Resolve AI, the production-operations startup backed by Greylock and Lightspeed Venture Partners, today announced a sweeping expansion of its platform that introduces always-on background agents, a redesigned investigation architecture, and a shared workspace where engineers and AI agents collaborate in real time on live incidents.
The centerpiece of the release is a new multi-agent investigation system developed by Resolve AI’s in-house research lab. Instead of deploying a single AI agent to diagnose a production failure — analogous to a lone engineer pulling an on-call shift — the platform now dispatches a coordinated team of specialized agents that pursue multiple hypotheses in parallel, independently verify each other’s conclusions, and construct complete causal chains from root cause to symptom. The company says the architecture delivers more than a twofold improvement in root cause accuracy on its internal evaluation benchmarks compared to earlier versions of its platform.
“Think of a single agent being on call, the way a human would be,” Resolve AI CEO and co-founder Spiros Xanthos told VentureBeat in an exclusive interview ahead of the announcement. “We now have a team of agents that all work together, almost like a team of humans debugging an issue, and that has improved quality by 2x.”
The announcement arrives at a moment of acute tension in the software industry. AI-powered code generation has exploded in adoption, enabling engineering teams to ship dramatically more software than they could two years ago. But keeping that software running in production — debugging it when it breaks, monitoring it after deployment, auditing its health — remains overwhelmingly manual. For a company that raised a $125 million Series A at a $1 billion valuation earlier this year, Resolve AI is making a direct bet that the operational side of the software lifecycle is the next major frontier for AI investment.
Advertisement
What hundreds of real-world test cases reveal about the accuracy claim
Any accuracy claim from a startup warrants scrutiny, and Xanthos was candid about both the scale and limitations of the evaluation. The 2x figure comes from internal benchmarks, not a third-party audit, though the evaluation set was built to mirror the complexity that Resolve AI’s enterprise customers encounter daily.
“These are very hard, complex evals that we built over time to represent real-world examples,” Xanthos explained. “This is not customer data, but these evals represent difficult cases similar to what we’ve seen at some of the largest tech companies we work with.” He described the set as comprising hundreds of cases that reflect the kinds of production failures encountered at companies like Coinbase, Salesforce, DoorDash, and Zscaler — all named Resolve AI customers.
The practical impact of that accuracy gain is significant. Resolve AI’s agents now act as first responders for every on-call alert, typically triaging within five minutes before a human engineer even becomes involved. In previous public disclosures, the company has cited DoorDash reducing time to root cause by up to 87 percent. When asked to contextualize that figure, Xanthos described the typical baseline.
“When something goes wrong, it might take five to 10 minutes for a human to even get their laptop and connect,” he said. “The typical MTTR is in the tens of minutes, sometimes hours, depending on severity. So an improvement of 80-plus percent — four to five times faster — is actually huge. It’s something we’ve never achieved before with AI, tools, data, or observability.”
Advertisement
How AI agents fact-check each other to prevent hallucinated root causes
One of the core challenges in applying large language models to high-stakes production environments is their tendency to generate plausible-sounding but incorrect answers — a failure mode that, in the context of a live outage, could send an engineering team chasing the wrong fix while a service stays down.
Xanthos acknowledged this directly. “This is a very common issue with models out of the box,” he said. “They always try to give you an answer, and if they don’t have enough evidence, they’ll give you the best possible answer — which is likely to be wrong.”
Resolve AI’s countermeasure is a system of layered verification among its agents. Each agent investigating a hypothesis must cite every piece of evidence it relies on and present that evidence to another agent for independent review. The investigating agent must construct the full causal chain — from root cause to symptom — and peer agents actively attempt to disprove the theory by identifying gaps in the logic.
“Often, agents actually disprove those theories because they find gaps,” Xanthos said. “There are many layers of defense and agentic checks that allow Resolve to be very accurate and not mislead.”
Advertisement
Equally important, he said, is the system’s willingness to say it does not know. “The bar to actually saying ‘I have the answer’ is very high. In those cases, it will say, ‘This is the evidence I found. Here are three or four paths you can take from here, but I wasn’t able to fully prove that this is the problem.’ A system like this that operates in production cannot be a black box.” In domains where wrong answers carry operational consequences, calibrated uncertainty can be more valuable than confident outputs. For an AI system integrated into an incident-response workflow, confidently pointing engineers in the wrong direction during a customer-facing outage could compound the very harm it was designed to prevent.
Inside the new background agents that never go off-call
Beyond incident response, Resolve AI is introducing a new class of background agents designed to handle the continuous, often invisible operational work that engineering teams are expected to perform but struggle to sustain at scale.
These agents run on schedules or wake automatically in response to events — a new deployment, a fired alert, a merged pull request — and accumulate institutional knowledge from every investigation and human interaction over time. When an engineer opens the Resolve AI interface, agents have already been working: pre-investigating priority issues, monitoring deployments, auditing alert hygiene, flagging configuration drift, and surfacing cost anomalies.
Xanthos drew a distinction between background agents and the incident-response agents that have been Resolve AI’s primary offering. “You can now have these agents run in the background at all times — not only when a human asks an agent to debug a problem or when an alert fires,” he said. “A lot of our customers are now monitoring changes that land in production before they cause an issue. There’s an agent that monitors those all the time.”
Advertisement
He described these background agents as “general-purpose SRE agents that are available to every developer,” capable of handling tasks that range from monitoring infrastructure changes that might increase cloud costs to performing post-incident follow-up work like generating code fixes based on incident learnings. The concept addresses a structural problem in software operations: the daily tasks required to keep production systems healthy — monitoring deployments, investigating alerts, tracking changes across complex environments — are critical but reactive and manual. Engineering organizations know this work needs to happen, but it competes for attention with feature development. Automated agents that perform this work continuously could shift teams from reactive firefighting to proactive operational management.
The shared workspace where engineers and AI agents investigate together
The third major component of the release is what the company calls a shared investigation surface — a workspace where engineers and AI agents work from the same live evidence during an active incident. Reports update dynamically as investigations evolve. Every finding is inspectable. Engineers can explore side investigations without interrupting the primary workflow. Source queries are pullable and modifiable in place, evidence is embedded directly into the workspace, and remediation actions can be triggered from the same interface without switching tools.
“Think of it as an interface to all the production tools, but also an interface where humans and agents can collaborate with each other — or agents with agents,” Xanthos said. “That’s what gradually leads to more trust and more automation, because you work with the agent, you teach it, you see the results.”
The company is also making its platform available as a REST API and an MCP (Model Context Protocol) server, enabling engineering teams to integrate Resolve AI into broader agentic workflows and infrastructure. According to Xanthos, this is already happening in practice. “A general-purpose agent that a company has built — when it comes to debugging, that agent could invoke Resolve,” he said. “Or somebody works on their coding agent on the laptop, and Resolve shows up there as an MCP. If there is some production-related activity, the coding agent can invoke it.” The interoperability play signals that Resolve AI sees itself not as a closed system but as a specialized node in a broader ecosystem of AI agents that will increasingly hand off tasks to one another — a pattern Xanthos compared to the open architecture of the web rather than the walled-garden model of an app store.
Advertisement
Why Resolve AI says it can outperform Datadog, PagerDuty, and the cloud giants
The agentic operations space has become crowded in the past year. Datadog, PagerDuty, and major cloud providers have all announced AI-augmented operations capabilities. When asked what separates Resolve AI from these incumbents, Xanthos pointed to the depth of the company’s technical foundation.
“We’re operating at the frontier here. There’s no blueprint for how you build a system like Resolve,” he said. He noted that he and co-founder Mayank Agarwal co-created OpenTelemetry, the most widely adopted open-source project in observability, which now serves as the de facto standard for collecting metrics, logs, and traces from modern software systems.
Xanthos also highlighted the company’s recent AI Lab, led by a researcher he described as the former post-training lead for Meta’s Llama models. “He managed to combine deep expertise of production observability with AI and models, and I think that’s very unique,” Xanthos said. “I don’t believe any other company, whether it comes from an observability background or it’s a startup, has all of that together.”
The company’s structural defenses, according to Xanthos, include a full environment model that Resolve builds for each customer, a memory system that learns within the customer’s specific production environment, and its multi-agent architecture. The lab is now post-training frontier models on production-specific data — the kind of procedural knowledge that experienced engineers use to debug production issues but that does not appear in standard model training sets. This approach reflects an increasingly common pattern among AI application companies: using frontier foundation models as a base layer but investing heavily in domain-specific fine-tuning, retrieval, and agent architectures to achieve accuracy levels that general-purpose models cannot reach alone.
Advertisement
How outcome-based pricing changes the economics of AI in production
Resolve AI’s pricing model departs from traditional enterprise software licensing. The company sells credits that are consumed when agents perform work — an outcome-based approach that ties cost directly to value delivered.
“We’re not selling software,” Xanthos said. “The way you buy and use Resolve is by buying credits that are consumed when Resolve performs an action. It’s outcome-based. Only when Resolve troubleshoots an alert — that’s the only time that it consumes credits.”
He addressed the cost question head-on, arguing that Resolve AI is actually cheaper than the alternative of building a similar system from scratch using frontier models and MCP integrations. “If you were to take Opus or GPT-5.4 and try to build a solution like Resolve with MCPs, we measured — you actually end up consuming a lot more in tokens than what you have to pay Resolve, because our system is very optimized in terms of context, in terms of how it reads time-series data.”
As for the always-on background agents, Xanthos said their continuous nature does not inherently add to cost. “The background agent doesn’t mean it does intensive work all the time. It means that it can be there; you can give it any task you want. A lot of these tasks are triggered based on some action — an alert happens, somebody merges a PR, and you want to see if it has an impact on production.” For enterprise customers in regulated industries — the Coinbases and Zscalers of the world — data residency and security are non-negotiable. Resolve AI accommodates this with a flexible deployment model: the data plane sits wherever the customer’s existing tools already live, while the inference layer can run as a standard SaaS deployment or inside a customer-specific VPC. “We designed Resolve to work with the large enterprises where security standards are the highest,” Xanthos said. “There are many measures we take to ensure Resolve is secure, including not retaining data.”
Advertisement
Why engineering leaders are slowly learning to trust AI agents with production systems
The question of whether engineering teams will trust AI agents to take autonomous action in production — rolling back a deployment, adding capacity, generating a pull request — is one of the defining cultural challenges of this technology wave. Xanthos drew an analogy to autonomous vehicles.
“For us to allow a car to drive on its own on the street, we have to prove that it’s safer than a human. Agents in production is a very similar concept,” he said. He acknowledged that not every customer is comfortable with agents taking automated action, but described a gradient of trust that he expects to evolve rapidly.
“There is a set of actions that are relatively risk-free that most tech companies probably are comfortable having an agent take, and probably there is another set of actions for which the human has to approve,” he said. “But as quality keeps climbing the way we see at Resolve, I would say we’re going to cross the threshold this year where most of the actions will be taken by an agent automatically.”
He described the typical adoption arc: companies begin with agents providing recommendations, then a human decides whether to press the button. Over weeks or months, trust builds incrementally. “I don’t think this is a problem where we just let the agents run wild from the beginning,” Xanthos said. The incremental approach mirrors how enterprise technology adoption has always worked — from cloud migration to container orchestration, organizations move at the speed of trust, not the speed of capability.
Advertisement
The argument that AI-generated code is making the production crisis worse, not better
Perhaps the most provocative argument in Resolve AI’s thesis is that the explosion of AI-generated code is actually intensifying the production-operations problem. In a recent LinkedIn post, Xanthos framed the dynamic in stark terms, arguing that engineering leaders who celebrate faster code shipping without investing in production operations are effectively having their senior engineers “subsidize velocity” through increased incident-response burden.
In his interview with VentureBeat, he returned to this theme. “Now that coding agents are producing code, we produce a lot more code that we’re less familiar with — humans are less familiar with — so you need the AI to be the defense,” he said.
This framing positions Resolve AI not merely as a productivity tool but as a necessary counterweight to the AI coding revolution. As organizations deploy more code, written by tools that their engineers may not fully understand, running against production systems those engineers did not build, the argument is that the operational complexity — and the consequences of failure — will grow proportionally. On the Stack Overflow Podcast last October, Xanthos put numbers to this claim, estimating that engineers spend upwards of 70 percent of their time maintaining and troubleshooting production systems rather than building new features. “We’re facing a new crisis where we’re building faster than we can operate,” he said in that conversation.
Resolve AI was founded in early 2024 by Xanthos and Agarwal, who first met during their PhD programs at the University of Illinois and have worked together for more than a decade. Xanthos previously co-founded Pattern Insight (acquired by VMware) and Omnition (acquired by Splunk), where the pair helped create OpenTelemetry. The company raised a $35 million seed round from Greylock in 2024, followed by the $125 million Series A led by Lightspeed at a $1 billion valuation earlier this year. Named customers include Coinbase, DoorDash, MSCI, Salesforce, MongoDB, and Zscaler.
Advertisement
Xanthos’s long-term vision is expansive. “Over the long run, once agent ability surpasses that of a human software engineer, the end result is a lot more technology and a lot more software,” he said. “It’s not actually fewer people working on it. It’s technology becoming cheaper, becoming more accessible, producing a lot more technology for the benefit of the world.”
That vision will take years to realize. But the more immediate promise of today’s announcement comes down to something every on-call engineer understands viscerally: the 2 a.m. page, the scramble for a laptop, the frantic search through dashboards and logs for an answer that might take minutes or might take hours. Resolve AI is betting that the next time that alert fires, a team of agents will have already investigated, verified, and documented the root cause before the engineer’s phone even lights up. For a profession that has long measured its nights by mean time to resolution, the question is no longer whether AI can help — it is whether engineers will let it.
After Independence Day, director Roland Emmerich and producer Dean Devlin could have easily kept feeding the multiplex more flag-waving alien carnage. Instead, with Robert Rodat’s Saving Private Ryan pedigree tempering the “U-S-A!” machinery, The Patriot became something more grounded, darker, and far better researched than anyone had reason to expect. It is still historical fiction with Hollywood fingerprints all over it, but compared to the glorious cheese fountain of ID4, this is a far more serious and satisfying story about American grit, personal loss, and the ugly cost of revolution.
Mel Gibson’s Benjamin Martin is not based on one specific individual but a composite of militia who fought so bravely to win our freedom from Great Britain some two-and-a-half centuries ago. As a veteran of the French and Indian War, specifically “the wilderness campaign,” he’s not only witnessed but perpetrated unspeakable horrors and is reluctant to see his people thrust into another bloody conflict. But when the war comes for him and his family, this expert in guerilla warfare answers the call and helps turn the tide. Beyond the relatively minor indignities such as the Tea Act Monopoly, the Stamp Act and the Writs of Assistance, we’re shown the atrocities visited upon the colonists by the British forces, a brutality born of arrogance, and it’s hard not to be invested in the struggle by the climactic battle.
My collection of The Patriot discs on DVD and Blu-ray.
A quick glance over my left shoulder at the shelf marked M through Z tells me that, damn, I must really like this movie. I’ve owned The Patriot on five different five-inch discs (see photo), and that was before Sony dropped its new and improved SteelBook 4K edition, once again serving up both the R-rated theatrical cut and the longer unrated version. On the 2018 UHD disc release, just the theatrical cut was in Atmos and 4K, but only in HDR10, with the unrated version in 1080p/5.1. This SteelBook set now offers both cuts each on its own BD-100 platter, in 4K, Dolby Vision and Dolby Atmos, with breathtaking cover artwork by Paul Shipper.
Say what you like about ol’ Mad Max here, but he is one intense thespian, and his closeups convey his grief, his conviction, and the inner demons that fuel Martin’s resolve. Not every shot is razor-sharp, but nonetheless it looks like we’re watching the Revolutionary War, the roll-up and the aftermath through an immaculate window. A light, consistent layer of film grain accentuates the 2.39:1 image, and the high dynamic range delivers consummate detail in the many period-authentic low-light scenes, illuminated by candles, fireplaces or campfires. Even the glisten of a brocaded epaulet is preserved. Whether through the use of filters, magic hour scheduling or post-production wizardry, Caleb Deschanel’s Oscar-nominated cinematography has an enticing golden glow, and the colors are significantly upgraded from the 2018 4K, with only the slightest perceptible drop in picture and sound quality at the inserts within the unrated cut.
Subtle LFE for hoofbeats and fireworks early in the story lull us into false resignation, until the realities of warfare are fully unleashed. Showoff scenes don’t come much better than Gabriel’s rescue–this is the one I routinely use to test speakers, in particular my surrounds and sub–conveniently located at the start of Chapter 5 on the theatrical version, 36:42 into the film. Our vantage point shifts frequently amid the chaos but the hard placement of off-camera voices keeps us in the middle of the action, with frequent booming gunshots all around. (I plan to watch it again after this review publishes, just because.) The whiz and impact of cannonballs are also outstanding. The active overhead channels bring a wonderful sense of spaciousness throughout, and dialogue scenes that proved challenging on past editions are now crystal-clear. The Patriot boasts a John Wiliams score, lesser-known despite its Oscar nomination, and it amplifies the excitement exponentially.
The extras are all ported from past editions, some dating back almost 26 years, spread across the two platters. The theatrical version carries an enjoyable Emmerich/Devlin audio commentary in addition to brisk featurettes about the production and the historical fact. Interestingly, the comprehensive deleted scenes section–13 minutes total with optional commentary–is located on the unrated disc, even though most of that footage has been integrated back into the movie to create the longer cut. Also on this disc are vignettes devoted to the visual effects and concept art.
Advertisement
The timing of the new The Patriot 4K SteelBook is curious, too late for Memorial Day and too early for The Fourth, but take it from me: This one would make a terrific Father’s Day gift. (Father’s Day feature incoming, but this one deserved its own review.) With top scores for the movie, audio and video, this one gets our highest recommendation.
Movie Details
STUDIO: Sony Pictures Home Entertainment
FORMAT: Ultra HD 4K Blu-ray (June 9, 2026)
THEATRICAL RELEASE YEAR: 2000
ASPECT RATIO: 2.39:1
HDR FORMATS: Dolby Vision, HDR10
AUDIO FORMAT: Dolby Atmos with TrueHD 7.1 core
LENGTH: 165/175 mins.
MPAA RATING: R/Unrated
DIRECTOR: Roland Emmerich
STARRING: Mel Gibson, Heath Ledger, Jason Isaacs, Chris Cooper, Tchéky Karyo, Joely Richardson
Which? discovers ‘potentially lethal’ phone chargers are being sold by major retailers
Cheap chargers pose serious risks of fire, electric shock and more
Consumer group demands stricter government regulation
British consumer watchdog Which? has revealed many of the third-party phone chargers available to consumers could present “potentially lethal” risks – and they’re often hidden in plain sight.
Nine of the 15 chargers tested by Which? posed serious electric shock risks, while eight also presented potential fire or explosion hazards, but more worryingly, many were available from popular and trustworthy high-street and online retailers like Amazon, B&Q and Debenhams.
The timeliness is also of note, because the findings come seven whole years after Which? first warned about dangerous counterfeit and low-quality chargers, suggesting the problem remains widespread despite repeated warnings.
Latest Videos From
These are the reasons cheap chargers can fail basic safety tests
According to the report, many of the chargers tested failed basic testing because the internal electrical components were positioned too close together, insulation was inadequate, high-voltage stress tests caused failures and plug pins did not meet British Standards requirements.
Advertisement
The most prevalent concern, then, was that the defective products could cause electrical arcing, where electricity jumps between components, leading to electric component failures in the best-off cases, but electric shocks, overheating, fire and explosions in the most severe cases.
Among the examples given by the group was a counterfeit Apple USC-C 35W Power Adaptor sold for £11.99 – a not-at-all similar mock of Apple’s £59 charger. The researchers discovered arcing noises after just 10 seconds, and upon further investigation, found modelling clay inside the charger.
They believe it was added to make the device heavier, making it feel more ‘premium’. A second, separate model, sold via Debenhams, also included modelling clay within.
Advertisement
Sign up to the TechRadar Pro newsletter to get all the top news, opinion, features and guidance your business needs to succeed!
Besides suspiciously cheap chargers from Amazon (£2.99 and £3.99), eBay (£2.10 and £2.80) and AliExpress (£1.30 and £5.69), Which? also found that a more expensive £10.99 charger sold via B&Q was subject to fire, electric shock and explosion risks.
And even the chargers that passed safety tests, including models sold via Temu and Shein, weren’t fully legitimate. They still lacked the required markings and importer details, making them illegal in the UK.
“Badly designed electricals like these can have life-altering – even fatal – consequences,” Head of Consumer Protection Policy Sue Davies commented.
Advertisement
(Image credit: Which?)
Which? calls for greater enforcement over online marketplaces
Which? argues that marketplace operators have now become a major route through which unsafe and illegal imports can reach UK consumers, because they often act as intermediaries for third-party sellers.
However, despite the implementation of the UK’s Product Regulation and Metrology Act in July 2025, the consumer group says implementation has been slow. Under the law, the government can place obligations on online marketplaces.
Advertisement
Which? is therefore advocating for stronger enforcement powers and greater accountability for third-party seller listings.
“By making online marketplaces legally responsible for unsafe products, the government can set a world-leading standard for product safety in the digital age,” Davies added.
As for consumers, they’re being advised to buy from recognized brands and be wary of very cheap big-brand chargers. UK citizens should also look for the CE or UKCA marks and importer details.
Looking ahead, Which? has presented UK Department of Business and Trade Minister Kate Dearden a petition, with 150,000 signatories, calling for the government to regulate online marketplaces and fine them for breaches.
Rivian engineers took everything they learned from the larger R1 models and applied it to a vehicle sized for more people and more driveways. The result is the R2, a two-row electric SUV that starts well below sixty thousand dollars for loaded early versions and dips into the mid-forties for simpler single-motor models arriving next year. That pricing alone sets it apart from bigger adventure-focused rivals while still delivering real capability.
The body looks just like you’d expect from a car with a specific function, with no unnecessary frills and a nice, boxy form that declares its intent without being too huge for its own good. At 186 inches long, it’s around the size of a Honda CR-V in terms of footprint, but its stretched-out wheelbase makes it feel a little longer. The ground clearance is 9.6 inches, and the approach and departure angles are adequate for off-road adventures.
2 Mercedes toy cars for kids – LEGO Speed Champions Mercedes-AMG G 63 & Mercedes-AMG SL 63 vehicle playset for boys and girls ages 10 and up and…
2 driver minifigures – Each buildable vehicle toy comes with a driver minifigure wearing a Mercedes outfit so kids can role-play fast-paced races
Authentic Mercedes design – Each collectible car model features design details from the real life versions, including front grilles, hoods, wheels…
Stepping inside, you get the idea that the cabin is a huge open space with useful touches all over. The back seats provide ample leg and headroom and are exceptionally comfortable even on extended trips. The materials used are an excellent mix of nice everyday goods and some very smart eco-friendly choices, such as birch trim created from repurposed birch and a headliner built from ocean-rescued plastic. There are also spacious door pockets for holding water bottles, and the rear liftgate glass lowers to make loading heavier items easier.
Advertisement
The controls are one of the car’s most noticeable elements. There is a large, wide touchscreen that handles all of the normal features like as navigation, media, and vehicle settings, and it responds rapidly. You also get two huge halo dials on the steering wheel that allow you to change the climate, radio, drive modes, mirrors, and other settings with a few twists, pushes, pulls, or tilts. This strategy immediately became popular among reviewers, who considered it more user-friendly than scrolling menus while driving. One thing they needed was smartphone mirroring, but the native apps fulfill the most of your demands, and the interface is constantly updated with new features.
Under the hood, there’s an 88-kilowatt-hour battery and some extremely efficient motors. The top dual-motor Performance model generates 656 horsepower and accelerates to 60 mph in 3.6 seconds flat. Acceleration is powerful and immediate, but never uncomfortable. With 350 horsepower, the single-motor rear-wheel-drive car is a great option for people seeking for something a little more efficient. If you’re curious about how far you’ll get on a charge, the car’s range can reach 345 miles on the most efficient settings, thanks to its extraordinarily light weight (about 5,000 pounds) and sophisticated aerodynamic architecture that decreases the drag coefficient to 0.3.
Charging is quite speedy, with up to 230 kilowatts charging the battery from 10% to 80% in around 29 minutes under optimum conditions. You can also do bidirectional charging of up to 11 kilowatts, which allows you to power tools or even transmit electricity back to your home when the power goes out. Production of these things has already begun at Rivian’s plant in Normal, Illinois, with the intention of delivering the higher trims to clients this summer first.
The custom APU at the core of Sony’s PlayStation 5 hasn’t just been quietly powering these game consoles, but also made their way onto cryptomining cards around 2023 which are called the BC-250. The APUs on these boards differ from the one found in the PS5 most notably by having two out of eight CPU cores disabled, along with many compute units (CUs) of the iGPU. Now apparently it seems that you can re-enable these CUs per instructions by [duggasco] if you’re feeling adventurous.
The BC-250’s AMD APU in all its glory. (Credit: Lowest Logan, YouTube)
As stated in the project’s README, BC-250 boards come with only 24 out of 40 CUs enabled, but this is not a permanent (e-fuse) thing. Instead you can write to two hardware registers during the GPU driver initialization, something which can be added to for example the Linux kernel module parameters.
Since many of these APUs likely had cores and CUs disabled due to them failing QA during PS5 APU manufacturing, there’s a good chance that some of the CUs truly are bad. Yet as we saw with the AMD Phenom II X3 with a supposedly bad fourth core back in the day, sometimes demand for the ‘defective’ part is high enough that good parts get mixed in as well.
Thus people like [Lowest Logan] decided to give it a shot, demonstrating the use of the patch with Bazzite Linux on a BC-250 system. After a reboot the system does indeed list 40 CUs as being enabled, and running Furmark shows a big boost in performance without any glitches or fire. There is of course thermal throttling, but that is due to the default cooling solution not being designed for running it at full blast.
Advertisement
Incidentally the real PS5 has only 36 active CUs, so this technically makes these unlocked APUs more powerful. With the water cooling solution demonstrated by [Lowest Logan] the thermal throttling is also resolved, showing that you can get a pretty nice gaming system out of these old cryptomining boards if you happen to win the silicon lottery.
Even the most well-intentioned edtech can fall short if it does not meet students where they are. After several years studying the usability of edtech for teachers, the research team at ISTE+ASCD turned its attention to students — examining how the technical and pedagogical design of digital tools shapes their learning experiences.
In partnership with In Tandem and Sesame Workshop, researchers spoke with high school students across the United States to understand how they actually use edtech in real learning contexts. The findings identify five areas that matter most to students and offer guidance for educators and product designers seeking tools that are intuitive, meaningful and engaging.
A full framework and guidance for edtech buyers and product providers will be released in 2026.
This article was sponsored by ISTE+ASCD and produced by the Solutions Studio team.
Diorama111 released a fresh build just days ago that shows what happens when extreme miniaturization meets real remote control hardware. He started with an ordinary KATO 1/150 scale Toyota ProBox, the kind of plastic model that sits on N-gauge train layouts or display shelves. Most builders stop at painting and weathering. He kept going until the little van could roll under its own power, steer on command, and light up properly in both directions.
The finished model is only a few centimeters long, but at true scale, it is roughly 150th the size of a real Toyota ProBox. The exterior still looks like it just came off the production line, but everything inside had to be modified just to fit. This entailed designing new parts, modifying old ones, or just building them from scratch to allow for movement and electronics to fit within. Power is supplied by a tiny lithium-polymer cell measuring 9 by 9 by 4 millimeters. It sits on top of a custom control board and provides power to the complete arrangement. The duration is purposefully brief, just long enough to test it on a tabletop or in a diorama before it needs to be recharged quickly via the underside connectors.
A DC motor in the back propels the vehicle, and watch gears are used to slow it down significantly, allowing it to move at a suitable rate for its size. A brass-bushed axle is used to reduce friction, resulting in smooth acceleration and excellent control at low speeds. As a result, when observed from above, it looks to be moving steadily rather than bouncing about haphazardly.
Steering was a challenge because the conventional solution would not have worked. So a second DC motor was added to drive a lead screw, and a rod-and-linkage system was adapted to convert linear motion into front wheel spin. Two photoreflector sensors continuously monitor the motor shaft and screw, providing feedback to the controller so that it knows exactly where the wheels are and can give true proportional steering rather than just on/off. This all fits into the compact chassis, leaving room for the battery and receiver.
The radio signal is sent via infrared because a small surface-mount IR receiver module is contained within the body. The matching handheld transmitter contains two analog joysticks that provide you independent control over the throttle and steering. It appears delicate enough to execute figure-eights, smooth turns, and faultless parking on a cutting mat. Lighting only adds to the illusion, with a few tiny surface-mount LEDs concealed beneath the headlight and taillight lenses. A hair-thin magnet wire runs through the roof and body, transferring power without adding mass. The headlights turn on as you go forward. When you change into reverse, the rear lights brighten. Direction sensing guarantees that the lights function in the same way as a full-size car does.
Building this micro car required a great deal of patience because each stage required extreme focus. The interior was removed to provide way for the construction process, and each mechanical part had to be precisely positioned and tested before being altered repeatedly until everything fit together neatly while still leaving room for the batteries and cables. The underside of the chassis contains terminals for charging the battery and reprogramming the controller, making it possible to do so without disassembling anything. [Source]
NTU researchers spent seven years building a magnetic robot just 4.4 millimeters long. The compact machine performs five surgical functions through external control alone. It travels across soft tissue, cuts when required, dispenses medicine, gathers samples, and creates localized heat. Work on the project took place in the School of Mechanical and Aerospace Engineering under Associate Professor Lum Guo Zhan.
Nicholas Yong Wei Foo, a PhD student, led the project’s hands-on development. Alumni and regulars, including Dr. Chelsea Shan Xian Ng and Yu Xuan Yeoh, made substantial contributions during the design and testing phases. The researchers published their findings in Advanced Materials after being funded by NTU funds, A*STAR, and the NHG Health group. NTUitive has received a technology disclosure and is poised to proceed. Almost every other magnetic robot of this size can only perform one or two tricks; this one can perform five, owing to a programmable core module.
With this set, based on the legendary Transformers Autobot, adults build LEGO Optimus Prime figure that converts from robot to truck and back
Relive the Transformers saga with accessories, including the ion blaster, Autobot Matrix of Leadership, an Energon axe, Energon cube and jetpack
This unique collectible robot model makes a great home décor piece – top off the build with a unique display plaque with Optimus Prime facts
The module’s magnetic state can be changed with a few external coils; just magnetize, demagnetize, or reverse the direction, and each condition corresponds to a specific tool or action. The innovative aspect is that zones in the body are designed to only illuminate active areas while leaving the rest dark. According to tests, a complete flip takes less than a second. Getting around is the first trick, but it is far from the only one. This robot can crawl on soft, uneven ground that replicates surfaces found deep within the body. It can also rotate along its long axis, which is important for achieving perfect alignment when things become tight or the surface starts to slope and fold. There’s also a blade that pops out for cutting, and in lab tests, it effortlessly sliced through chicken liver and gelatin models, or at least models designed to imitate the inside of a human.
Advertisement
When it comes to medicine delivery, the robot’s dispensing arm simulates drug delivery by using preloaded particles. The operator merely needs to maneuver it to the right spot and then press the release. The robot collects samples by utilizing a gripper to grab and secure tissue until it is ready for laboratory analysis. Test results showed that the gripper performed effectively on the models tested.
Then there’s heat generation, which is a different magnetic process that warms up a small region by creating high-frequency alternating fields that force the particles inside to warm up. The idea is that this might be used for targeted therapy, raising the temperature in a specific area without damaging the surrounding tissue.
The robot’s body is a clever combination of two flexible silicones, PDMS and Ecoflex, with magnetic particles that are barely five micrometers across. It has just enough give to bend and twist as needed while yet preserving enough shape to complete the task. Because there are no batteries, wires, or electronics inside the robot, all guidance and tool activation occurs from the outside. Because the robot is designed to be tiny and simple, all commands are given from the doctor’s control station.
Lab studies on actual biological samples showed that the design choices worked successfully. The robot completed all of the tasks on the samples, and subsequent tests comparing the materials to human skin cells found that more than 99% of the cells were still alive and active, which is optimistic for future development. According to Associate Professor Lum, most magnetic robots this size can only do one or two tasks. [Source]
A big civil rights deadline that impacts schools and vendors will hit this month.
Federal law has required accessibility for people with disabilities for decades, says Glenda Sims, chief information accessibility officer at Deque Systems, a company that specializes in digital accessibility.
But two years ago, the federal government finally gave schools a way to measure whether their websites, mobile apps and digital content were accessible under law when it released a “final rule.”
In essence, the final rule updated 2024 Title II of the Americans with Disabilities Act, a federal law concerning equal opportunity, setting out standards for public institutions around website and mobile app accessibility. When the deadline was put in place, disability experts told EdSurge that the rules provided clarity for schools and edtech vendors, and also set a ticking clock for when they would have to make changes. The rule set varying deadlines for school districts and state and local governments — in April 2026 or April 2027, based on population size.
Advertisement
On April 24, the first deadline will hit. By then, institutions have to make their web content and mobile apps comply with Level AA of the Web Content Accessibility Guidelines (WCAG) 2.1, a widely recognized accessibility standard that includes accommodations such as a minimum contrast ratio and a requirement for audio descriptions.
But with the well-advertised deadline just days away, schools are well behind schedule.
Some advocates worry that digital accessibility is being swept up in broader political trends. So, what happens when the deadline hits?
Not Ready for Prime Time
Only 14 percent of districts had completed the accessibility updates required by law, according to a survey from the National School Public Relations Association released last December. The survey also found fewer than half of districts prioritized digital accessibility or had procedures for vetting vendor accessibility, which is required by the rule.
Advertisement
It’s not just about course content, but also the apps that a school may use, says Sambhavi Chandrashekar, global accessibility lead at D2L, a company that runs a widely used learning management system. “I doubt if a single K-12 district in the U.S. or anywhere else has an inventory today of all the web apps and forms and content that they have that are not accessible,” Chandrashekar says.
NEWSLETTERS
STAY AHEAD IN EDUCATION.
Sign up for EdSurge newsletters for timely news, insights and analysis.
Figuring that out requires performing an audit, which most schools likely haven’t done and which can be expensive, she adds.
Advertisement
At EdSurge’s request, AAAtraq, a company that sells disability-related legal compliance services, surveyed around 20 of the largest schools across a number of states — in California, Colorado, Florida, Illinois, New York, Texas and Washington state. Many school websites and online PDFs failed along “basic accessibility fundamentals,” based on a benchmark the company uses to assess legal exposure. Alt text was missing, there was not enough color contrast and many websites didn’t have an accessibility statement, the company reports. The company found that 88 percent received an “F,” the lowest possible grade.
“Title II should have been a wake up call,” said AAAtraq CEO Lawrence Shaw in an emailed comment, referring to the major disability law behind the “final rule.” Yet many schools, including some of the largest in the country, have left themselves open to legal action.
Digital Exhaustion
Schools’ relationship to technology has also changed since two years ago, from rushing to embrace it to trying to limit it.
Advertisement
These days, beset by digital exhaustion and regret over the reach of tech into children’s lives, schools have sought to restrict screens in schools.
But it’s important for schools and lawmakers to distinguish between meaningful tech and doomscrolling on social media, says Luis Pérez, senior director of disability and accessibility for CAST, a digital access advocacy group. Students are under more pressure to manage their own attention, Pérez says, but those with disabilities and multilingual learners rely on certain digital tools, such as text-to-speech and adjustable text sizing to navigate daily learning. When used correctly, digital tools that expand accessibility can foster a sense of belonging, especially for underrepresented groups.
He worries that screen time laws that lump all screens together could make digital accessibility harder.
K-12 schools may be having the toughest time. Universities are usually more prepared for digital accessibility than state or local governments, which run K-12 public schools, says Sims of Deque. That’s partly because students with disabilities represent a more identifiable group in universities and that allows them to advocate for accommodation, she says.
Advertisement
These schools are heavily reliant on vendors for accessibility, Sims says.
It doesn’t help that there’s uncertainty at the moment.
Old Rules, New Rulers
While the accessibility deadline is still in place, the intentions of the federal government have become murky.
Last year, the Department of Justice signaled that it might issue a new “interim final rule” that would impact the deadline. And recently, the Office of Information and Regulatory Affairs — a federal agency that is usually not involved with accessibility — has been holding meetings on the rule, as “credible rumors” have circulated that the rule is in danger of getting delayed or scrapped.
Advertisement
Yet, the federal government has not publicly released information about its intentions, according to Jarret Cummings, senior adviser for policy and public relations at Educause.
The Office of Information and Regulatory Affairs did not immediately respond to a question from EdSurge about whether a delay is expected.
However, some documents related to the meetings are publicly accessible, giving a glimpse into what they are hearing.
A group representing more than 800 Minnesota cities argued in written testimony that none of the Minnesota cities that would be impacted by the rule are fully compliant with the law. The letter states that the cost of compliance would squeeze small government budgets. In a similar argument, testimony from the National Association of Counties estimated that it would cost small counties about $32,000 to fix problems with accessibility on their sites, and large counties as much as $700,000.
Advertisement
Cummings’ organization, Educause, has also argued that two years was not enough time for most higher-ed institutions to make changes. It suggested that the government alter the timeline.
In contrast, Mark Riccobono, president of the National Federation of the Blind, testified that the rulemaking process has been ongoing for decades, with ample time for comment. The bill represents a compromise that clarifies rules, while reducing the burden of those under the law by providing exceptions and generous timelines, Riccobono argued.
Politically, the national mood has changed since the rule was issued a couple of years ago.
The affiliation of accessibility with diversity, equity and inclusion has politically backfired under the Trump administration. The administration has shredded grants it has identified with “radical” DEI ideology, and mass firings have gutted agencies like the Education Department, which the administration is actively trying to dismantle.
Advertisement
For students with disabilities, it means that there’s no guarantee of federal support, even when a federal complaint is filed.
“I would say that so many of the places that were reasonably staffed… have been reduced to almost bare bones, nothing. And so even if there are complaints coming in, there’s no way to truly handle them,” says Sims, of Deque.
Indeed, mass firings have led to 90 percent of all student civil rights complaints, including from students with disabilities, being dismissed by the federal government in the second half of last year, according to a nonpartisan government watchdog report published in January.
In the absence of federal help, people with disabilities have turned to the courts. There were more than 3,000 accessibility lawsuits filed in federal court last year, according to legal analysis of court data.
Advertisement
Long-term Goals
Pérez of CAST maintains that advocates should keep on track, focusing on long-term strategy, no matter what happens at the federal level. Accessibility benefits everyone, regardless of their background or disability status, he says.
Sims, of Deque, has also made a “business case” for considering accessibility during the design of products, suggesting that as schools embrace accessibility, the vendors that can show they build accessibility into their products will be rewarded.
Some hope that artificial intelligence tools will help students with disabilities access information on their own, and point toward tools like Aira, an AI tool that aids in remote video interpretation for people with visual impairment.
But even there, disability law experts insist that the federal rule hasn’t actually changed. “The rule is the rule until it isn’t,” wrote Lainey Feingold in early March.
Tuesday’s patch bundle also fixed MiniPlasma, a separate vulnerability disclosed by Nightmare Eclipse. Microsoft said in an email that the vulnerability is tracked as CVE-2020-17103, a vulnerability Microsoft first fixed six years ago. That means MiniPlasma was the result of a regression or an incomplete patch in its initial form. The company is in the process of updating Tuesday’s bulletin to note the republication.
Microsoft has yet to release patches for other vulnerabilities disclosed by Nightmare Eclipse. The company did provide manual instructions for mitigating YellowKey, a vulnerability that allows attackers to defeat Bitlocker full-disk encryption. That could be a boon when attackers have physical access to a device (the precise scenario Bitlocker is designed to protect against). The company has yet to fix the underlying cause of the vulnerability.
The status of other vulnerabilities disclosed by Nightmare Eclipse are also unclear at the moment. The researcher named one vulnerability, present in Windows Defender RedSun. Another, named BlueHammer, is also a local privilege escalation flaw that provides SYSTEM rights.
Over the past few months, Nightmare Eclipse has taken multiple potshots at Microsoft. The specific criticisms remain unclear, but many make references to complaints about the company’s vulnerability disclosure program. Microsoft, in turn, has publicly railed against the researcher for “not responsibly” disclosing the vulnerabilities and made a vailed reference to the possibility of pursuing legal action. After a public backlash, Microsoft later relented and vowed no such legal action would occur.
Advertisement
On Tuesday, Nightmare Eclipse published exploit code for a new Windows vulnerability. It’s a race condition that targets Defender.
Tuesday’s patch batch included fixes for roughly 200 vulnerabilities. Notwithstanding the appearance that MiniPlasma was fixed, two of them were also confirmed as zero-days.
Post updated to include information Microsoft provided after initial publication of this post.
HiFiMAN Arya WiFi arrives as the more affordable sibling to the HE1000 WiFi, bringing open-back planar magnetic drivers, WiFi streaming, Bluetooth connectivity, and a very different approach to wireless listening in the high-end headphone category.
At $1,449, the Arya WiFi is less expensive than the HE1000 WiFi ($2,699), but nobody should mistake it for an inexpensive experiment. This is still a premium audiophile headphone with a very specific promise: better wireless sound quality by moving beyond the usual Bluetooth-first approach and letting listeners stream over WiFi at home or in the office.
That raises the question many people will ask right away: if the HE1000 WiFi delivered an exceptional listening experience but also revealed an operational issue that gave us pause, has HiFiMAN solved enough of the user-experience problem with the Arya WiFi to make the concept feel ready for prime time?
Because at this price, clever engineering is not enough. The Arya WiFi has to sound excellent, feel comfortable, connect reliably, and prove that WiFi headphones are not just another audiophile science project with ear pads.
Advertisement
HiFiMAN Arya WiFi (left) vs. HiFiMAN HE1000 WiFi Headphones (right)
Arya WiFi Specs & Technology
Inside the left earcup of the HiFiMAN Arya WiFi is the circuitry that allows the headphone to operate without a traditional source component, DAC, headphone amplifier, or cable hanging off your desk like audiophile spaghetti.
The core of the system is HiFiMAN’s 8mm HYMALAYA Mini DAC, which is specified at 0.0055% THD+N and 105dB channel separation. HiFiMAN also integrates the wireless streaming hardware and headphone amplification inside the earcup, which is rather impressive considering the Arya WiFi remains an open-back planar magnetic headphone weighing 452 grams.
The broader technology package is very similar to the HE1000 WiFi. Both models support WiFi, Bluetooth, and USB Audio operation, along with PCM playback from 44.1kHz to 768kHz, native DSD64 to DSD512, and Bluetooth codec support for SBC, AAC, aptX, aptX HD, and LDAC.
The key difference is the driver. The Arya WiFi uses HiFiMAN’s NEO Supernano Diaphragm Gen.2, while the HE1000 WiFi uses HiFiMAN’s nanometer-thickness diaphragm. HiFiMAN positions the HE1000 WiFi as the higher-resolution model, but the Arya WiFi shares much of the same wireless architecture at a significantly lower price. That makes the comparison more interesting than the price gap might suggest.
Advertisement
The Arya WiFi uses HiFiMAN’s planar magnetic driver platform with Stealth Magnets, a design intended to reduce wave diffraction as sound passes through the magnet structure. HiFiMAN says the result is lower distortion and improved clarity, but as always, the listening section is where the marketing language either earns its keep or quietly exits through the side door.
WiFi is the real point of difference here. Bluetooth remains convenient, but even the better codecs are still limited by available bandwidth and device support. By adding WiFi streaming, the Arya WiFi can support lossless and high-resolution playback in a way that conventional Bluetooth headphones cannot.
For listeners using TIDAL, Qobuz, Roon, or local high-resolution files, that matters. For Spotify users, the benefit is less about lossless playback and more about connection flexibility, convenience, and avoiding some of Bluetooth’s usual compromises.
Advertisement. Scroll to continue reading.
You can also use the Arya WiFi in wired mode via USB-C, or connect it to your playback device via Bluetooth thanks to the embedded Qualcomm QCC5181 chip, which is LDAC-compatible for up to 96kHz playback.
Advertisement
Design & Comfort
The HiFiMAN Arya WiFi follows the same general design language as the HE1000 WiFi, but with a different finish. The Arya WiFi wears a blacked-out look, while the HE1000 WiFi uses the more upscale brown and metallic color scheme. Black is the safer choice and probably the easier one to live with, unless your listening room already looks like a cigar lounge with a DAC budget.
The Gen.2 headband is comfortable and offers useful adjustment, although it does not provide the same full cup rotation found on some passive Arya and HE1000 models. That limits portability and flat storage, but this is still a better design than HiFiMAN’s older Sundara-style headband or the more basic strap-free assemblies used on models such as the HE400se and HE6se V2.
Build quality is good, but the Arya WiFi does not suddenly become a jewelry-grade headphone because WiFi has joined the party. There is still visible plastic, and the materials do not feel as premium as the price might suggest. That said, at $1,449 and positioned well below the HE1000 WiFi, those compromises are easier to accept.
Comfort remains one of the Arya WiFi’s strengths. The large open-back earcups, suspended headband, and familiar HiFiMAN pad shape make it easy to wear for longer sessions, provided you are comfortable with a firmer-than-average clamp. It is not featherlight, but the weight is distributed well enough that the Arya WiFi never feels like punishment for choosing wireless planar magnetic headphones.
The button layout is the same as the HE1000 WiFi. From top to bottom, there is a volume rocker, a function button for switching between WiFi, USB, and Bluetooth modes, and a power button. The buttons use different colors to indicate operating mode, charging status, and battery state. It is straightforward once you learn the color system, although there is still a short adjustment period.
The layout itself is simple and intuitive, and the buttons are easy enough to find by touch. My main issue was with the voice prompt, which was extremely quiet on my review sample. I cannot say whether that is specific to this unit or representative of the product more broadly, but it made mode confirmation less useful than it should be. Volume adjustment also felt too coarse for my liking, with each step creating a larger change than I would prefer.
Advertisement
HiFiMAN keeps the included accessories minimal. The Arya WiFi includes a USB-C to USB-A cable roughly six feet long, and that is essentially it. There is no display case like the one included with the HE1000 WiFi. At $1,449, that does not change the performance case for the headphone, but it is worth noting for buyers expecting a more premium unboxing experience.
Setup & Usability
The first thing I tried to do after unboxing the Arya WiFi was connect it to my home WiFi network. That process proved more frustrating than expected.
Setup requires a specific sequence of button presses on the earcup, entering URLs on a smartphone, navigating web pages that feel unfinished, and entering more than one password before the headphones can join the network. If everything works the first time, the process may only take a few minutes. In my case, it repeatedly failed at the final stage when entering my router password, for reasons that were never clear. After roughly 20 minutes, I was eventually able to complete the connection.
That matters because WiFi is the feature that separates the Arya WiFi from a conventional wireless headphone. If users struggle with the setup process, some will likely default to Bluetooth or USB-C operation instead. Those modes are useful, but they also reduce the value of buying the WiFi version in the first place.
HiFiMAN already has the GAIA app, and this seems like the obvious place to guide users through WiFi setup in a clearer and more polished way. A step-by-step setup flow inside the app would be preferable to asking users to rely on a web-based process or a support video. For a company selling a $1,449 wireless planar headphone built around WiFi streaming, the setup experience needs to feel more mature.
Advertisement
Advertisement. Scroll to continue reading.
Battery life was also a concern. HiFiMAN rates the Arya WiFi at 6.5 to 7.5 hours in WiFi mode and up to 23 hours over Bluetooth. In my testing, WiFi playback landed closer to 5 to 6 hours, which is below the rated figure but not dramatically far off. Bluetooth battery life was more disappointing, coming in around 12 hours, well short of the claimed 23 hours. Battery life will vary with volume level, codec, connection stability, and usage conditions, but that gap is large enough to be noted.
Listening
The Arya WiFi and HE1000 WiFi look closely related, feel very similar on the head, and share much of the same internal wireless architecture. The obvious question, then, is whether they sound the same.
The answer is almost, but not quite. The two models are clearly cut from the same cloth, and the Arya WiFi gets closer to the HE1000 WiFi than the price difference might suggest. That makes the less expensive model a more interesting proposition than expected.
Among the wireless headphones I have heard, HiFiMAN’s WiFi models are among the most technically capable, even when used over Bluetooth. For this review, however, most of my listening was done in WiFi mode, using a mix of Spotify streams and high-resolution FLAC files. That matters because WiFi playback is the main reason these headphones exist, and it is the mode that gives the Arya WiFi its best chance to separate itself from conventional Bluetooth designs.
Advertisement
Bass
The Arya WiFi has a slightly elevated bass response, but it is not tuned like a consumer noise-cancelling headphone trying to win a fistfight in the low end. The bass sits just above the rest of the spectrum, giving music some added weight, warmth, and physical presence without overwhelming the midrange.
It is not the fastest or most controlled bass I have heard from a planar magnetic headphone, but that comparison mostly involves high-end wired models with dedicated amplification. Against other wireless headphones, the Arya WiFi is much stronger. Bass notes start and stop cleanly, texture is easy to follow, and there is very little sense of bloat or overhang.
Sub-bass extension is also impressive for an open-back wireless headphone. The Arya WiFi can reproduce very low information with useful audibility, including material around the 20Hz region, although there is some slight roll-off as it approaches the lowest octave. On Hans Zimmer and James Newton Howard’s “Why So Serious?,” the deep rumble around the 3:30 mark was clearly present, which is not something every open-back headphone manages convincingly.
Midrange
Moving up from the bass, the transition into the lower midrange is smooth and well controlled. Male vocals do not sound thickened or pushed back by excess upper-bass energy, which helps the Arya WiFi maintain good clarity through the lower registers.
The midrange itself is slightly relaxed compared to the bass and treble, but I would not call the Arya WiFi a strongly V-shaped headphone. It has some extra energy at the frequency extremes, while the center of the presentation remains smooth, open, and even-handed.
Advertisement
That character works in its favor with vocals and acoustic instruments. The Arya WiFi does not push singers forward in the way some midrange-focused headphones do, and listeners who prefer a more relaxed presentation may find it easier to live with over longer sessions. Compared with something like the Sennheiser HD600, the Arya WiFi is less intimate and less midrange-forward, but it still presents vocals and instruments with convincing timbre and very good tonal balance.
On Dominique Fils-Aimé’s “Birds,” breath detail, vocal texture, and the space around her voice were easy to hear without sounding exaggerated. The Arya WiFi does not put the singer right in your lap, which is probably healthier for everyone involved, but it does a strong job of preserving the natural shape and presence of the performance.
Treble
The Arya WiFi follows the HE1000 WiFi with a treble balance that is more forgiving than some of HiFiMAN’s older or brighter-leaning models, including the Arya Organic and HE6se V2. It still sounds open and detailed, but it avoids the sharper edge that can make long listening sessions feel like a test of character.
Advertisement. Scroll to continue reading.
Advertisement
There is enough presence in the upper registers to show what the planar magnetic drivers can do, particularly with cymbal decay, small percussion details, and the air around instruments. Fine detail is easy to follow, but the Arya WiFi does not push treble information forward just to create the impression of resolution.
“La lune” by L’Impératrice is one of my regular treble test tracks because of the faint triangle hit that runs through much of the song. The Arya WiFi keeps that detail audible within the mix without turning it into a distraction. It has the clarity to reveal the upper-frequency information, but enough restraint to keep the presentation smooth and listenable.
Soundstaging & Imaging
HiFiMAN’s egg-shaped open-back headphones are known for producing a wide, spacious soundstage, and the Arya WiFi continues that pattern. The presentation is larger than average for a wireless headphone, with good width and a convincing sense of height, although it does not sound quite as expansive as some of the wired Arya variants, including the Arya Organic and Arya Unveiled.
Imaging is the stronger part of the presentation. The Arya WiFi places sounds with very good precision, and that matters more to me than absolute stage size. On TOOL’s “Chocolate Chip Trip,” the headphone made it easy to follow the individual layers, percussion hits, and spatial effects without turning the mix into a blur. That track can become a mess on less capable headphones. The Arya WiFi kept it organized.
Advertisement
Detail retrieval is also strong. The Arya WiFi does not leave much feeling hidden, especially in WiFi mode with high-quality source material. The HE1000 WiFi still has a slight edge in resolution and low-level information, but the gap is not large. You are most likely to notice it during direct comparisons, not during normal listening. That is probably the healthier way to experience music anyway.
The Bottom Line
The HiFiMAN Arya WiFi makes a strong case for itself because it delivers much of the HE1000 WiFi experience at a significantly lower price. It does not quite match the more expensive model for resolution or refinement, but the gap is small enough that many listeners will find the Arya WiFi the more sensible buy.
Sound quality is the main reason to consider it. The Arya WiFi offers a spacious presentation, strong imaging, extended bass, smooth mids, and a treble balance that is detailed without becoming aggressive. In WiFi mode, it feels like a meaningful step beyond most Bluetooth headphones, especially with lossless or high-resolution material.
The problems are not sonic. The WiFi setup process needs to be more polished, and battery life is only fair in WiFi mode and disappointing over Bluetooth based on my testing. Those issues matter because they affect daily use, and they make the Arya WiFi feel less finished than it should at $1,449.
Still, if you can live with the setup process and the need for frequent charging, the Arya WiFi is the better value in HiFiMAN’s WiFi headphone lineup. The HE1000 WiFi remains the more capable headphone, but the Arya WiFi gets close enough where the extra money becomes harder to justify.
Advertisement
Pros:
Built-in WiFi streaming supports higher-quality wireless playback than conventional Bluetooth headphones
Can also be used over USB-C or Bluetooth
Comfortable fit with good weight distribution for longer listening sessions
Spacious soundstage, strong imaging, and excellent detail retrieval for a wireless headphone
More affordable than the HE1000 WiFi while retaining much of the same wireless architecture
Cons:
Build quality is good, but the visible plastic and limited accessories do not fully match the price
First-time WiFi setup is more complicated and less polished than it should be
Battery life fell short of HiFiMAN’s stated figures in testing
Bluetooth mode is useful, but does not deliver the main benefit of buying the WiFi version
You must be logged in to post a comment Login