AMD’s CUDA Implementation Built On ROCm Is Now Open Source

Michael Larabel writes via Phoronix: While there have been efforts by AMD over the years to make it easier to port codebases targeting NVIDIA’s CUDA API to run atop HIP/ROCm, it still requires work on the part of developers. The tooling has improved such as with HIPIFY to help in auto-generating but it isn’t any simple, instant, and guaranteed solution — especially if striving for optimal performance. Over the past two years AMD has quietly been funding an effort though to bring binary compatibility so that many NVIDIA CUDA applications could run atop the AMD ROCm stack at the library level — a drop-in replacement without the need to adapt source code. In practice for many real-world workloads, it’s a solution for end-users to run CUDA-enabled software without any developer intervention. Here is more information on this “skunkworks” project that is now available as open-source along with some of my own testing and performance benchmarks of this CUDA implementation built for Radeon GPUs. […]

For those wondering about the open-source code, it’s dual-licensed under either Apache 2.0 or MIT. Rust fans will be excited to know the Rust programming language is leveraged for this Radeon implementation. […] Those wanting to check out the new ZLUDA open-source code for Radeon GPUs can do so via GitHub.

Read more of this story at Slashdot.

F-Zero Courses From a Dead Nintendo Satellite Service Restored Using VHS and AI

An anonymous reader quotes a report from Ars Technica: Nintendo’s Satellaview, a Japan-only satellite add-on for the Super Famicom, is a rich target for preservationists because it was the home to some of the most ephemeral games ever released. That includes a host of content for Nintendo’s own games, including F-Zero. That influential Super Nintendo (Super Famicom in Japan) racing title was the subject of eight weekly broadcasts sent to subscribing Japanese homes in 1996 and 1997, some with live “Soundlink” CD-quality music and voiceovers. When live game broadcasts were finished, the memory cartridges used to store game data would report themselves as empty, even though they technically were not. Keeping that same 1MB memory cartridge in the system when another broadcast started would overwrite that data, and there were no rebroadcasts.

As reported by Matthew Green at Press the Buttons (along with Did You Know Gaming’s informative video), data from some untouched memory cartridges was found and used to re-create some of the content. Some courses, part of a multi-week “Grand Prix 2” event, have never been found, despite a $5,000 bounty offering and extensive effort. And yet, remarkably, the 10 courses in those later broadcasts were reverse-engineered, using a VHS recording, machine learning tools, and some manual pixel-by-pixel re-creation. The results are “north of 99.9% accurate,” according to those who crafted it and exist now as a mod you can patch onto an existing F-Zero ROM. […] Their work means that, 25 years later, a moment in gaming that was nearly lost to time and various corporate currents has been, if not entirely restored, brought as close as is humanly (and machine-ably) possible to what it once was.

Read more of this story at Slashdot.

Threads Starts Testing Trending Topics Section

In a post on Threads today, Meta CEO Mark Zuckerberg announced that the Twitter rival is rolling out a new test in the U.S. that shows a list of popular topics that other people are discussing. It’s a core feature of Twitter (now “X”) that should help make it easier to find new posts and users on Threads. The Verge reports: The list will appear on the search page and in the For You feed, allowing you to browse through posts about a particular topic. In addition to testing a way to save posts, Meta recently said it will add controls to limit political content recommendations on Threads. Instagram head Adam Mosseri says Threads uses machine learning AI systems to determine what goes on the list. It takes into account how many people are talking about a certain topic, as well as how many people have interacted with posts about it.

Read more of this story at Slashdot.

Early Blood Test To Predict Dementia Is Step Closer As Biological Markets Identified

Researchers have made significant progress toward developing a blood test that can predict the risk of dementia up to 15 years before clinical diagnosis. The Guardian reports: Hopes for the test were raised after scientists discovered biological markers for the condition in blood samples collected from more than 50,000 healthy volunteers enrolled in the UK Biobank project. Analysis of the blood identified patterns of four proteins that predicted the onset of dementia in general, and Alzheimer’s disease and vascular dementia specifically, in older age. When combined with more conventional risk factors such as age, sex, education and genetic susceptibility, the protein profiles allowed researchers to predict dementia with an estimated 90% accuracy nearly 15 years before people received clinical confirmation of the disease.

For the latest study, blood samples from 52,645 UK adults without dementia were collected and frozen between 2006 and 2010 and analyzed 10 to 15 years later. More than 1,400 participants went on to develop dementia. Using artificial intelligence, the researchers looked for connections between nearly 1,500 blood proteins and developing dementia years later. Writing in Nature Aging, they describe how four proteins, Gfap, Nefl, Gdf15 and Ltbp2, were present in unusual levels among those who developed all-cause dementia, Alzheimer’s disease or vascular dementia. Higher levels of the proteins were warning signs of disease. Inflammation in the brain can trigger cells called astrocytes to over-produce Gfap, a known biomarker for Alzheimer’s. People with raised Gfap were more than twice as likely to develop dementia than those with lower levels.

Another blood protein, Nefl, is linked to nerve fibre damage, while higher than normal Gdf15 can occur after damage to the brain’s blood vessels. Rising levels of Gfap and Ltbp2 was highly specific for dementia rather than other brain diseases, the scientists found, with changes occurring at least 10 years before people received a dementia diagnosis. The researchers are speaking to companies to develop the test but said the cost, currently at several hundred pounds, would need to come down to make it viable.

Read more of this story at Slashdot.

Amazon Hides Cheaper Items With Faster Delivery, Lawsuit Alleges

A class-action lawsuit alleges (PDF) that Amazon manipulates its platform through a biased algorithm to favor the “Buy Box” for items that generate higher fees for Amazon, often leading consumers to overpay for products that could be obtained cheaper and just as quickly from other sellers on the platform. Ars Technica reports: The lawsuit claims that a biased algorithm drives Amazon’s “Buy Box,” which appears on an item’s page and prompts shoppers to “Buy Now” or “Add to Cart.” According to customers suing, nearly 98 percent of Amazon sales are of items featured in the Buy Box, because customers allegedly “reasonably” believe that featured items offer the best deal on the platform.

“But they are often wrong,” the complaint said, claiming that instead, Amazon features items from its own retailers and sellers that participate in Fulfillment By Amazon (FBA), both of which pay Amazon higher fees and gain secret perks like appearing in the Buy Box. “The result is that consumers routinely overpay for items that are available at lower prices from other sellers on Amazonâ”not because consumers don’t care about price, or because they’re making informed purchasing decisions, but because Amazon has chosen to display the offers for which it will earn the highest fees,” the complaint said.

Authorities in the US and the European Union have investigated Amazon’s allegedly anticompetitive Buy Box algorithm, confirming that it’s “favored FBA sellers since at least 2016,” the complaint said. In 2021, Amazon was fined more than $1 billion by the Italian Competition Authority over these unfair practices, and in 2022, the European Commission ordered Amazon to “apply equal treatment to all sellers when deciding what to feature in the Buy Box.” These investigations served as the first public notice that Amazon’s Buy Box couldn’t be trusted, customers suing said. Amazon claimed that the algorithm was fixed in 2020, but so far, Amazon does not appear to have addressed all concerns over its Buy Box algorithm. As of 2023, European regulators have continued pushing Amazon “to take further action to remedy its Buy Box bias in their respective jurisdictions,” the customers’ complaint said.

Read more of this story at Slashdot.

Computer Simulations of Atlantic Ocean Currents Finds Collapse Could Happen in Our Lifetime

An anonymous reader shared this report from the Associated Press:

An abrupt shutdown of Atlantic Ocean currents that could put large parts of Europe in a deep freeze is looking a bit more likely and closer than before as a new complex computer simulation finds a “cliff-like” tipping point looming in the future. A long-worried nightmare scenario, triggered by Greenland’s ice sheet melting from global warming, still is at least decades away if not longer, but maybe not the centuries that it once seemed, a new study in Friday’s Science Advances finds.

The study, the first to use complex simulations and include multiple factors, uses a key measurement to track the strength of vital overall ocean circulation, which is slowing. A collapse of the current — called the Atlantic Meridional Overturning Circulation or AMOC — would change weather worldwide because it means a shutdown of one of key the climate and ocean forces of the planet. It would plunge northwestern European temperatures by 9 to 27 degrees (5 to 15 degrees Celsius) over the decades, extend Arctic ice much farther south, turn up the heat even more in the Southern Hemisphere, change global rainfall patterns and disrupt the Amazon, the study said. Other scientists said it would be a catastrophe that could cause worldwide food and water shortages.

“We are moving closer (to the collapse), but we we’re not sure how much closer,” said study lead author Rene van Westen, a climate scientist and oceanographer at Utrecht University in the Netherlands. “We are heading towards a tipping point.” When this global weather calamity — grossly fictionalized in the movie “The Day After Tomorrow” — may happen is “the million-dollar question, which we unfortunately can’t answer at the moment,” van Westen said. He said it’s likely a century away but still could happen in his lifetime. He just turned 30.

“It also depends on the rate of climate change we are inducing as humanity,” van Westen said.

Read more of this story at Slashdot.

San Francisco Mob Lights Driverless Waymo Car on Fire

Last night in San Francisco’s Chinatown, “A person jumped on the hood of a Waymo driverless taxi and smashed its windshield…” reports the Verge, “generating applause before a crowd formed around the car and covered it in spray paint, breaking its windows, and ultimately set it on fire.”

The fire department arrived minutes later, according to a report in The Autopian, but by then flames had already fully engulfed the car…. Waymo representative Sandy Karp told The Verge via email that the fully autonomous car “was not transporting any riders” when it was attacked and fireworks were tossed inside the car, sparking the flames…

The fire takes place against the backdrop of simmering tension between San Francisco residents and automated vehicle operators… Just last week, a Waymo car struck a cyclist who had reportedly been following behind a truck turning across its path.

The “burnt-out husk of the electric Waymo Jaguar” appears in a video posted on YouTube, according to the article. “Another set of videos posted by software developer Michael Vendi gives a view into the scene as it played out and the fire grew.”

San Francisco’s 49ers play in the Super Bowl this afteroon, so last night’s celebrations for Chinese New Year could be followed by additional celebrations tonight. Police Chief Bill Scott is already urging residents to behave responsibly. “Please don’t light anything on fire.”

Read more of this story at Slashdot.

California Bill Would Require Computer Science For High School Graduation

At a press conference last week, a California Assemblymember joined the State Superintendent of Public Instruction in announcing a bill that, if passed, would require every public high school to teach computer science. (And establish CS as a high school graduation requirement by the 2030-31 school year.)

Long-time Slashdot reader theodp says he noticed posters with CS-education advocacy charts and stats “copied verbatim” from the tech giant-backed nonprofit Code.org. (And “a California Dept. of Education news release also echoed Code.org K-12 CS advocacy factoids.”)

The announcement came less than two weeks after Code.org CEO Hadi Partovi — whose goal is to make CS a HS graduation requirement in all 50 states by 2030 — was a keynote speaker at the Association of California School Administrators Superintendents’ Symposium. Even back in an October 20 Facebook post, [California state assemblyman] Berman noted he’d partnered with Code.org on legislation in the past and hinted that something big was in the works on the K-12 CS education front for California. “I had the chance to attend Code.org’s 10th anniversary celebration and chat with their founder, Hadi Partovi, as well as CS advocate Aloe Blacc. They’ve done amazing work expanding access to computer science education… and I’ve been proud to partner with them on legislation to do that in CA. More to come!”

Read more of this story at Slashdot.

In Big Tech’s Backyard, a California State Lawmaker Unveils a Landmark AI Bill

An anonymous reader shared this report from the Washington Post:

A California state lawmaker introduced a bill on Thursday aiming to force companies to test the most powerful artificial intelligence models before releasing them — a landmark proposal that could inspire regulation around the country as state legislatures increasingly tackle the swiftly evolving technology.

The new bill, sponsored by state Sen. Scott Wiener, a Democrat who represents San Francisco, would require companies training new AI models to test their tools for “unsafe” behavior, institute hacking protections and develop the tech in such a way that it can be shut down completely, according to a copy of the bill. AI companies would have to disclose testing protocols and what guardrails they put in place to the California Department of Technology. If the tech causes “critical harm,” the state’s attorney general can sue the company.

Wiener’s bill comes amid an explosion of state bills addressing artificial intelligence, as policymakers across the country grow wary that years of inaction in Congress have created a regulatory vacuum that benefits the tech industry. But California, home to many of the world’s largest technology companies, plays a singular role in setting precedent for tech industry guardrails. “You can’t work in software development and ignore what California is saying or doing,” said Lawrence Norden, the senior director of the Brennan Center’s Elections and Government Program… Wiener says he thinks the bill can be passed by the fall.
The article notes there’s now 407 AI-related bills “active in 44 U.S. states (according to an analysis by an industry group called BSA the Software Alliance) — with several already signed into law. “The proliferation of state-level bills could lead to greater industry pressure on Congress to pass AI legislation, because complying with a federal law may be easier than responding to a patchwork of different state laws.”
Even the proposed California law “largely builds off an October executive order by President Biden,” according to the article, “that uses emergency powers to require companies to perform safety tests on powerful AI systems and share those results with the federal government. The California measure goes further than the executive order, to explicitly require hacking protections, protect AI-related whistleblowers and force companies to conduct testing.”

They also add that as America’s most populous U.S. state, “California has unique power to set standards that have impact across the country.” And the group behind last year’s statement on AI risk helped draft the legislation, according to the article, though Weiner says he also consulted tech workers, CEOs, and activists. “We’ve done enormous stakeholder outreach over the past year.”

Read more of this story at Slashdot.

Nvidia is Forming a New Business Unit to Make Custom Chips

An anonymous reader shared this report from Reuters:

Nvidia is building a new business unit focused on designing bespoke chips for cloud computing firms and others, including advanced AI processors, nine sources familiar with its plans told Reuters. The dominant global designer and supplier of AI chips aims to capture a portion of an exploding market for custom AI chips and shield itself from the growing number of companies pursuing alternatives to its products.

The Santa Clara, California-based company controls about 80% of high-end AI chip market, a position that has sent its stock market value up 40% so far this year to $1.73 trillion after it more than tripled in 2023. Nvidia’s customers, which include ChatGPT creator OpenAI, Microsoft, Alphabet, and Meta Platforms, have raced to snap up the dwindling supply of its chips to compete in the fast-emerging generative AI sector. Its H100 and A100 chips serve as a generalized, all-purpose AI processor for many of those major customers. But the tech companies have started to develop their own internal chips for specific needs. Doing so helps reduce energy consumption, and potentially can shrink the cost and time to design.

Nvidia is now attempting to play a role in helping these companies develop custom AI chips that have flowed to rival firms such as Broadcom and Marvell Technology, said the sources, who declined to be identified because they were not authorized to speak publicly…
Nvidia moving into this territory has the potential to eat into Broadcom and Marvell sales.

Read more of this story at Slashdot.