Latest News - Page 639

UL Solutions, the creators of the popular 3DMark benchmark suite for PC gamers, is expanding its professional range of UL Procyon benchmarks with the arrival of the Procyon AI Image Generation Benchmark. Built around the Stable Diffusion AI model, this new benchmark measures the generative AI performance of a modern GPU.

In the benchmark results, part of UL's announcement, we see the GeForce RTX 4060 Ti deliver a score of 1080 - not that we have any reference to know how good or bad that is. Regardless, it's a powerful tool that supports multiple inference engines (Intel OpenVINO, NVIDIA TensorRT, and ONNX runtime with DirectML).

It also measures CPU and GPU temperatures, clock speeds, and usage. However, don't expect this to appear on Steam like 3DMark. Procyon AI Image Generation Benchmark is aimed at the professional space; it costs $5000 USD for an annual site license.

AMD recently re-launched the Radeon RX 7900 GRE, a GPU that outperforms the Radeon RX 7800 XT and NVIDIA's GeForce RTX 4070 while being priced the same as the latter. However, a hard lock (which AMD has said was unintended) on memory speeds at launch limited the GPU's overclocking potential.

This meant that an overclocked Radeon RX 7800 XT could be pushed to reach Radeon RX 7900 GRE performance (while drawing more power) with no option for the RX 7900 GRE to do the same. Still, it's an impressive GPU, with a price that brings some more competition to the stacked mid-range graphics card market that now has four viable choices - the Radeon RX 7800 XT ($499), GeForce RTX 4070 ($549), Radeon RX 7900 GRE ($549), and GeForce RTX 4070 SUPER ($599).

There's some good news for those eyeing a Radeon RX 7900 GRE. The memory overclocking limits have been removed courtesy of the latest Radeon drivers, Adrenalin Edition 24.3.1.

NVIDIA showcased a new AI-powered technology that could revolutionize game development and the PC modding community. Described as a "virtual 3D printer," LATTE3D turns text prompts into 3D objects within seconds. The company notes that the 3D creations are created in a popular format that can be slotted into virtual environments for games or other applications with a few clicks.

"A year ago, it took an hour for AI models to generate 3D visuals of this quality - and the current state of the art is around 10 to 12 seconds," said Sanja Fidler, vice president of AI research at NVIDIA. "We can now produce results an order of magnitude faster, putting near-real-time text-to-3D generation within reach for creators across industries."

NVIDIA adds that by running on a single GPU like the NVIDIA RTX A6000, 3D shapes, animals, and objects are created instantly.

Bethesda has taken to its social media channels to celebrate the 30th anniversary of The Elder Scrolls: Arena, the very first Elder Scrolls game.

The Elder Scrolls: Arena was released in 1994, and now 30 years later the franchise is one of the biggest open-world RPG series' of all time, with The Elder Scrolls: Skyrim being one of the most sold games of all time at 60+ million copies.

In celebration of how far the series has come, Bethesda took to its X account with a message outlining the impact of each game, with Bethesda writing that Arena "started it all, with the entirety of Tamriel generated for you to explore freely", Daggerfall introduced "political intrigue" and "a detailed character system," Morrowind introduced the franchise to consoles and modding, Oblivion brought players to the "heart of the empire," and Skyrim enabled epic battles with dragons.

The first privately constructed lunar lander to ever successfully land on the surface of the Moon has permanently faded, never to wake up again and communicate with its creators back on Earth.

The first US company to achieve a Moon landing was Houston-based Intuitive Machines, which partnered with NASA to construct the Odysseus lander, later nicknamed Odie. On February 22, the lunar lander descended to a location near the lunar south pole, and while Intuitive Machines announced the landing was a success, company CEO Steve Altemus later explained Odie was lying on its side. The phone-booth-sized lander tipped over from one or more of its landing legs, which were caught up on a rock during entry and snapped off.

Despite its undesired orientation, Odie was able to carry out its mission objectives to a reasonable degree, but unfortunately, its active time on the lunar surface has come to its end with Intuitive Machines officially announcing via its X account that the solar-powered spacecraft operated on the lunar surface for seven Earth days, "then went silent after the sun went down at its landing site". The company behind the historic spacecraft said, "This was the expected length of the lander's surface mission."

Microsoft's new DirectSR (Direct Super Resolution) tech for DirectX was created to make it easier for developers to integrate and add support for popular upscaling technologies like NVIDIA's DLSS, AMD's FSR, and Intel's XeSS. Supporting both Windows and Xbox, it doesn't replace existing methods and plugins for adding something like DLSS into a game - instead; it's all about making the game data that all upscaling tech uses available in a single catch-all location.

Once DirectSR is added to a game, it becomes simple to add DLSS or FSR support, with Microsft describing the process as "pick which one you want." Integrating it into DirectX means it becomes part of the pipeline, with the added ability for developers to update games with newer versions of upscale tech via a simple file swap.

Interestingly, as AMD's FSR works with all graphics cards and was created as a general-purpose program for upscaling, Microsoft has integrated the FSR 2.2 code path as the base algorithm for DirectSR. So, the new DirectX update will make it trivial for developers to add FSR support to a game, including the recently announced FSR 3.1 update.

Although Unreal Engine 5 powers several high-profile games currently in development, several studios and publishers still use their own proprietary technology. This includes EA's Frostbite engine, which powers everything from Battlefield to the upcoming Dragon Age Dreadwolf, and Capcom's RE engine, developed for Resident Evil but most recently seen in the popular open-world action RPG Dragon's Dogma.

According to industry insider Dusk Golem on X, Capcom's engineers had to add open-world functionality to the engine to support the game's scope. This is a significant upgrade compared to the version found in the recent Resident Evil remakes.

Interestingly, Capcom will use this new and improved RE engine in two upcoming releases: Monster Hunter Wild and Resident Evil 9. Resident Evil going open-world for its next mainline entry would be a significant departure for the series. Dusk Golem quickly adds that the team aims to ensure Resident Evil 9 "retains the series DNA."

During a recent interview with former CNN anchor Don Lemon, Tesla CEO Elon Musk revealed the long-awaited Tesla Roadster will feature "rocket technology".

The contentious interview with Musk covered various topics that the billionaire is involved in, with one being the highly anticipated release of the Telsa Roadster, which was originally unveiled back in 2017. The Tesla Roadster is the company's most-delayed vehicle, originally being unveiled as a concept back in 2017, slated for release in 2020, and now we are in 2024. Musk has said these delays can be attributed to production issues.

However, Musk recently said in February that the EV sports car was going to arrive next year and would feature a second-generation design that will be "something that's never existed before". Notably, Tesla's CEO said the new design will enable the Roadster to achieve 0-60mph in under 1 second. Furthermore, the new Roadster will be designed jointly between Tesla and SpaceX to create "something that's not really a car" and also features "rocket technology."

Real-time ray-tracing is not only hardware intensive to the point where it needs dedicated hardware to run complex calculations for advanced lighting effects but also heavy on VRAM usage. This has led to situations where 8GB video cards like the GeForce RTX 4060 Ti run into bottlenecks in heavy RT workloads.

According to a recently published patent filed by Microsoft, it is all about using level-of-detail (LOD) to improve ray-tracing performance and quality as needed. Different LOD settings and rendering exist in games, where objects in the distance are rendered with less geometry or detail to boost performance. This also applies to things like texture quality and foliage, which look their best the closer your in-game character is to a particular spot.

Ray-tracing involves a new "residency map for a sub-tree corresponding to a bounding volume hierarchy of objects," where ray-tracing detail is calculated based on an LOD system. Distant ray tracing will be of lower quality, but greater detail will be available up close, freeing up resources like VRAM.

Supercomputers are incredible pieces of technology for a myriad of reasons, and one of those reasons is the unlocked capability to run highly complex computer simulations.

Astronomers and researchers alike have been using supercomputers to run simulations of astronomical objects in an attempt to learn more about how they work, came to be, and evolve. One of those objects is neutron stars, or the leftover corpse of a star that has expelled all of its fuel and exploded in a fiery supernova.

While the material on the outside of the star is blasted into space, the core of the neutron star collapses, forming an extremely dense sphere that is only approximately 12 miles in diameter. However, this small ball has as much mass as 2.3 of our Sun's. Scientists don't have a full understanding of neutron star properties, with a prime query being how matter can be that dense.

In November, it was announced that NASA astronauts lost a bag while conducting a spacewalk, and now one of those astronauts who returned to Earth recounted their first-hand experience.

It is commonplace for astronauts aboard the International Space Station (ISS) to conduct spacewalks outside of the floating laboratory. These spacewalks involve various mission objects, such as repairing instruments and replacing old ones, both of which require tools to do so. During a spacewalk on November 2, 2023, NASA astronauts Jasmin Moghbeli and Loral O'Hara were separated from their tool bag, and funnily enough, reports at the time stated it was possible to see the bag with a pair of binoculars in Earth's orbit.

Now, Moghbeli, European Space Agency (ESA) astronaut Andreas Mogensen and Japan Aerospace Exploration Agency (JAXA) astronaut Satoshi Furukawa have returned back to Earth's surface thanks to SpaceX's Crew Dragon capsule. The Crew-7 astronauts sat down for an interview where they discussed the difficulties of their stay aboard the ISS and the challenges of returning back to Earth, such as the slow adaptation to gravity.

Sony's PlayStation VR2, currently exclusive to the PlayStation 5 console, will soon add official PC support. In an official PlayStation Blog post, Sony announced it was "currently testing the ability for PS VR2 players to access additional games on PC." Progress looks like it's well underway, with iVRy posting on X that the latest PSVR 2 firmware update adds support for NVIDIA GeForce hardware - previously limited to AMD and Radeon GPUs.

iVRy is a platform that enables non-PC VR headsets and devices to be used with Steam. After cracking PSVR 2 hardware authentication, it made an adapter available to those wanting to use the headset on a PC - specifically SteamVR. The PSVR 2 headset is impressive; for $549 USD, you get a 4K OLED display with a per-eye resolution of 2000 x 2040 pixels, up to 120 Hz support, and a wide field-of-view (FOV) of 110 degrees.

Official PC support will undoubtedly increase the device's sales; however, recent reports indicate that Sony has halted PSVR 2 production due to lackluster sales.

NVIDIA is absolutely dominating the AI GPU business with a purported 90% of the AI market taken by Team Green, but now a coalition of tech companies, including Google, Intel, and Qualcomm, are fighting back... they want to take down CUDA a notch or three.

In a new report by Reuters, the site says that Google, Intel, and Qualcomm have "plans to loosen NVIDIA chokehold by going after the chip giant's secret weapon: the software that keeps developed tied to NVIDIA chips," which is CUDA. Reuters continues, adding "they are part of an expanding group of financiers and companies hacking away at NVIDIA's dominance in AI".

Vinesh Sukumar, Qualcomm's head of AI and machine learning, told Reuters: "We're actually showing developers how you migrate out from an NVIDIA platform".

TinyCorp has just announced that it will offer new TinyBox AI systems with either AMD or NVIDIA AI hardware inside. AMD systems will start at $15,000, while NVIDIA hardware inside will start at $25,000.

The company offers its new TinyBox AI systems with 6 x AMD Radeon RX 7900 XTX graphics cards inside, starting at $15,000 per system. Meanwhile, TinyCorp offers a TinyBox AI system with 6 x NVIDIA GeForce RTX 4090 graphics cards, with this monster AI system starting at $25,000.

TinyCorp has been experiencing issues with AMD hardware, posting about its issues and new AI systems on X. The company explained that it was going to sell just AMD-powered systems, but due to various issues, it was forced to offer NVIDIA chips for AI GPU usage.

Intel's next-generation Battlemage "Xe2-HPG" GPUs have been spotted, running with 20-core and 24-core GPU variants inside the SiSoft Sandra database. Check them out:

Intel is expected to roll out its next-gen Battlemage GPU later this year, filling out the Xe2-HPG and Xe2-LPG variants: Xe2-HPG will be discrete graphic cards, while Xe2-LPG will be integrated GPUs inside upcoming processors like Intel's next-gen Lunar Lake CPUs.

The new Battlemage Xe2-HPG GPUs were spotted by "miktdt" with the 20-core Xe2-HPG GPU featuring 160 Vector Engines and 2560 ALUs, while the 24-core Xe2-HPG GPU packed 192 Vector Engines and 3072 ALUs. Both of the Xe2-HPG GPUs featured 8MB of L2 cache, which is half of the cache Intel uses on Alchemist Xe-LPG GPUs with 32 cores. Both of the GPUs also packed 12GB of VRAM on what seems to be a 192-bit memory interface.

The original Apple iPhone released in 2007 has been sold at auction for an incredible $130,027.20 in a factory-sealed retail package.

Apple's original 4GB iPhone is considered a "Holy Grail" amongst iPhone collectors, with its extreme scarcity because of its limited production. Apple released the iPhone in 4GB and 8GB versions on June 29, 2007, but the 4GB sales were slow compared to 8GB, hence why there are not many 4GB iPhones out in the wild. The 4GB iPhone was also discontinued shortly after, with Apple pulling it just two months after its release on September 5, 2007.

LCG Auctions hosted the original 2007 Apple iPhone Factory Sealed (1st Release, 4GB) smartphone at a starting price of $100,000 following an extremely rate 4GB iPhone hitting auction and selling for a record-breaking $190,373. Another 4GB iPhone model also appeared during the LCG Fall Premier Auction in October, selling for an impressive $133,435.

Chinese robotics firm LimX Dynamics has just shown off the latest developments of its new Biped Robot P1, testing the AI-powered robot and its ability to walk through mountainous forest terrain as well as being hit with a huge piece of wood. Check it out:

The new Biped Robot P1 uses reinforcement training to respond to the real-world, where there are moving objects and disruptions in its past -- a forest setting with leaves, wood, and dirt is great -- allowing LimX Dynamics to train its AI algorithms. You can see the AI robot kicking up dirt as it walks, walking into huge bunches of leaves and getting out of it without falling over.

A human operator is also there pulling at it, and it doesn't fall... as well as being hit by a rather large and heavy piece of wood, the Biped Robot P1 doesn't fault. LimX Dynamics took to the Tanglang Mountain range in Shanghai, China to test out its robot, moving around and navigating through rough terrain meant that the P1 had to be continuously aware of its surroundings, pumping that into its AI algorithm, and responding to it through reinforcement training.

Intel will unleash its new LGA 1851 socket later this year, where it will be home for its next-gen Core Series 200 "Arrow Lake" processors... but it looks like we'll be seeing some Meteor Lake-PS processors also slotting into that new LGA 1851 socket.

In a new post on X, hardware leaker "momomo_us" spotted a new Intel Meteor Lake-PS processor tested for the new LGA 1851 socket, meaning Meteor Lake will be first into the new socket, not Arrow Lake. Meteor Lake was originally launched in late 2023 as a laptop-focused chip, inside, featuring an NPU for AI workloads.

An upcoming iBase MI1002 motherboard is described as an "industrial motherboard with Intel Meteor Lake PS" with a render showing off the new LGA 1851 socket. Intel hasn't made its new Meteor Lake-PS processors official just yet, but if they're like their Alder Lake-PS brethren, we should expect Meteor Lake-PS processors to be aimed at the IoT market.

AMD is currently developing its 5th-generation EPYC "Turin" CPUs based on the next-generation Zen 5 and Zen 5c architecture, and now new leaks tease models with up to 160 cores and a huge 500W TDP.

The company will reportedly have at least 20 variants of its 5th Gen EPYC "Turin" CPUs based on Zen 5 and Zen 5c core architectures, with Zen 5 codenamed "Nirvana" and Zen 5c codenamed "Prometheus". AMD's new EPYC "Turin" processors have been designed to feature drop-in compatibility with existing 4th Gen EPYC processors on the SP5 socket (LGA 6096) with support for faster DDR5-6000 memory.

The news on AMD EPYC "Turin" CPUs is coming from leaker "YuuKi_AnS" on X, who posted a bunch of specifications for AMD's family of 5th Gen EPYC "Turin" processors. Starting with the EPYC 9015 with 8 cores and 16 threads with 32MB of L3 cache and a 155W TDP, we fly pretty high up to the flagship EPYC 9845 with 160 cores and 320 threads with 640MB of L3 cache and a huge 500W TDP.

Broadcom has been silently working on what appears to be one of the largest processors ever made, with 12 stacks of HBM memory, making this mystery XPU a bigger beast than NVIDIA's just-announced Blackwell B200 AI GPU.

In some new photos posted to X by our friend Patrick Moorhead, founder of top-ranked, technology analyst and advisory firm Moor Insights and Strategy. He snapped a photo with Frank Ostojic, who runs Broadcom's custom silicon group, and their new third-gen XPU design from a large "consumer AI company". You can see the picture of the XPU above.

Broadcom's mysterious third-gen XPU design features 12 stacks of HBM memory, which makes it bigger than NVIDIA's new Blackwell B200 AI GPU that the company unveiled at GTC 2024 last week.