partnering with Axon, the multi-billion dollar police technology company behind tools like Tasers and AI-generated police reports. Drones create unique issues given their vantage for capturing private property and unsuspecting civilians, as well as their capacity to make other technologies, like ALPRs, airborne. 

Functional, Even Without Funding 

We’ve seen cities decide not to fund a tool, or run out of funding for it, only to have a company continue providing it in the hope that money will turn up. This happened in Fall River, Massachusetts, where the police department decided not to fund ShotSpotter’s $90,000 annual cost and its frequent false alarms, but continued using the system when the company provided free access. 

In May 2025, Denver’s city council unanimously rejected a $666,000 contract extension for Flock Safety ALPR cameras after weeks of public outcry over mass surveillance data sharing with federal immigration enforcement. But Mayor Mike Johnston’s office allowed the cameras to keep running through a “task force” review, effectively extending the program even after the contract was voted down. In response, the Denver Taskforce to Reimagine Policing and Public Safety and Transforming Our Communities Alliance launched a grassroots campaign demanding the city “turn Flock cameras off now,” a reminder that when surveillance starts as a pilot or time‑limited contract, communities often have to fight not just to block renewals but to shut the systems off.

 Importantly, police technology companies are developing more features and subscription-based models, so what’s “free” today frequently results in taxpayers footing the bill later. 

Gifts from Police Foundations and Wealthy Donors

Police foundations and the wealthy have pushed surveillance-driven agendas in their local communities by donating equipment and making large monetary gifts, another means of acquiring these tools without public oversight or buy-in.

In Atlanta, the Atlanta Police Foundation (APF) attempted to use its position as a private entity to circumvent transparency. Following a court challenge from the Atlanta Community Press Collective and Lucy Parsons Labs, a Georgia court determined that the APF must comply with public records laws related to some of its actions and purchases on behalf of law enforcement.
In San Francisco, billionaire Chris Larsen has financially supported a supercharging of the city’s surveillance infrastructure, donating $9.4 million to fund the San Francisco Police Department’s (SFPD) Real-Time Investigation Center, where a menu of surveillance technologies and data come together to surveil the city’s residents. This move comes after the billionaire backed a ballot measure, which passed in March 2025, eroding the city’s surveillance technology law and allowing the SFPD free rein to use new surveillance technologies for a full year without oversight.

Free Tech for Federal Data Pipelines

Federal grants and Department of Homeland Security funding are another way surveillance technology appears free to, only to lock municipalities into long‑term data‑sharing and recurring costs. 

Through the Homeland Security Grant Program, which includes the State Homeland Security Program (SHSP) and the Urban Areas Security (UASI) Initiative, and Department of Justice programs like Byrne JAG, the federal government reimburses states and cities for “homeland security” equipment and software, including including law‑enforcement surveillance tools, analytics platforms, and real‑time crime centers. Grant guidance and vendor marketing materials make clear that these funds can be used for automated license plate readers, integrated video surveillance and analytics systems, and centralized command‑center software—in other words, purchases framed as counterterrorism investments but deployed in everyday policing.

Vendors have learned to design products around this federal money, pitching ALPR networks, camera systems, and analytic platforms as “grant-ready” solutions that can be acquired with little or no upfront local cost. Motorola Solutions, for example, advertises how SHSP and UASI dollars can be used for “law enforcement surveillance equipment” and “video surveillance, warning, and access control” systems. Flock Safety, partnering with Lexipol, a company that writes use policies for law enforcement, offers a “License Plate Readers Grant Assistance Program” that helps police departments identify federal and state grants and tailor their applications to fund ALPR projects. 

Grant assistance programs let police chiefs fast‑track new surveillance: the paperwork is outsourced, the grant eats the upfront cost, and even when there is a formal paper trail, the practical checks from residents, councils, and procurement rules often get watered down or bypassed.

On paper, these systems arrive “for free” through a federal grant; in practice, they lock cities into recurring software, subscription, and data‑hosting fees that quietly turn into permanent budget lines—and a lasting surveillance infrastructure—as soon as police and prosecutors start to rely on them. In Santa Cruz, California, the police department explicitly sought to use a DHS-funded SHSP grant to pay for a new citywide network of Flock ALPR cameras at the city’s entrances and exits, with local funds covering additional cameras. In Sumner, Washington, a $50,000 grant was used to cover the entire first year of a Flock system — including installation and maintenance — after which the city is on the hook for roughly $39,000 every year in ongoing fees. The free grant money opens the door, but local governments are left with years of financial, political, and permanent surveillance entanglements they never fully vetted.

The most dangerous cost of this “free” funding is not just budgetary; it is the way it ties local systems into federal data pipelines. Since 9/11, DHS has used these grant streams to build a nationwide network of at least 79–80 state and regional fusion centers that integrate and share data from federal, state, local, tribal, and private partners. Research shows that state fusion centers rely heavily on the DHS Homeland Security Grant Program (especially SHSP and UASI) to “mature their capabilities,” with some centers reporting that 100 percent of their annual expenditures are covered by these grants. 

Civil rights investigations have documented how this funding architecture creates a backdoor channel for ICE and other federal agencies to access local surveillance data for their own purposes. A recent report by the Surveillance Technology Oversight Project (S.T.O.P.) describes ICE agents using a Philadelphia‑area fusion center to query the city’s ALPR network to track undocumented drivers in a self‑described sanctuary city.

Ultimately, federal grants follow the same script as trials and foundation gifts: what looks “free” ends up costing communities their data, their sanctuary protections, and their power over how local surveillance is used.

Protecting Yourself Against “Free” Technology

The most important protection against “free” surveillance technology is to reject it outright. Cities do not have to accept federal grants, vendor trials, or philanthropic donations. Saying no to “free” tech is not just a policy choice; it is a political power that local governments possess and can exercise. Communities and their elected officials can and should refuse surveillance systems that arrive through federal grants, vendor pilots, or private donations, regardless of how attractive the initial price tag appears. 

For those cities that have already accepted surveillance technology, the imperative is equally clear: shut it down. When a community has rejected use of a spying tool, the capabilities, equipment, and data collected from that tool should be shut off immediately. Full stop.

And for any surveillance technology that remains in operation, even temporarily, there must be clear rules: when and how equipment is used, how that data is retained and shared, who owns data and how companies can access and use it, transparency requirements, and consequences for any misuse and abuse. 

“Free” surveillance technology is never free. Someone profits or gains power from it. Police technology vendors, federal agencies, and wealthy donors do not offer these systems out of generosity; they offer them because surveillance serves their interests, not ours. That is the real cost of “free” surveillance.

Originally posted to EFF’s Deeplinks blog.

Filed Under: alpr, dhs, drones, facial recognition, grants, law enforcement, surveillance

Companies: flock, flock safety, lexipol, motorola

At first glance, the Google Pixel 10a looks almost identical to last year’s Pixel 9a.

It keeps the same familiar silhouette, the same 6.3-inch display size and the same Google Tensor generation powering everything under the hood.

So what’s actually changed?

Google hasn’t torn up the A-series blueprint. Instead, the Pixel 10a focuses on refinement — small but meaningful upgrades that improve day-to-day usability without dramatically altering the formula.

The most noticeable physical tweak is around the back. The Pixel 10a adopts a completely flat rear panel, with the camera bar sitting flush rather than protruding slightly. It’s a subtle adjustment, but it means the phone now lies flat on a table without wobbling, slides into pockets more easily and looks cleaner overall.

Advertisement

Up front, the 6.3-inch Actua display is now 11% brighter. Resolution and size remain unchanged, but the brightness bump should translate to better visibility outdoors and punchier HDR playback. It’s not headline-grabbing, but it’s one of the more practical upgrades.

Battery life figures are broadly similar, with Google quoting over 30 hours of standard use and up to 120 hours with Extreme Battery Saver enabled. Charging speeds, however, have been improved, now at 30W. It’s not transformative, but any reduction in time spent plugged into a wall is a positive at this price point.

Durability also gets a boost. The Pixel 10a now carries an IP68 rating for water and dust resistance, paired with upgraded Corning Gorilla Glass 7i and a refined aluminium frame. That’s a notable step forward and arguably one of the most tangible improvements this year, though it’s just playing catch-up to most of the £499/$499 competition.

Advertisement

Google has also leaned further into sustainability. The 10a features recycled cobalt for the first time in the A-series, alongside recycled copper, gold and tungsten, a 100% recycled aluminium frame and an 81% recycled plastic back. It doesn’t alter the in-hand feel, but it pushes the environmental credentials forward.

There are new colour options too, with Lavender, Berry, Fog and Obsidian replacing last year’s palette.

Advertisement

On the software side, Satellite SOS arrives on the A-series for the first time, allowing users to contact emergency services without Wi-Fi or mobile coverage. It’s a meaningful addition that elevates the phone’s safety credentials.

Camera hardware remains familiar, with a 48MP main sensor and a 13MP ultrawide. The upgrades come via software, including Auto Best Take, Camera Coach and expanded Add Me support. As ever, Google is relying on computational photography rather than new sensors.

Advertisement

The Pixel 10a runs on the Tensor G4, the same chipset as the Pixel 9a, enabling Gemini Live, Circle to Search, AI photo editing, Call Screen and Hold For Me. It’s capable, but not a generational leap for existing Pixel users.

Overall, this is a refinement update rather than a reinvention. A flatter design, brighter display, faster charging, tougher build, Satellite SOS and expanded AI tools form the core changes. Everything else remains reassuringly familiar.

For new buyers, it’s arguably the most complete A-series device yet. For Pixel 9a owners, the upgrades may feel incremental rather than essential. At £499/$499, though, Google is clearly betting that steady polish still wins in the mid-range market.

Advertisement

Advertisement

SQUIRREL_PLAYLIST_10208265

If you tuned in to China’s 2026 CCTV Spring Festival Gala looking for traditional lion dances and nostalgic tunes, you may have done a double-take when what greeted you was a squad of humanoid robots performing kung fu, synchronized moves, and comedy sketches with more precision than most of us manage during family reunions.

It was not just spectacle; it was industrial policy with flair

This year’s gala, which is China’s equivalent of the Super Bowl meets cultural heritage broadcast, featured everything from high-speed martial arts sequences to choreographed routines done by humanoid robots from leading local makers like Unitree Robotics, Galbot, MagicLab and Noetix. 

Watching them flip, strike poses, and dance isn’t just futuristic entertainment; it’s a deliberate signal about where Chinese tech wants to sit at the global table.

From props to protagonists

Just last year, robot appearances were charming but clunky: think awkward “dances” that needed human help to keep upright.

The 💜 of EU tech

The latest rumblings from the EU tech scene, a story from our wise ol’ founder Boris, and some questionable AI art. It’s free, every week, in your inbox. Sign up now!

Instead, this year, they executed complex actions; backflips, martial arts inspired routines, even comedic timing that was surprisingly sharp for machines. C

lips of the robots went viral almost immediately, flooding social platforms and dominating international tech feeds.

Chinese state media and commentators have leaned into the moment as proof of rapid progress in humanoid robotics, placing it squarely within China’s “AI+manufacturing” industrial ambitions.

While some viewers showered praise on the displays, others grumbled that the new lineup made the gala feel more like CES Lite than a cosy celebration of culture, and yes, robots “stealing the Year of the Horse thunder” is now a real complaint.

Why robots at the Gala?

The optics are unmistakable. China’s robotics sector, already responsible for a lion’s share of worldwide humanoid robot production, is eager to tell a simple story: we don’t just build hardware in factories, we animate it with AI brainsthat can perform with finesse.

Whether it’s a crowd-pleasing kung fu sequence or a scripted comedy routine, these robots have become ambassadors of technological prowess on very prime cultural real estate.

But let’s not overstate their practical chops just yet. Despite impressive stage routines, robotics experts point out that most current humanoids are still best at pre-programmed movements and lack true autonomous adaptability in unpredictable environments.

In other words, they’re amazing performers on cue, not yet ready to be your personal caregiver or industrial line worker without a lot more training and development.

Still, the spectacle accomplishes something important: it thrusts humanoid robots into the public imagination while signalling to investors, startups and rival nations that robot development is now prime time tech theatre, not just a research lab curiosity.

Whether you see this as a fun blend of culture and innovation or a high-stakes display of national tech strategy, China’s robotics presence at the gala is now part of a broader conversation about where AI embodied in hardware might go next.

It’s less about whether robots will “replace us” and more about how they’re being introduced into narratives that billions of people watch together, shaping perceptions and expectations about the future.

China’s robots danced their way into billions of screens this Lunar New Year, and they did it with style, precision, and a well-timed cultural wink.

The Pixel 10a might be Google’s latest affordable phone, but on paper it feels more like a gentle refresh than a true upgrade – especially when last year’s Pixel 9a is now cheaper than ever. 

Both phones share the same core specs, the same Google-made Tensor chip and a near-identical design, yet one of them regularly dips well below its original asking price while the other launches at £499/$499. 

Unless you’re absolutely set on having the newest model for the sake of it, the Pixel 10a has to work much harder than this to justify its existence.

SQUIRREL_PLAYLIST_10208265

The Pixel 10a is a very familiar phone

You’ll be forgiven if you can’t tell the difference between the Pixel 10a and last year’s Pixel 9a when looking at the spec sheet – the two phones are practically the same in key areas. 

Advertisement

That extends to the chipset, with both the Pixel 9a and 10a sporting the same Tensor G4 chipset – a chipset that first debuted on the Pixel 9 collection in 2024. It was a solid addition to the Pixel 9a in 2025, as it represented the latest in Google’s processing power, but the absence of the G5 found in the Pixel 10 series means the same can’t be said here. 

There’s no justification for Google not to include the chipset, aside from a way to protect the sales of the regular Pixel 10. The entry-level Pixel is often compared to the ‘a’ alternative, and much of the time, the ‘a’ alternative is the better buy. With an older chipset, that recommendation is no longer quite as easy. 

Advertisement

It would’ve been an easier pill to swallow had Google introduced new game-changing features to balance that decision, but there isn’t really much else new about the Pixel 10a. 

Generally, it looks a lot like last year’s Pixel 9a, sporting the same flat-edged, rounded-corner design with a completely flat rear – a stark contrast to the bar-style camera housing present on the flagship Pixel 10 range. It would’ve been easy for Google to score a win in this department by including support for PixelSnap magnetic accessories, but that’s sadly not the case. 

There’s also an identical camera offering with a 48MP primary lens and a 13MP ultrawide on the rear and a 13MP selfie snapper up front, with no discernible changes to the underlying hardware. 

Advertisement

Now, Google could’ve improved its image processing to provide a tangible boost, but I’d argue that the company’s processing tech was already pushed to the limits by the 9a’s hardware last year. It’ll be interesting to see how, if at all, it differs in actual performance. 

Advertisement

You likely won’t see a battery life boost either; it uses the same 5100mAh battery as its predecessor, with Google claiming the same ‘over 30 hours’ battery life estimate. Another downside to not using a newer, more energy-efficient chipset.  

Upgrades that don’t exactly get the heart racing

Despite all the above, the Pixel 10a does come with a few key upgrades – but these are what I’d describe as bog-standard niceties rather than trying to truly tempt consumers to the newer phone.

That’s best exemplified by the only tangible design difference aside from the new Berry finish and slightly improved IP68 dust- and water-resistance: a completely flat rear camera housing. 

You might think ‘well, that was the case with the 9a too’, but there was an ever so slight bump around the edges of the module. That’s gone this year. It does mean that the 10a should sit completely flush on a table and not wobble at all, but the 9a’s wobble isn’t exactly egregious compared to some phones. 

Advertisement

Advertisement

There are also improvements in the screen department; it’s still a 6.3-inch 120Hz AMOLED screen, but Google has slimmed down the infamously thick bezels of the latest ‘a’ series handset, giving it a more premium look in line with 2026 mid-range rivals. It’s also brighter at 3000nits, though that’s only a 300nit jump compared to the Pixel 9a.

The Pixel 10a should also charge a little faster than its predecessor despite using the same 5100mAh cell, thanks to boosted 30W charging, though again, this is only a 7W increase compared to the 9a and far from true fast-charging mid-rangers like the Xiaomi 15T Pro and its 90W charging. 

Google is also keen to point out that the Pixel 10a will be ‘the first’ in the ‘a’ series to get flagship Pixel 10 AI features like Auto Best Take and Capture Coach – but the ‘first’ wording there suggests that the Pixel 9a may get the features in the future. 

After all, it’s not like Google can rely on the usual excuse of needing the latest Tensor chipset to power the features – if they’re on the 10a, they can definitely run on the 9a. That’d be a massive win for existing Pixel 9a winners, but it’d mean one less reason to opt for the newer, more expensive model. 

Advertisement

Advertisement

The Pixel 9a seems like the smarter buy

With that all in mind, it seems like the Pixel 9a is the smarter buy – especially when you can pick it up for as little as £345/$399 at the time of writing, compared to the £499/$499 price tag of the newer Pixel 10a. 

The Pixel 9a offers an oddly familiar experience to the new model, complete with the same chipset, RAM configuration and storage, same design, same camera hardware and same battery. 

Sure, the Pixel 10a has a slightly better screen, a reduction of the already unnoticeable camera bump and slightly faster charging, but without the Pixel 10’s G5 chipset, it’ll remain an unusually hard sell for what is usually one of the best mid-range smartphones around.

If you are tempted by the newer model, you can pre-order the Pixel 10a now via the Google Store ahead of release on 5 March. 

Advertisement

At $999, the iFi Audio NEO iDSD 3 lands in a price bracket that’s starting to feel less taboo in 2026, and that’s not an accident. In a Head-Fi market crowded with competent boxes chasing the lowest possible number, iFi is playing a slightly different game: value through consolidation. The NEO iDSD 3 isn’t trying to be the cheapest DAC/headphone amp on your desk. It’s trying to be the one that makes you stop shopping. With serious output power, lossless Bluetooth, full-tilt hi-res playback, and both analog and digital flexibility in a single chassis, iFi is betting that “affordable” no longer means stripped down, but the last box you are likely to need for many years. Which is probably not a thing for some people, but it’s a concept worth thinking about.

Yes, you can spend less with brands like Shanling, FiiO, Schiit Audio (once you start stacking boxes), Topping, or Fosi Audio. The argument isn’t whether $1,000 is cheap—it isn’t. The real question is whether any of those alternatives deliver this level of power, codec support, DAC performance, and feature depth in one box without meaningful compromises. That’s where the NEO iDSD 3 makes its case: not as a bargain-bin hero, but as a compact, do-everything desktop component that treats flexibility and sound quality as non-negotiable.

What the iFi NEO iDSD 3 Actually Is

The iFi Audio NEO iDSD 3 is a compact desktop DAC and headphone amplifier designed to cover a lot of ground without forcing users into a stack of separate boxes. At $999, it sits squarely in the middle of a crowded category, but its pitch is straightforward: combine high-power headphone amplification, full hi-res wired playback, and current-generation lossless Bluetooth support in a single, flexible component that works equally well on a desk or in a traditional hi-fi system.

At its core is iFi’s custom Burr-Brown DAC stage, supporting PCM up to 768kHz and DSD512, paired with revised internal circuitry aimed at improving accuracy rather than adding coloration. JVCKENWOOD’s K2HD Technology is included to restore harmonic content in digital recordings, while the headphone stage delivers up to 2,532mW RMS (5,551mW peak).

Wireless is treated as a primary input, not an afterthought. Bluetooth 5.4 with aptX Lossless, LDAC, and LHDC allows the NEO iDSD 3 to function as a serious wireless DAC. Analog-domain tuning is handled via XBassII and XSpace, for users who want subtle low-frequency reinforcement or a broader presentation without relying on DSP.

Practical details round out the package. The iFi Nexis app enables OTA firmware updates and remote control, and a 2-inch rotating color display allows the unit to be positioned vertically or horizontally depending on system layout.

The NEO iDSD 3 doesn’t try to redefine desktop audio. It focuses on consolidation: strong output power, modern wireless support, broad format compatibility, and real-world usability in one box. Whether that justifies $1,000 depends on your specific priorities, but in terms of scope, it’s clear what iFi is aiming to deliver.

Connectivity

The NEO iDSD 3 is designed to handle both modern digital sources and older analog gear without adapters or workarounds. Digital inputs include USB-B 3.0, S/PDIF coaxial, and S/PDIF optical, covering computers, streamers, TVs, and disc players with digital outputs.

For system integration, the NEO iDSD 3 offers balanced XLR and single-ended RCA line outputs, selectable as fixed or variable, allowing it to function as a DAC, preamp, or headphone amp depending on the setup. A BNC clock input is also included for users running an external master clock.

One notable change from earlier models is the removal of the 3.5mm analog input, replaced by a single-ended RCA line input. This makes the NEO iDSD 3 far more practical for traditional sources such as a turntable with a built-in phono preamp, an external phono stage, a tuner, or older CD players without digital outputs. Yes, that includes legacy gear that still works perfectly fine but predates modern digital connectivity.

Power and Physical Specs

The headphone amplifier section is built to cover a wide range of loads. Output power reaches 2,532mW RMS (5,551mW peak) from the 4.4mm balanced output, with the 6.3mm single-ended output also providing substantial drive. Output impedance is kept low at ≤1Ω, with iEMatch available for sensitive earphones, allowing the unit to work equally well with IEMs, dynamic headphones, and demanding planar designs.

Physically, the unit is compact but solidly built, measuring 214 x 158 x 41mm (8.4 x 6.2 x 1.6 inches) and weighing 916g (2.0 lbs).

The Bottom Line

The NEO iDSD 3 is a DAC and headphone amplifier that focuses on doing fewer things well, rather than trying to be an all-in-one. What’s unique is the combination of high headphone output power, lossless Bluetooth support, broad digital format compatibility, and a proper RCA analog input in a compact chassis. That last point matters if you’re connecting an external phono stage, a turntable with a built-in phono preamp, or older sources that don’t offer digital outputs.

What it doesn’t include is a built-in streamer or phono stage, by design. If you want network playback, you add one upstream—such as iFi’s ZEN Stream 3 or NEO Stream 3—and keep the DAC and amplifier roles clearly defined.

The NEO iDSD 3 is best suited for listeners who want one capable desktop or rack-friendly component to handle headphones, active speakers, wired sources, and modern wireless playback, without stacking multiple boxes or paying for features they won’t use.

For more information: ifi-audio.com

Related Reading:

The sub $1,000 network streaming market is no longer short on options. Between WiiM, Eversolo, Bluesound, and Cambridge Audio, buyers can choose from a wide range of compact streamers that handle modern services cleanly and integrate easily into existing systems. Most now cover the basics well. What separates them is how they integrate into real world systems, the flexibility they give users, and their performance in everyday listening.

That is where iFi Audio positions the $999 NEO Stream 3 and $399 ZEN Stream 3. Both are built around a modern streaming platform that supports Qobuz Connect, TIDAL Connect, Spotify Connect, and Apple AirPlay 2, allowing users to stream directly from native apps without workarounds or proprietary control layers. Setup is handled through a straightforward IoT process, and ongoing updates are managed through iFi’s Nexis platform rather than requiring users to chase firmware manually.

In practical terms, the NEO Stream 3 is designed for listeners who want a single box that can act as the digital front end of a system, handling streaming duties and conversion without external hardware. The ZEN Stream 3 takes a different approach, focusing on being a quiet and reliable network transport for systems that already have a DAC the owner likes and does not want to replace. Both models lean on improved power design and noise reduction to keep the signal path stable, an area where iFi has traditionally focused its engineering effort.

But is that enough to compete below $1,000? Let’s take a look at what both offer straight out of the box.

NEO Stream 3 and ZEN Stream 3 Share the Same Core Streaming and Noise Control Architecture

Both the NEO Stream 3 and ZEN Stream 3 are built on the same core streaming and noise-control platform, with differences that are deliberate rather than cosmetic. At a baseline level, both models receive refinements to their power supplies, including upgraded polymer capacitors for lower noise delivery and the continued use of ELNA Silmic II capacitors in the audio path. The goal here is consistency and stability, not tonal revoicing, with iFi focusing on cleaner power and predictable behavior in long-term use.

Setup and system management are also shared. Both streamers now include iFi’s updated IoT hardware, enabling faster initial configuration and smoother day-to-day operation.Firmware updates and system control are handled through either a browser based interface or the iFi Nexis app, streamlining setup and ongoing maintenance compared with earlier manual update processes. Exclusive Modes return on both units as well, allowing users to disable unused background processes during playback to minimize potential noise sources.

Noise mitigation remains a defining design priority across both products. Each includes iPurifier2 technology on the S/PDIF outputs and Active Noise Cancellation on the USB ports, specifically aimed at reducing interference from connected storage devices or computers when feeding an external DAC.

The NEO Stream 3 goes further by retaining the OptiBox optical isolation system from the original NEO Stream, which isolates the wired network connection and prevents network-borne electrical noise from entering the audio system. The ZEN Stream 3 does not include OptiBox as one its connectivity features.

On the software side, both units use the latest version of iFi’s ultra-resolution streaming engine. This platform supports Qobuz Connect, TIDAL Connect, Spotify Connect, and AirPlay 2, with improvements focused on stability, smoother web radio playback, and a cleaner interface with deeper configuration options. Control is handled through native apps rather than forcing users into a proprietary ecosystem. Support for Spotify Lossless, however, is still not finalised but we’re told it’s coming.

Where the two models clearly diverge is in system role. The NEO Stream 3 is designed as an all-in-one digital front end, combining network streaming with an integrated DAC capable of handling up to 768 kHz PCM and DSD512. It is intended for users looking to modernize an existing hi-fi system without adding external digital components. The ZEN Stream 3, by contrast, is a dedicated network transport. It supports up to 384 kHz PCM and DSD256 and is meant to slot into systems where a preferred external DAC is already in place.

Both models also incorporate K2 technology developed in collaboration with JVCKENWOOD. This processing is designed to restore harmonic information often lost during recording, mastering, or encoding. Two modes are available: K2, which preserves the original file resolution, and K2HD, which optionally upsamples PCM content to 192 kHz 24-bit. With this generation, K2 processing is no longer limited to internal DACs. Any external DAC connected to either streamer can benefit from K2 processing, while the NEO Stream 3 adds the ability to apply K2HD upsampling internally to PCM material below 192 kHz.

NEO Stream 3

The NEO Stream 3 is a combined network streamer and DAC designed to function as a complete digital source component. It supports Wi-Fi and wired networking via RJ45 Ethernet as well as iFi’s M12-X optical network input. Digital inputs include USB-A for storage or DAC use and a front-mounted USB-C port. Digital outputs are comprehensive, with dual USB-A, S/PDIF optical, S/PDIF coaxial, AES/EBU, and I2S, allowing the unit to operate either as a DAC or as a dedicated digital transport. Analog outputs are provided via a balanced 4.4 mm connection and single-ended RCA.

Internally, the NEO Stream 3 uses a Burr-Brown DAC stage derived from the NEO iDSD 2 and supports high-resolution audio up to 768 kHz PCM and DSD512. Balanced output voltage is rated at 4 V RMS, with 2 V RMS available from the RCA outputs.

Power is supplied via a DC input supporting 9 to 15 volts, with internal regulation using upgraded polymer capacitors and ELNA Silmic II capacitors. Power consumption is under 0.5 watts at idle and approximately 14 watts at maximum load.

Setup and ongoing updates are handled through a simple browser-based interface or the iFi Nexis app, made possible by the new IoT hardware, so there is no need to fuss with manual update routines. With a compact chassis measuring 214 x 151 x 41 mm (8.4 x 5.9 x 1.6 inches) and weighing just 1 kg (2.26 lbs), the NEO Stream 3 is easy to place on a desk or equipment shelf without demanding much space.

Comparison

Tip: There was never a NEO Stream 2 product.

ZEN Stream 3

The ZEN Stream 3 is designed as a dedicated network transport for systems where digital conversion is handled elsewhere. It focuses entirely on getting a clean, stable digital signal out to an external DAC rather than duplicating functionality already present in many higher-end systems. Networking is handled via Wi-Fi or RJ45 Ethernet, with two USB-A ports available for local storage or DAC output.

Digital output options include two USB-A outputs for DAC connection and a coaxial S/PDIF output, each incorporating iFi’s iPurifier and Active Noise Cancellation technologies to reduce electrical noise before the signal reaches the DAC. There are no analog outputs and no internal DAC, which keeps the signal path simple and aligned with its role as a transport.

High-resolution support extends up to 384 kHz PCM and DSD256, handled by the same next-generation streaming engine used in the NEO Stream 3. Streaming is managed directly through native apps with support for Qobuz Connect, TIDAL Connect, Spotify Connect, and AirPlay 2, avoiding reliance on proprietary control software. K2 processing is included to restore harmonic information lost during recording or encoding and is applied before the digital output, allowing connected DACs to benefit without altering their internal architecture.

Power delivery has been revised with upgraded polymer capacitors and ELNA Silmic II capacitors to improve stability and reduce noise under load. Power is supplied via a DC input supporting 9 to 15 volts, with idle consumption under 6 watts and a maximum draw of approximately 10 watts. Firmware updates and system setup are handled through a rear USB-C service port, using either a browser-based interface or the iFi Nexis app, enabled by the updated IoT hardware platform.

Physically, the ZEN Stream 3 is compact and lightweight, measuring 158 x 100 x 35 mm (6.2 x 3.9 x 1.4 inches) and weighing 578 g (1.27 lbs). It is sized to sit easily on a desktop or equipment shelf alongside an external DAC, making it a practical drop-in upgrade for your existing system.

Comparison

Tip: There was never a ZEN Stream 2 product.

The Bottom Line

With competition this strong, iFi is not trying to win by simply piling on features. The strategy here is differentiation through system role, execution, and the details that tend to matter after the initial setup. That includes easy integration into existing systems, app-native control rather than closed ecosystems, and a design that stays below the psychological $1,000 threshold while still targeting more demanding listeners.

That approach comes with tradeoffs. Brands like WiiM have shown that a comprehensive, frequently updated operating system can win over a large segment of the market, especially when paired with features such as room calibration and broad device compatibility. At this point, support for the major streaming platforms is largely table stakes across the category.

Where these products are ultimately judged is not on spec lists or feature counts, but on how smoothly they operate day to day, whether firmware updates are painless rather than disruptive, and how they actually sound once they are part of a real system. iFi is betting that its focus on noise control, power integrity, and flexible system roles will resonate. Whether that is enough to stand out in a crowded sub-$1,000 field will be up to listeners to decide.

Price & Availability

• NEO Stream 3 – $999
• ZEN Stream 3 – $399

For more information: ifi-audio.com

Related Reading:

Chinese robotics leader Unitree took full advantage of the nation’s Lunar New Year celebrations this week to show off the impressive skills of its G1 humanoid robot.

A video (top) of the event shows numerous G1 robots participating in what Unitree described as “the world’s first fully autonomous humanoid kung fu performance.” There’s a spot of breakdancing in there, too.

Performing alongside kids from the Tagou Martial Arts School for the Spring Festival Gala on China Central TV, the robots displayed incredible agility and coordination, moving at around 3 meters per second while performing flips, table vaults, somersaults, and rapid formation changes, blending martial arts with robotics innovation.

The robots were recently upgraded with improved, more dexterous hands, which during this week’s performance supported rapid switching and stable gripping of martial arts props such as nunchaku, Global Times reported.

Unitree founder and CEO Wang Xingxing said the robots executed many of their moves while running at high speed, claiming it as a first for high-dynamic, highly coordinated cluster-control technology.

Wang described the innovations as “very practical” and said they will “facilitate large-scale group deployment of robots in the future.”

The 132-centimeter-tall G1 robot was unveiled in May 2024 and made available three months later for $16,000. The bipedal bot is targeted for research, education, entertainment, and light industrial applications, with researchers, students, and developers encouraged to program and customize the the robot for different tasks.

Unitree also has a full-size humanoid robot called H1, which is about 180 centimeters tall. The H1 is more robust and powerful and is aimed at industrial deployment, while the more compact G1 prioritizes agility and affordability, and is geared more toward research and entertainment.

Unitree is aiming to ship in the region of 20,000 humanoid robots this year, nearly four times last year’s number.

It’s competing with a slew of robotics companies in China, the U.S., and beyond, with all of them racing to find meaningful and manageable roles for their humanoid robots in industrial settings.

The G1 also hit the headlines earlier this month when it took on an autonomous walking challenge in deep snow and brutally cold conditions.