Latest News - Page 789

A new report from Reuters suggests that OpenAI are considering developing its own in-house chips to power the future of AI-tool queries.

The creators behind the immensely popular ChatGPT are reportedly tossing up between acquiring a company to supply GPUs or making one in-house. The report states that OpenAI CEO Sam Altman has prioritized acquiring more AI chips, which will improve the speed and reliability of the API. Additionally, acquiring a new chip source may dramatically reduce running costs for OpenAI, as an analysis by Stacy Rasgon from Bernstein Research suggests that each query on ChatGPT costs OpenAI 4 cents.

Furthermore, ChatGPT received more than 100 million monthly users in its first two months of launching, which equates to millions of queries per day. If OpenAI's chatbot reaches a tenth of Google's total queries per day it would cost the company $48.1 billion in GPUs and $16 billion a year on chips for future years. At the moment NVIDIA is dominating the AI chip-making market, with OpenAI using approximately 10,000 NVIDIA GPUs to power its AI tools.

Follow-up observations of a luminous fast blue optical transient (LFBOT), a massive bright flash of light, one of the brightest in the universe, that are typically traced back to the arms of a spiral galaxy was photographed by NASA's Hubble Space Telescope.

These massive bursts of bright light are a mystery, with the first being detected in 2018, and then more being detected once per year. What are they? Researchers originally believed they are supernova, the result of a star exhausting all its fuel and dying. However, data indicates these LFBOTs can be 100 times brigher than a supernova, and only hung around for a few days when a typical supernova can be detected for weeks.

The latest of these LFBOTs that was detected is called "the Finch" and was measured by the The Gemini South telescope in Chile, which found its temperature was 20,000 degrees Celicus - not hot enough to be a supernova. Additional measurements revealed the event took place 3 billion light years away within a galaxy that Hubble took aim at.

SpaceX CEO Elon Musk appeared via video link at the International Astronautical Congress (IAC) in Baku, Azerbaijan, where he gave an update on the recent capabilities of Starship, and expectations for the future.

Musk can be seen in the above video discussing SpaceX's roadmap for its Starship launch vehicle, ahead of the company's highly anticipated second orbital attempt that is currently unscheduled. Musk explains that before Starship is caught by the chopstick arms equipped to its Mechazilla, the launch tower located at Starbase, the world's largest and most powerful rocket will attempt a landing at a precise location within the Pacific Ocean.

The reasoning behind this is simple: there are fewer people, no people, out in the middle of the Pacific Ocean, and if something goes wrong with Starship, it would be better if it went wrong over the ocean than over civilization. Musk projects that SpaceX will be able to "catch the booster within the next year, or maybe less than a year," and the upper stage of the rocket, or the ship, "toward the end of next year".

Intel's next-gen flagship Core Ultra 9 185H mobile processor offers 16 cores and 22 threads of Meteor Lake CPU power, with frequencies hitting 5.1GHz... perfect for next-gen gaming laptops. But, it'll arrive later than the Core Ultra 5 and Core Ultra 7 mobile processors.

Intel will be launching its next-gen Meteor Lake CPUs later this year, with the rebranding to take place with this family of CPUs that will be starting out as the new Core Ultra 100 series processors. The start of this rebranding is coinciding with the release of Meteor Lake, with the Core Ultra 5 125H and Core Ultra 7 155H mobile processors rumored for a December 14 launch.

In total, we should see Intel announce 4 new Meteor Lake-H processors, with the Core Ultra 5 and Core Ultra 7 laptop processors launching first -- in December -- while the flagship Core Ultra 9 185H mobile processor boosting up to 5.1GHz will be launching in January. If that happens, it looks like Intel will have something very new to show off at CES 2024 which kicks off in Las Vegas on January 9 and runs through to January 12.

During his interview at the 74th International Astronautical Congress, Elon Musk discussed the possibility of life outside of Earth and whether he has encountered such evidence during his space fairing ventures.

The CEO of SpaceX was awarded the World Space Award and appeared at the event via video link to discuss some of the developmental changes of Starship, the world's largest and most powerful rocket that has ever taken flight. Musk explained that the design behind Starship is purposeful, with the primary objective of providing a reusable transportation method to set up colonies of humans on the surface of the Moon, and ultimately Mars.

Additionally, Musk touched on the hot-button topic of aliens, beginning the conversation with, "People often ask me if I'm seeing any evidence of aliens - and I unfortunately have seen no evidence of aliens yet." Musk jokingly added, "We are the aliens". This take on Musk is nothing unusual as the SpaceX CEO has long pointed out the lack of substantial evidence for extra-terrestrial life.

Astrophysicists are on the brink of seeing the universe right before the Big Bang, and it's all thanks to NASA's immensely powerful James Webb Space Telescope (JWST).

Astrophysicist Richard Ellis from the University College London spoke at New Scientists Live, where he explained the raw power of the Webb telescope and how close astrophysicists are from seeing the very first stars and galaxies forming - a feat that is currently breaking the standard model of the nature of the universe, including its evolution.

The preliminary results from Webb's observations of the early universe reveal there are fewer and brighter galaxies than the Lambda cold dark matter model states possible. Webb is capable of looking further back in time than any other space telescope, and according to Ellis, Webb can see as far back as just 400 million years after the occurrence of the Big Bang.

A team of researchers has published a new study confirming the age of the oldest fossilized human footprints discovered in North America.

The new study was published in the journal Science and details fossilized human footprints in White Sands National Park, New Mexico, which were originally discovered in 2021. These footprints have since undergone thorough dating through various methods, with each of the methods arriving at the same approximate figure of 21,500 years. Leading to the conclusion that humans were present during the Last Glacial Maximum that occurred around 20,000 years ago, during the last phase of the Pleistocene epoch.

So, how are the scientists so sure about the age of the footprints? Scientists performed radiocarbon dating on the seeds of an aquatic plant named Ruppia cirrhosa, which was discovered embedded within the footprints. These results suggested the footprints were between 21,000 and 23,000 years old. Next, the team took 75,000 pollen grains, along with quartz, which is also within the fossilized footprint. Radiocarbon dating on the pollen put its age between 22,600 to 23,400 years ago.

Bethesda's Todd Howard is surprised by how far people have taken the custom shipbuilding in Starfield, an area of the game that allows players to flex their creativity to create something new or something based on pop culture. When the game launched last month, we highlighted some of our favorite creations, including ships from Star Wars, Futurama, and some fun stuff like Thomas the Tank Engine.

Thanks to mods that remove the physical size limitation of ships on PC, we've seen the shipbuilding go one step further - and potentially reach the absolute end-game thanks to this new faithful recreation of a giant Star Destroyer from Star Wars in the game. Created by Reddit user DotElectronic7174, they note that due to the size and scope of the ship, the in-game shipbuilder struggles to rise above 15 FPS.

However, performance is fine when the massive Star Destroyer is rendered in-game orbiting above a planet.

Intel has been in the semiconductor chip manufacturing game for a while. Recently, it's been looking to expand its foundries and production to the point where it could make cutting-edge chips for other companies, a territory we normally associate with TSMC and Samsung.

Currently, Intel is building and ramping up manufacturing in North America, Mexico, and Germany - and these foundries will be doing a lot more than simply creating chips for the latest generation of Core processors. Earlier this year, Intel and Arm announced a partnership to build mobile SoCs on Intel's 18A process node, and we've even heard from the likes of NVIDIA stating that it's open to working with Intel to produce its hardware.

All of this makes the recent statement from Darren Grasby, the executive vice president for strategic partnerships and president of AMD EMEA, a little shocking. His words were harsh when asked if Intel would succeed in its ambitious plans to build global foundries and develop and create chips for multiple companies. To say the least!

Last week, AMD released a special AMD Software: Adrenalin Edition Preview Driver for RDNA 3 owners to check out the company's new AMD Fluid Motion Frames (FMF) technology in 20 optimized DirectX 11 and 12 games. An impressive list that includes Star Wars Jedi: Survivor, Resident Evil 4, Starfield, and The Last of Us Part 1.

This is the driver-level implementation of AMD's answer to NVIDIA DLSS 3 Frame Generation, allowing Radeon owners to enable Frame Gen in games that don't feature native support for it. Over the weekend, AMD updated the Adrenalin Edition Preview Driver to add support for Radeon RX 6000 Series GPUs - which is great to see, as there are a lot of RDNA 2 cards out in the wild.

Like with the original preview release, this is described as an "early first look" by AMD, with some caveats. Fluid Motion Frames are automatically disabled during fast motion and, as such, "may not offer the optimal experience in fast-paced competitive titles." However, it's still exciting to see Frame Generation at the driver level like this.

Assassin's Creed: Mirage is now out in the wild, and the folks over at Digital Dreams have cranked up Ubisoft's latest title at the glorious 8K resolution, running at 60FPS.

Digital Dreams ran Assassin's Creed: Mirage at 8K (so 7680 x 4320) at 60FPS+ on NVIDIA's flagship GeForce RTX 4090 graphics card, with the ReShade preset enabled. The rest of the hardware is beefy, too: AMD's Ryzen 9 7950X processor is installed into an AORUS X670 AX motherboard with Kingston DDR5-6000 memory to boot.

Chromatic Abberation was disabled for the test, with "Beyond All Limits" ray tracing added... and still, it doesn't make Assassin's Creed: Mirage look out-of-this-world, but you can't knock that it doesn't look fantastic cranking along at 8K 60FPS+ on that level of hardware.

Sony is hard at work on its next-generation PlayStation 6 console, which should be released sometime in 2027 or so... but if you can believe it, rewinding back the clock to June 2018... I remember driving back home from Melbourne to Adelaide and had a call from an industry insider who exclusively revealed to me that Sony was working directly with AMD on the Navi GPU architecture for its next-gen PlayStation 5 console. We have been all over the console APU releases since, and now we're looking at the PlayStation 6.

The folks over at Yanko Design noticed a new job listing at Sony, that saw a position open for someone who would help the company "contributing to identifying and developing the technology portfolio of future PlayStation platforms". Once that happened, PlayStation 6 concept renders were created... and man, they look absolutely wild.

I don't think Sony would ever go for this radical design for the next-gen PlayStation 6 console, but I do love it... the way that the controller slots into the side of the PS6 would be such a fantastic idea. It is functional, and it looks awesome at the same time.

It seems like it's the week for information on Intel's next-gen Arrow Lake platform, which will arrive after the 14th Gen Core "Raptor Lake Refresh" family of processors that are only a few days away now. Today, we're hearing about alleged power limits on Arrow Lake, so let's dive right in, shall we?

Intel's next-gen Arrow Lake-S processors with 8 P-Cores are based on the Lion Cove CPU architecture, while the 16 E-Cores are based on the new Crestmont CPU architecture: a grand total of 24 cores and 32 threads. This is no change to the current-gen Core i9-13900K and upcoming Core i9-14900K, by the way.

Onto the power limits: 125W baseline power seems to be the standard for Intel's upcoming unlocked CPUs, while the unlocked "K" series processors have a PL1 rating of 125W, while the specific SKU in question on these charts has a PL2 rating of 177W.

One of the best gaming CPUs money can buy just got cheaper, with Newegg discounting the AMD Ryzen 7 7800X3D from its $449 MSRP down to $399, and with the promo code available, down to just $349.

AMD's kick-ass Ryzen 7 7800X3D processor features 8 cores and 16 threads based on the Zen 4 architecture, with 96MB of L3 cache and its snazzy second-generation 3D V-Cache technology from AMD. There's also the AMD Ryzen 7 7700X processor for $25 less, but I don't know why you'd buy that when the Ryzen 7 7800XD is available for just $25 more. It's a no-brainer for AM5 motherboard owners.

If you decide on the AMD Ryzen 7 7800X3D, there are multiple benefits here: it's cheap, it goes into any AM5 motherboard, and just works. There's no need to overclock it, as you're going to be loving the performance squeezed out of it at stock... and you won't need some huge, expensive AIO liquid cooler. The AMD Ryzen 7 7800X3D processor will run on an air-cooler, which makes it an even better proposition for the best value for money gaming CPU money can buy... and now, it's available for $349.

ASUS is gearing up for the release of one of the most interesting -- and useful -- graphics cards of the last few years, with its GeForce RTX 4060 Ti with an SSD slot turning up at the EEC (the Eurasian Economic Commission).

In the listing at the ECC, it looks like ASUS is preparing 2 different cards: a faster model with factory overclocking, and then one that isn't overclocked out of the box. It looks like ASUS will be launching it under their DUAL family of graphics cards, so we should expect the ASUS GeForce RTX 4060 Ti DUAL (overclocked) and then the ASUS GeForce RTX 4060 Ti DUAL (not overclocked) in the near future, both rocking the SSD slot on-board.

Earlier this year, ASUS China teased the new GeForce RTX 4060 Ti graphics card powered by a Samsung 980 PRO 2TB SSD which was mounted to the rear of the card.

ASUS has just outed its new Hyper M.2 x16 Gen5 expansion card, which is capable of taking on 4 x M.2 SSDs that will use the total bandwidth provided by the super-fast Gen5 x16 slot on your compatible motherboard. It boasts a boatload of speed, and it's cheap as chips.

The new ASUS Hyper M.2 x16 Gen5 expansion card can handle up to 512Gbps (64GB/sec) of bandwidth when the 4 x M.2 SSDs are installed into the expansion card, and you've got RAID0 enabled. You can't get this speed with any old drives, as you'll need next-gen M.2 SSDs that are capable of pumping out 16GB/sec through each drive... see, we told you this wasn't going to be easy.

However, ASUS is making its new Hyper M.2 x16 Gen5 expansion card compatible with multiple M.2 form factors including 2242, 2260, 2280, and 22110. It's also backward compatible, meaning you can chuck PCIe 4.0 M.2 SSDs into it, or you can wait for those next-gen PCIe 5.0 M.2 SSDs that crank up to 16GB/sec each so you can meet that juicy maximum bandwidth of 64GB/sec. Not that you'd need that in a gaming system, that is.

Intel's new 14th Gen Core "Raptor Lake Refresh" family of processors is only a few days away now, but now a Spanish retailer has listed both the Core i9-14900K and Core i5-14600K with some EU pricing for us to gander at.

Spanish CoolMod has listed 4 out of 6 new Raptor Lake Refresh processors on their website, and while the pricing isn't final -- we really should pay attention to this pricing as it's much closer to reality. Starting with the new flagship Core i9-14900K processor, which is listed at €799.96, which is €160 more expensive than the current-gen Core i9-13900K processor.

Moving onto the Core i5-14600K which is priced at €429.95, which is €88 more expensive than the Core i5-13600K processor. If you're looking at no integrated graphics, than the KF models are €190 and €70 more expensive for the Core i9-14900KF and Core i5-14600KF processors, respectively.

NVIDIA is reportedly planning a new high-end GeForce RTX 40 series card release, with rumors suggesting that the company is tinkering with a GeForce RTX 4080 Ti -- or GeForce RTX 4080 SUPER -- graphics card that will fall into a "similar price range" to the GeForce RTX 4080 that's on the market today.

We're hearing the news of the GeForce RTX 4080 Ti from MEGAsizeGPU, with the new GPU set to arrive sometime in early 2024... which smells like CES 2024 in January. We could see the GeForce RTX 4080 Ti or GeForce RTX 4080 SUPER, although a "SUPER" branded release would feel weird on its own against non-Ti and Ti class GPUs that NVIDIA already has filling out its GeForce RTX 40 series family of graphics cards.

Now, what's going to be inside? Rumor has it that NVIDIA's upcoming GeForce RTX 4080 Ti will be based on the AD102 GPU GPU, with a bit more beef to it than the GeForce RTX 4080. We're hearing that it'll have a 380W to 420W TGP, as well as 20GB of GDDR6 memory on a 320-bit memory bus, which is a bigger upgrade over the 16GB of GDDR6X memory on a 256-bit memory bus on the regular GeForce RTX 4080 graphics card.

If you've been following news around Microsoft this week, you may well have seen an alarming suggestion that the next version of Windows will be subscription-based - the good news is that this isn't true.

Let's rewind a bit to where this rumor came from. It originated from German tech site Deskmodder, which discovered some entries in a configuration file in a Windows 11 preview build (from the Canary channel).

Those entries mentioned things like subscription status and subscription type, leading to some speculation that this might be the groundwork for some kind of subscription model coming to Windows.

PAX Australia returned to Melbourne this weekend, a celebration of gaming, tabletop, esports, and pop culture that saw thousands brave the unseasonably chilly weather (to be fair, Melbourne's weather is always all over the place) to take in the sights, play the latest games (or some classics), or sit in on one of the many panels. I was keen to get my hands on a pre-release version of Super Mario Bros. Wonder at the massive Nintendo booth and check out a stack of locally grown indie games. However, PC Gaming was the show's star - and many of the top brands were on hand with local builders and partners to showcase the latest hardware.

Plus, in the case of Yamaha, showcase one of their guitars turned into a PC. Seeing this one from a distance, I had no idea it was secretly a gaming PC - mainly because it was wedged between an actual drum kit and electric pianos (this is Yamaha, after all). It was only when I saw the liquid cooling and fans on the back that I realized it was indeed a gaming PC with I/O ports on the side and hardware hidden in the guitar's casing.

Like many at PAX this year, Yamaha partnered with the PC design gurus over at Aftershock to create this very cool fusion of music and technology. With a ZOTAC GeForce RTX 2080 SUPER, it might not be the most powerful rig on our list, but it's certainly a show-stopper. And yes, a playable instrument, too.