Latest News - Page 162

NVIDIA's RTX Remix platform has been a game-changer for the PC modding community, giving developers and enthusiasts access to tools that let them upgrade classic PC games with modern path-traced visuals. Plus, access to RTX Technologies like DLSS Super Resolution, Frame Generation, Reflex, and neural rendering.

One of the highest-profile RTX Remix projects currently in development is Half-Life 2 RTX, which looks and feels like a modern path-traced remake of Valve's iconic shooter, with impressive path-traced lighting. With over 125 RTX Remix mods and a vibrant community creating tools and plug-ins, RTX Remix is getting upgraded in 2026 with the arrival of RTX Remix Logic.

With 2025 bringing path-traced volumetric (like smoke) and particle effects, RTX Remix Logic builds on this by giving modders the ability to "remix" a game's visuals based on game logic. For example, in Half-Life 2 RTX, opening the door to Ravenholm could then randomly pick a time of day and type of weather, totally transforming the game. Yes, Half-Life 2: Winter.

NVIDIA has just announced DLSS 4.5, and within this update comes a bunch more tools to improve PC gaming performance, depending on your available hardware.

NVIDIA has announced at CES 2026 the deployment of DLSS 4.5, and how it has managed to increase multi-frame generation (MFG) up to 6x, enabling the possibility of reaching 4K at 240Hz with Path Tracing enabled. To achieve 6x MFG NVIDIA made several improvements to how the model interprets input frames, resulting in the quality of the generated frames improving.

Additionally, NVIDIA explained via a press briefing the introduction of Dynamic Multi Frame generation, which NVIDIA described as working similarly to an automatic transmission. When enabled the technology will detect when a graphically intensive scene is occurring and automatically increase frame generation to compensate for the reduction in FPS. NVIDIA said that by enabling this players will be able to experience a "continuous refresh rate." With Dynamic MFG enabled fewer generated frames will occur as the workload on the system reduces.

NVIDIA has outlined some of the advancements it has made in bringing local AI to as many users as possible, and according to the company one of the main issues with connecting PC users with local AI is the size of the AI models themselves.

NVIDIA explained during a press briefing for CES 2026 that significant improvements have been made in reducing the size of AI models, and maintaining an appropriate level of intelligence. According to NVIDIA, small AI models that were shrunk down from its larger Cloud-based counterparts were about two years behind in intelligence levels, but just in the last six months significant improvements have been made, doubling the intelligence of these small language models.

NVIDIA told the press these smaller models are getting close to matching their Cloud counterparts, and that generally the intelligence gap is closing between the two. NVIDIA added that it believes 2026 will be the "breakout year" for AI PCs, as other metric indicators such as the first months downloads for most of its PC models have jumped 10x in a year, with the top four tools also doubling in popularity during the same period.

At CES 2026, NVIDIA not only announced DLSS 4.5 for PC gaming but also made it available now in over 250 games and apps. This is all possible via the NVIDIA App, which offers a DLSS Override tool that lets users choose which model to use for Super Resolution upscaling.

NVIDIA's second-generation Transformer model for DLSS 4.5 features five times the computational power of the baseline DLSS 4 and takes full advantage of the more advanced Tensor Cores in GeForce RTX 40 and 50 Series GPUs. However, like with DLSS 4, users will be able to select the new Transformer Model, the older model, or DLSS 3.5's CNN model for upscaling.

And if you want to check out DLSS 4.5's improved image fidelity alongside reduced ghosting and shimmering, all you need to do is go to the NVIDIA App's Settings, opt in to access Beta or Experimental features, and DLSS 4.5 will be primed and ready to go.

Although we didn't get any new GeForce RTX GPU announcements from NVIDIA at CES 2026, not that we were expecting any based on the current memory crisis, but we did get the announcement and launch of DLSS 4.5.

Building on DLSS 4, which launched alongside the RTX Blackwell-powered GeForce RTX 50 Series in 2025, DLSS 4.5's second-generation Transformer model for Super Resolution features five times the computational power of the original DLSS 4's Transformer model. NVIDIA calls it its most sophisticated AI upscaling model so far, trained on a significantly expanded data set.

Now, as we and others have already covered, DLSS 4's Super Resolution model, which improves performance by upscaling from lower resolutions to 1080p, 1440p, or 4K, is (or was) the gold standard, offering impressive image clarity. So then, how good can DLSS 4.5 be?

NVIDIA talked about something really cool during CES 2026, with new gameplay being unlocked thanks to on-device AI.

The company talked about how on-device AI is helping unlock totally new gameplay, using Inzoi and PUBG as recent examples of using it for AI citizens in Inzoi, and AI teammates in PUBG. NVIDIA said it has worked with Total War developer Creative Assembly on reimagining Total War: PHARAOH Advisor with its NVIDIA ACE technology.

The Total War Advisor can help you get used to the gameplay mechanics, asking it questions along the way with a prompt, and the advisor will help you.

ASUS ROG has unveiled its new line up of products scheduled to be showcased at CES 2026, and three of them are brand new OLED gaming monitors that are pushing the envelope forward for the pixel technology.

Introducing the ROG Swift OLED PG27UCWM and PG34WCDN, two new QD-OLED gaming monitors that have received the Tandem RGB OLED upgrade, improving color reproduction and accuracy. For example, the PG27UCWM, a 27-inch QD-OLED gaming monitor that can either run in 4K 240Hz or 1080p at 480Hz, features the Tandem RGB OLED strip introduces improvements to the subpixel layout of the panel, resulting in sharper text edges and much more accurate color reproduction.

Additionally, Tandem RGB OLED introduces higher brightness levels for both color and white light, while also maintaining consistent brightness levels across the panel. ASUS explains the result of the Tandem OLED is no color fringing, distortion, or image sticking. The PG34WCDN also comes with the RGB strip, which ASUS claims can deliver a 27% larger color volume, especially at higher luminance levels.

NVIDIA founder and CEO Jensen Huang proudly took the stage at CES 2026, unveiling the company's next-generation Rubin AI platform.

NVIDIA's new Rubin AI platform is the successor to its dominant Blackwell AI chips, with Rubin being the first extreme-codesigned, six-chip AI platform, with Jensen adding that it's now in full production. NVIDIA is aiming to "push AI to the next frontier" with Rubin, not just offering far more computing power, but slicing the cost of generating tokens to around 1/10 of Blackwell, making large-scale AI "far more economical to deploy".

The use of extreme codesign means that designing all of the components together is essential because scaling AI to gigascale requires tighter integration innovation between chips, trays, racks, networking, storage, and software to remove bottlenecks. This massively reduces the costs of training and inference, added Huang.

ASUS has unveiled its new ROG Zephyrus G16 and G14 gaming laptops at CES 2026, powered by Intel's new Core Ultra Series 3 "Panther Lake" CPUs.

The new ROG Zephyrus G16 (GU606) and ROG Zephyrus G14 (GU405) gaming laptops sport the new Intel Panther Lake CPUs with up to 16 cores and up to 50 TOPS from the NPU for local and in-game AI workloads.

ASUS lets you configure the smaller ROG Zephyrus G14 with up to an NVIDIA GeForce RTX 5080 Laptop GPU, while the bigger G16 laptop can be tweaked up to the RTX 5090 Laptop GPU. Max GPU TGPs have also been increased by up to 23% through manual mode, from 120W to 130W on the G14, and 130W to 160W on the G16.

ASUS ROG announced its new collaboration at CES 2026 with Hideo Kojima-founded Kojima Productions, for a new lineup of KJP peripherals, including a new custom ROG Flow Z13-KJP gaming tablet.

The company says that at the heart of its new collaboration is "Ludens", the belief that humans aren't just thinkers, but players and creators. The new collaboration combines the spirit of the ROG motto with "For Those Who Dare with KOJIMA PRODUCTIONS" vision From Sapiens to Ludens.

ASUS says that Ludens are the ones who experiment, mod, create, and challenge what games - and gaming hardware - can be. By uniting cutting-edge ROG hardware with KOJIMA PRODUCTIONS visionary storytelling, this partnership is a tribute to that spirit: everything we build, every detail we design, is for gamers everywhere - For Ludens Who Dare.

When it comes to the new 16-pin PCIe power connector for modern GPUs, we've seen no shortage of cables melting and connection points on both the GPU and PSU sides damaged by overheating. It's a definite concern when you get into the realm of flagship cards like the GeForce RTX 5090, which is rated at 575W but can draw up to 1000W over two 16-pin PCIe power connectors, as seen in MSI's new GeForce RTX 5090 32G LIGHTNING Z.

And when it comes to safe power delivery and safety measures for MSI's new beast, and all other GPUs that use the 12V-2x6 connector, MSI's new MPG Ai1600TS, MPG Ai1300TS, MAG A10000PLS, and MAG A1200PLS PCIe 5 power supplies feature an innovative solution.

These PSUs, announced at CES 2026, offer active protection against overheating and cable melting, thanks to GPU Safeguard and Safeguard+ technology, which includes the world's first active, instant protection for power supplies.

ASUS has just unleashed its new ROG G1000 at CES 2026, a next-level desktop gaming PC that has the best PC gaming hardware money can buy, and show-stopping innovation. Check it out:

At the core of the new ASUS ROG G1000 desktop gaming PC is the built-in AniMe Holo fan, which is the world's first holographic fan system inside of a pre-built gaming PC, and it's capable of projecting customizable holographic visuals on your PC.

The AniMe Holo module is housed inside of an independent chamber that features a hinge-door design, making sure that airflow isn't interfering with the rest of the components thermal path, helping reduce overall system noise by limiting vibration from the AniMe Holo system.

AMD has officially launched its new Ryzen AI 400 series CPUs ready to power the next wave of AI desktops and laptops, with the flagship Ryzen AI 9 HX 475 processor leading the charge.

AMD announced its new Ryzen AI 400 series processors will be available starting in Q1 2026 from partners including Acer, ASUS, Dell, HP, Lenovo, GIGABYTE, MSI, and more. Inside, they all feature AMD Zen 5 cores, RDNA 3.5 GPU cores, XDNA 2-based NPUs, ROCm support, and they're Copilot+ PC ready.

The flagship Ryzen AI 9 HX 475 processor rolls out with 12 cores and 24 threads of Zen 5 processing power that boosts at up to 5.2GHz, it has 36MB of cache, 60 TOPS from the NPU for AI workloads, 16 CUs of RDNA 3.5 GPU power, and memory speed support of up to 8533MT/s.

AMD has officially unveiled the new replacement for the best-in-slot gaming CPU, the AMD Ryzen 7 9800X3D, as the company has unveiled a new selection of CPUs at CES 2026.

The CPU AMD is now putting forward as its fastest gaming-focused CPU to date is the AMD Ryzen 7 9850X3D, a refresh of the Zen 5 architecture that comes with very similar specifications when directly compared to the 9800X3D, but with performance metrics that exceed it. According to AMD the Ryzen 7 9850X3D is a 8-core 16 thread CPU that features a 5.6GHz boost clock, 104MB cache (L2 + L3), and a 120 TDP.

The only difference between the 9850X3D and its predecessor is the boost clock total, with the 9850X3D being 0.4GHz faster than the 9800X3D's maximum boost clock of 5.2GHz. The remaining specifications are exactly the same, but given the increase in boost clock we can expect modest gains across both single and multithreaded tasks. Leaked performance testing between the two CPUs indicates an approximate 5% performance gain with the 9850X3D.

AMD has just introduced its new Ryzen 7 9850X3D processor at CES 2026, offering up to 60% more gaming performance than Intel's far more power-hungry Core Ultra 9 285K.

The new Ryzen 7 9850X3D is now the (new) fastest gaming processor on the planet, with AMD also announcing the best gaming and creation processor -- the new Ryzen 9 9950X3D. AMD provided some in-house benchmarking comparing its new Ryzen 7 9850X3D against the Core Ultra 9 285K, where it comes out on top across the board.

The 8 cores and 16 threads of Zen 5 processing power comes with a beefier 5.6GHz CPU boost clock speed, the same 104MB of L2+L3 cache and 120W TDP, and on the current AM5 socket. The biggest difference between the 9800X3D and 9850X3D is the clock speed increase (5.2GHz versus 5.6GHz) giving it an even bigger edge over the 285K.

When it comes to picking up a CPU purely for PC gaming, AMD's Zen 5-powered Ryzen 7 9800X3D with its 3D V-Cache technology is widely considered the best choice if all you're looking for is maximizing your frames. Well, at CES 2026, AMD has officially confirmed that the Ryzen 7 9850X3D is coming, and it will indeed be faster than the Ryzen 7 9800X3D.

The configuration for the new Ryzen 7 9850X3D remains virtually the same, aside from the CPU Boost Clock increasing to 5.6 GHz from 5.2 GHz on the baseline Ryzen 7 9800X3D. The rest of the specs remain unchanged, as the CPU still sports 8 Zen Cores with 16 Threads, 104MB of cache, and a 120W TDP rating.

As part of its CES 2026 presentation, AMD confirmed that the new Ryzen 7 9850X3D is on track for a Q1 2026 release; however, no pricing was given. That said, we should expect it to launch with a higher price than the Ryzen 7 9800X3D's $479 USD MSRP.

AMD has kicked off CES 2026 by pulling the curtain back on a new family of processors designed to power the ambitious exponentially evolving future of artificial intelligence software.

Team Red gave details on the new AMD Ryzen AI Max and AI Max Pro series processors, which will be coming in the form of five SKUs. At the top of the list is the AMD Ryzen AI Max+, which sports 16 cores 32 threads with up to 5.1 GHz max boost clocks, 80MB of cache, 50 peak TOPS, 40 graphics cores, and 60 TFLOPS. The Max+ will be coming in three SKUs, the 395, 392, and 388.

The Ryzen AI Max has two SKUs, the 390 and 385, with the 390 being the beefier out of the two with 12 cores and 24 threads, a boost clock of 5.0GHz, 50 TOPS, 32 graphics cores, and 48 GPU TFLOPS. AMD is also touting the Max+ series of the processors to have 1.4x faster AI performance in tokens per second, 1.8x faster at multitasking, 1.8x faster at content creation, and 1.6x faster at gaming.

At CES 2026, AMD announced a new lineup of client processors, including the Ryzen AI 400 Series, which will power a wide range of laptops and even desktops this year. Building on the success of the Ryzen AI 300 Series, these CPU, GPU, and NPU chips are Copilot+ and AMD ROCm-ready with the integrated AMD XDNA 2 NPU offering up to 60 AI TOPS of performance.

The 60 AI TOPS is only available in the flagship Ryzen AI 9 HX 475 model, which includes 12 Zen 5 Cores with 24 Threads, a Max Boost clock speed of 5.2 GHz, 36MB of Cache, and 16 RDNA 3.5 Compute Units. AMD notes that the integrated Radeon GPU features an impressive GPU Boost Clock speed of 3.1 GHz, which should deliver faster gaming performance than previous generations.

As part of its announcement, AMD provided benchmark results and performance comparisons between the Ryzen AI 9 HX 470 (which is also a 12 Core 24 Thread processor but with 55 TOPs) and the Intel Core Ultra 9 288V running at an efficient 30W.

After a year of rumors and even a small teaser a year ago, MSI has finally brought back its flagship LIGHTNING series of graphics cards at CES 2026 with the new GeForce RTX 5090 32G LIGHTNING Z. It's an absolute beast, with cutting-edge liquid cooling, a default 800W OC mode, and an Extreme 1000W mode for those wanting the absolute fastest gaming GPU on the planet.

The MSI GeForce RTX 5090 32G LIGHTNING Z will be limited to 1300 units when it launches next month, and it will introduce several world-first features. Powered by two 16-pin PCIe connectors, the new GeForce RTX 5090 32G LIGHTNING Z is also the world's first GPU to include a full copper, full cover cold plate covering all components, which is one of the reasons why it's the world's first GPU to reach 1000W.

A liquid-cooled GPU, the main unit includes a vibrant 8-inch display (the GPU will ship with a complete kit for vertical mounting), another world's first, and a rigid carbon fiber backplate. As an all-in-one liquid-cooled GPU, it also sports a next-gen pump that increases liquid flow by an impressive 71% and pressure by up to 45% when compared to MSI's GeForce RTX 5090 SUPRIM LIQUID card.

Intel's new Core Ultra Series 3 "Panther Lake" CPUs were unveiled at CES 2026 today, the first AI PC processors fabbed on its new in-house Intel 18A process node, designed and manufactured on American soil.

The new Panther Lake CPUs are a big moment for Intel in many ways, with one of the major ones being that Intel Foundry fabbed the new Core Ultra Series 3 processors on US soil, a big win for the semiconductor arm of Intel against TSMC. The new Intel 18A process is the most advanced semiconductor process ever developed and manufactured in the United States.

The compute die, which houses the CPU and NPU, is fabbed on Intel 18A, but the IO and GPU aren't -- they're either fabbed on TSMC or possibly on the Intel 3 process node (especially for lower-tier SKUs). This detail was released through leaks in October 2025, but now Panther Lake is (nearly) officially here.