• Thunderobot uses SSD storage to help its 64GB laptop run huge AI models
• The M7000 can reportedly run GPT-OSS-120B despite having 64GB of memory
• Phison's aiDAPTIV+ technology turns part of the SSD into an AI cache

Thunderobot has released the AI Master M7000, a 16-inch mobile workstation built around AMD's Ryzen AI Max+ 395 processor.

The M7000 comes with 64GB of LPDDR5X-8000 memory and Radeon 8060S integrated graphics for demanding workloads.

Its 768GB drive incorporates an 85GB caching layer designed to help AI workloads handle models exceeding available system memory.

AI workloads get an unusual storage assist

The M7000 uses LPDDR5X-8000 memory, while its processor provides 16 cores, 32 threads, boost speeds reaching 5.1GHz, and an XDNA 2 neural processor.

AMD rates the neural hardware at 50 TOPS, while the Radeon 8060S configuration reaches as many as 40 compute units for graphics and accelerated workloads.

Thunderobot says the device can run GPT-OSS-120B locally, despite having 64GB of physical memory available for applications and operating-system requirements.

Its unified memory architecture allows processor and graphics resources to access shared memory, supporting workloads that frequently move substantial model data.

The additional cache comes from Phison’s aiDAPTIV+ technology, which uses part of the internal storage to hold data associated with artificial intelligence workloads.

Rather than functioning as conventional RAM, this arrangement keeps selected information on the SSD when unified memory capacity becomes restrictive during demanding workloads.

Such information can include Key-Value cache, less-active mixture-of-experts components, and model data that applications may require again during processing.

Phison has also shown this storage approach with computers carrying considerably less memory, including a 32GB laptop running the same 120B model.

The technology addresses a limitation faced by large local models, where available memory can restrict the performance of certain workloads.

A 160W mobile workstation with unusually high specifications

Thunderobot gives the M7000 a 160W operating profile supported by two fans and three heat pipes for sustained processing.

The device also uses a 2560×1600 display running at 165Hz, with 500-nit brightness and complete sRGB coverage for color-sensitive work.

Power comes from a 99Wh battery, while the chassis weighs 2.45kg despite its workstation-oriented hardware and cooling requirements.

The display and cooling configuration add considerable hardware capability without changing the notebook’s emphasis on local computing.

For connectivity, this device includes OCuLink, USB-C supporting 40Gbps transfers, HDMI 2.1 FRL, 2.5Gb Ethernet, and a UHS-II microSD reader.

OCuLink provides another high-speed connection for compatible external hardware, giving users added expansion options beyond the laptop's built-in resources.

Thunderobot lists the M7000's Chinese retail price at 24,999 yuan, which converts to roughly $3,733 based on current exchange rates.

That figure places the device well beyond typical gaming budgets, despite the company's product lineup historically centering on gaming hardware.

This device instead reads as an artificial intelligence laptop, given how conspicuously its marketing and branding lean toward AI use cases.

Yet it remains surprising that the caching arrangement still falls short of matching genuine additional system memory in practice.

Via Videocardz

• Cache-to-Cache lets separate AI models exchange internal information without generating text
• A learned Fuser converts one model’s internal data for another
• C2C uses selective gating to control which layers receive information

Researchers from Tsinghua University have published a paper describing a technique that lets separate AI models exchange information without producing any text.

The method, called Cache-to-Cache (C2C), has already been accepted at ICLR 2026 and ships with open-source code available to developers.

It targets a specific inefficiency present whenever multiple language models work together inside a shared pipeline.

Skipping words entirely

When two AI models cooperate today, one has to turn its thinking into written sentences before the other can read them.

That writing step takes real computing time and throws away small details buried inside the first model's raw thinking process.

Every AI model keeps a working memory of everything it has processed so far, known technically as a cache.

C2C skips typed language entirely by letting one model pass that working memory straight into a second model's memory bank.

A small assistance program called a Fuser handles this handoff, reshaping and rotating the information so the second model can actually use it.

Different AI models store their memories using completely different internal layouts, sizes, and structures from one another.

Simply dumping one model's raw memory into another would likely confuse it or cause its answers to fall apart.

To prevent that, C2C includes a smart filter that decides which pieces of incoming memory are worth absorbing immediately.

Some internal layers accept the new information right away, while other layers keep reasoning independently without any outside interference.

According to the researchers, this setup makes AI models run between 100% and 150% faster during shared collaborative tasks.

That upper figure works out to roughly two and a half times quicker than the usual back-and-forth typing process.

The team also reports accuracy gains as high as 14.2% when models work together instead of operating entirely alone.

Compared against older setups where models still communicate through typed text, accuracy reportedly improved by 3.1% to 5.4%.

Why does this method have limits

This approach currently works only with open-weight models, since it requires direct access to a model's internal cache and layer structure.

Most popular AI tools, the kind ordinary people chat with online, hide those internal details completely from outside users.

That means everyday apps like certain chatbots cannot use this shortcut unless their own creators build it in privately.

Nobody outside these companies currently knows for certain whether anyone has started using a similar method internally.

The research team argues that typed language has always slowed machines down since it forces them to think like humans do.

That argument deserves some caution, since the same team that built the system also ran every test proving it works well.

Whether this speeds things up as much as claimed will depend on other external testing of the system.

• FEVM FA80G mini PC packs a desktop Ryzen 9 9950X3D inside a 2.6-liter chassis
• The FA80G pairs AMD’s gaming-focused processor with Nvidia’s 24GB mobile flagship
• Nvidia gives the RTX 5090 Laptop GPU 150W inside this compact system

FEVM has unveiled a new compact desktop system pairing the Ryzen 9 9950X3D processor with a 24GB RTX 5090 Laptop GPU.

The pairing sits inside FEVM's FA80G chassis, a compact system built around AMD's AM5 desktop platform rather than mobile silicon components.

The flagship processor contains 16 cores, reaches 5.7GHz and carries 144MB of cache through AMD’s 3D V Cache technology.

Ryzen 9 9950X3D and RTX 5090 pack serious power

FEVM allows the processor to receive up to 90W, while the accompanying RTX 5090 Laptop GPU can receive up to 150W.

Nvidia’s mobile graphics processor has 10,496 CUDA cores, 24GB of GDDR7 memory and 896GB/s memory bandwidth.

It also supports hardware ray tracing, DLSS 4 and Multi Frame Generation, giving the FA80G several features associated with current high-end PC gaming.

The desktop processor is equally important, as the 9950X3D is designed for high-performance gaming workloads rather than conventional mobile computing.

This mini PC also supports up to 128GB of DDR5 memory via two slots, while two M.2 2280 connectors support drives reaching 4TB.

FEVM separates processor and graphics cooling, assigning four heat pipes to each component across the device’s internal thermal arrangement.

The company rates full load noise at 42dB, with CPU temperatures below 85°C and GPU temperatures below 80°C.

For connectivity, this device includes HDMI 2.0, HDMI 2.1, two DisplayPort 2.0 outputs, USB-C, USB-A, an SD reader and dual 2.5GbE.

FEVM has the hardware for a serious M5 Ultra gaming comparison

In comparison to Apple’s top M5 Ultra Mac Studio, the FA80G takes a more conventional high-end PC approach to gaming hardware.

The M5 Ultra offers up to 36 CPU cores, 80 GPU cores, 512GB unified memory and 1.2TB/s memory bandwidth.

Apple also includes hardware-accelerated ray tracing, Neural Accelerators and a 32-core Neural Engine within its highest M5 Ultra configuration.

Those specs give the Mac Studio substantial graphics and compute resources for professional rendering, video production and artificial intelligence workloads.

For gaming, however, the FA80G CPU and GPU combo fits the Windows PC gaming environment more directly.

It also supports DLSS 4, ray reconstruction, frame generation and Multi Frame Generation across its RTX 50 series gaming hardware.

Apple supports advanced gaming technologies through Metal, MetalFX and its Game Porting Toolkit, including tools for bringing Windows games onto Mac.

That means the M5 Ultra is capable of modern gaming, but developers must still account for Apple’s Metal-based gaming environment.

The FEVM instead combines established Windows gaming hardware with Nvidia’s dedicated graphics stack and AMD’s gaming-focused desktop processor.

For a device specifically judged by gaming hardware, that makes the FA80G’s specification unusually aggressive for a 2.6-liter system.

The M5 Ultra remains a much larger compute platform in memory capacity and integrated resources, but the FEVM is built differently for gaming

The flagship configuration of FEVM's FA80G costs ¥22,299 (about $3,300), before adding memory and storage to the bare-bone system.

Via Videocardz

• A chatbot invented entirely false cargo details about a Chinese vessel
• Armed troops prepared to board the ship before anyone checked the source
• Military aircraft were already active when officials caught the error

A false intelligence assessment nearly triggered a US military invasion of a Chinese vessel after AI-assisted analysis wrongly identified its cargo.

The report circulated during the Iran conflict and prompted preparations to intercept the vessel before officials discovered serious problems with its underlying information.

Sources told CNN that armed personnel were preparing to board while military aircraft supported preparations, creating a potentially dangerous confrontation between major powers.

AI-assisted analysis produced the mistaken assessment

The episode began when an analyst examined intelligence concerning the ship's manifest and used a chatbot to interpret information.

Sources said the system combined publicly available material with classified signals information before producing an incorrect assessment about the vessel's cargo.

The analyst subsequently used AI again to prepare the findings in a standard intelligence format that was distributed among military officials.

Officials only examined the underlying information shortly before the planned operation, revealing that the reported cargo identification was wrong.

The source described the assessment as “entirely false,” while CNN reported that another source said it “almost started a war.”

The exact material that the chatbot misidentified has not been disclosed publicly, leaving important details about the original error unavailable.

The Pentagon and the military command responsible for Pacific special operations did not provide comments on the incident.

The episode comes as the US military expands its use of AI tools across intelligence, operations, administration, and battlefield decision-making.

In January, Defense Secretary Pete Hegseth announced an AI strategy intended to accelerate deployment across military missions and organizational functions.

“We will unleash experimentation, eliminate bureaucratic barriers, focus our investments and demonstrate the execution approach needed to ensure we lead in military AI,” Hegseth said.

The strategy includes programs for intelligence work, operational planning, enterprise functions, and wider access to generative AI systems.

The department later said more than 1.3 million personnel had used GenAI.mil, generating tens of millions of prompts within five months.

The incident exposes a difficult verification problem

The reported mistake involved more than an inaccurate chatbot response because its output entered a process used for consequential military decisions.

Once incorrect information appeared in a formal assessment, officials receiving the document had to determine whether its underlying evidence supported the conclusion.

There is no information on which verification procedures were applied before the assessment reached personnel preparing the maritime operation.

It also remains unclear whether the analyst knew that the chatbot had introduced an incorrect interpretation before the report circulated.

According to CNN, different military organizations use different AI systems, while officials have described varying practices for checking information produced by them.

That fragmented environment makes it difficult to identify a single verification process governing every AI-assisted intelligence product across the military.

AI tools can make military analysis faster and easier, yet their potential for significant errors requires strict human verification before operational decisions.

That need becomes more important when intelligence concerns a rival country, where an incorrect assessment involving a military vessel could trigger retaliation.

A mistaken attack could then escalate tensions between the countries involved, potentially producing a wider armed confrontation or prolonged political and military conflict.

• Two artificial whiskers give this tiny drone a working sense of touch
• Pressure sensors at the whisker base detect surface contact instantly
• Onboard software processes tactile data using only 34 kilobytes of memory

A research team from Delft University of Technology in the Netherlands has developed a tiny drone that uses artificial whiskers to navigate where conventional sensors struggle.

The system relies on physical contact rather than cameras or satellite positioning, allowing the aircraft to detect nearby surfaces through touch.

The researchers designed the technology for drones weighing below 100 grams, where adding substantial sensing equipment can quickly consume available capacity.

Tiny whiskers give the drone another way to sense surroundings

The device carries two slender filaments near its nose, angled upward and connected to miniature pressure sensors underneath.

When either filament touches an object, pressure changes provide information that helps the aircraft estimate its position relative to nearby surfaces.

That information allows the drone to move around obstacles, maintain contact with structures, and construct a basic representation of surrounding spaces.

The approach takes inspiration from rodents and other mammals, which use facial hairs to move through confined environments with limited visibility.

“Here, we aim to equip drones with rich tactile sensing — not for manipulation in the air, but for a novel concept of tactile navigation,” said Salua Hamaza, Associate Professor of Aerial Physical Interaction and Embodied Intelligence.

The system needed to be lightweight, process rapidly, and consume limited energy before tactile navigation could become practical aboard small aircraft.

The sensing system runs on only 34KB

A major hurdle was airflow, since passing wind can easily disturb whisker readings and confuse the pressure sensors beneath them.

To solve this, the researchers wrote a lightweight processing program that separates turbulence from genuine surface contact in real time with millimetric precision.

The complete onboard software occupies just 34 kilobytes of memory, keeping the computational requirements unusually low for an autonomous sensing system.

Traditional navigation software built for larger drones typically demands considerably more computing power than this compact whisker-processing system requires today

“We wanted to show that touch does not have to come at the cost of size or computational power,” said Chaoxiang Ye, a PhD candidate and researcher at Deft.

The small size allows the drone to interpret environmental information without depending on external processing.

The approach will not suit the fastest drones, but for small rescue units, sensing through touch may prove genuinely valuable.

This device is especially useful for environments where smoke, dust, darkness, or damaged infrastructure could interfere with cameras and other optical equipment.

One potential application of this drone is in collapsed buildings where rescue operations can prove to be difficult because of instability and tight spaces, and regular aerial sensing may be unreliable

Small drones could enter restricted spaces ahead of human responders, using physical contact to gather information about nearby structures and possible pathways.

The system could also provide another sensing method when GPS signals are unavailable or visual information becomes difficult to interpret.

Via TomsHardware

• Russia is testing ammunition that splits apart to confront small drones
• Russian testing found separated elements reaching critical components inside drones
• The ammunition fits standard rifles instead of requiring specialized counter-drone weapons

The Russian military is developing a multi-bullet cartridge designed to give standard Kalashnikov rifles a new way to engage small drones during combat.

The ammunition separates into several elements during flight, creating multiple projectiles from a single round fired through a compatible rifle.

Military expert Yury Knutov said the design is intended for troops who may encounter drones without having access to specialized firearms.

A multi-part cartridge for standard rifles

The reported cartridge uses the 5.45mm format and can be loaded into a standard AK-12 magazine without requiring a different rifle platform.

Its unusual construction keeps several elements together before firing, after which they separate while travelling toward the aerial target.

The concept has drawn comparisons with a Matryoshka doll because multiple components are contained within one projectile before separating in flight.

Pervy Tekhnichesky reported in April that Kalashnikov Concern had developed and tested the ammunition as a potential option for engaging small drones.

The reported trials included individual shots and bursts, providing an initial assessment of how the multi-part projectile performed against unmanned aircraft.

Some of the separated elements reportedly reached batteries, electronic boards, and power-related components during those tests.

Those components are critical to drone operation, making their reported involvement in the testing relevant to the cartridge's intended aerial application.

The ammunition is therefore being developed specifically around small unmanned aircraft rather than as a general replacement for ordinary rifle rounds.

Different performance against personnel

The cartridge has different limitations when used against personnel, particularly when protective equipment is involved during an engagement.

Body armor can reduce its penetrating effect, while hitting an exposed area would still depend on accurate placement of the separated elements.

That distinction keeps the ammunition's reported purpose focused on small aerial targets rather than conventional engagements involving protected ground personnel

“Our specialists have made bullets that separate into several parts, which are used against drones,” said Yury Knutov, a military expert.

“You just need two magazines, one loaded with regular bullets, the other with these special multi-bullets.”

Internal testing reportedly showed that components reached internal drone parts, but that does not establish overall battlefield effectiveness.

As drone warfare continues expanding across the ongoing conflict between Russia and Ukraine, new countermeasures keep emerging from both sides of the front line.

Military analysts frequently propose innovative solutions on paper, yet battlefield conditions often reveal weaknesses that laboratory testing simply cannot anticipate.

The new multi-bullet promises theoretical advantages that still require genuine combat validation before earning credibility.

Until Russian forces deploy this ammunition extensively under real combat pressure, its actual reliability and effectiveness in battle cannot be ascertained.

Independent verification, therefore, remains essential before military observers can responsibly assess whether this technology delivers meaningful battlefield advantage.

Via 1.ru

• Tether's African model supports 19 languages while running directly on local devices
• Tether removed 96% of low-quality training material before building AfriSLM
• EuroNano supports 90 translation directions while occupying only 36MB

Tether AI Research has unveiled a new set of translation systems that work entirely offline, requiring no connection to any network.

The release includes QVAC TranslatePsy-AfriSLM, built for 19 African languages, alongside QVAC TranslatePsy-EuroNano, built for nine European languages.

Each model processes translations directly on the device, keeping personal data local rather than sending it to remote cloud servers.

Translation designed for devices with limited resources

The African release includes Hausa, Amharic, Yoruba, Lingala, Swahili, Igbo, Zulu, Somali, Oromo, Malagasy, Kinyarwanda, Xhosa and Afrikaans.

It also handles Wolof, Luganda, Nyanja, Shona, Tswana and Southern Sotho, covering languages used across several regions of Africa.

Tether says those 19 languages collectively account for about 50% of the continent's population — the figure represents language reach rather than individual users.

AfriSLM contains 800 million parameters, yet Tether says the LLM surpassed Qwen3.5-122B-A10B, TranslateGemma-27B and NLLB-3.3B.

The comparisons were measured using FLORES-200, BOUQuET and SMOL, although benchmark performance does not establish identical results across every practical translation situation.

Tether attributes the results partly to a data-screening system that discarded as much as 96% of material judged unsuitable.

The company also released smaller and larger versions, including 0.8B, 2B and 4B parameter configurations for different hardware requirements.

EuroNano takes another approach, using English as an intermediary language while connecting European languages through 90 possible translation routes.

That package occupies only 36MB, which Tether says represents about 94% less storage than a comparable Firefox offline arrangement.

The compact footprint means translation can remain available after connectivity disappears, without requiring separate downloads for every language combination.

That distinction matters for field workers, travellers and local applications where storage capacity and network availability can both impose practical constraints.

Why Tether is starting with African languages

Tether's African focus also connects with the physical infrastructure it has developed across parts of Sub-Saharan Africa.

The company has installed solar-powered kiosks that provide phone charging, battery exchanges and access to digital financial services in communities.

Those locations could eventually provide another route for distributing educational material, agricultural information and other locally translated content without depending on continuous connectivity.

“Four billion people were left behind by the traditional financial system, and the most powerful technology of our age has repeated that failure,” said Paolo Ardoino, CEO of Tether.

“Language should not determine who can benefit from artificial intelligence. Open translation models like these are a step toward a future where education and AI tools reach hundreds of millions of people who have neither reliable connectivity nor access to expensive systems.”

The same technology could also operate alongside QVAC MedPsy, Tether's smaller model intended for healthcare-related applications on local devices.

“A mother could get real medical information she understands, instead of guessing. A child could learn in their own language. That is the future we are building through QVAC,” Ardoino added.

AfriSLM is currently available through Hugging Face, and the associated research has also been accepted for presentation at EMNLP 2026.

• MORENA beat 26 tested models despite having only 1.5 billion parameters
• The model from Vambo AI was built specifically around African languages from scratch
• MORENA uses fewer tokens to represent African text than major rivals

Vambo AI has released MORENA, a 1.5B parameter language model covering 12 languages spoken across Africa, plus English, French and code.

The developers say it beats models from Google, Meta, Alibaba and HuggingFace on African language modelling and translation while being up to 8x smaller.

On a key benchmark, this LLM scored 1.408 bpb — the best of 26 models tested — beating its closest rival, which carried over five times as many parameters yet still scored only 1.423 bpb.

Training without a borrowed foundation

The model covers Nigerian Pidgin, Igbo, Yoruba, Hausa, Kiswahili, ChiShona, isiZulu, isiXhosa, Kinyarwanda, Setswana, Afrikaans, and isiNdebele.

Most projects serving these languages continue training an existing Llama variant, which forces them to keep a vocabulary meant for English and programming text.

Vambo AI instead fixed the tokenizer, the data mixture, and the language list before training began, after comparing several vocabulary sizes for cost and efficiency.

MORENA's vocabulary encodes African text with 1.39 times fewer tokens than Gemma 3 and 1.53 times fewer than Llama 3.2 on identical passages.

African text still costs 0.249 tokens for each byte against 0.234 for English, roughly 6% more, and the team cannot fully explain the difference.

The nearest competitor, Lugha-Llama-8B, an Africa-adapted Llama 3.1 variant, scores 1.423 bpb and loses to MORENA in eight languages out of 12.

The 12B Gemma system, by comparison, consumes roughly 11 times the compute of MORENA while producing a weaker score on the same text.

The chat-tuned instruct version scores 1.441 bpb, trailing Lugha-Llama-8B overall but leading it in five shared languages while using about 20% of its parameters.

The instruct model, after seeing three sample translations, renders English into five languages from Africa at 45.8 chrF++, statistically level with one dedicated translation system.

It trails a larger translation model by about 1.4 points, while general models of similar size typically land in a range from 9 to 14.

A family built on 22,000 GPU hours

MORENA used 251.7B tokens during pretraining, followed by another 63B tokens during its mid-training stage.

Over 22,000 A100 GPU hours were reportedly required during development, with the associated computing resources valued at roughly $40,000.

According to the developers, this project was supported by UNDP, AIHub4SD, and CINECA, providing computing resources and other assistance during development.

This project also includes smaller releases, with versions containing 0.5B and 0.2B parameters available alongside the principal 1.5 billion model.

The 0.5B variants reportedly exceed every external system tested across 11 of the 12 languages covered by the project.

The 0.2B nano version is intended for speech-recognition rescoring, keyboard applications, and text normalisation tasks.

This nano version also costs 58 times less than Lugha-Llama-8B for each byte processed and generates 104 tokens each second on one A100 GPU.

A CPU-compatible build of the model also runs offline on a laptop, which suits settings where reliable GPU access is not guaranteed.

That gives the developers several model sizes for different computing requirements, rather than relying exclusively on the largest release.

MORENA also supports conversational applications, translation, and connections to other AI tools through its instruction-focused release.

Via Glitch Front

• TENKU Note Pro AI uses a full Windows 11 Pro setup inside its tiny convertible chassis
• The 960g notebook folds completely around its touchscreen for tablet use
• Intel's Core Ultra 5 125U brings dedicated AI hardware to TENKU's compact computer

TENKU has launched a compact Windows 11 Pro computer in Japan, combining Intel processing hardware with a touchscreen and convertible design.

The Note Pro AI uses a Core Ultra 5 125U processor with 12 cores and 14 threads, alongside hardware dedicated to selected AI workloads.

At 960g, the laptop combines a 10.51-inch screen with a physical keyboard while remaining considerably smaller than conventional full-size notebooks.

A convertible design built around Intel's mobile processor

The device uses a 360-degree hinge, allowing its keyboard section to fold behind the screen when tablet operation becomes more suitable.

Its touch panel recognizes ten simultaneous inputs, while an optional digital pen supports 4,096 pressure levels for writing and drawing applications.

The IPS panel produces 1,920 x 1,280 pixels, providing a 3:2 display area for Windows applications and productivity software.

Intel's Core Ultra 5 125U also includes an integrated NPU, which supports functions such as background processing and noise reduction during video calls.

The system includes 16GB of LPDDR5 memory and a 512GB M.2 SSD for applications, documents, and other locally stored data.

Wireless connectivity covers 802.11a, b, g, n, ac, and ax standards, with Bluetooth 5.2 available for compatible accessories and peripherals.

Battery capacity, dimensions and pricing

Under the hood, the device packs a 3,800mAh battery that is smaller than many smartphone batteries today, yet TENKU rates it for 17 hours.

According to the company, the chassis of this notebook measures 243mm across, 177.6mm from front to back, and 17.5mm thick.

Its compact chassis leaves less internal space than larger notebooks, while retaining the hardware required for its convertible configuration and cooling system.

On the software end, this device comes installed with Windows 11 Pro, giving the computer access to desktop applications rather than restricting its use to tablet-oriented software.

The available AI functions depend on software support, meaning the processor's dedicated hardware does not automatically provide every possible feature across Windows.

TENKU has listed the Note Pro AI in Japan, and it is currently retailing for about 176,800 yen (about $1,127) including tax.

There is also a launch promotion that cuts the price by 25% through Highbeam's official online store and physical outlets.

The 25% discount takes the promo retail price to 132,600 yen (about $845), and the offer ends on September 24 2026.

Amazon Japan is also offering a separate 15% reduction during the same promotional period, giving buyers another discounted purchasing channel until September 24.

Via Hermitage Akihabara (originally in Japanese)

• Chinese universities are cutting traditional courses because of AI
• Translation courses face pressure as universities reassess their employment value
• Photography programmes are being merged as AI changes creative production

The Central Academy of Fine Arts has dropped courses in illustration, photography and translation, arguing that AI has made them poor value for money.

Fashion design degrees are also being scrapped, with universities saying the future rests on close cooperation between people and machines in creative work.

The changes follow a state campaign to direct learners toward chips, robotics and AI, industries the authorities regard as vital for national prospects.

Universities are cutting courses at scale

From 2021 to 2025, Chinese universities closed or paused 12,200 undergraduate programmes and launched roughly 10,200 fresh programmes.

The restructuring affected over 30% of undergraduate programmes nationwide, and arts, humanities and language courses absorbed the heaviest cuts.

A survey of 70 universities found cuts to five translation majors, eight Japanese majors and five German majors among language programmes.

Alex Shi launched her creative career with a degree that trained her to translate and interpret English, and says such courses could soon vanish, as the cuts overlook the wider skills underlying translation, including communication and critical thinking, which extend beyond producing accurate text.

Shi concedes that AI could bring translation accuracy almost to perfection, yet maintains that interpreting demands far more than converting words between languages.

“There’s still that personal style that’s always been essential in translation,” said Shi.

The Communication University of China has reportedly closed five programmes, among them visual communication design, comics and photography, according to a report on its overhaul.

“Traditional photography major can no longer exist as an independent discipline… today everyone can be a self-media creator and recorder,” said Liao Xiangzhong, top official at the Communication University of China.

Liao later clarified that the majors were merged into other disciplines instead of being cancelled outright, with photography absorbed by film and television photography.

New film roles and hybrid courses

New roles such as prompt engineer and AI animator are emerging in film and animation, where AI is also creating fresh openings alongside the closures.

One industry worker says designing characters or scenes once took a month, but with AI, a complete project can now be delivered in that time.

Degree programmes in intelligent imaging art, intelligent audiovisual engineering, and intelligent engineering and creative design have been launched by some schools.

Universities are also launching hybrid courses in which artistic instruction sits beside AI training, while some standalone degrees are being scrapped across the sector.

That shift is changing what employers expect from graduates entering translation, design, photography and film production.

For students, traditional qualifications may increasingly need technical skills that allow them to work alongside automated creative systems.

The result is a university system where some familiar degrees disappear while new AI-focused programmes take their place.

Via Al Jazeera English

• A M4 Mac Mini has been rebuilt into a battery-powered handheld computer
• The custom handheld keeps Apple’s desktop motherboard and original cooling system intact
• A removable Mac module sits inside the custom 3D-printed handheld enclosure

Apple's M4 Mac Mini has been transformed into a handheld Mac with a 7-inch 120Hz touchscreen, physical controls, keyboard, and battery.

Shing Solutions stripped the desktop computer down to its motherboard and cooling system before rebuilding it inside a custom 3D-printed enclosure.

The resulting device, dubbed Mac Nano, keeps the full macOS functionality while replacing the desk-bound form factor with something designed for use in two hands.

The Mac Mini becomes a removable computing core

The project retains Apple's original heatsink and blower fan instead of replacing them with smaller cooling hardware to save internal space.

That decision allows the M4 processor to retain its factory thermal arrangement while the motherboard moves into a completely different enclosure.

Inside the new chassis, the computer module sits on rails, allowing it to slide out without disturbing the screen, keyboard, or controllers.

The arrangement means the central hardware can theoretically be removed for servicing or replaced while the surrounding handheld computer remains assembled.

A 7-inch touchscreen provides the main display, with its 120Hz refresh rate giving the portable machine a smoother interface and gaming experience.

It is also equipped with a compact QWERTY keyboard sitting beneath the display, with an integrated touchpad providing cursor control.

The keyboard originally worked wirelessly, but its battery was removed, and the keyboard was rewired to draw power directly through USB.

This modification reduces the number of rechargeable components while allowing the keyboard to receive power directly from the handheld computer.

The internal layout therefore combines Apple's desktop motherboard with controls and display hardware that were never designed specifically for this computer.

Nintendo controls and batteries complete the handheld device

The Mac Nano also includes physical gaming controls, using circuit boards taken from Nintendo Joy-Con charging grips and mounted inside both sides of the enclosure.

Those controls are wired directly into the Mac instead of using Bluetooth, but macOS does not natively support wired Joy Con controllers.

To make the controls usable, the project relies on Steam, whose controller mapping system can translate their input for compatible games.

Connecting the keyboard, controllers, touchscreen, and other peripherals also created a USB bottleneck inside the compact enclosure.

Shing Solutions addressed that problem by combining two miniature USB splitter boards into a custom three-port hub.

The Mac Nano is powered by rechargeable cells connected to a custom power delivery system, allowing it to run without staying plugged into an outlet.

The enclosure also incorporates handles, giving the device a grip that accommodates all its components.

Unlike a conventional Mac Mini, the Mac Nano therefore combines desktop Apple silicon with the components needed for genuinely untethered operation.

Via Yanko Design

• Pensacola fishermen pulls large unidentified military drone from the Gulf
• The aircraft surfaced only miles from Eglin’s extensive military testing zones
• Authorities recovered the drone without publicly identifying its origin or mission

A Pensacola-based fishing charter pulled a large military-style drone from Gulf of Mexico waters over the weekend, surprising both crew and onlookers.

Hot Spot Fishing Charters discovered it during a routine six-hour trip roughly 15 to 30 miles offshore from Pensacola Beach.

Authorities arrived later that same day and retrieved the aircraft, though details about the recovery process itself remain largely unavailable to the public.

An unusual interruption

The waters near this stretch of coastline sit close to Eglin Air Force Base, one of the largest military test ranges nationwide.

That base oversees restricted airspace and offshore zones used for trials of advanced weapons systems, including hypersonic missile technology in recent years.

The exact model of the drone remains unconfirmed, though local station WEAR-TV suggested the aircraft resembles an MQM-172, a drone flown for weapons testing and aircrew training.

The delta-wing shape of that platform also resembles Iran's Shahed-136 strike drone, a design widely recognized for its distinctive triangular silhouette.

Such resemblance likely explains why a civilian fishing crew would pause and take a closer notice of an unfamiliar aircraft floating offshore.

WEAR-TV described the discovery only as a large drone pulled from Gulf waters, providing little additional identifying information beyond that.

A brief video clip, shared widely across social media platforms, shows crew members hauling the large aircraft aboard their vessel.

The charter company regularly runs six- to eight-hour offshore excursions, giving anglers access to waters far beyond the immediate coastline.

Those extended trips routinely place crews within range of restricted military zones tied to ongoing weapons trials and training missions.

Eglin's long association with cutting-edge programs, including hypersonic missile trials, adds extra intrigue to any unexplained hardware found nearby offshore.

Lingering questions

No agency has issued a public statement claiming the aircraft or explaining how it drifted into civilian waters.

Regulations at Eglin explicitly bar small unmanned aircraft systems from operating within the base's land reservation and its surrounding boundaries as well.

Aircraft used for weapons testing sometimes conclude their operational service in open water, particularly following live-fire exercises conducted offshore near the base.

Whether this particular drone came from a specific Eglin program or another branch's testing effort remains an open question for now.

Investigators have not confirmed whether the aircraft drifted naturally with Gulf currents or was intentionally released during a recovery attempt.

Images circulating online appear to show visible damage along the fuselage, consistent with an extended period spent adrift at sea.

Residents have expressed curiosity about whether similar recoveries might happen again given the area's ongoing military testing activity.

It usually takes days or weeks for government officials to comment on unexplained hardware recovered along the Gulf coastline, especially when it's military.

Via Gadget Review

• Hackers are using blockchains to keep malware instructions available after servers disappear
• AI is making blockchain-based malware infrastructure easier for less experienced hackers
• Blockchain traffic is difficult to block without disrupting legitimate cryptocurrency services worldwide

Hackers are increasingly hiding malware instructions inside public blockchains, creating communication channels that can survive the removal of conventional infrastructure.

New figures from Chainalysis claim malicious blockchain activity increased 440%, with daily entries rising from 2.06 to 11.1 after newer AI systems emerged.

The technique gives attackers another way to maintain communication with compromised computers without relying entirely on conventional servers controlled by hosting providers.

Blockchain networks become malware dead drops

Blockchain dead drops use transaction data or smart contracts as lookup points, allowing infected computers to retrieve commands, addresses, or configuration information.

Because blockchain records are distributed across networks, removing a conventional server does not erase information already stored on the ledger.

A North Korean-linked operation associated with UNC5342 uses TRON and Aptos as alternate routes before retrieving encrypted instructions through the BNB Chain.

Its malware can check one network, switch to another when necessary, and retrieve updated addresses without receiving another malware package.

Iranian actors suspected of links to the country's intelligence ministry have embedded encoded routing information inside Bitcoin transactions used for malware retrieval.

Russian-speaking cybercriminals have also commercialized the technique, using Polygon contracts to provide blockchain-backed infrastructure for malware campaigns operated by different customers.

One related operator controls more than 50 BNB Chain resolver contracts while also conducting activity involving fraudulent tokens and clipboard-monitoring malware.

These operations show how blockchain records can function as persistent lookup infrastructure rather than merely serving their conventional financial and transactional purposes.

AI lowers the technical barrier

Chainalysis said the sharp increase in this malicious activity followed the arrival of high-capacity Chinese open models, which placed fewer restrictions on malware development requests.

Before those systems appeared, building reliable blockchain-based malware infrastructure required expertise across malicious software, cryptocurrency networks, and distributed communication systems.

AI tools can reduce that knowledge barrier by helping less experienced operators understand unfamiliar technologies and produce components needed for blockchain communication.

In the second quarter of 2026, state-linked groups accounted for roughly two-thirds of newly observed activity.

Those groups also represent about half of overall observed activity, indicating that blockchain-based malware infrastructure extends beyond conventional cybercriminal operations.

Defenders face difficulties because blocking blockchain traffic could also disrupt legitimate wallets, decentralized applications, exchanges, and decentralized finance services used worldwide.

Attackers can further complicate disruption by operating their own blockchain nodes, reducing dependence on external providers that defenders might otherwise pressure or disable.

Some operators have hidden server addresses inside wallet identifiers without usable private keys, then used zero-value transfers to trigger malware retrieval.

Those transactions leave public records that investigators can examine, potentially providing useful clues even when attackers attempt to conceal their infrastructure.

"While the exploitation of blockchain by state-linked organizations such as North Korea is becoming more sophisticated, on-chain records left by attackers can actually serve as important clues to track them," said Kwon Jun-hyeok, General Manager of Chainalysis Korea.

"Tracking these traces and identifying attackers and related infrastructure through blockchain intelligence will become increasingly important in responding to new cyber threats."

• Huawei’s Ascend 960 roadmap depends on scale, memory, and faster interconnects
• UnifiedBus connects thousands of processors while reducing communication bottlenecks inside AI systems
• The Atlas 960 roadmap reaches 15,488 processors with 34 PB/s bandwidth

Huawei is expanding its Ascend platform with larger systems designed to handle AI models containing up to 10 trillion parameters.

The company is advancing UnifiedBus, an interconnect architecture intended to connect processors, memory, and storage with lower communication overhead.

That approach gives Huawei another way to compensate for hardware limitations that have constrained individual Ascend accelerators against Nvidia's products.

Huawei expands the Ascend 960 system

Huawei's Atlas 960 SuperPoD roadmap calls for systems containing as many as 15,488 Ascend 960 processors, up from its earlier 8,192 processor configuration.

The company says the system can deliver 30 EFLOPS of FP8 computing and 60 EFLOPS using FP4, alongside 4,460 TB of memory capacity.

Its interconnect bandwidth is specified at 34 PB/s, while Huawei expects training performance to reach 15.9 million tokens per second.

Huawei has also accelerated development of the Ascend 960 family, with the 960DT scheduled for Q1 2027 and the 960PR planned for Q3 2027.

The company says both versions will arrive earlier than previously expected, while Huawei says performance improvements have exceeded the original roadmap.

"We're evolving our Ascend chip series on a one-generation-a-year cycle," said David Wang, deputy chairman and rotating chairman of Huawei.

"In 2028 and 2029, we will roll out the Ascend 970 and 980 chips, respectively. Thanks to the Tau (τ) Scaling Law, not only will their compute specifications continue to double, but you can also expect to see huge improvements..."

UnifiedBus connects 4,096 NPUs inside the Atlas 960E

UnifiedBus is Huawei’s interconnect architecture for linking processors, memory, and other components within large AI computing systems.

Huawei applies UnifiedBus in the Atlas 960E, a 960 variant that ties up to 4,096 NPUs and gives them access to shared memory across the SuperPoD.

It combines UnifiedBus with Hi-ONE optical technology, extending high-speed optical connections deeper into the computing system.

The approach uses near-packaged optics, or NPOs, placing optical components closer to processors to improve transmission speeds and reduce energy consumption.

Huawei says this design helps the SuperPoD move data between thousands of processors without relying entirely on conventional optical connections outside the computing system.

The Atlas 960E can use approximately 5,500 Hi-ONE units instead of about 48,000 conventional 800G optical modules.

This configuration can reduce power consumption by more than 550 kilowatts across the processor interconnection system.

The architecture also gives the Atlas 960E 8 EFLOPS of FP8 computing performance while Huawei says it supports models containing up to 10 trillion parameters.

Atlas 960E systems can also connect into SuperClusters containing 512,000 NPUs, with the architecture ultimately supporting million-processor configurations.

The larger systems are important because Huawei cannot rely solely on individual accelerator performance to narrow the gap with Nvidia.

Its strategy combines more processors with faster communication, allowing thousands of chips to operate within a shared computing environment.

UnifiedBus therefore becomes important because thousands of processors must exchange data continuously during demanding AI training and inference workloads.

• Canva ProSuite connects Affinity, Cavalry, Flourish, and Leonardo into one ecosystem
• Affinity and Cavalry stay free as ProSuite gains over 100 additions
• Scene Decomposition uses AI to split flattened pictures into separate editable components

Canva is giving Affinity a major set of upgrades as it tries to make its professional creative tools more useful against established competitors.

The free software now adds photography, design, automation and file compatibility features alongside deeper connections with Cavalry, Canva and other creative products.

The changes form part of Canva ProSuite, which brings Affinity, Cavalry, Flourish and Leonardo together while keeping Affinity and Cavalry available without payment.

Affinity gets a major photography and design update

The photography changes are particularly extensive, adding batch importing and an improved RAW processing system for users handling large collections.

Preset looks can now be applied across multiple images, while new correction tools address mixed illumination and camera movement during editing.

Scene Decomposition uses AI to separate flattened pictures into editable components, allowing photographers to modify individual elements without rebuilding images manually.

A dedicated Astrophotography Studio adds specialized controls for photographers working with night skies and astronomical subjects during detailed image processing.

Affinity is also expanding the range of files it can handle, including support for bringing InDesign documents directly into the application.

The company says this gives users access to major industry file formats while reducing obstacles when moving projects between different creative applications.

Integration with Canva Brand Kit and Canva Sheets also adds a route for connecting design work with brand assets and structured information.

Other additions cover vector editing, typography, brushes, noise reduction, high dynamic range processing and export controls, giving professional users more options.

Sony ARW6 support, OpenColorIO display emulation, and improved output matching also extend Affinity's capabilities for photography and color-managed production workflows.

The release includes more than 60 Affinity additions, while Canva says ProSuite brings over 100 additions and changes across its connected products this year.

Canva connects Affinity with motion and AI tools

Canva is also connecting Affinity with Cavalry, its software for creating motion graphics and animations, as part of the new ProSuite.

Cavalry receives more than 50 changes, including new shape controls, lighting effects, graph editing improvements, and expanded SVG handling for motion designers.

The integration lets selected elements within Cavalry projects become editable inside Canva, while keeping the underlying animation protected.

This arrangement could allow teams to make limited changes without repeatedly sending modification requests back to motion designers and animators across team workflows.

Affinity and Cavalry can also connect with AI assistants through ProSuite's AI Connectors, allowing repetitive operations to be handled from natural-language instructions.

“We're genuinely excited to be releasing such a huge swathe of new tools for professional designers, animators, photographers and other advanced users,” said Duncan Clark, CEO of Pro Design at Canva.

“We asked professionals what they actually needed to power their creative work…there’s a lot of noise right now about AI replacing human creatives, but we believe professional craft matters more than ever. We want the Canva ProSuite to be where that craft lives and thrives.”

Affinity's new features, the Cavalry improvements and tighter connections across Canva's products give the free software a significantly larger professional toolkit.

• Snap is taking its AR glasses directly into the enterprise market
• Nvidia, Salesforce, and AWS are joining Snap's push into workplace computing
• Snap wants Specs to become a computer rather than another camera accessory

Snap is taking its augmented reality glasses into the enterprise market through partnerships with Nvidia, Salesforce and Amazon Web Services.

The company is looking beyond its established consumer business for workplace applications that could generate demand for its new hardware.

Snap’s CEO Evan Spiegel wants businesses to view Specs as a computing platform rather than another pair of glasses designed primarily for taking pictures.

Snap looks beyond its consumer business

Snap's Specs glasses cost $2,195, putting the device into a category where business customers will need clear reasons to purchase it.

The company says the glasses can support field technicians by displaying digital instructions while workers inspect or repair equipment.

They can also allow employees to interact with colleagues while accessing information through the glasses during workplace tasks.

Microsoft and Meta spent years developing augmented and virtual reality hardware without producing products that became widely adopted consumer devices.

Microsoft's HoloLens and Meta's Quest line have faced challenges turning immersive computing into a mainstream hardware business.

The company believes that the physical size of previous devices was an important problem for users, arguing that Specs is better suited to workplace use.

Snap is approaching Specs through practical workplace applications while attempting to expand beyond the consumer market that has driven its business.

The company has not disclosed how many enterprise customers are committed to using Specs or how much revenue its business partnerships could generate.

Nvidia, Salesforce and AWS join the enterprise push

Snap is collaborating with Nvidia for access to an open-source extended reality AI platform that businesses can use when developing applications for immersive devices.

Salesforce is also part of the initiative, while Amazon Web Services provides another major technology relationship for Snap's enterprise effort.

“They’ve got an open-source XR AI platform that a number of businesses are already building on top of,” Spiegel said.

“The goal with both Salesforce and Nvidia is to make it really easy for businesses to get the benefits of those agentic platforms.”

These partnerships are intended to connect Specs with software and AI services that businesses already use.

Snap is also developing Specs Intelligence, an AI service designed to operate across the glasses, iPhones and Mac computers.

Spiegel said the company has not finalized pricing, although the service is expected to have free access alongside other usage-based options.

The system is intended to let users complete tasks through an AI assistant while moving between the glasses and other Apple devices.

These efforts may give Snap a route toward making Specs useful for work beyond the consumer applications that have defined its business.

Via CNBC

• Rune wants unused solar electricity powering GPUs instead of going to waste
• RELIC connects computing equipment directly to functioning solar generation sites
• The system operates without grid connections or conventional data center construction

A San Francisco company named Rune has introduced a modular compute unit built to operate directly on unused solar power.

Called RELIC, the equipment attaches to functioning solar installations without needing any connection to the electrical grid or new construction work.

The launch came alongside a $40 million Series A financing round for Rune, lifting the firm's overall capital raised to $53.5 million total.

Skipping the grid bottleneck

Solar facilities routinely waste as much as 20% of the electricity they generate, adding up to more than 50 TWh yearly nationwide.

RELIC pulls that idle current straight from the source, avoiding metering charges, utility fees, and the grid limitations slowing AI growth.

"Every solar plant is a latent data center. The power is already there, sitting idle while AI labs wait years for grid connections that may never come," said William Layden, Co-Founder and CEO of Rune.

Rune states that RELIC can be installed in about 60 minutes and deliver working GPU capacity within six weeks of a contract.

That timeline contrasts sharply with conventional data center construction projects, which frequently stretch across several years before reaching full operational status.

The company reports that RELIC lowers non-compute infrastructure spending by 85%, saving roughly $620 million across a 100 MW deployment.

RELIC operates natively on direct current — solar's original output form — sidestepping energy losses that occur when converting power to alternating current.

It needs no extensive site preparation, leaves no visible trace across the landscape, and operates entirely without consuming any water.

Rune's initial working installation sits at a 200 MW solar facility located in Texas, mounted on existing infrastructure without any modifications.

Addressing a system under strain

Traditional data center construction cannot keep pace with the surging demand pouring in from artificial intelligence developers and infrastructure buyers everywhere.

Rune's approach addresses that shortfall by placing compute equipment directly beside existing renewable generation instead of pursuing new construction projects.

“Demand for AI infrastructure is growing faster than it can be delivered. Rune deploys compute directly behind the meter at operating solar farms, converting clean power into rapidly scalable AI capacity,” said Santo Politi, Founder and General Partner of Spark Capital.

“Every renewable asset in operation becomes a potential deployment site. Rune is the fastest path to new capacity we have seen."

Many data centers currently running were adapted from designs built for earlier computing needs rather than for artificial intelligence workloads specifically.

“AI’s power constraint isn’t generation, it’s resource allocation,” said Varun Palivela, Co-Founder and CTO of Rune.

“The industry is retrofitting data centers designed for a different era rather than rethinking the architecture itself. We built an entirely new technology stack from the ground up, coupling compute directly to clean generation so AI can scale here on Earth.”

The approach leaves Rune testing whether unused solar capacity can support AI computing at the speed required by growing demand.

• Samsung is shifting additional DDR5 and SSD production to outside manufacturers
• The company needs internal factory space for advanced HBM production
• DDR5 module assembly remains simpler than advanced HBM packaging processes

Samsung Electronics is shifting all additional production of conventional memory modules onto outsourced partners going forward.

The company is directing capacity growth for DDR5 modules and SSDs toward external firms, reserving its in-house capacity for the more advanced HBM packaging process.

Industry sources say this reflects Samsung's mounting need to use scarce back-end capacity more efficiently.

A deliberate pivot toward outsourced capacity

“Samsung Electronics lacks space to expand existing back-end production lines needed to manufacture semiconductors for artificial intelligence (AI), so it is reallocating space and equipment at packaging plants in Cheonan and Onyang to advanced packaging lines, including HBM,” a person familiar with the matter said.

“The company has decided to handle most increases in conventional memory module production through outsourcing rather than expanding its own capacity.”

Samsung has reportedly pressed its outsourced semiconductor assembly and test partners to expand capacity for DDR5 modules and SSDs.

The company has even gone a step further by asking partners to accelerate capacity expansions that were already part of its existing plans.

Several partners are already scaling up in response, including Dreamtech, Hanyang Digitech, and SFA Semicon.

Dreamtech began mass producing DDR5 modules in India last November and approved further equipment investment in June.

Its Noida facility, previously used for smartphone circuit board assembly, will eventually reach an annual capacity of over 50 million units.

Hanyang Digitech has invested more than 31.5 billion won in 2026 so far to automate its existing memory module plant located in Vietnam.

SFA Semicon is relocating DDR5 testing and assembly equipment from Samsung's Onyang campus to a new site in the Philippines.

That transition begins in November, with a second phase following during the second quarter of 2027.

Limited space drives the broader manufacturing shift

Samsung's own capacity constraints stem partly from a new HBM plant under construction at its Onyang campus, valued at 6 trillion won.

That facility will take multiple years to finish, leaving little room for conventional module expansion in the meantime.

A separate $1.5 billion back-end processing plant is also under construction in Thai Nguyen Province of northern Vietnam.

That facility will handle testing for older memory types, including DDR4, low-power LPDDR4, NAND flash, and universal flash storage.

DDR5 modules require mounting prepackaged DRAM and NAND chips onto circuit boards using surface-mount technology.

That process remains considerably less complex than advanced packaging methods used for HBM, making outsourcing comparatively straightforward for partners.

Micron previously pursued a similar route, shifting conventional memory production externally while concentrating internal resources on higher-margin products.

Whether SanDisk eventually follows a comparable path remains uncertain given how differently each company's manufacturing footprint and product mix are currently structured.

Via The Elec

• Mouse Computer has fitted Ryzen AI processors into a laptop weighing just 981g
• The X2 uses a 12.2-inch display with 1920 by 1200 resolution
• A free SODIMM slot allows memory expansion from 16 GB to 32 GB

Japanese electronics maker Mouse Computer has introduced a compact laptop called the X2, a compact notebook weighing 981g.

The device pairs a 12.2-inch display with AMD processors, offering configurations that scale from modest to genuinely capable performance.

The laptop frame measures 277.1 mm wide, 210.2 mm deep, and 19 mm tall, retaining a single unoccupied SODIMM slot for later RAM upgrades.

Specifications built around three AMD chip options

Buyers can choose between three distinct AMD APUs, each aimed at a different performance and price bracket for portable computing.

The base option, a Ryzen 5 216, combines two full Zen 4 cores with four smaller Zen 4c cores in a six-core layout, which should perform similarly to AMD's existing Ryzen 5 8540U based on available benchmark comparisons circulating online.

A step up brings the Ryzen AI 5 340, part of AMD's Krackan Point lineup, also arranged in a six-core configuration.

Multicore results suggest this chip lands close to Intel's Core Ultra 7 165H and edges past the Core Ultra 5 325.

The top configuration, a Ryzen AI 7 350, uses eight cores split evenly between four Zen 5 and four Zen 5c units, which approaches Intel's Core Ultra 9 185H in combined multi-threaded and synthetic benchmark scores, according to early comparisons.

The portable display panel runs at 1920 by 1200 resolution and reportedly covers 100% of the sRGB color gamut.

A 180-degree hinge, 60 Wh battery, and 500 GB NVMe SSD round out the internal hardware package.

Connectivity includes gigabit LAN, Wi-Fi 7, Bluetooth 5, USB Type-A ports, HDMI output, and a microSD card slot for storage expansion.

A 2 MP webcam with a privacy shutter supports Windows Hello, while Windows 11 Home ships preinstalled on every unit.

Battery life reaches roughly 12.5 hours during video playback and about 25 hours in idle mode, measured under JEITA 3.0 standards.

Pricing stays confined to the Japanese market for now

The Ryzen AI 5 variant, equipped with 16 GB of RAM and 500 GB of storage, costs ¥239,800, or roughly $1,545.

The higher Ryzen AI 7 configuration carries the same memory and storage allotment but costs ¥259,000, or about $1,670.

Sales of this device will begin sequentially from October, though the rollout remains confined to the Japanese market via the company’s online store for now.

The laptop is currently available in three colour options — Cosmo Blue, Frost Gold, and Pearl White.

Both versions ship with single-channel memory by default, though the empty SODIMM slot allows an upgrade path from 16GB to 32 GB.

A compatible 16 GB DDR5-5600 module currently retails for around $229 through Amazon's storefront.

Interested customers elsewhere would need to rely on import services to acquire a unit, adding cost and complexity to the process.

Via Mouse | GDM (originally in Japanese)

• Samsung has backed a tiny Dutch chipmaker targeting Nvidia
• Euclyd raised $231 million before shipping its first commercial chip
• Euclyd plans to sell complete AI systems directly to enterprises but will not deliver its first hardware until 2028

Samsung has joined a group of investors backing a small Netherlands-based chip designer aiming to challenge Nvidia's dominance in artificial intelligence hardware.

The startup, called Euclyd, closed a funding round worth $231 million, drawing support from several firms beyond the South Korean electronics giant.

According to Bernardo Kastrup, Chief Executive of Euclyd, commercial hardware from the company will not reach customers until 2028

A chip built to rival Nvidia's grip

Euclyd's Series A round totalled €200 million and was co-led by Somerset Capital Partners, the Scaleup Europe Fund, and Innovation Industries, alongside Samsung.

In an exclusive interview with CNBC, Kastrup said that the capital would support two separate lines of business going forward.

One involves selling physical hardware and rack systems directly to enterprises wanting to run their own AI inference securely on-site.

The other involves licensing the underlying intellectual property to firms that want to build their own chips using Euclyd's designs.

Founded in 2024, the company is developing a chip system built around a different architecture than conventional graphics processors, covering both the processor and memory layers.

Nvidia's GPUs were originally built for video games but became central to AI training and inference and pushed the company to become the world's most valuable.

That dominance now covers the market for the very highest performing chips available today, though rivals are emerging quickly.

“AI is becoming a foundation of economic growth, scientific discovery and national competitiveness, but its potential will remain constrained unless we fundamentally change the infrastructure beneath it,” Kastrup said.

Why Samsung's involvement matters beyond cash

Kastrup revealed that Samsung's contribution extends well past the money it committed to the round.

"Samsung can help us in more ways than money. They are one of the biggest memory manufacturers in the world,” said Kastrup. “They do a lot of engineering, they know a lot about systems, they know the supply chain, they have a huge network."

Although Euclyd expects to commence shipping its chips in two years, the company plans to serve thousands of enterprise customers by 2030.

The company's systems remain unproven at any meaningful commercial scale so far, leaving real-world performance an open question.

Several major technology firms are pursuing similar paths away from dependence on Nvidia's processors for AI workloads.

OpenAI announced in August that its first in-house AI chip, named Jalapeño, delivered industry-leading speed and efficiency.

Google, Amazon Web Services, and Meta are each developing their own chips for internal AI use as well.

Whether a small Dutch startup with no shipped product can meaningfully dent Nvidia's position remains genuinely uncertain given the scale involved.