Thermaltake has introduced a revised version of its Steel Shadow Tough liquid cooler series in Europe, featuring subtle but significant changes. In Europe, the new series will be named LA(S) and available with 240mm and 360mm radiators.

As noted by Cowcotland, the core design, including the large, magnetically attached top panel for easy installation, appears to be the same. However, the fans have received a noticeable upgrade, operating within a speed range of 600 RPM to 2500 RPM and delivering 79.3 CFM airflow and 3.45 mmH2O static pressure. Moreover, the fans incorporate a daisy-chaining mechanism that simplifies PWM and RGB connection.

The most significant difference between the LA (240 and 360) and the LA-S (240 and 360) AIO liquid coolers is the display on the latter. While the LA series has a water block cover with the Thermaltake logo and lighting, the LA-S uses a 2.4-inch display with CPU metrics.

The Thermaltake LA series starts at €74.90 for the LA240 model and increases to €84.90 for the LA360. As for the LA-S series, the LA240-S model goes for €84.90 and the LA360-S model for €94.90.

KitGuru says: Considering their feature set, the new Thermaltake AIO CPU coolers sure look cheap. Would you go for the LA or the LA-S AIO liquid coolers?

Sharkoon has expanded its Skiller headset lineup with two new offerings: the SGH20 USB and the SGH25. Both headsets aim to provide immersive audio and comfortable wear, catering to different user needs and budgets, even if both are still entry-level.

The Skiller SGH20 USB headset presents a design with dynamically changing RGB lighting. The headset uses 50mm drivers and connects via USB for a straightforward setup with PCs. For extended gaming sessions, the SGH20 has soft synthetic leather ear cushions designed to seal around the ears, while the cushioned headband aims to distribute weight evenly. The flexible microphone can be easily positioned for clear voice transmission.

Priced higher, the Skiller SGH25 offers a different set of features. It also uses 50mm drivers and a USB connection, but instead of synthetic leather, it uses cloth for the ear cushions. As such, it remains comfortable even in warmer environments. Moreover, the SGH25 features an inline controller integrated into the cable, providing volume and microphone control, and circumaural earcups to isolate noise and provide a snug fit. The adjustable headband is designed to adapt to various head sizes and shapes. Another extra of the SGH25 over the SGH20 is the detachable microphone.

The Skiller SGH20 is priced at €19.90, while the Skiller SGH25 is positioned as a more budget-friendly option, costing €19.99.

KitGuru says: If you were to get an entry-level gaming headset, would you consider the new Sharkoon Skiller SGH20 and SGH25?

MSI has another great game bundle available. This time around, MSI has partnered with PlayStation Studios to offer up free codes for The Last of Us Part 2 Remastered on PC. These codes will be redeemable to those picking up an MSI QD-OLED gaming monitor. 

Those planning to buy a QD-OLED monitor soon might want to look at MSI's lineup. In addition to the monitor, you can get a digital copy of The Last of Us Part II Remastered, adding value to your purchase.

In the marketing material, MSI shows two of its most recent models, the MSI MAG 272URX QD-OLED and the MSI MAG 322URX QD-OLED, so we know those will be eligible for the free game code. However, the full list of eligible products has not yet been revealed.

In order to claim your code, you will need to register your purchased monitor with MSI. Once your purchase has been confirmed, MSI will send you the code for the game, redeemable on Steam.

KitGuru says: Planning on getting an MSI QD-OLED monitor? Will you be waiting to take advantage of the new game bundle? 

Astro Bot, the acclaimed platformer, emerged as the dominant force at the 2025 BAFTA Games Awards, securing five wins, including the prestigious Best Game of the Year. Still Wakes the Deep, a psychological horror title, also enjoyed a successful night, earning three BAFTA awards, two of which recognised outstanding acting performances.

The awards saw intense competition, wiSenua'sa's Saga: Hellblade II leading the nominations with 11. Astro Bot and Still Wakes the Deep followed closely with six nominations each, alongside Call of Duty: Black Ops 6, Lego Horizon Adventures, and Final Fantasy VII Rebirth. The nominations encompassed games released between November 25th, 2023, and November 15th, 2024.

Below you can find the list of nominees for GOTY at the BAFTAs:

Best Game:

• Astro Bot – Winner
• Balatro
• Black Myth: Wukong
• Helldivers 2
• The Legend of Zelda: Echoes of Wisdom
• Thank GoodneYou're’re Here!

Beyond celebrating the year's best games, the ceremony also recognised significant contributions to the industry. Bridgerton star Adjoa Andoh presented Alec Newman with the BAFTA for Performer in a Leading Role for his portrayal of Cameron McLeary in Still Wakes The Deep. Additionally, Japanese composer Yoko Shimomura was honoured with the BAFTA Fellowship, the highest accolade presented at the BAFTA Games Awards for her iconic work on series like Kingdom Hearts, Final Fantasy XV, and Mario & Luigi.

KitGuru says: Do you think the winners deserved their respective awards? What other games should have won?

It's known that Nvidia has been using Micron and Samsung GDDR7 chips to power the RTX 50 series. However, a new player has now joined their ranks, with SK Hynix now supplying GDDR7 memory for newly manufactured RTX 50 series graphics cards. 

It's important to clarify that Nvidia doesn't actually use the memory chips. While board partners ultimately procure and install the memory chips, Nvidia often selects the memory vendors. This practice stems from cost-efficiency, as purchasing GPUs and memory in bundles can be more economical.

According to MEGAsizeGPU, Nvidia's GeForce RTX 50 GPUs now support memory from three manufacturers: Samsung, SK Hynix, and Micron.

To identify the memory vendor on a specific graphics card, users can run GPU-Z, which provides detailed information about the installed memory chips. You can also disassemble the card to see what memory modules are installed. However, this may void the warranty or damage thermal pads, so it's better to stick with GPU-Z.

KitGuru says: With one more memory manufacturer producing memory chips for Nvidia, do you think the availability of these cards will improve?

Gigabyte has registered its upcoming Radeon RX 9060 XT graphics card variants with the RRA (South Korean Radio Research Agency), signaling their readiness for release. Moreover, it registered various RTX 5060 Ti cards, but to our surprise, all were 16GB variants.

There's a notable disparity between Gigabyte's planned offerings for Radeon and GeForce GPUs. As spotted by @harukaze5719, Gigabyte intends to provide a diverse range of GeForce RTX 5060 Ti models, including Eagle, Windforce, Gaming, Aero, and even Aorus Elite options. However, its Radeon RX 9060 XT lineup is more limited, featuring only Gaming models, one with 16GB VRAM and another with 8GB. Additionally, the GeForce RTX 5060 Ti SKUs do not include any 8GB variants.

Image credit: @harukaze5719

The GeForce RTX 5060 Ti and the Radeon RX 9060 XT are expected to be available in 16GB and 8GB memory configurations. However, the memory technology differs: the RTX 5060 Ti will use GDDR7, while the Radeon RX 9060 XT will use GDDR6. Both cards are projected to employ a 128-bit memory bus, resulting in a bandwidth of 448 GB/s for the former and 320 GB/s for the latter.

The GeForce RTX 5060 Ti is anticipated to launch next week, while a firm release date for the Radeon RX 9060 XT is still unknown.

KitGuru says: Out of the two upcoming mid-range GPUs from Nvidia and AMD, which one are you the most curious about? Are you planning on getting one of them?

Corsair has launched its new, browser-based tool for firmware updates, offering a streamlined alternative to iCUE. This move aims to simplify the perceived complexity of firmware updates, particularly for users less familiar with PC peripherals.

Corsair states that this web-based utility is a response to customer feedback, providing a more convenient method for updating firmware on compatible Corsair mice, keyboards, and other devices. While iCue remains the primary software for accessing the full range of customisation options, including lighting control and advanced settings, this new tool caters to users prioritising quick and easy firmware upgrades.

Corsair emphasises that the web-based updater ensures users have access to the latest firmware versions. Users can also opt to provide their email addresses to receive notifications about future firmware releases. The initial release of the web-based firmware updater supports a selection of Corsair devices, with Corsair promising to expand compatibility to include both new and existing devices and introduce new features in the near future.

The web-based tool is optimised for Chrome and Chromium-based browsers (such as Edge and Opera) and is compatible with Windows and Mac operating systems. Corsair is also currently testing Linux support.

Discuss on our Facebook page, HERE.

KitGuru says: Do you own Corsair peripherals? Will you keep using iCue to update the firmware, or will you start using the web-based firmware updater?

The configuration of Nvidia's upcoming lower-end GeForce RTX 50 series GPUs is still mostly unconfirmed, but rumours and reports have given us an idea of what to expect. The latest report comes from newly found shipping manifests, which show the memory bus of a GB206-based graphics card.

While previous reports linked PG152 board designs to the rumoured RTX 5060 Ti, RTX 5060, and RTX 5050 “Blackwell” GPUs, the reported “128-bit memory bus” specification for these cards has drawn scepticism from some. Many had hoped that Nvidia would opt for a wider 192-bit bus for its next-generation budget-friendly options, rather than seemingly revisiting design choices from the GeForce RTX 4060 series.

Image credit: VideoCardz

Unfortunately, it seems Nvidia has indeed chosen the 128-bit memory bus for the upcoming entry-level graphics cards. VideoCardz found shipment manifests with the IDs “PG152 SKU 25” and “PG152 SKU 10”, believed to correspond to GeForce RTX 5060 and GeForce RTX 5060 Ti graphics cards, respectively. The manifest's description linked these to a 128-bit memory bus.

Despite the 128-bit memory bus, these “Blackwell” GPUs are expected to leverage GDDR7 memory, increasing the bandwidth compared to the predecessors. This upgrade pushes bandwidth to 448 GB/s, a significant improvement over the 272/288 GB/s the older models offer. Based on the latest reports, the Nvidia RTX 5060 series is expected to be announced next week, with a release following soon.

Discuss on our Facebook page, HERE.

KitGuru says: Should Nvidia have chosen a 192-bit memory bus for the RTX 5060 series GPUs, even if it came with a price increase?

Enthusiasts delving into the overclocking potential of AMD's latest Radeon RX 9070 series found something we haven't seen for some time. In a throwback to the days of widespread BIOS modding, it appears that some Radeon RX 9070 non-XT GPUs can be flashed with the BIOS from their XT counterparts, increasing power limits and unlocking more performance.

As PC Games Hardware (via VideoCardz) reported, this BIOS flashing effectively unlocks higher clock speeds and a greater TBP. However, it's crucial to understand that this modification does not alter the core count or memory specifications, which are currently identical across both SKUs.

By modding the BIOS of these cards, users can overclock these cards more aggressively. One modder claims that the flashed card achieves performance levels comparable to or exceeding the reference RX 9070 XT. The performance uplift is substantial, with users reporting gains of 15-20% compared to the stock RX 9070 non-XT.

While the performance gains are enticing, there are potential drawbacks. Reports indicate that ULPS (Ultra Low Power State) mode may become dysfunctional, and some users have experienced system freezes. Furthermore, there are hardware considerations, such as the power delivery system, which might not be adequate for the extra that will go through it. If you want to go through this route, ensure that your card can handle it.

KitGuru says: If you had an RX 9070, would you mod it with the RX 9070 XT BIOS? 

The heavily rumoured remake of Bethesda's beloved 2006 RPG, The Elder Scrolls IV: Oblivion, might be coming soon. According to the latest reports, the game is ready to be launched and could be shadowdropped later this month.

The report from Jeff Grub from GiantBomb matches the previous rumours and claims that Bethesda was considering a simultaneous reveal and release strategy for March or April. Note, however, that there was also an earlier report pointing to a release in June.

Reports indicate that Virtuous is developing the supposed Oblivion remake. Moreover, it's supposed to be a “full remake” using Unreal Engine 5 with significant overhauls to various gameplay mechanics. The game's existence was initially hinted at in documents from the 2023 Microsoft vs. FTC lawsuit related to Microsoft's acquisition of Activision Blizzard. These documents listed an unnamed Bethesda title referred to as “Oblivion Remaster”, slated for release in fiscal year 2022, which didn't happen.

KitGuru says: Would you like to see a TES: Oblivion made with the current technology? How would you idealise it?

Reports on the Nvidia GeForce RTX 5060 Ti are increasing, with the latest being benchmark entries on the Furmark database. Nvidia's board partners are reportedly already distributing samples, so we expect to see a few more benchmark database leaks in the coming weeks.

One of the earliest leaks has surfaced from FurMark. Richard Huynh (via VideoCardz) discovered a new entry within the benchmark's database with the code 10DE-2D04, identifying an Nvidia GB206 GPU. Subvariant data further points to HP and Gigabyte as the originators of this information. VideoCardz found two other entries with the same PCI ID, suggesting these are also an RTX 5060 Ti.

Image credit: Richard Huynh

FurMark's focus is typically on thermal stress testing rather than performance benchmarking, making its appearance as an early leak somewhat unusual. Notably, one of the tested cards was power-limited to 61% TDP (116W), meaning that at its 100% TDP, it would be 190W. Given that the RTX 5060 Ti is rumoured to have a default TDP of 180W, this sample might be from a board partner who may have a 10W higher power limit out of the box.

Overall, the scores show a performance level that is not significantly higher than the RTX 4060 Ti, if not pretty much the same. However, this might not reflect the final results, as various factors like driver versions or platform limitations might be limiting the card's capabilities. Although the release date of the RTX 5060 Ti hasn't been confirmed yet, it is expected to launch within the following weeks.

Discuss on our Facebook page, HERE.

KitGuru says: How will the RTX 5060 Ti compare against the RTX 4060 Ti when it releases?

A new connection standard might be on the horizon. Developed by a consortium of Chinese companies, General Purpose Media Interface (GPMI) leverages USB Type-C and USB Type-B connectors to deliver high bandwidth for next-generation displays and integrated power delivery.

As reported by HKEPC (via VideoCardz), GPMI is positioned as a future-proof, all-encompassing technology designed to supersede existing audio and video interfaces. The standard's core emphasis is on providing a comprehensive solution, integrating power delivery within the cable to address power supply limitations. However, GPMI's capabilities extend beyond power, encompassing data transfer, fast wake-up functionality, bidirectional connections, and other features.

The standard is targeted at a wide range of applications, spanning consumer electronics, radio and television broadcasting, smart homes, security monitoring systems, automotive systems, and industrial equipment. Notably, the initial focus appears to be on the consumer electronics sector.

The USB Type-C implementation offers up to 96 Gbps, surpassing the latest DisplayPort 2.1 standard's bandwidth while supporting up to 240W of power delivery. The USB Type-B variant significantly expands these limits, doubling bandwidth and power capacity to 192 Gbps and 480W, respectively. The standard's bidirectional nature allows simultaneous data transfer alongside audio and video transmission. For example, a set-top box, typically relying on WiFi for network connectivity and HDMI for audio/video output, could potentially use GPMI for network data reception and display output, eliminating the need for separate connections.

GPMI is being championed by the Shenzhen 8K Ultra High Definition Video Industry Collaboration Alliance, which comprises over 50 companies, including TCL, Hisense, and Huawei. Moreover, the Type-C variant has already obtained the approval of the USB Association, meaning we might see devices using the standard very soon.

Discuss on our Facebook page, HERE.

KitGuru says: Do you think other manufacturers should use the GPMI standard? Will it ever become the norm for USB-C and USB-B cables and connectors?

Nvidia has made two of its proprietary technologies open-source: Flow, a real-time, grid–based fluid simulation library, and PhysX, a GPU acceleration library for physics.

While Nvidia initially released the PhysX library on GitHub a few years ago, it wasn't fully open source. As Wccftech highlights, Nvidia has now updated its PhysX GitHub repository to include the complete SDK source code, encompassing its GPU acceleration capabilities. In addition to PhysX, Nvidia is open-sourcing its Flow technology, providing the complete GPU compute shader version of the SDK.

PhysX recently garnered attention due to performance discrepancies in some games on Nvidia's latest GeForce RTX 50 series hardware, as these GPUs lack support for 32-bit PhysX. However, Nvidia has indicated that it does not intend to reinstate compatibility for the older 32-bit library.

Now that it has become open-source, developers might develop a 64-bit PhysX library to run on newer hardware or a translation layer so that newer architectures can interpret the instructions. Hopefully, now that it has become open-source, someone capable will find a way to run PhysX on modern hardware with decent performance.

Discuss on our Facebook page, HERE.

KitGuru says: Would you like it if the RTX 50 series could run PhysX properly, or is it something that doesn't bother you?

CPU-Z validation results for Q1 2025 reveal an emerging trend in CPU preferences, with eight-core processors gaining significant traction. Data shows that eight-core CPUs now represent 24.7% of all validations, marking a substantial 32.6% increase compared to previous figures. Conversely, six-core processors have declined, accounting for only 22.5% of validations, a 6.9% decrease.

This surge in eight-core CPU adoption reflects the growing demand for multi-threaded performance across various computing environments, from professional workstations to high-end gaming systems. These market share figures (via VideoCardz) also offer insights into shifts in customer allegiance among CPU manufacturers. While Intel maintains its market dominance with 56.3% of validations, AMD has made significant strides, capturing 43.7% of the market share, a notable 16.6% increase from the previous year.

The rising popularity of specific models is likely a key driver behind the eight-core trend. One such model would be the Ryzen 7 9800X3D, which emerged as the most popular CPU based on CPU-Z validations.

Another interesting point in CPU-Z's data is that 27-inch displays are now more popular than 24-inch monitors, following a 7% growth in a year. There has also been a shift in GPU brand popularity. Nvidia appears to have lost 6.4% of its market share in a year, leaving it at 67.9%, while AMD now has 31.1%, up 16.6% from the previous year. On the other hand, Intel controls 0.7% of the GPU market (presumably for discrete GPUs), representing a 27% rise in a year.

Discuss on our Facebook page, HERE.

KitGuru says: How many cores does your CPU have? 

Nvidia is reportedly gearing up to launch its GeForce RTX 5060 graphics cards, with information from Board Channels suggesting a potential release schedule. As previously indicated, Nvidia is expected to introduce the RTX 5060 series this month. This new lineup is anticipated to include the RTX 5060 Ti in both 16GB and 8GB configurations and the standard RTX 5060.

On the Board Channels forum (via VideoCardz), it has been suggested that Nvidia intends to maintain pricing parity with the previous generation. The RTX 4060 Ti launched at $399 for the 8GB variant, while the 16GB variant carried a $100 premium. If Nvidia adheres to this precedent, it may necessitate phasing out older models or implementing price reductions to remain competitive.

It's important to approach these pricing rumours cautiously, as pricing information can fluctuate significantly close to launch. However, Nvidia's strategy of mirroring the RTX 40 series SKU selection makes this pricing approach plausible. Still, depending on the pricing of competing products, like AMD's upcoming Radeon RX 9060 XT, which is also rumoured to offer 16 GB and 8 GB memory configurations, the RTX 5060 Ti could be forced to be sold at a lower price point.

Currently, pricing details for the non-Ti RTX 5060 remain undisclosed. The coming weeks will likely reveal more concrete information about Nvidia's pricing strategy and how it will position the RTX 5060 series against AMD's offerings.

Discuss on our Facebook page, HERE.

KitGuru says: The RTX 5070 and RTX 5070 Ti held similar price tags to their last-gen counterparts. Do you think that trend will continue for the RTX 5060? 

The ongoing saga of potential collaboration between Intel and TSMC has taken another turn, with a new report suggesting a “preliminary agreement” for a joint venture involving Intel's chip fabrication plants (fabs).

This news, reported by The Information (via Reuters), revisits a familiar narrative. This latest development shares a bit more about the potential collaboration, with TSMC reportedly considering acquiring a 20% stake in Intel's fabs.

Conflicting reports have emerged regarding the ownership structure of the new entity formed from the current Intel Foundry. Some sources indicate that “Intel and other U.S. chip companies” would retain a controlling interest, but the specifics remain hazy. TSMC's influence over Intel's fabs under this agreement is also unclear. While TSMC has a superior track record in recent years, it's unlikely it could instantly overhaul Intel's manufacturing processes. Reversing Intel's course in chip fabrication, particularly with processes like the nearly finalised 18A node, would likely be a multi-year undertaking.

Initial rumours have been followed by reports of TSMC engaging with other industry giants like Nvidia, AMD, and Broadcom about participating in such a venture. However, the involvement of these other firms in the current “preliminary agreement” remains unclear.

It's worth noting that Intel's Direct Connect event, focused on its foundry business, is scheduled for April 29th. Any potential announcements regarding collaborations or changes to Intel Foundry could be unveiled at this event.

Discuss on our Facebook page, HERE.

KitGuru says: Do you think TSMC will partner with Intel? Will other semiconductor companies join this venture?

Microsoft has announced the general availability of hot patch updates for enterprise customers using Windows 11 Enterprise 24H2 on x64 systems. Hotpatching in ARM-based devices running Windows 11 Enterprise is also in the works and currently in public preview.

Hotpatching enables the application of OS security updates in the background, eliminating the need for restarts. This is achieved by patching the in-memory code of running processes, eliminating the downtime associated with traditional updates. The deployment of these updates is controlled through Windows Autopatch policies created in the Microsoft Intune console. Devices managed by these hotpatch-enabled policies will receive hotpatch updates every quarter.

To be eligible for hotpatching, devices must meet specific requirements:

• A Microsoft subscription (Windows 11 Enterprise E3, E5, or F3, Windows 11 Education A3 or A5, or a Windows 365 Enterprise subscription)
• A Windows 11 Enterprise 24H2 PC with the latest baseline update
• An x64 AMD64 or Intel CPU
• Virtualization-based Security (VBS) enabled
• Microsoft Intune for managing hotpatch update distribution

Microsoft notes that hotpatch updates for Arm64 devices are currently in public preview. However, administrators can temporarily disable CHPE support via registry (Path: “HKLM\SYSTEM\CurrentControlSet\Control\Session Manager\Memory Management”, Dword Key Value: “HotPatchRestrictions=1”) to receive the functionality ahead of the official release. For the remaining eligible devices, hotpatching can be turned on or off within the Microsoft Intune admin center by navigating to Devices > “Windows updates” > “Create Windows quality update policy”.

This technology is not entirely new to the Microsoft ecosystem. Windows Hotpatch capability was initially introduced for Windows Server Azure Edition core virtual machines and later became widely available in February 2022 for systems running Windows Server 2022 Datacenter: Azure Edition. Since then, Microsoft has added hot patching to various preview builds of operating systems but has yet to arrive in the mainstream segment. Hopefully, hot patching will eventually be added to Windows 11 Home and Pro editions.

Discuss on our Facebook page, HERE.

KitGuru says: Do you have a system running Windows 11 24H2? Would you like it if it had hotpatch support?

New rumours suggest AMD might be preparing to expand its RDNA 4 graphics card lineup with a new addition. These claims have emerged following reports of low stock for the non-XT Radeon RX 9070 in China, fuelling speculation that AMD might be clearing the way for a new model.

According to ITHome (via VideoCardz), a Radeon RX 9070 GRE is in development. While this initially points towards a China-exclusive release, it's worth noting that AMD has previously offered GRE versions to the global market, as seen with the RX 7900 GRE. Benchlife also corroborated these reports, stating that its sources claim AIB partners are presently preparing the Radeon RX 9070 GRE and are ready for mass production. If so, expect the first RX 9070 GRE graphics cards to emerge in the Chinese market shortly.

Currently, detailed specifications for this potential RX 9070 GRE remain scarce. However, ITHome and Benchlife speculate it will feature a 192-bit memory bus and a 12 GB GDDR6 configuration. A hypothetical RX 9070 GRE would effectively slot between the RX 9070 and the RX 9060 series in terms of performance and memory configuration, while also aligning with the memory specifications of the Nvidia RTX 5070.

Discuss on our Facebook page, HERE.

KitGuru says: How much slower than the RX 9070 do you think this supposed RX 9070 GRE will be?

MediaTek has unveiled its latest processor, the Kompanio Ultra 910, designed to elevate Chromebook performance. This new chipset emphasises AI capabilities, enhanced power efficiency, and high-speed connectivity, incorporating advanced computing technologies to drive productivity, creativity, and multimedia experiences.

Built on TSMC's 2nd-generation 3nm process, the MediaTek Kompanio Ultra 910 processor aims to deliver exceptional power efficiency, enabling Chromebooks to achieve all-day battery life while providing a significant performance boost. The CPU architecture features an eight-core configuration, comprising a high-performance Arm Cortex-X925 core clocked at up to 3.62 GHz, three Cortex-X4 cores, and four Cortex-A720 cores. This core combination is coupled with 12 MB of L3 cache and 10 MB of system-level cache (SLC).

The 11-core Immortalis-G925 MC11 GPU handles graphics processing, including ray tracing support, enhancing gaming and graphics-intensive applications. The Kompanio Ultra 910 also integrates LPDDR5X RAM (8533 Mbps) to support its processing power. Early benchmark results indicate strong performance, with a Geekbench 6 single-core score of 2600 and a multi-core score of 8000.

The Kompanio Ultra 910 uses an 8th-generation NPU, significantly boosting on-device AI performance. This enables real-time AI image and video generation and achieves up to 50 TOPS AI performance. AI acceleration also improves overall power efficiency by offloading tasks from the CPU and GPU.

The Kompanio Ultra 910 also features Bluetooth 6.0 (dual engine) and Wi-Fi 7, which can support speeds up to 7.3 Gbps. Moreover, this SoC can simultaneously drive up to three 4K displays: one integrated and two external. Lastly, it supports 4K 10-bit video decoding and encoding in HEVC, AVC, VP9, and AV1 formats.

Discuss on our Facebook page, HERE.

KitGuru says: Although it looks like a capable SoC in single-core and AI workloads, its multi-core score doesn't come near what many expect from a proper, high-end laptop CPU. However, it should be more than enough given what most Chromebook users do in their systems.