AMD unveiled a suite of new data centre products at its Advancing AI event in San Francisco on Thursday, claiming they will outperform Nvidia’s competing hardware across AI training and inference. The announcements included the MI455X AI accelerator, the Helios server rack that packs 72 of those chips into a single system, and the Venice generation of Epyc server processors built on TSMC’s two-nanometre process. AMD predicted the total market it is pursuing will reach $2 trillion by 2030, with AI accelerators alone accounting for more than a trillion dollars of that figure.

Investors were unimpressed, sending AMD shares down about four percent during the presentation despite the stock having more than doubled in value this year. The scepticism reflects the gap between AMD’s ambitions and Nvidia’s dominance: Nvidia’s Vera Rubin platform is already in full production and shipping to customers including OpenAI, CoreWeave, and Microsoft. AMD CEO Lisa Su has delivered a remarkable growth run, but Wall Street’s expectations were already baked into a stock that had risen roughly 145 percent year to date heading into the event.

Su brought executives from major AI companies on stage to endorse AMD’s hardware. OpenAI’s Sachin Katti said his company expects to deploy Helios “at massive scale” as it scrambles to add infrastructure capacity. The event came a day after AMD announced it would invest up to $5 billion in Anthropic and deploy two gigawatts of MI450 series GPUs in Helios racks to run Claude, with the first gigawatt shipping in the first half of 2027.

AMD also announced a partnership with Cerebras, whose wafer-scale chips specialise in ultra-fast AI inference, to combine Helios racks with Cerebras servers for customers who need the lowest possible latency. AMD hardware will handle the work of deciphering queries while Cerebras chips provide rapid answers. Cerebras CEO Andrew Feldman said the combined product will ship from Cerebras-owned data centres starting in the fourth quarter and will beat a similar offering Nvidia is assembling from its recent Groq acquisition.

The Venice Epyc processor, meanwhile, is AMD’s answer to Nvidia’s push into server CPUs with its Vera chip, which Nvidia has claimed gives it a performance edge. AMD said Venice will keep it well ahead of Vera, setting up a direct contest that independent benchmarks have yet to settle. The rivalry now spans accelerators, server processors, and complete rack-scale systems, with both companies pitching end-to-end data centre solutions rather than individual components.

SYSTEMS

The enemy of my enemy is my friend

GPUs are great for training, but for inference, you need a heavy dose of speedy memory to churn out the tokens. AMD has tapped Cerebras Systems to develop a disaggregated compute platform combining Instinct GPUs with the chip startup’s SRAM-powered AI accelerators. The goal: to deliver ultra-low-latency inference for agentic workloads.

The collaboration, announced on stage during AMD CEO Lisa Su’s Advancing AI keynote Thursday, closes a gap in AMD’s portfolio that cost Nvidia $20 billion to acquihire from Groq back in December.

Cerebras CEO and cofounder Andrew Feldman is no fan of Nvidia, having previously denigrated the GPU giant as a mere AI arms dealer. And unlike GPUs, Cerebras’ wafer scale engines (WSE) don’t rely on HBM4 but instead use on-chip SRAM that’s orders of magnitude faster.

This has made Cerebras one of the fastest inference providers in the world, with output speeds often exceeding 2,000 tokens a second.

By running compute-heavy prompt processing operations on AMD’s Instinct GPUs and offloading the memory intensive token generation to Cerebras’ WSE accelerator, the duo aims to achieve higher interactivity without compromising on throughput or cost to do it.

“What you have with Instinct and the Helios rack is you have the leader in performance and memory capacity. And you marry that with our Wafer Scale Engine, which is the leader in SRAM and in memory bandwidth, and that combination allows us to deliver a solution that is unmatched,” Feldman said on stage.

Neither company has shared specific figures, but the combination is expected to boost the number of tokens per second generated per watt of electricity consumed by as much as 5x.

If any of this sounds familiar, Cerebras’ accelerators fill the same role as the Groq 3 LPUs (Language Processing Units) announced alongside Nvidia’s Vera Rubin rack systems at GTC in March.

But where Nvidia needs two thousand Groq LPUs worth of SRAM to serve a trillion-parameter model like Kimi K2.5, AMD and Cerebras will need at most a few dozen.

The combined offering will be available in Cerebras Cloud later this year, but may not be AMD’s last deal with the upstart.

“There are lots of ways to get workload-specific acceleration done, and I think Cerebras has a very interesting technology. It works very well with Helios,” Su said during a press conference following the keynote. “The idea of our open ecosystem is frankly that we will work with a number of different companies that may have technology that could be useful.”

“You can expect that we’re going to do more workload disaggregation going forward,” she added. ®

On Thursday morning, Meta published an ad to its official Instagram account with the message that artificial intelligence will not “make us less connected” or “leave us behind.” It was set to the 1972 David Bowie track “Five Years,” which is about, among other things, the end of the world.

“Call us optimists, call us dreamers, call us whatever the hell you want—but we’re betting on people, and we like those odds,” the narrator says, as a toddler plays with a butterfly, two parents laugh with their infant, and Jalen Brunson waves at the Knicks Championship Parade. (The ad also takes special care to highlight use cases for its controversial smart glasses that are not, say, harassing women while surreptitiously recording them.)

The snippet of “Five Years” used in the ad includes the following lyrics:

I heard telephones, opera house, favorite melodies

I saw boys, toys, electric irons and TVs

Advertisement

My brain hurt like a warehouse, it had no room to spare

I had to cram so many things to store everything in there

And all the fat, skinny people

And all the tall, short people

Advertisement

And all the nobody people

And all the somebody people

I never thought I’d need so many people

Perhaps understandably, it skips other parts of the song, including the opening:

News had just come over
We had five years left to cry in
News guy wept and told us
Earth was really dying

When reached for comment, Meta spokesperson Francis Brennan said he “would like to highlight what David Bowie himself said about ‘Five Years’ and its meaning: having optimism for the future.” Brennan sent along a 1972 quote about the song that he sourced to BowieBible.com.

“People are so incredibly serious and scared of the future that I would wish to turn the feeling the other way, into a wave of optimism,” reads the quote, which is itself sourced from NME Magazine. “If one can take the micky out of the future, and what it is going to be like … It’s going to be unbelievably technological.”

Nevertheless, BowieBible.com also notes that the song “speaks of a dystopian nightmare in which Earth enters its final five years, for reasons undisclosed, and the ensuing panic, violence, and attempts at redemption.”

BowieBible.com also quotes from a 2018 interview with drummer Mick Woodmansey: “Well, we chatted about it, and we knew it was about the end of the world and pretty depressing,” Woodmansey said of the song. “I remember when we recorded it, he was actually in tears. He was doing the vocal and he was actually crying.”

According to Jeremy D. Larson, the deputy director of Condé Nast sister publication Pitchfork, taking songs out of context for commercials is a time-honored tradition. He points to a Carnival Cruise ad that uses an Iggy Pop song about heroin addiction. (He also says that since Warner Chappell Music owns Bowie’s catalog, they are the ones with discretion over where his songs end up.)

Ford will use Apple Maps as the built-in navigation system for its next generation of electric vehicles, replacing the Google-based software it has run since 2023. The integration uses Apple’s new MapKit for Automotive SDK and will launch with Ford’s $30,000 midsize electric pickup in 2027, according to a joint announcement from both companies on Wednesday. Ford will also feed Apple’s road-level mapping data into the development of its next-generation BlueCruise hands-free driving system.

The move is a notable reversal. Ford partnered with Google in 2021 to bring Android Automotive to millions of its vehicles, making it one of the highest-profile adopters of Google’s in-car software platform, but for its most important new product line Ford is now turning to Apple instead. CEO Jim Farley called the electric pickup a vehicle that “redefines what advanced technology can be,” and said Ford is “honoured to be among the first to embed” Apple Maps directly into a car.

The system will offer turn-by-turn directions with natural language search, real-time traffic and incident data, and EV-specific route planning that accounts for charging station locations, estimated range, and battery preconditioning. That last feature automatically warms or cools the battery before arriving at a fast charger, which can cut charging times significantly. Apple’s senior vice president of services, Eddy Cue, said the SDK lets automakers build a navigation experience that matches the look and feel of their own vehicle software rather than simply mirroring the Apple Maps app.

Beyond navigation, Ford’s Latitude AI team will use Apple’s road-level data to build a more seamless on-ramp-to-off-ramp experience for BlueCruise, the New York Times reported. That means Apple’s mapping data will help the hands-free driving system understand where highway interchanges, merges, and exits are located. Google’s Android Automotive platform has been gaining traction across the industry, but Ford’s decision to pick Apple for its flagship EV platform suggests the competition for in-car software is far from settled.

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!

The electric pickup launching on Ford’s Universal Electric Vehicle platform will be built at Louisville Assembly Plant in Kentucky, using LFP batteries and megacasting to hit the $30,000 target. CarPlay will remain available on Ford vehicles alongside the native Apple Maps integration. Ford has not said whether its existing models running Android Automotive will eventually switch to Apple Maps or whether the change applies only to UEV-based vehicles going forward.

A new Dolphin X remote access trojan claims to use an AI-powered profiling feature to score and rank infected users, helping cybercriminals identify which victims should be targeted first.

The malware was analyzed by Varonis Threat Labs researcher Daniel Kelley, who spotted it being advertised on a cybercrime forum by a vendor using the alias “Kontraktnik,” promoting it as an all-in-one remote access trojan.

According to Varonis, the operator panel lists 329 features across ten categories, including a credential-stealing feature that claims to target more than 300 applications.

However, one of its notable features is an “AI Profiler” that analyzes information collected from infected computers and assigns each victim a risk score.

“Beyond credential collection, the panel includes a surveillance tab containing the AI Profiler. The seller describes it as an ‘AI behavioral profiler with app usage tracking, risk score, and daily summary,’” explains Varonis.

Varonis obtained the Dolphin X operator panel and analyzed it in an isolated lab, noting they examined the malware builder and its network traffic rather than executing a live Dolphin X agent on an infected computer.

AI Profiler ranks victims for attackers

Credential-stealing malware can allow attackers to steal credentials for hundreds, if not thousands, of online accounts, making it difficult to manually review them all for high-value targets.

Dolphin X’s AI Profiler claims to automate this process by acting as a sorting system that scores, categorizes, and ranks infected computers so that the attackers know which are the most high-value to target further.

The operator panel claims that the AI Profiler can process victims’ application usage, risk scores and tags, browser domains, and installed software to produce ranked profiles.

These scores are given to attackers in daily summaries containing ranked victim profiles, allowing them to prioritize machines that may provide access to valuable accounts, cryptocurrency, corporate networks, cloud environments, or production systems.

“In practice, the feature appears designed to help operators triage victims,” explains Kelley.

Varonis researcher Daniel Kelley confirmed to BleepingComputer that the AI Profiler is present in the operator panel and discovered technical strings supporting the profiling workflow, including Auto-Start AI Profiler, ProfilerStart, ProfilerGetData, risk_score, risk_factors, and categoryusage.

The researcher said these strings indicate that the profiling workflow is actually included and that the panel can process the data needed to rank victims.

However, Varonis could not determine what artificial intelligence engine is being used to produce the rankings without analyzing a live Dolphin X malware sample.

The malware also operates as a credential stealer, with the operator panel showing that it targets more than 300 applications, including 9 Chromium and Gecko browsers, 100 cryptocurrency wallet extensions, 65 desktop crypto wallets, 10 password managers, and more than 30 cloud command-line tools.

Dolphin X also claims to steal .env files, SSH keys, cloud access tokens, browser login data, cryptocurrency wallet information, and other developer credentials.

As Varonis analyzed the Dolphin X operator panel, builder, and related network traffic rather than a live malware sample executing on an infected machine, the malware’s advertised collection capabilities were not independently confirmed by the researcher.

Artificial intelligence has become a popular tool among threat actors, with it being used to launch cybercrime services such as SpamGPT and AI agents conducting autonomous cyberattacks.

Dolphin X platform instead uses AI to solve an operational problem by processing large amounts of stolen data and automatically sorting infected users into highest-value victims.

Security teams log 54% of successful attacks and alert on just 14%. The rest move through your environment unseen.

The Picus whitepaper shows how breach and attack simulation tests your SIEM and EDR rules so threats stop slipping by detection.

Get the whitepaper

A Florida man is suing OpenAI and CEO Sam Altman, alleging the company’s chatbot, ChatGPT, provided him with incorrect and “dangerous” medical guidance regarding a health condition that nearly became fatal.

According to the lawsuit, Scott Winters, a 55-year-old pastor, often turned to ChatGPT for medical advice when he experienced health symptoms. The suit alleges the chatbot eventually diagnosed Winters with dysautonomia, a nervous system disorder, and “developed a personalized recovery plan with specific instructions for Scott on how to heal his body.”

But on July 13, 2025, Winters experienced a “massive pulmonary embolism due to multiple blood clots in both of his lungs that brought him to the brink of death,” according to the lawsuit, which has been shared by multiple publications. Doctors’ notes later contradicted ChatGPT’s alleged diagnosis.

“I do not believe he has any dysautonomia (self-diagnosed from the internet),” one note read, according to the complaint.

The suit alleges that ChatGPT went beyond its scope by providing medical advice, and that the chatbot “should have escalated crisis safeguards and directed” Winters to medical personnel. Instead, it allegedly “dissuaded him from seeking that help and instead offered him unqualified diagnoses and personalized treatment plans,” which the suit alleges led to the “life-altering medical emergency.”

OpenAI spokesperson Drew Pusateri told CNET in part that “ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis or treatment. Treating chatbots as the whole story behind people’s medical decisions or outcomes oversimplifies a much bigger challenge, and risks getting in the way of people accessing powerful new tools that can aid them in their health journey.”

According to the suit, the chatbot did at least on one occasion provide a disclaimer that read, “it is important to consult a healthcare provider to identify the underlying cause and receive appropriate treatment.” However, the suit alleges it subsequently began giving medical advice without a disclaimer.

“The lengthier Scott’s interactions with the tool, the fewer disclaimers it included in its messages,” the suit alleges. Winters is seeking financial damages based on these allegations.

Winters was using an older ChatGPT model, GPT-4o, which OpenAI has since retired due to concerns that the bot was too agreeable and could possibly cause harm. The same model was named in a separate lawsuit filed by a Canadian woman whose daughter died by suicide after interacting frequently with the chatbot. That case is still in litigation.

(Disclosure: Ziff Davis, CNET’s parent company, in 2025 filed a lawsuit against OpenAI, alleging it infringed Ziff Davis copyrights in training and operating its AI systems.)

AI and health advice

A recent study published in the scientific journal Nature Medicine found that when people interact with an AI chatbot for health diagnoses, it misses more than half of all medical diagnoses. This is despite the fact that large language models can now score comparably to passing the US Medical Licensing Exam on medical knowledge benchmarks.

The study notes that missed diagnoses may be at least partially attributed to users not providing enough information regarding symptoms.

According to a survey released in January and conducted by OpenAI itself, three in five US adults say they’ve used AI tools for health-related information in the past three months.

Corin Cesaric-Epple

Editor

Corin Cesaric-Epple is an Editor at CNET covering home and kitchen tech and meal kits, and reporting regularly on artificial intelligence. She earned her bachelor’s degree in journalism from the University of Missouri-Columbia. Before joining CNET, she covered crime at People Magazine and national and international news for NBC Local Television Stations.

See full bio

Death of Gaming: Just like any other company dealing in hardware products built from off-the-shelf components, Valve is facing some truly difficult choices when it comes to pricing. The company tried to fend off the chip shortage caused by the AI boom with a massive price increase, but potential customers apparently responded by removing themselves from the buying equation.

A recent report by Boiling Steam might confirm what most enthusiasts have always suspected: an overpriced Steam Deck is never going to sell well. In fact, Valve’s handheld machine might have already suffered a disastrous drop in sales. Despite the price hike, Valve is earning far less money than it did before.

Boiling Steam used the same methodology it employed to estimate Steam Machine sales, tracking data points from Steam’s own global list of best-selling games and products. The outlet compared the Steam Deck’s ranking throughout 2025, when it was often listed fairly high on the sales chart. Things changed in 2026, when Valve decided to retire the older – and cheaper – LCD models and offer only the OLED models at a significantly higher price.

At first, OLED Steam Decks appeared to sell out despite the $300 price hike. According to new estimates from Boiling Steam, sales have now undergone yet another dramatic change. Valve is selling 82% fewer units than it did in 2025. Steam Deck sales have allegedly fallen from 11,000 – 18,000 units per week to an average of 1,400 – 3,000 units per week.

Valve tried to offset the massive increase in component prices by passing the supply-chain surcharge on to customers. According to Boiling Steam’s estimates, the company is now suffering from a 72% drop in revenue as well. Despite the price hike, Valve might have significantly diminished the Steam Deck’s appeal as an enticing option for handheld gaming enthusiasts.

Boiling Steam speculates that the Steam Deck was an interesting option when it sold for $650 but is a complete waste of money when priced at around $1,000 for the 1TB model. Furthermore, affluent gamers looking for a handheld at that price point will likely gravitate toward more powerful machines, such as the MSI Claw 8 EX AI+ or the Asus ROG Ally X.

If the publication’s methodology checks out, we might soon see a dramatic collapse in Steam Machine sales as well. The IT industry is currently going through a series of massive transformations, which means it’s probably too early to say that the Steam Deck is completely finished. After all, Valve engineers recently confirmed that things can always get worse.

Elon Musk told investors on Tesla’s second-quarter earnings call that Optimus will be “the hardest product to scale manufacturing that we’ve ever made at Tesla,” warning that the humanoid robot has no existing supply chain and that every component is new. He said he wanted to “calibrate people correctly” on expectations, acknowledging that the initial production ramp will follow a normal curve but with a flat early stretch that could last longer than investors hope. Tesla also quietly removed references to Optimus mass production from its Q2 shareholder presentation, after using that language as recently as the first quarter.

The admission marks a sharp change in tone from previous quarters. Last year Musk said he would be “shocked” if Tesla was not producing 100,000 Optimus robots per month within five years, and in January 2025 he projected 50,000 to 100,000 units for 2026. By January of this year he acknowledged that zero Optimus robots were doing useful work in Tesla’s factories, with the more than one thousand Gen 3 units deployed across Fremont and Gigafactory Texas collecting training data rather than performing productive tasks.

Musk was candid about the engineering barriers, saying Optimus needs at least human-level dexterity before it can be useful outside controlled factory settings and calling the human hand “an incredible thing” that no robotics company has managed to replicate. Tesla must also solve memory, logic, and chip packaging challenges while securing AI processors from partners including Samsung and TSMC. The robot contains roughly 10,000 unique parts, and unlike Chinese competitors that can tap an existing smartphone supply chain for actuators and precision motors, Tesla is building its manufacturing base from scratch.

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!

The financial backdrop adds pressure. Tesla reported Q2 revenue of $28 billion, up 26 percent, but adjusted earnings of 33 cents per share badly missed the 51 cents analysts expected, and operating income fell 57 percent. Capital expenditure jumped to nearly $6 billion in the quarter, part of Tesla’s $25 billion full-year spending plan, pushing free cash flow to a negative $1 billion and sending shares down about seven percent in pre-market trading on Wednesday.

Musk’s compensation deal ties part of his payout to delivering one million Optimus robots within a decade, a target that now sits uncomfortably against his own warnings about how difficult scaling will be. Tesla has converted its Fremont Model S and Model X lines for Optimus production and is building a dedicated facility near Gigafactory Texas, but the company has stopped promising when volume output will begin. Tesla’s China president has described the Shanghai Gigafactory as a “golden key” to mass-producing Optimus, though that factory has not yet begun robotics manufacturing either.

‘Ottava’ has been given market authorisation for use in multiple general surgery procedures within the upper abdomen.

Healthcare giant Johnson & Johnson has been granted ‘de novo’ approval by the US Food and Drug Administration (FDA) for what the company said is the world’s first table-integrated soft tissue robotic surgical system.

‘Ottava’ has been given market authorisation for use in multiple general surgery procedures within the upper abdomen, including gastric bypasses, gastrectomies, appendectomies and hiatal hernia repair.

When integrated into an operating theatre, the system’s four robotic arms occupy up to 50pc less space than ‘boom- and cart-mounted’ solutions for such environments, the company said. The system can also be programmed to help simplify set-up and breakdown of an operating theatre, or manipulate the positioning of a patient during a procedure.

Ottava also features a curated range of surgical tools and has its own “digital ecosystem” to support surgeons working with it, according to Johnson & Johnson.

“We are pioneering a new category of surgical robotics with Ottava,” said Tim Schmid, executive vice-president and worldwide chair of medtech at Johnson & Johnson.

“This is the start of the next era in surgery as we deliver not just a new surgical robotics system, but a catalyst for fundamentally better surgical care that is informed by technology, guided by clinical insight and delivered with care.”

The company said that Ottava – which will initially launch commercially with select US customers while aiming to advance into other use cases and regulatory jurisdictions – is purpose-built for present-day surgical procedures and environments while building a pathway to “support more digitally enabled, data-driven surgery in the future”.

Hani Abouhalka, the company’s group chair for surgery and medtech, said: “By opening up the way we think about architecture, data and the very workflows of surgery itself, we have designed Ottava to deliver clinical and economic benefits, and support surgeons and hospitals as the future of surgery unfolds.”

Earlier this week, US neural engineering company Science Corporation received EU approval for its vision-repairing medtech implant, Prima, with FDA approval processes also ongoing.

Don’t miss out on the knowledge you need to succeed. Sign up for the Daily Brief, Silicon Republic’s digest of need-to-know sci-tech news.