AMD RDNA3 GPU Revealed With Two Eight-Pin Connectors

This week AMD will finally unveil its chiplet-based RDNA3 GPUs. The big question is whether AMD will follow Nvidia’s lead with power-guzzling, fire-breathing GPUs, and at what price? The gaming community has been split on the RTX 4090. Some appreciate its record-breaking performance, while others have lambasted its size, power requirements, and its ability to melt power cables. However, it seems AMD might be taking a different approach with RDNA3. One of its new GPUs has been photographed, and though it’s bigger than its predecessors, it sports only two eight-pin power connectors. That seems kind of quaint in the current environment. It also flies in the face of rumors that AMD would be unleashing a monster GPU in the same vein that Nvidia has with the 4090.

Photos of an AMD engineering sample GPU have appeared on Overclock3D.net. The card is shown next to its RDNA2 sibling, and we can draw two firm conclusions from it. First, as far as power draw goes, it’s safe to assume this is a 375W GPU. It can pull 150W from each power cable, and 75W from the PCIe slot. That’s considerably less than the RTX 4090’s 450W total board power (TBP). That will likely come as good news for a lot of gamers while being a cause for concern to others. After all, AMD has stated previously that board power consumption would be going up for RDNA3. That lead a lot of people, us included, to assume it meant it was really going up. However, if it’s 375W that’s only 45W more than the current RX 6950 XT. The original 6900 XT was a 300W card though, so 75W is a decent amount.

The prototype board is certainly larger than its sibling but not humungous. (Image: Overclock3d.net)

The other conclusion we can draw is the card is actually a reasonable size. It looks like a regular dual-slot design, which actually seems small now. It’s almost, dare we say, kind of cute? With the RTX 4090, some people had to buy a new case just to make it fit. That’s because some of the partner boards are outlandishly massive, with 3.5 slot designs and unprecedented width. It’s actually exacerbated the adapter bending issue as the connector is where the side panel is on some mid-towers.

Another explanation for its size is it could be the second-tier RX 7900 XT. According to rumors, AMD is launching two GPUs this week. There will be a 7900 XT and a 7900 XTX flagship card. It’s possible the XTX version might have a higher power draw. That would be similar to some of the RX 6950 XT partner boards, which featured three connectors instead of the two found on the AMD version. It’s also worth remembering that the noted overclocking site Igor’s Lab posted a mockup of the card’s PCB previously. They said it would have three eight-pin connectors based on what their sources told them. That could very well happen with the full-sized XTX board.

Also, the pictured board is clearly an engineering sample. Its naked PCB is exposed, as it’s lacking a backplate. This suggests we shouldn’t rush to judgment just yet on its features. However, it has the same red “racing stripes” shown in a teaser during the Zen 4 launch (top). That seems to indicate this is the real deal.

(Image: AMD)

There could also be a huge plot twist here: AMD has found a way to compete with the RTX 4090 with 75W less power. If that’s the case, the company has made some truly revolutionary efficiency gains. It’s already stated its goal is a 50 percent improvement in performance-per-watt for RDNA3. It calls that “industry-leading efficiency,” and it might be right about that. Nvidia isn’t exactly a slouch, though. After all, it found a way to keep TBP for the RTX 4090 the same as the 3090 Ti despite it having more than triple the number of transistors.

Still, AMD had a PR campaign for its current-gen cards stating they were more efficient than Nvidia’s. Generally speaking, it’s right about that, so it’s possible AMD could have the upper hand in terms of efficiency with RDNA3. The big question is whether that will matter to gamers.

Now Read:

• AMD Engineer Confirms RDNA3 AMD Noise Suppression For Radeon GPUs Accidentally Leaks
• MSI Yanks Video Showing Off AMD’s New AM5 Socket (Updated)
• AMD’s RDNA3 7900XT Flagship to Offer 384-bit Memory Bus, 24GB of VRAM

Battery-Powered Shoe Attachment Boosts Walking Speed by 250 Percent

(Photo: Shift Robotics) What if you could walk really fast? Okay, not Sonic the Hedgehog fast…and not Quicksilver fast, either. Think more along the lines of nearly three times faster than you already cruise through the park or through a dying shopping mall.

Moonwalkers claim to help you accomplish that. Designed by a team of Carnegie Mellon University engineers who banded together to found Shift Robotics, these battery-powered attachments strap onto almost any pair of shoes to give you an enviable speed boost. Instead of free-wheeling like roller skates, Moonwalkers’ eight polyurethane wheels work with a set of built-in sensors to switch between “lock” and “shift” modes, which prevent the wheels from spinning when the wearer is navigating stairs, using public transit, or otherwise requiring full motion control. These modes also help the wearer stop within one meter even at top speed, which is said to be 250 percent faster than the wearer’s normal walking speed.

The attachments’ chassis are made entirely from aluminum to prevent crushing and assist in thermal management. The 300-watt electric motor, which powers the wheels for up to six miles of active use per charge, is fully sealed to protect against water and debris ingress. According to Shift, this is what allows Moonwalkers to navigate puddles and sidewalks that are in less-than-perfect condition. Because Moonwalkers are designed to match the wearer’s gait, there’s said to be zero learning curve, which can’t be said for conventional equipment like roller skates and rollerblades.

It’s hard to avoid wondering if Moonwalkers are a solution to a problem that doesn’t exist. After all, a full range of motion is still required to use them; they don’t serve as a mobility aid for those who can’t already walk, and since they use wheels, Moonwalkers are rendered pointless on most unpaved surfaces. (You certainly can’t bring them on a hike or to the beach, where walking is arguably more exhausting.) Shift Robotics seems to be positioning its attachments as a way to make city life a bit more efficient: “With Moonwalkers, you can pick up your dry cleaning across town, carry those grocery bags a little easier, grab those last-minute dinner items much quicker, or whatever else with much more ease,” its Kickstarter page reads.

But at $799 to $1,299 per pair (depending on the Kickstarter campaign’s progress), the cost of that added efficiency is pretty steep. This means Moonwalkers’ target audience is quite small: Frugal budgeters, rural dwellers, and those who like to stop and smell the roses need not back this project.

Now Read:

• New Transparent Solar Cells Could Help Scientists Create Energy-Generating Windows
• Harvard Scientists Invent Gripping Robot With Soft, Inflatable Tentacles
• Scientists Develop Mouthguard Input Device That Responds to Bite Force

AMD Rumored to Launch Two High-End RDNA3 GPUs This Week

It’s been a time of glorious bounty in the PC hardware universe unlike any other. We’ve seen powerful new CPU architectures from both Intel and AMD launched within weeks of each other. We also witnessed the unabashedly potent RTX 40-series launch. It’s been an exciting time to be a PC gamer.

Now there’s just one more highly anticipated arrival that will effectively wrap up the party: AMD’s next-gen GPUs. These will truly be something special: the first chiplet-based GPUs for consumers. The 7900X was an obvious product, but the existence of a second, more-powerful GPU is the big news here. It seems AMD is following Nvidia’s strategy of launching two high-end cards first. As you recall, Nvidia “unlaunched” the third 40-series GPU previously.

The latest rumors indicate AMD will be launching the RX 7900 XTX and the RX 7900 XT this week. Yes, that is a lot of Xs. It’s also surprisingly bringing back the XTX moniker, which it hasn’t used since 2006 for the X1900 XTX. Both GPUs will seemingly take on the RTX 4090 at the top of the stack for RDNA3. Both GPUs will be Navi31, which is the big die of the family. The XTX version will be the full version of the chip, with the XT version being slightly cut-down. The flagship should come with 24GB of 20Gb/s memory on a 384-bit memory bus, which aligns with previous rumors. This should give it almost a terabyte of memory bandwidth. That number doesn’t include the benefits of its infinity cache.

This fan-made render of the flagship Navi31 GPU looks tantalizing. (Image: @Technetium)

It’s expected to offer up to 96MB of cache this time, according to TechSpot. That’s a small reduction from the 128MB it offered on the 6900 XT. However, it could be using its V-Cache technology to vertically stack some of it too. It was previously rumored it might offer as much as 384MB, which would be truly nuts. The XTX card is reported to boast 12,288 streaming processors. That’s more than double the 5,120 found in the previous flagship; the Radeon RX 6900 XT. Its power consumption is still a big unknown here. Previously AMD stated it’ll be going up this generation. Since the 6900 XT was a 300W card, we can comfortably predict it’ll be 375W or so.

The cut-down XT version will use 20GB of GDDR6 memory across a narrower 320-bit bus. It’ll allegedly offer 800GB of memory bandwidth, which is more than the RTX 4080 16GB, not accounting for infinity cache benefits. It’s possible this card could land in between Nvidia’s top GPUs though. It’ll offer 10,752 stream processors, more than double that of the 6800 XT’s 4,680.

Notably, neither GPU will be using the newfangled 12VHPWR connector. Instead, it’s using the tried-and-true eight- and six-pin connectors we all have now. This will come as good news for those who have been following GPU news. Nvidia is currently embroiled in a controversy concerning its 4-into-1 PCIe power adapters melting. It’s supposedly caused by a suboptimal soldering job on the wires inside the plug that goes into the card. The hullabaloo caused an AMD executive to confirm on Twitter it wasn’t using the 12-pin connector.

Sadly, we must report that although AMD is announcing these GPUs on Nov. 3, it might be another month before they reach retail. Noted tipster Greymon55 recently posted that they won’t go on sale until December. They were originally supposed to be offered two weeks after the reveal, in late November. Pricing and board power are still TBD, but we shall find out soon enough.

Now Read:

• AMD Engineer Confirms RDNA3 AMD Noise Suppression For Radeon GPUs Accidentally Leaks
• MSI Yanks Video Showing Off AMD’s New AM5 Socket (Updated)
• AMD’s RDNA3 7900XT Flagship to Offer 384-bit Memory Bus, 24GB of VRAM

UFO Sightings Are Usually Foreign Spying or Space Junk, Government Sources Say

(Photo: Stefan-XP/Wikimedia Commons) Many UFO sightings recorded over the last several years can likely be attributed to far more ordinary circumstances than originally thought. Instead of extraterrestrials, most strange aerial activity can be chalked up to foreign countries’ surveillance operations and space debris, say government officials.

The last year has seen the US government actually begin to pay attention to possible alien activity. In 2021 the Office of the Director of National Intelligence conducted an unprecedented preliminary assessment of Unidentified Aerial Phenomena (UAPs), or the government’s term for UFOs. The assessment aimed to determine how federal agencies could begin investigating, classifying, and addressing 144 UAP sightings from the previous 17 years. Most of these sightings weren’t explainable at the time, but this week, intelligence agencies will provide Congress with an update…and it’s a little less flashy than anticipated.

According to a handful of government officials who spoke in confidence with The New York Times last week, many UAP sightings from the last decade or so are bound to be the result of foreign surveillance or space junk. Some incidents have officially been attributed to Chinese surveillance, which the country conducted with “relatively ordinary drone technology.” Others are thought to be associated with China but haven’t formally been classified as such. Based on what people close to the matter told the New York Times, China has a vested interest in learning how the US trains its military pilots. The best place to figure that out, Chinese intelligence agencies seem to believe, is high in the sky.

NASA’s BARREL balloon admittedly looks a bit odd from the ground. (Photo: NASA/Goddard/BARREL/Brett Anderson/Wikimedia Commons)

Other sightings have been attributed to space junk, or retired satellites and other equipment that remain in orbit. Even less exciting is the fact that some “flying saucers” have reportedly been weather balloons, which sometimes look a bit otherworldly from down on the ground.

But if so many UAPs are actually the result of familiar, Earth-based activity, why have they been considered UAPs for so long? “In many cases, observed phenomena are classified as ‘unidentified’ simply because sensors were not able to collect enough information to make a positive attribution,” Defense Department spokesperson Sue Gough told The New York Times. Impractical sensor optics and illusions caused by water have also made certain aerial phenomena look a lot more unique (and worrisome) than they actually are. According to Gough, the Pentagon—and likely NASA, thanks to its new independent UAP study—is working to build upon its analysis methods in the future so that normal activity doesn’t get stuck in UAP status for so long.

Now Read:

• An Astronomer Is Privately Funding a Search for Alien Technology on the Ocean Floor
• SETI Researchers Have Begun Scanning the Backgrounds of Images for Alien Signals
• Possible Alien World Discovered Thanks to Orbiting Debris

Perseverance Rover Prepares to Drop Off Mars Samples for Return Mission

The Perseverance rover is a well-equipped robot with a gaggle of cameras, multiple spectrometers, and even a little box that makes oxygen. You can’t get every possible scientific instrument on a Mars-bound rover, though. To really understand the red planet, we need to get samples back to Earth, and Perseverance is preparing to take the next step in making that happen. NASA and the ESA have agreed on a location for Perseverance to deposit the first sample cache, which could be retrieved a few years down the line in the NASA-ESA Mars Sample Return Campaign.

During its time on Mars, Perseverance will analyze numerous samples with the tools at its disposal, but the team is also carefully curating a collection of samples that will come back to Earth. The rover was designed with an innovative sample caching system, which packages up rock cores in pristine metal tubes that will protect them from contamination on the return journey. So far, Perseverance has collected 14 rock-core samples in the tubes. The robot has several dozen sample tubes at its disposal.

NASA and the ESA have agreed that the first batch of samples will be deposited at a site known as Three Forks, near the base of the ancient river delta in Jezero Crater. The mission to retrieve the samples is still evolving, so the team has to make some guesses about how the Return Campaign will work.

A few months ago, the agencies updated the plan to drop a second rover that was supposed to fly with the return vehicle. Now, Perseverance will be the primary means of getting samples to the Mars Ascent Vehicle (MAV). It will also have a pair of helicopters based on Ingenuity’s wildly successful design. The tube cache at Three Forks will act as a backup in the event Perseverance cannot rendezvous with the MAV or an issue pops up in the sample caching system.

While Perseverance drops off its first collection of samples, engineers back on Earth have begun the process of testing hardware for the return campaign. In what is known as “Phase B,” the team is working to develop prototypes that will eventually become the final flight hardware, which will hopefully have no defects or software glitches. There’s enough that can go wrong without hardware failures. After landing in Jezero Crater, the MAV will deposit the recovered tubes into a rocket that blasts them into orbit. At that point, an ESA spacecraft will have to pick them up and make its way back to Earth. If all goes as planned, the samples could be back on Earth as soon as 2033.

Now read:

• What Ingenuity and Perseverance Have Discovered on Mars So Far
• Perseverance Mars Rover Finds Exciting Organics in Ancient River Delta
• Perseverance Has Pebbles Stuck in its Sample Caching System

Musk Completes Twitter Takeover, Promises It Won’t Become a ‘Hellscape’

It finally happened — after months of snide public statements and legal wrangling, Elon Musk has acquired Twitter for the original $44 billion offer. Musk wasted no time firing several executives, including CEO Parag Agrawal, almost the instant the deal was done. Musk has hinted at some major changes to the way Twitter works, including the end of lifetime bans, but he promises it won’t become a “free-for-all hellscape.” It may not be possible to thread that needle, though.

According to Musk, who has a penchant for grandiose overstatements, he purchased Twitter because he wants to help humanity. Also among those immediately fired by Musk was Twitter’s head of policy, trust, and safety, Vijaya Gadde. Musk has harshly criticized Twitter’s handling of numerous disputes, which sometimes involved temporary and lifetime bans on controversial figures. Reports now indicate that Musk plans to end the practice of permanently banning users.

Conservative users, who largely believe Twitter discriminates against them, are cheering the news. Musk has spent the past few years adopting right-wing talking points around COVID and criticizing Democrats, so they do have reason to be hopeful. Should Musk follow through on his apparent intention to reverse lifetime bans, that could mean the return of former president Donald Trump, who was permanently banned from Twitter following the January 6th Capitol riot. Trump has since launched his own social network known as Truth Social, but it’s apparently floundering. Returning to Twitter is probably what he’s wanted all along, anyway.

Dear Twitter Advertisers pic.twitter.com/GMwHmInPAS

— Elon Musk (@elonmusk) October 27, 2022

Musk says he plans to make some changes to the company’s suspension policy quickly, so accounts banned for harassment and misinformation could reappear soon. In addition to Trump, conservative political actors on Twitter are begging Musk to reinstate the accounts for Infowars and its founder Alex Jones, as well as Project Veritas, Milo Yiannopoulos, and election conspiracy theorist Mike Lindell.

We might soon get a preview of how, if at all, Musk will attempt to reign in misinformation on his new social network. As the deal closed, news broke of a break-in at the home shared by US Speaker of the House Nancy Pelosi and her husband Paul Pelosi. While the Speaker was in Washington at the time, reports say Paul was badly injured and is currently recovering in the hospital. Conservatives on Twitter have started alleging this attack was a false flag aimed at influencing the upcoming election — they have compared Pelosi to actor Jussie Smollette, who was convicted of concocting a fake assault story several years ago. It’s trending right alongside news of the break-in. Should this become a new election conspiracy theory, the old Twitter would have taken action, but will Musk?

Right now, it sounds like Musk wants to have it both ways. He’s committed to running Twitter as a “digital town square” that allows diverse opinions, but he also says it must be “warm and welcoming to all.” Regardless of what Musk says during this pivotal time for the company, it seems inevitable that Twitter will spend more time targeting bots than misinformation. Some people may prefer that, but it could make the site, if not a hellscape, at least a lot more combative. Musk got himself into this mess, and he’ll be blamed for whatever happens.

Now read:

• SpaceX Fires Employees Behind Musk Open Letter
• Elon Musk Says He Can Legally Cancel Twitter Deal
• Elon Musk Offers to Buy Twitter

These Haptic Gloves for the Metaverse Require ‘Airpack,’ Cost $495 Per Month

(Photo: HaptX Inc.) Virtual reality (VR) experiences can be impressively realistic. In fact, some are so convincing that users forget they’re seeing something fictional or end up using VR to treat certain phobias. But what if you could take this realism a step further by physically feeling the items you encounter in-game? That’s the question behind HaptX Inc.’s new haptic gloves, which use tactile responses to make VR experiences feel more immersive.

HaptX Gloves G1 incorporate two kinaesthetic feedback gloves and a wireless “Airpack,” the latter of which can either be worn by the user like a backpack or set on a table nearby. The Airpack is responsible for generating compressed air and controlling the air’s flow, both of which are essential to conducting detailed haptic feedback. The gloves themselves, which come in four sizes, contain hundreds of microfluidic actuators that displace the skin with each pulse. This gives the user the sensation that the objects they’re touching and interacting with in the virtual space are real.

The product is meant for the “enterprise metaverse,” or VR experiences created by or dedicated for corporate entities. As HaptX suggests in the video above, some larger organizations have recently expressed interest in conducting training within the metaverse, while others believe it could transform the way people shop. If that interest endures the metaverse’s shaky future and turns into true commitment, companies might be able to forgo physical training materials and interactive product displays in favor of virtual ones.

Whether that’d be financially viable depends on the company. HaptX Gloves G1 cost $5,495 per pair with a $495 monthly fee. The fee is part of the unavoidable HaptX Subscription program, which includes service, maintenance, and the company’s software development kit (SDK). The HaptX SDK is said to enable developers to incorporate Unreal Engine and Unity plugins, utilize C++ API, and take advantage of features already built into the product, like advanced vibrotactile feedback and a haptic multiplayer feature that allows multiple users to “feel” the same objects. (Sound familiar?)

HaptX argues its Gloves G1 are far more accessible than their predecessor, the HaptX Gloves DK2, which cost over $10,000 per unit, which is pricey even for the enterprise VR scene. Still, despite the HaptX Gloves G1’s cost and physical bulk, it’s impressive that users can feel different textures, manipulate virtual materials, and otherwise physically experience simulations. VR cat cafe, anyone?

Now Read:

• Meta Has to Force Its Metaverse Developers to Use Horizon Worlds
• Meta to Clash With Apple Over ‘Open or Closed’ Metaverse
• Interpol Launches ‘Police Metaverse’ to Train Officers on Virtual Crime Fighting