MSI's new Claw A8 is powered by AMD
That's a bit less than the 32GB of memory that came with the Intel-equipped Claw 8 AI Plus released late last year, but it still has an 8-inch full HD display, a 120Hz refresh rate, and a 1TB M.2 SSD. The AMD-powered Claw A8 will also come in two colors: white and lime green.
MSI had a bit of a rough start with the release of its original Claw handheld in March 2024, but it seemed to redeem itself with the launch of the MSI Claw 8 AI Plus. Lenovo also snapped up one of AMD's new handheld-focused chips for the Legion Go S, but it uses the weaker Z2 Go version that my colleague Sean Hollister said 'can't compete' with its predecessor's Z1 Extreme.
There's a new MSI Claw 8 AI Plus 'Polar Tempest' edition, too, which features an up to Intel Core Ultra 7 258V processor and a 2TB NVMe SSD. It also has what MSI calls a 'glittering' white coating. MSI didn't reveal a release date or price for either model, but it will likely be in the same ballpark as the standard Intel-powered MSI Claw 8 Plus, which Best Buy lists as costing $999.99.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
an hour ago
- Yahoo
Japan sees bright future for ultra-thin, flexible solar panels
Japan is heavily investing in a new kind of ultra-thin, flexible solar panel that it hopes will help it meet renewable energy goals while challenging China's dominance of the sector. Pliable perovskite panels are perfect for mountainous Japan, with its shortage of flat plots for traditional solar farms. And a key component of the panels is iodine, something Japan produces more of than any country but Chile. The push faces some obstacles: perovskite panels contain toxic lead, and, for now, produce less power and have shorter lifespans than their silicon counterparts. Still, with a goal of net-zero by 2050 and a desire to break China's solar supremacy, perovskite cells are "our best card to achieve both decarbonisation and industrial competitiveness," minister of industry Yoji Muto said in November. "We need to succeed in their implementation in society at all costs," he said. The government is offering generous incentives to get industry on board, including a 157-billion-yen ($1 billion) subsidy to plastic maker Sekisui Chemical for a factory to produce enough perovskite solar panels to generate 100 megawatts by 2027, enough to power 30,000 households. By 2040, Japan wants to install enough perovskite panels to generate 20 gigawatts of electricity, equivalent to adding about 20 nuclear reactors. That should help Japan's target to have renewable energy cover up to 50 percent of electricity demand by 2040. - Breaking the silicon ceiling - The nation is looking to solar power, including perovskite and silicon-based solar cells, to cover up to 29 percent of all electricity demand by that time, a sharp rise from 9.8 percent in 2023. "To increase the amount of renewable energy and achieve carbon neutrality, I think we will have to mobilise all the technologies available," said Hiroshi Segawa, a specialist in next-generation solar technology at the University of Tokyo. "Perovskite solar panels can be built domestically, from the raw materials to production to installation. In that sense, they could significantly contribute to things like energy security and economic security," he told AFP. Tokyo wants to avoid a repeat of the past boom and bust of the Japanese solar business. In the early 2000s, Japanese-made silicon solar panels accounted for almost half the global market. Now, China controls more than 80 percent of the global solar supply chain, from the production of key raw material to assembling modules. Silicon solar panels are made of thin wafers that are processed into cells that generate electricity. They must be protected by reinforced glass sheets and metal frames, making the final products heavy and cumbersome. Perovskite solar cells, however, are created by printing or painting ingredients such as iodine and lead onto surfaces like film or sheet glass. The final product can be just a millimetre thick and a tenth the weight of a conventional silicon solar cell. Perovskite panels' malleability means they can be installed on uneven and curved surfaces, a key feature in Japan, where 70 percent of the country is mountainous. - Generating where power is used - The panels are already being incorporated into several projects, including a 46-storey Tokyo building to be completed by 2028. The southwestern city of Fukuoka has also said it wants to cover a domed baseball stadium with perovskite panels. And major electronics brand Panasonic is working on integrating perovskite into windowpanes. "What if all of these windows had solar cells integrated in them?" said Yukihiro Kaneko, general manager of Panasonic's perovskite PV development department, gesturing to the glass-covered high-rise buildings surrounding the firm's Tokyo office. That would allow power to be generated where it is used, and reduce the burden on the national grid, Kaneko added. For all the enthusiasm, perovskite panels remain far from mass production. They are less efficient than their silicon counterparts, and have a lifespan of just a decade, compared to 30 years for conventional units. The toxic lead they contain also means they need careful disposal after use. However, the technology is advancing fast. Some prototypes can perform nearly as powerfully as silicon panels and their durability is expected to reach 20 years soon. University professor Segawa believes Japan could have a capacity of 40 gigawatts from perovskite by 2040, while the technology could also speed up renewable uptake elsewhere. "We should not think of it as either silicon or perovskite. We should look at how we can maximise our ability to utilise renewable energy," Segawa said. "If Japan could show a good model, I think it can be brought overseas." hih/sah/stu/lb Solve the daily Crossword
Yahoo
2 hours ago
- Yahoo
China's new X-shaped rail gun design doubles firepower and improves range
Chinese researchers are experimenting with a new rail gun design to help overcome the firepower limitations of existing designs. They believe that stacking two rail guns in an X-shape in a single barrel could be the solution they need. A rail gun is a futuristic weapon that utilizes electromagnetic force, rather than explosives or gunpowder, to launch projectiles. Designs vary, but rail guns typically utilize electricity to generate powerful magnetic fields that accelerate a projectile to extremely high speeds, around Mach 7 (approximately 5,344 mph or 8,600 kph). Many nations around the world have experimented with the technology, with China developing its own version in 2018. This gun was mounted on the Haiyangshan, a Chinese Type 072III-class landing ship, and was seen as a first-of-its-kind at the time. Meanwhile, the U.S. has abandoned its rail gun program owing to cost, complexity, and the rise of hyper-sonic missiles. Japan is working on a smaller rail gun—essentially a technology demonstrator—that fires tiny 0.66-pound (300g) shells from ships. X marks the spot China's existing rail gun, which seemed insurmountable during its induction, has displayed its set of limitations over time. Being a naval gun, the weapon needs to pack a punch but this requires more power, which in turn leads to issues. Too much current melts the metal components, like the U-shaped armature (which helps carry current and launch the shell). Magnetic forces can tear apart melted metal mid-shot, likened to a "magnetic saw." This leads to the rails getting damaged, reducing performance. At present, the navy's rail gun is capped at firing 33-pound (15 kg) shells, which is not nearly enough to damage a ship—suggesting the weapon is not yet war-ready. In a bid to address the issue, a group of Chinese army researchers—led by associate professor Lyu Qingao from the Army Engineering University of PLA in Shijiazhuang—proposed a x-shaped 'double-decker' rail gun which could overcome power limitations. To this end, the design features each rail gun having its own power circuit, allowing them to operate independently but in parallel. The setup utilizes four rails and two U-armatures that work together. Patent pending Vertical magnetic fields don't interfere with each other, so stacking two rail guns does not reduce performance. You can think of it like twin engines on a rocket, delivering more power without one interfering with the other. This design, the team believes, will enable the weapon to launch a 132-pound (60 kg) shell at speeds of at least Mach 7. The improved ordinance size also happens to be four times more than the current Chinese Navy requirement. The team has already filed a patent for the new 'x-shape armature' technology, but has not conducted any live-fire tests as yet. If the trials prove successful, it should translate to ranges in the region of 248 miles (400 km), with the shell able to cross that distance in as little as 6 minutes. The team also believes the impact speed of the shell could exceed Mach 4 on target. However, researchers caution that the 'proximity effect' (interference between nearby electrical currents) may still cause issues with performance and reliability. To this end, real-world testing is necessary to demonstrate its effectiveness. Solve the daily Crossword
Digital Trends
3 hours ago
- Digital Trends
Are external GPUs still worth it in 2025?
External GPUs (eGPUs) have long promised a tantalizing compromise: portability when you need it, and desktop-class performance when you don't. But as laptop GPUs get faster and AI workloads move to dedicated NPUs, the value proposition of an eGPU dock feels less straightforward than ever. For gamers and creative professionals tied to thin-and-light ultrabooks, eGPUs still offer a way to transform a modest machine into a graphics powerhouse. Thunderbolt 5 and the fairly new Oculink interfaces now offer significantly improved bandwidth over their predecessors, reducing bottlenecks that once plagued external GPU performance. However, that doesn't mean all the original drawbacks have been solved. Recommended Videos Take for instance, Razer's newly launched Core X V2, which showcases the current state of external GPU (eGPU) technology by adopting the latest Thunderbolt 5 interface. This upgrade significantly boosts bandwidth up to 80 Gbps in both directions (bi-directional) or can utilize a 'Bandwidth Boost' mode for up to 120 Gbps for video-intensive tasks in one direction while maintaining 40 Gbps in the other. These speeds allow desktop-grade GPUs to perform more effectively over a cable than ever before. However, the Core X V2 also reflects modern trade-offs by dropping the built-in power supply and I/O ports found in earlier models requiring users to bring their own power supply unit and, optionally, buy a separate $390 dock for connectivity. At $350, it's also more expensive than its predecessor despite being more stripped down. On top of that, compatibility is now limited to devices running on Windows with either a Thunderbolt 4/5 or USB4 port while Apple Silicon Macs remain unsupported. One of the biggest arguments against eGPUs in 2025 is diminishing returns. While they still offer a dramatic performance uplift over integrated graphics, many premium laptops come with capable discrete GPUs that can handle most modern games at respectable settings and resolutions. For creators, the line between desktop and mobile GPUs has narrowed since render times and AI-assisted workflows are often bottlenecked more by CPU and memory than by pure GPU horsepower. Performance: Gains and Limitations Even with fast Thunderbolt links, eGPUs still lose some speed compared to an internal GPU slot. PC Gamer conducted a detailed performance analysis of eGPU setups using two fairly modern graphics cards including an RTX 4070 Ti and an RTX 4090. The testing was done across four different configurations including a Thunderbolt eGPU dock, an OCuLink eGPU dock (via Minisforum DEG-1), PCIe x8 (Beelink EX Dock), and a native desktop PCIe x16 connection. All tests were run on up-to-date hardware platforms with DDR5 memory and PCIe 4.0 support to ensure consistency. Benchmarks were performed at 1440p using ultra settings, measuring both average framerates and 1% lows to assess frame pacing and stutter. The goal was to evaluate how each connection standard affected GPU performance in real-world gaming scenarios. The RTX 4070 Ti results showed that Thunderbolt was by far the most limiting setup, running approximately 25% slower than OCuLink and showing pronounced frame stutter and poor 1% lows. OCuLink, despite also being limited to four PCIe lanes, performed noticeably better and approached desktop-like averages in many cases, though some minor stutter remained. Interestingly, the PCIe x8 configuration performed nearly identically to full PCIe x16 desktop usage, proving that eight lanes are sufficient for this mid-range GPU. In our own testing of the Minisforum DEG1, a $99 open-air eGPU dock using the OCuLink standard, we found that it manages to deliver good real-world performance with only modest impact even when paired with high-end GPUs. When tested with an RTX 4080 Super, Cyberpunk 2077 and Returnal ran nearly identically to a desktop setup, while demanding titles like Red Dead Redemption 2 and Forza Horizon 5 showed a 12% drop. Horizon Zero Dawn was the exception, with a ~35% performance decline, but this still outperforms much older eGPU systems. In benchmarks like 3DMark Time Spy on a 4060 Ti, the DEG1 achieved graphics scores between ~9,600 and 13,500, depending on the host, with PCIe‑Gen4 bandwidth measured around 6–6.7 GB/s—indicating its ability to preserve the majority of a card's potential in realistic scenarios. It is fair to conclude that eGPUs still carry notable bandwidth limitations, particularly when paired with high-end GPUs. While newer interfaces like Thunderbolt 5 and OCuLink improve average framerates, stuttering remains a major bottleneck that undermines the experience. As such, it is recommended that eGPU users are better off pairing their enclosures with upper mid-range cards which are less likely to be choked by the limited bandwidth and can still deliver solid performance without the drawbacks of stutter-heavy frame pacing. In short, expect roughly 70–80% of desktop performance from an external setup, and diminishing returns as you push the GPU size upward. If you plan to output to the laptop's own screen rather than an external monitor, prepare for even more overhead as some tests found the performance penalty shrinks with an external display because data doesn't have to double-hop back into the laptop. Cost and Portability A core drawback of eGPUs is the price of admission. The enclosure itself can cost hundreds of dollars. Razer's Core X V2 is $350 on its own, not including the GPU or even a power supply. Many older TB4 enclosures similarly went for $200–400. Then you must buy a desktop-class GPU which is a luxury on its own and, in most cases, an ATX power supply unless the enclosure includes one. Compare that to buying a new gaming laptop or compact desktop for a similar investment, you could get a machine that already has a GPU with no latency penalty. Portability is another concern. An eGPU setup is far from 'light and portable' especially the ones that allow you to install a desktop class GPU. You'd dock it at home or office, not toss it in a backpack on a plane. By contrast, a high-end gaming laptop contains its GPU internally and requires only one power cord. Some companies have tried truly portable eGPUs for instance, the ROG XG Mobile dock from Asus. While the previous version included a proprietary connector, the latest variant has moved to TB5 expanding compatibility to a wider range of devices. However, it is important to know that Asus could only make the dock portable by including laptop-class GPUs instead of the more powerful desktop counterparts . Compatibility quirks also add to the hassle. You need a Thunderbolt or USB4 port with proper support, with some Windows laptops still lacking a TB port entirely. Hot-plugging is generally supported on PC, but modern Macs with Apple Silicon simply do not support any sort of external GPUs. Additionally, driver issues can crop up. While Windows 11 handles many eGPUs better than older Windows 10 did, it's not flawless. In short, an eGPU is a power-up for your laptop, but only when docked, so you must accept that you're essentially carrying a desktop GPU in one giant box, plus another dongle for ports if needed. Alternatives and Outlook Given these trade-offs, many look at alternatives. Laptops have soldered or integrated GPUs that can't be swapped. Thus, in many cases, buying a new laptop with a better GPU is simpler than an eGPU. With every laptop refresh cycle, vendors cram more GPU power into slim machines, so a consumer might choose to buy a new $1500 gaming laptop instead of spending on an eGPU setup for their old ultrabook. Additionally, services like Nvidia GeForce Now, Xbox Cloud Gaming, and Amazon Luna have matured. These let users stream AAA games to any device without a local GPU. For gamers with good internet connection, cloud services can deliver 4K at 60–120 FPS with minimal local hardware. Creative professionals likewise have cloud rendering options (e.g. Blender or Adobe cloud rendering). The upside is no hardware purchase, only a subscription. The downside is latency and variable image quality which is unacceptable for competitive play or precision tasks. But for some, cloud gaming can entirely replace the need for an eGPU or even a powerful PC, especially on portable devices like tablets. External GPU docks are more of a specialty solution than a mass-market trend. They've never been as popular as once hoped, and advances in laptop hardware and cloud services have only chipped away at their appeal. That said, they aren't completely dead. For a subset of users say, a Linux mini-PC or Windows handheld, an eGPU might be the only way to run high-end games or CUDA workloads. Ultimately, eGPUs remain a niche tool, useful if you already have a compatible laptop and an extra graphics card you want to reuse. They offer a clear benefit (big jump in GPU power on demand) but demand tradeoffs in cost, weight, and complexity.
