TSMC to Produce 4nm Chips at Arizona Fab in 2024

Credit: Laura Ockel/ Unsplash, PCMag

(Credit: Laura Ockel/Unsplash) It was previously reported that Apple has been urging TSMC to begin producing some of its more advanced chips in America. This would allow Cupertino to claim it’s using American-made silicon in its devices, a first for Apple. A new report confirms this will indeed be happening, but TSMC won’t be making 3nm chips at first, contrary to earlier reports. Instead, it will be focusing on its N4 process, which should be online in 2024. By then, that will be far from bleeding-edge technology, but the company will eventually produce 3nm chips in Arizona as well.

News of TSMC’s plans comes from notable Apple reporter Mark Gurman and company of Bloomberg. Previously TSMC was expected to start production of 5nm chips in Arizona, but it has apparently jettisoned that notion in favor of N4 production, which is a more advanced version of its process. It still falls under the 5nm family of products but offers improved efficiency and performance. The company is investing $12 billion into the fab in order to get production started by 2024. It will reportedly announce its plans next week during a presser with President Joe Biden and Commerce Secretary Gina Raimondo. Apple CEO Tim Cook will be at the event as well.

An artist’s rendering of what fab in North Phoenix will look like when it’s complete in 2024. (Image: TSMC)

TSMC is rumored to be only taking baby steps with its new fab, however. It’ll supposedly be cranking out just 20,000 wafers a month, and Apple is expected to claim about a third of that output, which is table scraps for both companies. TSMC currently produces 1.3 million wafers a month from its Taiwan facilities, so the latest news reinforces the “symbolic gesture” label a Bloomberg opinion writer had called the move. Still, it will give companies access to an advanced node producing American-made chips. TSMC will also be adding a second plant nearby that will handle some 3nm production in the future. However, no timeline was given for when that facility would begin manufacturing.

The move by TSMC marks the first time the company will produce some of its more advanced products outside of Taiwan. This will likely give the company some resiliency against the worst-case scenario of a Chinese invasion of Taiwan. The US government also passed the CHIPS Act earlier this year to bolster efforts by companies producing semiconductors in the United States. The act provides billions in subsidies for both fab expansion and R&D. However, TSMC seems to be intent on keeping its most advanced node under its own roof in Taiwan. It’s just now beginning production at 3nm, and by 2024 that will likely be the main product it’s offering to partners such as Apple, AMD, and Nvidia.

Now read:

• Commerce Secretary Issues Stark Warning About Losing Access to Taiwan Fabs
• Intel Urged to ‘Explore Alternatives’ to Manufacturing Its Own Chips
• TSMC Suspends Advanced GPU Production for Chinese Startup

Valve’s Steam Deck Giveaway Is Already Experiencing Issues

(Image: Valve) Valve has some pretty grandiose ideas for spicing up this year’s Game Awards. On Wednesday the company tweeted that it would be giving away a Steam Deck every minute of the event. Considering the Game Awards is typically a two-to-three-hour affair, that’s anywhere from 120 to 180 Steam Decks…except we’re a week out from the event, and Valve is already experiencing some hiccups.

The giveaway is supposed to be simple: If you live in the United States, Canada, the United Kingdom, or European Union, you should be able to participate. All it takes to sign up is logging into Steam and navigating to the “Deck Drop” page. Click a big blue button to register, and that’s it—until the actual Game Awards, when you’ll need to tune in via Steam.tv.

In reality, things don’t appear to be that easy. Mere minutes after Valve’s announcement, people began reporting issues with the giveaway’s registration process. Multiple Twitter users responded to Valve’s tweet with screenshots alleging they didn’t reside within a participation-eligible region despite living in the United States. One of the giveaways’ registration requirements is having made a Steam purchase between Nov. 14, 2021, and Nove. 14, 2022, to validate one’s account and country of residence, which some hopeful participants may have missed, although it’s unlikely everyone who’s complaining online has refrained from buying a game through that window.

Those who haven’t been able to sign up for the giveaway can still get the exclusive Steam Pal digital sticker made for the event, according to multiple screenshots. It’s obvious, however, that the (admittedly adorable) dancing Steam Pal isn’t what folks are storming Steam for. Valve isn’t just giving away a Steam Deck per minute—it’s giving away a 512GB NVMe SSD Steam Deck, the most powerful (and expensive) of Valve’s handheld consoles. Although the entry-level and mid-tier Steam Decks became available for purchase without reservations in October, the 512GB version was reservation-only up until very recently, which only added to its perceived exclusivity. It’s only fair that, given the chance to get one of those premium editions for free, people are upset that a simple registration page isn’t working.

Valve responded to Twitter users’ distress Wednesday evening by saying it had fixed a few issues with the registration process. Still, people are reporting that they haven’t been able to sign up for the giveaway despite living in Arizona, Nebraska, Pennsylvania, New Jersey, and other parts of the US. Some Steam users have reported fixing the issue by closing Steam and reopening the registration page via the link Valve tweeted out, but as of this writing, others still haven’t been able to sign up.

Now Read:

• Valve Launches Repair Centers for Steam Deck
• Valve Plans ‘Next-Gen’ Steam Decks, With Multiple Games In Development
• iFixit Begins Selling Replacement Parts For Almost The Entire Steam Deck

Chemists Develop Process to Turn Previously Unrecyclable Plastics Into Propane

(Photo: Nick Fewings/Unsplash) Most of us have by now realized that conventional plastic recycling is more fantasy than reality. Despite our best attempts at separating out product packaging in the hope that some of it will be reused later, only 6 percent of the United States’ plastic waste is ever recycled. The rest of it ends up in landfills, in the ocean, or in other undesirable locations where it threatens local ecosystems and creates an unpleasant landscape.

Chemists are trying their hand at tackling that problem. A group of researchers at MIT have developed a process of turning previously unrecyclable plastics into propane, which can then be used as fuel or incorporated into a new product. Better yet, the process works for multiple types of plastic—a key element of any successful fight against plastic waste.

Disposable plastic goods don’t always share the same chemical makeup. Water bottles are made from a different type of plastic than milk containers, and your credit card is made up of a chemically different material than your red party cups. Sorting these materials not only takes time and money but also diverts some waste away from the recycling process entirely. But despite their differences, these plastics do have something in common: carbon and hydrogen, which make up a large portion of their chemical chains.

(Photo: Sigmund/Unsplash)

The chemists’ process, a type of hydrogenolysis, uses cobalt and zeolite (a grainy, porous substance) as a catalyst to break down polyethylene and polypropylene. In a reactor, the cobalt and zeolite work together to “edit” the plastics’ chemical makeup: The cobalt keeps the zeolite active as its pores limit the plastics’ molecular reactions. At the end of the process, a gas made primarily of propane emerges.

Propane has a number of uses, but its most recognizable is fuel, namely for household appliances, backyard grills, camping stoves, generators, and some farming equipment. It can also be used as a building block for new materials, like other plastics.

The process offers a way to recycle plastics that would otherwise be landfilled, but as with most things, it has its disadvantages. The reaction responsible for this chemical process is currently only done in small quantities, meaning it’ll need to be scaled up before it can be considered practical. Some plastics—like polyvinyl chloride, or PVC—would produce toxic gaseous byproducts if used alongside safer materials, so the process is far from perfect. Still, in a world with an increasingly problematic plastic waste problem, every bit helps.

Now Read:

• Viruses Extend Their Lives by Hitching a Ride On Microplastics
• Self-Propelled Robo-Fish Can Grab Microplastics Out of the Ocean
• Scientists Discover Plastic-Eating Worms That Digest Styrofoam

San Francisco Legislators Approve Use of Lethal Police Robots

(Photo: Max Fleischmann/Unsplash) The San Francisco Police Department, one of the country’s largest municipal law enforcement agencies, recently revised its equipment use policy to include its surprisingly diverse robot stockpile. Though most of its robots disarm bombs or conduct reconnaissance in dubious conditions, some ground-based robots can be used to carry active explosives. Inevitably, this led policy writers to wonder: Should law enforcement be allowed to engage its robots in lethal force? According to San Francisco’s Board of Supervisors, the answer is yes.

The city’s legislative body has voted to approve SFPD’s use of deadly robots, according to a report by the Washington Post. Tuesday’s Board meeting contained a “heated debate” centering on officers’ right to deploy robots with explicitly fatal intent. While some argued this capability would be helpful in active shooter scenarios, others painted the use of lethal police robots as “militaristic.” In the end, the Board voted 8 to 3 in favor of allowing lethal robot deployment in “extraordinary” circumstances.

“Robots will only be used as a deadly force option when risk of loss of life to members of the public or officers is imminent and outweighs any other force option available to SFPD,” the policy reads. The policy requires a second Board vote and the mayor’s approval before becoming city law.

Though not included in the policy above, the Post reports the Board made a final amendment Tuesday night to require that a supervisor approve the robots’ lethal use prior to deployment. This is standard across most other equipment listed on the policy, like SFPD’s semi-automatic rifles, robotic loudspeakers, and some SWAT tools. But for many, this is of little comfort.

The QinetiQ TALON, one of the SFPD’s many robots, has a long arm that the agency might use to carry explosives. (Photo: QinetiQ)

“We are living in a dystopian future, where we debate whether the police may use robots to execute citizens without a trial, jury, or judge,” civil rights and police misconduct attorney Tifanei Moyer told San Francisco’s Mission Local before the policy was voted through. “This is not normal. No legal professional or ordinary resident should carry on as if it is normal.”

San Francisco residents, legislators, police oversight groups, and even robotics scientists took to Twitter Wednesday morning calling for increased scrutiny over the policy. According to the Post article and other sources, SFPD has so far responded by assuring the public that it never planned to give its robots firearms—just bombs.

Though the policy does contain mentions of imminent danger and a lack of other force options, such verbiage has historically been flimsy at best. According to the organization Mapping Police Violence (which gets its data from government agencies, publicly-accessible media, and other sources), US police have shot and killed nearly 100 people confirmed to have been unarmed in 2022. Another 45 people have lost their lives despite a lack of knowledge regarding their ability to engage in lethal force. The police regularly make headlines for shooting people carrying water guns, airsoft guns, a hairbrush, or candy in a display of what officers themselves describe as an inability to distinguish threats from safety. Given the opportunity to share responsibility for lethal force with a robot, it’s difficult to say whether this ability will improve.

Now Read:

• Interpol Launches ‘Police Metaverse’ to Train Officers on Virtual Crime Fighting
• Amazon Gave Ring Footage to Police Without a Warrant or Owner Consent
• Police Are Using DNA to Generate Suspect Images, Sight Unseen

Nvidia’s Scrapped RTX 4080 12GB Will Return as the 4070 Ti: Report

(The Gigabyte RTX 4070 Ti will likely look very similar to its 4080 16GB. Credit: Gigabyte) Nvidia famously “unlaunched” the much-derided RTX 4080 12GB GPU a few weeks after it was announced. Instead, it chose to only move forward with its big brother, the RTX 4080 16GB. Despite both GPUs sharing the “4080” moniker, they have drastically different specs. Nvidia had said this was the reason for the confusing unlaunching. Now it appears the GPU will be reborn as the RTX 4070 Ti, according to regulatory filings from Gigabyte, and its AD104 die has also been photographed.

Gigabyte filed paperwork for its upcoming GPUs with the Eurasian Economic Commission (EEC), as reported by Videocardz. Although the card’s full specs aren’t listed, the titles include names such as GV-N407TEAGLE-12GD, so there’s no secret it’s a 4070 Ti. Overall Gigabyte will be offering a full suite of cards with eight models. This includes standard and “OC” variants of its Eagle, Gaming, and Aero, with the Aorus Elite and Aorus Master cards at the high end. Gigabyte usually offers a liquid-cooled model as well, but it’s not included in the EEC listing. The AD104 was built with TSMC’s 4N process. It’ll feature 7,680 CUDA Cores, 12GB of GDDR6X memory, 240 tensor cores, 60 ray tracing cores, and a 192-bit memory bus. This is one of its biggest differentiators from the 16GB card, which has a 256-bit memory bus. The RTX 3070 Ti also has a 256-bit bus.

The AD104 die is estimated to be 295mm², roughly half the size of the 608mm² AD102 die in the RTX 4090. The RTX 4080 16GB uses a 379mm² die named AD103. Prior to Ampere, the x104 dies were used in x80 GPUs, so this is a return to form for Nvidia. With Ampere, The RTX 3080 used the same AD102 die as the flagship RTX 3090, which is why there were so close in performance. For Ada Lovelace, Nvidia seems intent on putting more distance between its GPUs. Also, the AD104 die is about 100mm² smaller than its 392mm² Ampere equivalent.

The midrange-ish RTX 4070 Ti die, aka AD104. (Image: @MegasizeGPU)

It’s likely the reborn GPU will be launched at CES in January 2023. The timing is interesting as its main competitors from AMD will be launching on Dec. 13. This will give Nvidia plenty of time to figure out where its GPUs should land in pricing. AMD’s 7900 XTX and 7900 XT GPUs are $999 and $899 respectively. The company has also stated on the record that RTX 4080 is the target. Nvidia had previously priced the 4080 12GB at $899, so it’ll be interesting to see if it sticks with that price tag. The RTX 3070 Ti  Founders Edition cost $599 at launch (theoretically).

Nvidia is in a bit of a sticky situation with its GPU pricing. Nobody is complaining about the $1,599 RTX 4090 as it’s easily the most powerful GPU ever made. Even AMD has admitted it has no answer for it. The stepped-down RTX 4080 16GB is seen mainly as a flop, with inventory languishing on store shelves weeks after it launched. This is despite reports of a smaller launch inventory than the 4090. For anything other than flagship GPUs, customers are much more sensitive to price now, and the ongoing economic situation hasn’t helped matters. Nvidia is obviously aware of this, so there’s hope it could price the 4070 Ti competitively. To us, that sounds like a $799 or even $699 GPU, but we’ll have to wait until January to find out.

Now Read:

• Class Action Suit Against Nvidia Filed Over Melted RTX 4090s
• Nvidia Is Reportedly Ending Production of Its Most Popular Turing GPUs
• Nvidia RTX 4090 Sets New Records in Password Cracking Benchmarks

Microsoft Fixes Windows 11 22H2 Gaming Issues, Resumes Updates

(Credit: PCMag) It wouldn’t be a Windows update without some bugs, and indeed, there were some annoyances when Microsoft finally released the 22H2 update (formally the Windows 11 2022 Update). Among the issues was a possible drop in game performance, but gamers can rest easy: Microsoft has released a hotfix that takes care of the problem.

After the release of version 22H2 in late September, gamers complained about spotty performance in some titles. Part of the issue stemmed from Nvidia GPU drivers, which the company solved with a quick update. However, Microsoft noted there were deeper problems in the update. “Affected games and apps are inadvertently enabling GPU performance debugging features not meant to be used by consumers,” it said at the time.

Microsoft never provided a list of titles suffering from such performance degradation, but gamers online have cited issues with Call of Duty: Modern Warfare 2, Halo Infinite, and God of War, to name a few. When it became apparent the performance issue was widespread, Microsoft blocked 22H2 update notifications and automatic installations on systems with affected games. It dropped that limit last week, and now there’s a fix.

The KB5020044 update is rolling out now on all Windows 11 systems.

Anyone running the latest build can take a quick trip into the settings to check for updates to find the patch: KB5020044. This update addresses the GPU debug issue, which should eliminate the slowdowns and stutters that many were experiencing. That’s not all it does, though. There’s a full changelog to peruse, but the highlights include some improvements to OneDrive integration, and fixes for File Explorer crashes. It also combines Windows Spotlight with Themes on the Personalization page. Despite being probably the most important change in KB5020044, the GPU fix is listed way at the bottom of the changelog. Way to bury the lede, Microsoft.

The KB5020044 patch only takes a few minutes to install, but you will need to restart your PC to complete the process. Microsoft decided recently to end its practice of releasing one or two major updates for Windows each year. Going forward, it will release new features for Windows 11 when they’re ready. Hopefully, the focus on smaller incremental updates will decrease the number of major bugs that slip through the cracks.

Now read:

• PC Gamers Are Warming Up to Windows 11, Steam Survey Reveals
• Microsoft to Add Android 13 and New Features to Windows 11
• Microsoft Is Showing Ads for Its Own Products in Windows 11’s Sign-Out Menu

SpaceX Fires 11 Raptor Engines as It Prepares for Orbital Starship Test

SpaceX has seen enormous success with the Falcon 9 and Dragon combo, which are certified to carry both cargo and astronauts into space. Its follow-up will be the Starship, which has conducted a few short suborbital test flights. To leave Earth behind, though, Starship needs the Super Heavy booster. SpaceX has just conducted the longest-yet firing test of the Super Heavy’s Raptor engines, and nothing exploded.

When stacked atop the Super Heavy, Starship is the tallest rocket in the world, and it will also be the most powerful launch vehicle ever designed once it gets off the ground. Before that, SpaceX needs to ensure that the Super Heavy’s bank of 33 Raptor engines works correctly. Previously, the company lit 14 Raptor engines on Super Heavy, but they ran only a few seconds before shutting down. In the most recent test, 11 of the 33 boosters ignited for 12 seconds.

SpaceX didn’t elaborate on the purpose of this test, but it might be related to work on the rocket’s autogenous pressurization system, according to Ars Technica. Rather than use an inert gas like helium to pressurize the fuel tanks, Super Heavy uses gasses generated on the rocket to do the job. CEO Elon Musk said this test would take 20 seconds, though. It’s possible the 12-second firing was to test autogenous pressurization, or it could have been something completely different.

Next test is ~20 sec firing with max oxygen fill to test autogenous pressurization, possibly one more static fire, then orbital launch attempt

— Elon Musk (@elonmusk) November 15, 2022

Musk’s aerospace firm is taking an iterative approach to developing Starship, which he has said will eventually replace the Falcon 9 for all of SpaceX’s operations. At least initially, Starship will be utilized for a smaller number of important launches. NASA has contracted with SpaceX to develop the Starship HLS (Human Landing System) for lunar landings in the Artemis Program. Musk has also said that Starship is necessary to deploy the company’s next-gen Starlink v2 communication satellites.

SpaceX has previously said it intends to conduct an orbital test of Starship this year. However, it has yet to even light all 33 Super Heavy engines at the same time, and Starship doesn’t have enough power to reach orbit on its own. Even when the company has completed testing, it needs to secure a launch license from the US Federal Aviation Administration, and that is still in process. It’s possible Starship will try to head into space this year, but it looks increasingly likely the test will be pushed to early 2023.

Now read:

• Falcon Heavy Rocket Aces First Launch in 3 Years
• Musk Says SpaceX Could Land Humans on Mars in Just Five Years
• SpaceX Starship SN15 Lands in One Piece After High-Altitude Flight