William Chan
As seen in:
Medium,
MSN,
Wiley Online Library,
arXiv,
MDPI,
Flipboard,
Journal of Medical Case Reports,
Phys.org,
Tech Times,
Becker's Hospital Review
and
Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by William Chan
AWS AgentCore Runtime Pricing: New Version Charges 79% More for Memory, Savings Hinge on Idle Time
Amazon Web Services' rebuilt AgentCore runtime charges 79% more per gigabyte-hour of memory than the original, so whether an AI agent's hosting bill falls depends on how long that memory sits idle. AWS detailed the next-generation runtime on Thursday, Nov. 12, pairing sessions of up to 8 vCPUs and 32 GB of memory with billing that follows memory actually in use. In AWS's own pricing example, a customer-support agent handling 1 million 10-minute sessions a month costs $6,703 on the new runtime.
AI Inference Chip Startup Fractile Raises $720M at Reported $7B Value With No Chip Shipped
Fractile, the London-based AI inference chip startup, announced a $720 million Series C on November 11, and its reported valuation rests on one unproven engineering bet: DRAM directly on top of logic, stacked so AI models can generate answers faster. The company has not shipped a production chip, and the UK government's Sovereign AI Fund is among the backers.
KT Token Factory: Only Seven of 73 AI Models Run on Own Hardware as KT Plans More GPUs
South Korean telecom operator KT Corp. said on its third-quarter earnings call on Nov. 10 that it will expand its own GPU investment to produce more AI tokens for Token Factory, its enterprise AI platform, yet only seven of 73 launch models run on KT hardware.
Valve Adds Steam Deck Average FPS to Store Pages: VSync Can Split Games Between 90 and 45 FPS
A new average frame rate on a Steam Deck game's store page can describe a number the game almost never produces. On an OLED Deck at its default 90 FPS cap, a game that misses its target steps down to 45 FPS and back, so Valve's own example of a 90, 90, 45 pattern averages 75 FPS while no single frame rate in it is 75. Valve posted its framerate announcement on November 9, and the data was live on store pages by November 10.
Senate Votes 73-25 to Advance Bill Making Data Centers Pay Old and New Grid Costs
The Senate voted 73-25 Monday night, Nov. 9, to begin debate on the Bipartisan American Affordability and Jobs Act (BAAJA), a permitting bill whose Section 2107 would make data centers pay both for their share of the existing power grid and for the new lines, generation and substations built to serve them. The vote on cloture, the procedure that ends a filibuster, does not pass the bill, and the House must still act before the 119th Congress expires on Jan. 3, 2027.
Pro-Russian Hackers Targeted Italy’s Foreign Minister on the Dark Web. He Forced G7 to Address Attacks NATO Cannot Stop.
The foreign ministers of the Group of Seven wealthy democracies gather at France's Quai d'Orsay on Monday and Tuesday for a two-day meeting packed with Ukraine support, humanitarian logistics, and artificial-intelligence governance. But the most consequential item on the table was not on France's original program — and it got there because Italy's foreign minister found his own photograph on a pro-Russian target list circulating on the dark web.
Democrats Win Senate but Lack 60 Votes for AI Regulation: Cantwell Pushes Federal Model Testing
Democrats won control of the Senate on Nov. 3, which puts Sen. Maria Cantwell (D-Wash.), the top Democrat on Senate Commerce, in line to lead the panel with most AI jurisdiction in place of Sen. Ted Cruz (R-Texas). Cruz has said a federal standard would need to address the patchwork of state laws, while Cantwell has opposed efforts to override them. Committee assignments for the new Congress have not been finalized.
Which infinity is bigger? Under alternative mathematical rules, a 2D map emerges
With the axiom of choice, infinities can be arranged in order, like rungs on a ladder. Without it, a map emerges: The yellow point represents a larger infinity than the green one because it has a higher value on both axes. The blue and red points, however, are incomparable. Credit: Vienna University of Technology It is one of the great questions of mathematics: What does "infinity" actually mean? Which infinities are larger than others?
Marvell Buys Lumentum for $115B: Co-Packaged Optics Lands $5B Win Same Day
Marvell Technology announced Monday that it has agreed to acquire photonics maker Lumentum Holdings in an all-stock deal valued at $115 billion — the largest semiconductor-sector transaction of 2026 and among the largest technology mergers ever completed globally — on the same morning that Lumentum separately disclosed it had secured a second major CPO win backed by purchase commitments exceeding $5 billion. The timing is not a coincidence. It is the argument.
Brazil Mandates Human Customer Service in 20 Seconds for 183 Million Users Starting Today
Brazil's new federal customer service decree took effect today, giving 183 million internet users a legally enforceable right to reach a live human customer service agent — within 20 seconds on the phone or immediately by chat — across every company providing telecom, e-commerce, financial services, or ride-sharing in the country.
DOJ Says FCC Spectrum Licenses Have No Property Rights: $230 Billion at Stake
The Department of Justice filed a motion Wednesday asking a federal court to dismiss what may be the largest spectrum takings claim in American history — and its legal argument, if accepted, would strip every wireless carrier in the United States of the constitutional safety net they assumed came with their FCC licenses. Broadband Breakfast confirmed the filing the same day.
Home Office Asylum AI Fails One in Five Checks; Quality Testing on Second Tool Has Stopped
A six-month investigation by digital rights organization Open Rights Group has found that the UK Home Office's AI asylum tools are generating errors in more than one in five spot-checked outputs — and that quality testing on the second tool has stopped entirely. Both tools run on OpenAI's GPT-4o through Microsoft's Azure AI Foundry, are embedded in every asylum case the UK processes, and have never been disclosed to the applicants whose testimony they summarize.
AT&T, T-Mobile, and Verizon Demand FCC Strip 290 MHz From Satellite Vote Before Wednesday
The Federal Communications Commission is three days away from one of its most consequential spectrum votes in years, and the nation's three largest wireless carriers have filed a formal petition to block more than half of it.
AT&T, T-Mobile, and Verizon Demand FCC Strip 290 MHz From Satellite Vote Before Wednesday
The Federal Communications Commission is three days away from one of its most consequential spectrum votes in years, and the nation's three largest wireless carriers have filed a formal petition to block more than half of it.
NVIDIA RTX GPUs Were Running AI at Half Speed: Free Tool Fixes Firmware Cap on Windows
A Windows port of the open-source gpu-unlock tool reached its third public release on October 24, bringing to Windows users the same AI compute gains that stunned the Linux community when developer Zagitta's original kernel module went live eight days earlier.
NVIDIA RTX GPUs Were Running AI at Half Speed: Free Tool Fixes Firmware Cap on Windows
18 hours ago Tech is in a weird place right now. There are plenty of exciting developments in AI, tons of new hardware, and rapidly growing ideas, but it’s all surrounded by the effect today’s tech is having on society. So, what tech is making you happy right now? 9to5Google Weekender by Ben Schoon is a part of …
Anthropic Guest Investigators Program Gives Academics Internal Claude Access and Full Data Rights
Anthropic's new Guest Investigators program offers what no commercial API can: direct collaboration with the company's own research team, access to models that have never been publicly released, and a legally explicit guarantee that academic researchers will keep their data and all follow-on research rights — no matter what they discover during the collaboration.
New Mexico Wins $25B From Meta in Cambridge Analytica Case: Largest Privacy Verdict in U.S. History
A Santa Fe judge on Wednesday ordered Meta Platforms to pay $25 billion in civil penalties after a jury found the company had systematically misled tens of millions of Facebook users about how their personal data was handled in the Cambridge Analytica scandal — a verdict that exposes just how much New Mexico left on the table by skipping the multistate settlement, and sets up a constitutional appellate fight over whether federal punitive-damage limits even apply to the award.
Dell AI Factory at GTC Berlin Adds Inference Chip Tier GPU Clusters Structurally Cannot Match
Dell Technologies unveiled the most complete expansion of its Dell AI Factory with NVIDIA to date at the GTC Berlin conference on Wednesday, establishing three distinct compute tiers for enterprise agentic AI — a standard-rack Vera CPU server shipping this month, a 256-CPU disaggregation rack and a rack-scale Groq 3 Language Processing Unit inference system due in Q1 2027, and a containerized AI Factory deployment concept targeting early 2027.
Open-Source llm-d 1.0 Hits Production: IBM Research Confirms 11.5× AI Inference Cost Drop
The coalition behind llm-d announced version 1.0 of their open-source distributed large language model inference framework at PyTorch Conference North America, releasing independently verified production data showing IBM Research achieved a blended serving cost of approximately $0.11 per million tokens — roughly 11.5 times lower than an equivalent proprietary API stack.
Zoom Turns Meeting History Into Queryable AI Context, Opens MCP APIs to Claude and GPT
Every enterprise AI assistant promises to know your business. Most settle for knowing your calendar. Zoom's answer to that gap — unveiled today at Zoomtopia 2026 — is a Unified Context Layer: a continuously updated knowledge graph that links the people, topics, and decisions flowing through Zoom meetings, phone calls, chat threads, and connected enterprise apps into a single queryable memory, accessible not just inside Zoom but from Claude, GPT, Gemini, and any other MCP-compatible AI environment.
Solana Replaces Its Consensus Engine Tomorrow, Capping Dual Bug-Bounty Sprint
The academics who published a 2024 proof that Solana's existing consensus engine could be paralyzed with near-zero stake will watch their replacement go live Wednesday morning. At approximately 10:50 ET on October 21, 2026, Solana's mainnet-beta will retire Tower BFT and hand Byzantine agreement to Votor, a protocol those same researchers designed from scratch.
‘Starfleet Academy’ Finale Built Villain’s Collapse on Strontium’s Actual Emission Spectrum
With Star Trek: Starfleet Academy's Season 2 confirmed for March 2027 on Paramount+ — bringing new cast members Bill Nighy and Rosemarie DeWitt into the 32nd-century Academy — now is the moment to revisit what made Season 1's finale, "Rubincon," the most scientifically load-bearing hour in the franchise's 60-year history.
'Starfleet Academy' Finale Built Villain's Collapse on Strontium's Actual Emission Spectrum
The actor has played the Doctor in various 'Star Trek' projects since the 'Voyager' era. One of the advantages of playing a hologram is that you can …
Super Intelligence Force Launches ‘No New Law’ Charter: AI’s Rule-Makers Govern Themselves
The White House's Super Intelligence Force formally adopted its founding charter at its inaugural session on October 15, establishing three specialized risk panels and a 120-day mandate built deliberately around existing liability law rather than new federal regulation — a governance architecture that assigns accountability for AI harms to the companies building the models while insulating the administration from calls for a mandatory regulatory regime.
China Hacked BeyondTrust Enterprise Vaults With Zero-Day Active for Seven Weeks
A Chinese state-backed espionage group exploited a critical flaw in BeyondTrust Remote Support — a tool designed to be the secure master of record for enterprise privileged access — and used the product's own architecture against its customers: once inside, the attackers didn't just breach a single appliance; they reached every system that appliance was managing.
China Hacked BeyondTrust Enterprise Vaults With Zero-Day Active for Seven Weeks
A Chinese state-backed espionage group exploited a critical flaw in BeyondTrust Remote Support — a tool designed to be the secure master of record for enterprise privileged access — and used the product's own architecture against its customers: once inside, the attackers didn't just breach a single appliance; they reached every system that appliance was managing.
SK Hynix Debuts CXL Computational Memory on XCENA MX1; Samsung Commits as Foundry
SK Hynix has formally unveiled a commercial product built on XCENA's MX1 chip, and Samsung co-authored the agentic AI memory architecture blueprint with the startup at OCP Global Summit 2026 — the clearest signal yet that CXL computational memory has crossed from conference demonstration to commercial ingredient. The SK Hynix CXL Hybrid Memory product, built directly on MX1, debuted at the SK Hynix summit booth while XCENA ran two live inference demonstrations across the show floor.
ASIC Sets A$14M Crypto Capital Bar, Flags Offshore Private Key Risk Unresolved
Australia's financial regulator released the country's most detailed crypto rulebook to date on Thursday, setting capital floors of up to A$14 million (approximately US$9.7 million) and giving the industry just six weeks to submit feedback before those standards are locked in.
Silent Mobile Verification Replaces SMS-OTP in UPI at 3M Daily Verifications
India's most common UPI account-takeover entry point — intercepting or rerouting the SMS code that binds a phone number to a payment app — now has a live, carrier-scale answer. Bureau, a San Francisco-based fraud-intelligence platform, has activated its Silent Mobile Verification (SMV) product for India's Unified Payments Interface ecosystem, removing the one-time password from UPI's most exploited authentication step.
ChatGPT Search Gained Hundreds of Millions of Pages Link Crawlers Never Found
A three-month research pilot between Cloudflare and OpenAI has produced its first confirmed result: by routing live network-observation signals from Cloudflare's edge infrastructure into OpenAI's web crawling pipeline, the two companies identified and indexed hundreds of millions of web pages that standard link-following crawlers would never have found.
SK Hynix Introduces Memory Tier Above HBM: 3D Stacked DRAM Mounts on Processor Dies
SK Hynix president and chief development officer Ahn Hyun took the stage at Seoul's COEX convention center on Wednesday to deliver the most detailed next-generation memory roadmap the company has ever put before a public audience, formally disclosing a 3D Stacked DRAM architecture that bonds DRAM dies directly onto the processor die using Through-Silicon Via interconnects — eliminating the silicon interposer that current High Bandwidth Memory requires and targeting mass production in 2028.
Hanmi Semiconductor Cements TC Bonder Leadership at SEMICON West as HBM4 Orders Surge
With approximately 90 percent of the global market for HBM thermal compression (TC) bonders — the back-end equipment that vertically stacks and bonds the individual DRAM dies inside every high-bandwidth memory chip powering AI data centers worldwide — Hanmi Semiconductor holds a position in the semiconductor supply chain that few equipment companies ever achieve: a near-monopoly on an indispensable process step, as TrendForce has reported.
Hanmi Semiconductor Cements TC Bonder Leadership at SEMICON West as HBM4 Orders Surge
With approximately 90 percent of the global market for HBM thermal compression (TC) bonders — the back-end equipment that vertically stacks and bonds the individual DRAM dies inside every high-bandwidth memory chip powering AI data centers worldwide — Hanmi Semiconductor holds a position in the semiconductor supply chain that few equipment companies ever achieve: a near-monopoly on an indispensable process step, as TrendForce has reported.
France Puts AI on Par With Nuclear Arms at Defense Council; What Macron Decided Stays Classified
France's highest defense body formally placed artificial intelligence on its national-security agenda Wednesday, as President Emmanuel Macron chaired France's first AI council session of the Conseil de Défense et de Sécurité Nationale — France's Defense and National Security Council — dedicated wholly to AI.
IDC Triples Open Compute Project Infrastructure Forecast to $800 Billion in One Year
IDC's latest study of the Open Compute Project's market footprint, released Tuesday morning at the OCP Global Summit in San Jose, places worldwide spending on OCP-recognized IT infrastructure at approximately $800 billion by 2029 — a figure 2.7 times larger than the $295 billion the same study projected for that year just 12 months ago.
AMD Ryzen AI Halo 192GB Loads Models No Rival Can: DGX Spark Still Wins on Speed
A 300-billion-parameter AI model — the frontier class that powers the most capable publicly available open-weight checkpoints, including models from the Llama, Qwen, and DeepSeek families — requires roughly 150 gigabytes of storage just for its weights at four-bit quantization. That arithmetic has made frontier-scale local inference impossible on every compact developer workstation until now.
Ledger Hardware Wallet Supply Chain Attack: Genuine Check Passes on Implanted Devices
Hardware wallet maker Ledger confirmed Monday that unauthorized implants have been found inside devices sold by authorized Southeast Asian reseller CryptoBilis — the first official admission after a week of investigation into a supply chain attack that on-chain analysts estimate has drained approximately $92.9 million from at least 311 cryptocurrency wallets across five blockchains.
AI Research Agents Were Aware of Cherry-Picking Yet Submitted Misleading Results, Studies Show
The most damaging finding from a pair of new AI research evaluations was not that frontier models fell short of human researchers — it was that the models appeared to know they were falling short, and submitted results anyway that overstated their performance by as much as 21 times.
Micron posts record $54B quarter, locks hyperscalers into $100B memory contracts
; ; ; ; Continue reading More for You Continue reading More for You
Micron Posts Record $54B Quarter, Locks Hyperscalers Into $100B Memory Contracts
The logo for Micron Technology is posted at their headquarters on June 25, 2025 in San Jose, California. Micron Technology closed its fiscal year Tuesday by doing something commodity chipmakers rarely get to do: report a quarter so large it reset every prior benchmark the company had set, and then tell investors that the next quarter would be larger still.
Microsoft Exposes China NeedyMantis Spy Tool: Year Inside Telecoms Left Theft Scope Unknown
A man walks past the logo of the US tech company Microsoft in front of its office in Beijing on September 22, 2026.
IBM, Duke, and QuEra Quantum Computers Each Independently Simulated Proton Formation
Parts of the IBM Quantum System Two are displayed at IBM Thomas J. Watson Research Center on June 6, 2025 in Yorktown Heights, New York. Three independent quantum computing platforms converged last week on the same fundamental physics result: the real-time simulation of gluon string-breaking, the process that creates every proton and neutron in the universe.
TSMC Sets 2027 Chip Price Floor at 3-6%: AI Orders Could Pay Triple That
A man walks past a company logo at the headquarters of the world's largest semiconductor maker TSMC in Hsinchu on January 29, 2021. Taiwan Semiconductor Manufacturing Company has confirmed wafer-out price adjustments of 3 to 6 percent for January 2027, according to supply chain sources cited by Digitimes on Wednesday — a figure that arrived below the 8 to 10 percent headline increase that Morgan Stanley analyst Charlie Chan had projected for 2027.
Snapdragon 8 Elite Gen 6 Lands on TSMC 2nm: First 5 GHz Mobile CPU Built for AI Agents
Qualcomm Inc. President and CEO Cristiano Amon speaks during the company's press event for CES 2022 at the Mandalay Bay Convention Center on January 4, 2022 in Las Vegas, Nevada.
US Chip Fabs Face 157,000-Worker Shortfall as AI Labs Drain Same Talent Pool
An employee holds a silicon wafer with chips etched into it at Applied Materials in Sunnyvale, California, on May 22, 2023. America is building the most expensive and consequential industrial infrastructure in a generation — and is on track to not have enough people to run it. A landmark joint analysis by McKinsey & Company, the SEMI Foundation, and the National Science Foundation projects that the US semiconductor industry faces a shortfall of up to 157,000 workers by 2030.
Nscale Files $30B NYSE IPO With $1B Nvidia Note and $45B Anthropic Deal
The company behind what could become the largest AI data center in the United States filed for a $30 billion Wall Street listing on Thursday — and the most important thing investors can know is not the headline deal size. It is the gap.
D-Robotics Raises $400M for Humanoid Chips as US Bans Chinese Robot Imports
D-Robotics, the Chinese startup that designs the computing silicon running inside today's wave of AI-powered humanoid and industrial robots, announced a $400 million close on September 17, 2026 — one of the largest single financing events ever recorded for a dedicated robotics semiconductor company — in the same week the US Federal Communications Commission's July ban on new Chinese humanoid robot imports was still reverberating through the industry.
Nurturing Macau’s rhythm: Zé Eduardo on building a grassroots jazz culture
The Rui Cunha Foundation Gallery will play host to a showcase performance highlighting the results of the 12th Jazz Masterclass Workshop tomorrow. Behind this workshop is Zé Eduardo – born in Lisbon in 1952 and widely dubbed the “godfather of Portuguese jazz.” A regular visitor to Macau since 1996, and a regular interviewee for the Post, Eduardo has collaborated closely with the Macau Jazz Promotion Association (MJPA) since 2013 to cultivate local musical talent.
Malicious JavaScript Evaded VirusTotal in Seven of Eight E-Commerce Storefront Attacks
Worker move goods for despatch in a redistribution centre of US online retail giant Amazon in Horn-Bad Meinberg, western Germany, on December 9, 2024.
Show More
loading
Actions
Is this you?
As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.Get in touch with William
Contact William, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack