TechPowerUP NEWS

Texas Instruments (Nasdaq: TXN), a global semiconductor company that designs, manufactures and sells analog and embedded processing chips, and Silicon Labs (Nasdaq: SLAB), a leader in secure, intelligent wireless technology, today announced they have signed a definitive agreement under which Texas Instruments will acquire Silicon Labs for $231.00 per share in an all-cash transaction, representing a total enterprise value of approximately $7.5 billion.

The acquisition will create a global leader in embedded wireless connectivity solutions by combining Silicon Labs' strong portfolio and expertise in mixed signal solutions with Texas Instruments' leading analog and embedded processing portfolio and internally owned technology and manufacturing capabilities. The combined company will accelerate growth by better serving existing and new customers through enhanced innovation and market access.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AMD posted its record fourth quarter revenue of $10.3 billion in 2025, and during the earnings call, the company issued some guidance on the upcoming product portfolio. During the call, AMD confirmed that Valve's Steam Machine is on track and shipping early this year, while its custom SoC division that designs processors for PlayStation and Xbox consoles will deliver an RDNA 5-based SoC for the next-generation Xbox console. While the [Only registered and activated users can see links. Click Here To Register...] are confirmed, Xbox [Only registered and activated users can see links. Click Here To Register...] SoC is still largely a collection of rumored specifications. The "Magnus" SoC is rumored to feature the largest APU ever designed for a consumer console, with a 408 mm² chiplet design. Of this, 144 mm² is dedicated to the SoC built on TSMC's N3P node, while the GPU occupies 264 mm². The AMD chip is expected to include up to 11 CPU cores—three Zen 6 and eight Zen 6c—alongside a substantial GPU setup with 68 RDNA 5 compute units, four shader engines, and at least 24 MB of L2 cache. Memory might expand to 48 GB of GDDR7 on a 192-bit bus. A dedicated NPU is rumored to offer significant on-device AI performance, with reports suggesting up to 110 TOPS.

Dr. Lisa SuFor 2026, we expect semi-custom SoC annual revenue to decline by a significant double-digit percentage as we enter the seventh year of what has been a very strong console cycle. From a product standpoint, Valve is on track to begin shipping its AMD-powered Steam Machine early this year, and development of Microsoft's next-gen Xbox featuring an AMD semi-custom SoC is progressing well to support a launch in 2027.

[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Welcome to the second week of February and another packed week of new releases, even though none of them are AAA titles. The major release this week comes from the creators of Little Nightmares and to no surprise, we're talking a co-op horror adventure. The week kicks off with a bad bout of the Monday's, followed by some artifact hunting and a trip to Japan. Next up we have some motorcycle racing, followed by a sequel of a rather crazy game in space. This week we also end up with a Saturday release, which we decided to include. On top of all of that, we have a few extra games that didn't quite make the list, but that we reckon some of you might be interested in.

[Only registered and activated users can see links. Click Here To Register...] / This week's major release / Friday 13 February
The original creators of Little Nightmares & Little Nightmares II have returned to take you on a more terrifying journey than ever before. In this co-op horror adventure game, you play as a brother & sister who go through hell to rescue their missing friends. Exploring by boat and on land, you must use your wits to survive, work together to escape the hellish island, and the dark secret that haunts you. [Only registered and activated users can see links. Click Here To Register...][Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

The company that started the entire wave of AI infrastructure and development is now enjoying the fruits of its work. NVIDIA has deployed generative AI tools across its company to an astonishing 30,000 engineers. In a partnership with San Francisco-based Anysphere Inc., the company is getting a customized version of the Cursor integrated developer environment, which focuses on AI code design. This is important to note as NVIDIA's engineers are now reportedly producing as much as three times the code compared to the previous development pipeline, and we are now probably using NVIDIA's products or services that have been designed by AI guided by humans.

NVIDIA offers a range of mission-critical products that cannot afford to be as error-prone as most AI-generated code tends to be. This includes GPU drivers that support everything from basic gaming to large-scale AI training and inference operations. The company is likely enforcing strict guidelines for its newly generated code, with an extensive range of tests required before the code is deployed in production. This isn't the first time NVIDIA has utilized AI-assisted workflows in its products. The company has already implemented a dedicated supercomputer that has been [Only registered and activated users can see links. Click Here To Register...] (Deep Learning Super Sampling) for several years, and some [Only registered and activated users can see links. Click Here To Register...] using the company's internal AI tools.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

In an exclusive report for VideoCardz, Intel is reportedly targeting an LPDDR5X memory speed of 8,533 MT/s for its upcoming [Only registered and activated users can see links. Click Here To Register...] arriving in the second quarter for handheld gaming devices. After we learned that Intel is imposing certain [Only registered and activated users can see links. Click Here To Register...], it seems like the Core Ultra G3 will face similar mandates from the company to prevent OEMs from "cutting corners" and implementing slower LPDDR5X memory. For the new handheld-tuned Core Ultra G3 and G3 Extreme, that specification is now set to 8,533 MT/s, which is slightly below its flagship "Panther Lake" Core Ultra X SKUs that can support LPDDR5X memory running at 9,600 MT/s.

Presumably, Intel will require OEM partners and makers of the next-generation handheld consoles to use this 8,533 MT/s memory on both SKUs. These chips will feature a 14-core CPU configuration, including two P-Cores, eight E-Cores, and four LPE-Cores. A key selling point of these SoCs is the Arc integrated graphics, with the G3 Extreme offering 12 Xe3 cores and the standard G3 featuring 10 Xe3 cores. The G3 Extreme plans to run the Arc B380 iGPU with 12 Xe3 cores at 2.3 GHz, just 200 MHz below the flagship Core Ultra X9 388H's Arc B390. Essentially, G3 Extreme handhelds can expect gaming performance similar to that of the flagship SKU, albeit with two fewer P-Cores and a slightly lower GPU clock speed. The regular G3 maintains its CPU capabilities, but the GPU is reduced to a 10-core Xe3 IP called Arc B360, with a GPU boost frequency of 2.2 GHz, resulting in a notable decrease in both gaming performance and TDP.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Noctua has earned itself a reputation in the enthusiast space not only for its highly efficient air coolers, but also for its free mounting upgrade kit policy, which allows buyers of Noctua coolers to request a new mounting kit when a new motherboard socket from AMD or Intel launches, even years after they've purchased their original cooler. The company recently announced that it has shipped 500,000 of these free mounting upgrade kits since the project started way back in 2006 with the AMD AM2 socket. Noctua also promises that "as new sockets and architectures appear, we will continue to provide free mounting or upgrade kits where technically feasible."

Noctua explains that the ethos behind the program is to "ensure that a quality product remains useful for many years," and, in the blog post celebrating the achievement, it explains that change in approach meant rethinking the design of its coolers. New coolers designed by Noctua from 2006 onwards would need to be modular and standardized to maintain support across different platforms and generations. Noctua elaborated that part of the justification for the past and ongoing commitment to the free upgrade program is as much about the environment as it is about customer service and technical compatibility. Theoretically, 500,000 mounting kit upgrades means that up to 500,000 perfectly functional CPU coolers were kept out of the e-waste pile, conserving raw materials and cutting down on waste.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Intel's Arc graphics solutions initially had a rough start to life on the Linux side of things, but it seems as though the tide is changing. [Only registered and activated users can see links. Click Here To Register...] the new Panther Lake Intel Arc B390 iGPU in the [Only registered and activated users can see links. Click Here To Register...], comparing the new iGPU to the venerable [Only registered and activated users can see links. Click Here To Register...] in the AMD Ryzen AI 9 HX 370, and, surprisingly, Intel's iGPU consistently outpaces the Radeon iGPU, which isn't a surprise if you've been keeping track of early benchmarks on Windows. The publication tested the new Xe3 integrated graphics in an MSI Prestige 14 running the latest Mesa 26.0 drivers on Linux Kernel version 6.19 on Ubuntu 26.04.

When it comes to gaming, the only games in which the Xe3 iGPU was bested by the AMD competition were Counterstrike 2, running at 1920×1200, and Quake II RTX. In every other game, the Intel Arc B390 beat both the AMD competition and the Intel Core Ultra 7 155H and Core Ultra 7 258V significantly. Depending on the game and in-game settings, the Arc B390 was anywhere between 15% and upwards of 50%. Hitman 3 was a particularly impressive outlier, with the B390 managing to get over 50% more FPS than the nearest competitor and 64% more FPS than the Radeon 890M at 1920×1200 and low settings, with similar gains at medium quality settings, and a slightly less impressive performance improvement when bumping up to ultra settings.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Intel's "Arrow Lake Refresh" has not even been released, but the company has already canceled its flagship SKU planned for this refresh cycle, according to a report from [Only registered and activated users can see links. Click Here To Register...]. Two sources close to the media note that Intel's flagship Core Ultra 9 290K Plus might not roll out at all, despite the massive hype and leaked benchmarks indicating that Intel is releasing this CPU SKU as part of the "Arrow Lake Refresh" generation expected to arrive in March or April. Reportedly, Intel will instead focus on delivering value with its Core Ultra 7 270K Plus SKU, which carries 8 P-Cores and 16 E-Cores and a 5.5 GHz maximum turbo boost. For individual boosting frequency, P-Cores top out at 5.4 GHz, while the base runs at 3.7 GHz. For E-Cores, the boost frequency is set to a maximum of 4.7 GHz, while the base is set at 3.2 GHz.

As for a possible reason why Intel would cancel this SKU, the sources close to VideoCardz note that product overlap is the main issue, as the flagship Core Ultra 9 290K Plus would have the same core configuration as the Core Ultra 7 270K Plus, just with slightly higher clock speeds. Additionally, Intel already maintains a Core Ultra 9 285K SKU from the regular "Arrow Lake" family, meaning that the company would have three similar SKUs at the very top of the stack. This way, it would only have to maintain two products, which would simplify manufacturing and supply chain logistics, allowing Intel to spend more time preparing for the next-generation "Nova Lake" launch later this year.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

AMD Ryzen Threadripper Pro 9995WX HEDT processor with 96 cores and 192 threads comes with a default TDP of 350 W. However, heavy overclocking can bring the CPU to 1,300 W and requires a custom integrated heat spreader (IHS) that serves as a direct-die waterblock. In the latest endeavor by Geekerwan, the enthusiast created a custom fin structure inside the Ryzen Threadripper Pro 9995WX IHS that serves as a direct-die waterblock to achieve an impressive overclock of 5.325 GHz, drawing an astonishing 1,340 W during load, with the entire system drawing around 1,700 W. According to Geekerwan, he contacted ASUS China regional manager Tony Yu to experiment with different IHS designs before "ruining" the IHS of a $12,000 HEDT CPU. He then proceeded with trying a straight fin structure common in commercial waterblocks, but also conducted computer simulations that showed a curved, wavy S-shaped fin structure is the most efficient in capturing heat, as the coolant flows over a longer distance with minimal obstruction, resulting in 20% better cooling than the straight fin structure.

The IHS of the AMD Ryzen Threadripper Pro 9995WX processor is 4.1 mm thick, which left Geekerwan with about 2.0 mm of fin depth and about 2.1 mm for the structural integrity of the IHS, which is subject to a lot of water pressure. After a heavy 19-hour session of CNC milling, the result is a CPU that ran between 30-50°C, which is an amazing temperature under Cinebench 2026 load. The system also placed 7th in Cinebench R23, just behind an LN2-cooled AMD Ryzen Threadripper Pro 7995WX running at 6.2 GHz. Impressive heat dissipation and the massive 5.325 GHz clock on a 96-core system are also made possible with an industrial chiller, two Bosch water pumps from cars, and a 37-gallon water tank. You can check out the entire process below.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

MSI's most powerful GPU—the GeForce RTX 5090 Lightning Z—will come with an extreme price tag to match, as the company has listed its GPU for NT$165,000, which works out to about $5,200. The company noted this pricing in a 24-hour giveaway scheduled to begin on Monday, February 9, at 10:00 AM Taiwanese time, lasting until Tuesday, February 10. The listing has revealed that the card we previewed at the [Only registered and activated users can see links. Click Here To Register...] is not only a premium design but also a premium-priced product, as the supply is limited to only 1,300 samples. MSI advertises a factory boost clock of 2,730 MHz and an "Extreme Performance" OC profile of 2,775 MHz. Additionally, the GPU is capable of reaching 3,742 MHz, which is the fastest LN2 GeForce RTX 5090 GPU ever.

The MSI GeForce RTX 5090 Lightning Z will come with an 800 W power limit out of the box, while the "Extreme" power preset mode gives it a 1000 W power envelope on the 360 mm AIO water cooling. The extensive engineering involved in the PCB design along with a 40-phase VRM allows the GPU to sustain multi-kilowatt loads. The card uses 28 Gbps Samsung GDDR7 memory, which can be overclocked to 36 Gbps on LN2. Additionally, only LN2 is capable of taming the XOC BIOS, which comes with 2.5 kW of power load and will require extensive PCB modifications. For a product that costs $5,200, only extreme overclockers would dare to modify the card. For the rest of us mere mortals, MSI recommends a power supply with a capacity of 1600 W, providing ample room for basic overclocking without ruining the card.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

NVIDIA's upcoming "Vera Rubin" AI systems are scheduled for [Only registered and activated users can see links. Click Here To Register...] in the form of VR200 NVL72 rack-scale solutions that will power the next generation of AI models. However, not every memory maker of HBM4 qualified for a design win, as Micron has reportedly fallen out of the equation, with only Samsung and SK hynix left to supply the precious HBM4 memory. According to leaked institutional notes from SemiAnalysis, which tracks the supply chain in great detail, SK Hynix will represent about 70% of the HBM4 supply for VR200 NVL72 systems, with Samsung getting the remaining 30% of the supply. For a major memory maker like Micron, there is reportedly zero commitment for the supply of HBM4 memory.

Interestingly, this is not the end of Micron's share of memory in NVIDIA VR200 NVL72 systems. Instead of HBM4, the company will supply LPDDR5X memory for "Vera" CPUs, which can be equipped with up to 1.5 TB of LPDDR5X, making up for the lost share with the HBM4. It is possible that Micron didn't qualify for the significant [Only registered and activated users can see links. Click Here To Register...] that NVIDIA performed for VR200 NVL72, which went from the initial system target of 13 TB/s in March 2025 to 20.5 TB/s in September. However, at CES 2026, NVIDIA confirmed that the VR200 NVL72 system is now operating at 22 TB/s of bandwidth, marking a nearly 70% increase in system bandwidth, all derived from aggressive memory specification scaling that the company demanded from the memory makers.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Intel and AMD have reportedly warned customers in China about tightening server CPU supplies, with delivery times stretching significantly and prices already moving higher. According to Reuters, average prices for Intel server processors in China have increased by more than 10%, depending on contract terms, while delivery times for some Xeon models have extended to as much as six months. The situation appears most severe for Intel's fourth and fifth-generation Xeon processors, where backlogs are reportedly building fast. AMD is facing similar, though less extreme constraints, with some server CPU orders now quoted at eight to ten weeks for delivery. China remains a major market for both companies, accounting for over 20% of Intel's total revenue, and includes customers such as large server OEMs and cloud operators like Alibaba and Tencent. The supply crunch is being driven by a mix of factors, including surging demand tied to AI infrastructure and so-called "agentic AI" systems, which significantly increase CPU requirements alongside GPUs.

Manufacturing limits are also having a negative effect with Intel admitting supply issues linked to manufacturing and yield problems, while AMD relies on TSMC, which now gives priority to high-profit AI chips. Rising memory costs in China have put more pressure on the market, with buyers rushing to buy CPUs to avoid paying more for complete systems later. Both companies say they are working to improve supply. Intel told Reuters its inventory hit a low point in Q1 but expects gradual improvement through 2026, while AMD stated it remains confident in its ability to meet demand based on supplier agreements and its partnership with TSMC. For now, however, customers may still be facing higher prices and longer waits for server CPUs.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

Simon Lin, the chairman of Wistron, has gone on record to say that AI is not a bubble and that the growth of AI-related orders for the company will be higher in 2026 than those it had in 2025. Reuters is quoting him saying "We believe AI really does help all industries, so I don't think it's a bubble; I think it will mark a new era. A new AI era is arriving." Wistron is expecting AI-related orders to see them through all the way into 2027 and this year's growth is expected to be "significant."

Wistron is expecting to open its new US manufacturing facility this half of 2026, where the company will be producing AI servers on behalf of NVIDIA. Wistron is said to have a four-year contract with NVIDIA that is valued at as much as US$500 billion. The question here is if the industry is trying its best to make as much profit out of the current market trend, or if this really is a long-term business for companies such as Wistron. We've seen past bubbles where company leadership has denied it being a bubble, until it was, but maybe it's different this time around.[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

GALLOP INNOTEK today unveiled a new generation of silicone-free thermal gap pad, the GTP160-NS. This GTP160-NS is engineered to solve silicone out-gassing or silicone bleeding in sensitive electronics, optical systems, satellite technology, and medical devices.

With the non-silicone formulation, GALLOP INNOTEK uses different materials like acrylic or urethane polymers to solve the siloxane compatibility problem. This marks a major milestone in GALLOP's efforts to provide thermally conductive pads for silicone-sensitive applications.[Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...] [Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]

According to Nikkei Asia, some of the biggest PC makers like ASUS, Acer, Dell, and HP are exploring alternative memory suppliers amid industry-wide memory shortages, which are forcing PC OEMs to seek supply even from Chinese memory maker CXMT. Late last year, [Only registered and activated users can see links. Click Here To Register...] unveiled its homegrown DDR5-8000 and LPDDR5X-10667 memory modules at the 2025 China International Semiconductor Expo. This has likely prompted many OEMs to start finding alternatives to the traditional triad of SK Hynix, Samsung, and Micron, whose supply has been very limited outside AI accelerator workloads.

CXMT offers 12 Gb and 16 Gb LPDDR5X capacities, while DDR5 scales to 16 Gb and 24 Gb module formats. The 16 Gb DDR5 chips from CXMT measure 67 square millimeters, with a density of 0.239 Gb per square millimeter. The G4 DRAM cells are 20% smaller than CXMT's previous G3 generation. Reportedly, CXMT manufactures these chips using a 16 nm node, which is three years behind Samsung, SK Hynix, and Micron in manufacturing capabilities. However, CXMT is progressing quickly, and its DRAM modules adhere to the official JEDEC specifications and even exceeding the specification, making them ideal for OEM PCs depending on the use case.[Only registered and activated users can see links. Click Here To Register...]
[Only registered and activated users can see links. Click Here To Register...]

[Only registered and activated users can see links. Click Here To Register...]