TLDR Index

Search the archive

Full-text search across article titles and summaries. Refine with sector, year, content type, and reading-time filters.

Results

403 matches

TESLA'S AI5 CHIP ENTERS TRIAL PRODUCTION AT SAMSUNG'S TEXAS FAB

Samsung has started trial production of Tesla's AI5 chip on its 2nm node at the new Taylor, Texas foundry, fast-tracking prototype runs ahead of schedule to handle a wave of custom silicon orders, backed by Tesla's $16.5 billion manufacturing contract at the site.

2026.09.18
notateslaapp.com3 min readTLDR Hardware

STARSHIP FLIGHT 14 TARGETS FIRST FULL ORBITAL LAUNCH, NET SEPTEMBER 22

SpaceX has set Starship Flight 14 as its first attempt at a full orbital mission, currently targeted no earlier than September 22 (an earlier FCC filing had pointed to September 15), with Booster 21 and Ship 41. The flight is expected to deploy next-gen Starlink V3 satellites, which SpaceX says carry over ten times the capacity of current satellites for "fiber-like" speeds, marking the first time Starship has released operational payloads in orbit rather than flying a suborbital trajectory. Musk confirmed the upper stage won't attempt a tower catch this time, that capability is being saved for Flight 15, where he's given it a 50-60% success chance, while Booster 21 will most likely attempt a soft splashdown in the Gulf of Mexico following an anomaly during Booster 20's boostback and landing.

2026.09.16
arstechnica.com3 min readTLDR Hardware

AGILITY ROBOTICS UNVEILS DIGIT 5 HUMANOID BUILT FOR COOPERATIVELY SAFE WORK AT SCALE

Agility Robotics has introduced Digit 5, a next-generation humanoid robot featuring a 50-pound payload capacity, swappable end-effectors, and a rapid nine-minute recharge capability. The system incorporates a holistic safety architecture powered by onboard sensors and advanced safety controllers to autonomously slow down, avoid, or stop when people are detected nearby.

2026.09.16
links.tldrnewsletter.com2 min readTLDR Hardware

NVIDIA OPEN-SOURCES OSMO TO ORCHESTRATE TRAINING, SIMULATION, AND ROBOT TESTING FROM ONE YAML FILE

NVIDIA OSMO is an open-source, Kubernetes-native Physical AI orchestrator that connects data generation, RL training, simulation, evaluation, and hardware-in-the-loop testing across heterogeneous compute. Developers describe the pipeline in one YAML file, while OSMO schedules work across data-center GPUs, RTX simulation systems, cloud infrastructure, and edge platforms including Jetson and Arm. The system also exposes pipeline state, datasets, and available GPU capacity to coding agents.

2026.09.16
developer.nvidia.com6 min readTLDR Hardware

OPEN SOURCE SAFETY CONSORTIUM FOR PHYSICAL AUTONOMY LAUNCHED

Move fast and break things is great, until your robot hurts a person. Creating robots that can safely operate around people is incredibly important, but it is also incredibly time-consuming and expensive. The Open Source Safety Consortium aims to fix this using open source. The consortium is an industry collaborative initiative dedicated to defining standardized safety frameworks, reproducible verification procedures, and validation benchmarks for open-source physical AI software stacks. The organization is attempting to bridge the gap between open-source robotic software innovation and rigorous industrial safety certifications. The hope is that future robotics projects can use turn-key products and libraries to build safe robots out of the gate.

2026.09.16
opensourcesafe.com3 min readTLDR Hardware

WAYMO OPENS FULLY AUTONOMOUS LAS VEGAS RIDES WITH ITS 6TH-GEN DRIVER

Waymo started providing fully autonomous public rides in Las Vegas, expanding commercial deployment of its physical-AI driving stack. Las Vegas is among the first cities operating primarily with Waymo's Ojai vehicle platform powered by its sixth-generation Waymo Driver, across a nearly 24-mile service area. More than 100,000 people had registered interest.

2026.09.16
waymo.com4 min readTLDR Hardware

STRETCHABLE QLED HITS 16,000 PPI WITHOUT GOING DARK WHEN PULLED

Older stretchable screens only flexed the wiring, so fixed pixels thinned out and picture quality collapsed as you pulled. A UNIST/DGIST/IBS team printed intrinsically stretchable quantum-dot pixels at about 16,000 PPI and 53,300 nits that keep working at 1.65× stretch, so wearable and e-skin displays can stay outdoor-bright when bent.

2026.09.16
en.sedaily.com2 min readTLDR Hardware

META SAYS ITS THIRD-GEN IN-HOUSE AI CHIP BEATS NVIDIA ON EFFICIENCY, DEPLOYING BY EARLY 2027

Meta shares rose after Bloomberg reported the company plans to deploy its third-generation MTIA chip, internally called MTIA 450 or Arke, in data centers during the first half of 2027, with the next version, MTIA 500 (Astrid), set to finish design in about a month and reach data centers by the end of 2027. Twelve new sample chips arrived from TSMC on September 1 performing within 2-3% of simulations, and Yee Jiun Song, the VP overseeing Meta's custom silicon program, said each generation delivers better performance per watt and per dollar than current Nvidia offerings running the same AI models. Meta is developing the chips with Broadcom on design and TSMC on manufacturing, part of a broader push to diversify its silicon supply and reduce reliance on Nvidia.

2026.09.16
links.tldrnewsletter.com4 min readTLDR Hardware

ASTERA LABS' NEW LEO X CONTROLLER LETS MEMORY EXPANDERS PLUG DIRECTLY INTO AI ACCELERATOR FABRIC NETWORKS

Astera Labs' second-gen Leo 2 CXL memory controllers move from PCIe Gen5/CXL 2.x to Gen6/CXL 3.2 and double memory channels from 2 to 4, doubling both bandwidth and total capacity to up to 768GB of DDR4 or 4TB of DDR5 per expander, with new RAS features like hardware DIMM testing and soft-ECC on top of hardware ECC, aimed at economically reusing cheap, aging DDR4 pulled from retired servers rather than buying expensive new DDR5 amid the ongoing memory shortage.

2026.09.16
servethehome.com5 min readTLDR Hardware

MICRON SHOWS OFF 512GB DDR5 RDIMM, PACKING 12TB INTO A DUAL-SOCKET SERVER

Micron demonstrated a 512GB DDR5 RDIMM, doubling per-slot capacity by vertically stacking DRAM dies inside the package and connecting them with through-silicon vias, the same die-stacking technique HBM uses, letting a 24-slot dual-socket server hold 12TB without changing the module count or footprint.

2026.09.16
storagereview.com3 min readTLDR Hardware

TESLA SETS OCTOBER 1 UNVEILING FOR THE LONG-DELAYED SECOND-GEN ROADSTER

Tesla posted "Go for launch" with a "10.01" graphic, pointing to an October 1 reveal of the second-generation Roadster, first announced back in November 2017 and delayed repeatedly since. The Information recently reported the original 2017 design has been scrapped entirely for a new, unrevealed one, and the teaser graphic hints at the promised SpaceX-developed cold gas thrusters meant to let the car briefly fly. Even after next month's unveiling, Musk has previously said the Roadster won't reach actual production for another 12 to 18 months.

2026.09.14
techcrunch.com2 min readTLDR Hardware

ROBERT FRIEDLAND WARNS COPPER SUPPLY CAN'T KEEP UP WITH AI, EVS, AND REINDUSTRIALIZATION

Ivanhoe Mines founder Robert Friedland, on Bloomberg's Odd Lots podcast, argued the world isn't close to having enough copper for what's coming, with copper up 15% this year as demand from data centers, electric vehicles, and global reindustrialization compounds on top of the metal's already-tight supply. He framed US tariffs as a potential fix rather than just a cost, arguing that if a 30% tariff pushed copper from $6 to $7.80 a pound, domestic mining could again afford unionized US labor, pointing out that 10% of all copper ever mined came from Arizona before manufacturing and refining capacity was largely outsourced to China.

2026.09.14
links.tldrnewsletter.com5 min readTLDR Hardware

GEORGIA TECH RESEARCHER REVERSE-ENGINEERS APPLE'S NEURAL ENGINE FROM SILICON TO FIRMWARE

Every Apple chip since the A11 and M1 ships a Neural Engine that runs on-device AI across over 2.5 billion active devices, yet Apple has never published an instruction set, driver interface, or way to even confirm a computation ran on it. This guide bypasses Core ML entirely, reaching the ANE directly through the compiler, runtime, kernel driver, and firmware via a mix of live silicon measurement and static/dynamic reverse engineering, cross-checking both methods against each other.

2026.09.14
ane-guide.readthedocs.io5 min readTLDR Hardware

AMAZON BETS ON HOLLOW CORE FIBER TO KEEP A MILLION CHIPS IN SYNC

AWS is wiring AI data centers with hollow core fiber, an air-filled strand that moves light about 50% faster than glass and cuts latency roughly 30%. Amazon has already deployed it across more than 10 sites so million-chip clusters can stay synchronized as facilities spread out for power and land.

2026.09.14
en.sedaily.com5 min readTLDR Hardware

IONQ DEBUTS SIXTH-GEN SUPERION QPU, MOVING QUBIT CONTROL ONTO THE CHIP ITSELF

IonQ unveiled its sixth-generation quantum computing platform, Superion 256, built around Electronic Qubit Control (EQC), integrating the control electronics directly onto CMOS silicon using standard high-volume semiconductor fabrication rather than the racks of external control electronics trapped-ion systems normally depend on. That integration is what shrinks the whole system down to fit in a standard server rack running on typical data center power and cooling, while significantly cutting cost-per-qubit.

2026.09.14
ionq.com3 min readTLDR Hardware

RPI AND IBM PROPOSE REACH, A MEMORY CONTROLLER THAT CUTS HBM ERROR-CORRECTION OVERHEAD FOR LLM INFERENCE

Stronger HBM error correction usually means long-span codes that protect more efficiently, but coupling every small memory access to span-wide state and decoding at full HBM bandwidth is normally too costly to implement directly. REACH sidesteps that by exploiting a property specific to LLM decode workloads: reads are sequential, so they can be aggregated across a span cheaply, while writes are sparse, keeping parity-update traffic low. The controller uses standard inner codes to fix common errors and flag whichever chunks it can't resolve, then falls back to a long outer code reserved specifically for repairing those flagged erasures, avoiding the cost of running heavy decoding on every access.

2026.09.14
semiengineering.com2 min readTLDR Hardware

SPACEXAI QUIETLY BUILDS WHAT MAY BE THE NATION'S LARGEST GRID BATTERY AT ITS MEMPHIS SITE

Satellite imagery from July shows 720 Tesla Megapacks installed at SpaceXAI's Colossus 2 data center in Memphis, equating to roughly 2.8 gigawatt-hours and 720-1,400 megawatts depending on the exact model, with company and utility executives separately describing the completed system as 3.3 GWh or 2,000 MW of instant discharge, either of which would top the current largest US grid battery at California's Edwards & Sanborn plant.

2026.09.14
canarymedia.com5 min readTLDR Hardware

ZIPLINE'S DRONES NOW DIVE FOR SPEED AND CLIMB TO BRAKE, NO NEW HARDWARE REQUIRED

Zipline shipped a software update, "Dynamic Flight Trajectories," that replaces its drones' flat climb-cruise-descend flight profile with a wave shape: climb while already moving, dip to build speed, deliver from altitude, then climb again to shed speed before landing, letting gravity do work the motors used to handle. The company reports the median mission got 48 seconds shorter, left 40.6 watt-hours more in the battery on landing, and brought the hottest hover motor home 8.9°C cooler, all from a firmware download to the existing fleet.

2026.09.14
dronexl.co4 min readTLDR Hardware

APPLE DEBUTS IPHONE DUO, ITS FIRST FOLDABLE, AT $1,999

New CEO John Ternus led his first keynote, headlined by the iPhone Duo, Apple's first foldable, a passport-style device with a 5.4-inch outer display and a 7.6-inch inner display that automatically reflows apps based on how the phone is folded, starting at $1,999. Apple also unveiled the iPhone 18 Pro and Pro Max ($1,199/$1,299, up $100 from last year) with new A20 Pro processors, a variable mechanical aperture camera, and 24-30 hour battery life, alongside the Apple Watch Series 12 and Ultra 4, which add a 5-second heart rate sensor Apple claims is more accurate than a chest strap.

2026.09.11
finance.yahoo.com5 min readTLDR Hardware

OPENAI MAY DOUBLE-SOURCE NEXT-GEN AI CHIPS AT SAMSUNG, ALONGSIDE TSMC

OpenAI Korea GM Harrison Kim revealed the company has made real progress with Samsung Electronics on joint production and research for its next-generation chips, expanding a relationship that so far has been mostly about memory supply and enterprise software. OpenAI's current inference ASIC, Jalapeño, was designed with Broadcom and built at TSMC in under 18 months and is already in mass production there, so any Samsung involvement more plausibly targets its successor, which is reportedly approaching tapeout.

2026.09.11
tomshardware.com3 min readTLDR Hardware

TSMC'S AUGUST REVENUE JUMPS 53% TO A RECORD HIGH, PLANS HIGH-NA EUV ADOPTION FOR 2030

TSMC posted NT$514.8 billion ($16.1 billion) in August revenue, up 53.3% year-over-year and 10.1% from July, hitting a fresh monthly record as AI-driven chip demand keeps outpacing supply. It came alongside a separate announcement with Dutch equipment maker ASML: TSMC plans to bring ASML's High-NA EUV lithography into large-scale manufacturing starting in 2030 for its most advanced nodes, as AI accelerators push toward increasingly complex transistor architectures.

2026.09.11
cnbc.com3 min readTLDR Hardware

INTRODUCING CUDA RUST FOR HIGH-PERFORMANCE GPU KERNEL DEVELOPMENT

NVIDIA has announced CUDA Rust and introduced two specialized compilation tracks that allow developers to author and optimize GPU compute kernels directly in Rust. This release brings Rust's strong compile-time memory safety guarantees and modern tooling ecosystem to GPU-accelerated workloads. Developers implementing embedded vision, intensive sensor fusion, and real-time state estimation on mobile platforms can now exploit GPU compute power while minimizing memory-related crashes and runtime concurrency bugs.

2026.09.11
developer.nvidia.com5 min readTLDR Hardware

VLLM TRIPLES MINIMAX M3 INFERENCE THROUGHPUT ON AMD'S MI355X THROUGH SYSTEMATIC BOTTLENECK CHASING

AMD and Embedded LLM teams took MiniMax M3's day-0 vLLM implementation from 109.1 to 342.4 output tokens/s/GPU on AMD's Instinct MI355X, a 3.14x gain, by repeatedly asking five questions as the bottleneck kept shifting: what local shape reached each rank after tensor-parallel sharding, what work repeats every layer, which bytes actually move, whether a claimed fast path actually executed, and which queue grows once kernels stop being the limiter. Big wins came from fusing the model's shared expert into the same grouped GEMMs as routed experts (up to 30% faster at low concurrency), and reusing sparse-attention block selections across adjacent layers instead of recomputing per layer.

2026.09.11
vllm.ai6 min readTLDR Hardware

BESXAR FLIES FIRST SUBORBITAL SEMICONDUCTOR MANUFACTURING DEMO, BANKING ON FREE VACUUM IN SPACE

Besxar, a $9.35 million seed-funded startup, flew its first mission on July 5 aboard a SpaceX Falcon 9 alongside a Starlink batch, carrying its own semiconductor wafers plus test wafers from the University of Virginia and UT Austin. The company's pitch is simple: Earth-based fabs spend billions on pumps to create the vacuum semiconductors need, but in space vacuum is free, so Besxar wants to manufacture higher-quality, lower-cost wafers in orbit and return them to Earth to sell.

2026.09.11
payloadspace.com2 min readTLDR Hardware

GOOGLE COMMITS $15B TO FINLAND DATA CENTERS, EXPLORES NUCLEAR POWER FOR THE BUILDOUT

Google plans to invest at least $15 billion in Finnish AI infrastructure through 2028, its largest single investment in Europe, building on over 15 years of presence there. Alongside the buildout, Google signed a 22-year power agreement with Finnish utility Fortum and said it will explore business models that could support new nuclear reactors at Fortum's Loviisa site, tying the compute expansion directly to a long-term power source.

2026.09.11
cnbc.com3 min readTLDR Hardware

LG DENIES CLAIMS ITS SMART TVS CONSTANTLY RECORD AND UPLOAD AUDIO

Gamers Nexus published a two-hour investigation, done with security researchers MrBruh and uturn using packet capture and firmware analysis, alleging LG's smart TVs constantly log and upload user data even offline or in standby, scan local Wi-Fi networks for other devices, and record and store ambient audio through the TV's microphone for later retrieval, even when the TV appears off. LG called the claims untrue, saying its TVs only process voice data when the remote's voice button is pressed or its "Hi LG" wake word is detected after a user manually enables far-field voice recognition, with wake-word audio processed entirely on-device and deleted immediately if no match is found.

2026.09.11
tomshardware.com4 min readTLDR Hardware

CHINA CURBS HUMANOID IPOS AFTER UNITREE'S VOLATILE DEBUT

Chinese regulators are tightening approval for humanoid robotics startups seeking to go public, following a volatile trading debut from sector leader Unitree Robotics, as they try to cool a sector described as sizzling but rife with copycats and little real technological breakthrough. The China Securities Regulatory Commission's tighter posture is expected to affect other humanoid IPO hopefuls, reportedly including names such as AGIBOT and Galbot.

2026.09.09
theinformation.com3 min readTLDR Hardware

CALL FOR OPEN-SOURCE ROBOTICS BENCHMARKS

Robocurve, a public benefit company backed by YC, has opened a community bounty program to build reproducible open-source evaluation benchmarks for physical AI and autonomous robot systems. To make these benchmarks happen, it is giving away $20,000 and a pair of I2RT ABC Box robot arms to 25 teams who propose the best research benchmarks. Robocurve hopes to accelerate the development of Physical AI by increasing the number and diversity of dexterous robotics benchmarks.

2026.09.09
bounty.robocurve.org3 min readTLDR Hardware

BENCHMARKING LIDAR SLAM ON LIVOX MID-360 DATASETS

This open-source repository presents a comparative benchmark evaluating four popular LiDAR odometry and SLAM algorithms—KISS-ICP, DLIO, FAST-LIO2, and GLIM—on live aerial drone flight data collected with a Livox Mid-360 LiDAR and Pixhawk flight controller. Running directly on a Jetson Orin NX embedded computing platform, the project compares trajectory accuracy and map reconstruction quality against millimeter-grade VICON ground-truth data. These empirical findings provide developers and perception engineers with actionable performance data to choose and tune state estimation algorithms for compute-constrained autonomous platforms.

2026.09.09
github.com3 min readTLDR Hardware

XIAOMI 18 FOLD BEATS APPLE'S IPHONE FOLD TO MARKET BY A DAY, PACKING A TELEPHOTO LENS APPLE REPORTEDLY SKIPS

Xiaomi's 18 Fold launched in China a full 24 hours ahead of Apple's rumored iPhone Fold, packing a wide-style, passport-sized chassis with a 5.38-inch outer OLED and 7.58-inch inner display, both hitting 4,000 nits peak brightness with a 1-120Hz LTPO refresh rate and, notably, an RGB sub-pixel structure that's a first for any foldable. Xiaomi prioritized battery over thinness, fitting a 6,000mAh cell (with 67W wired charging) into a 5.02mm-unfolded body that's still slightly thicker than Samsung's Galaxy Z Fold8.

2026.09.09
stuff.tv3 min readTLDR Hardware

ROBOTOK: AN INTERNET-SCALE DATA ENGINE FOR HUMAN DEMONSTRATION VIDEO RETRIEVAL AND DEXTEROUS MANIPULATION LEARNING

Data collection for Physical AI is an expensive problem blocking the development of more advanced models. To get around this bottleneck, researchers at Rice University have developed RoboTok, an automated data engine that indexes, filters, and processes large-scale human demonstration videos from the internet to train manipulation policies. By extracting hand trajectories and object interactions directly from uncurated web video, the system learns complex, multi-stage manipulation skills without requiring physical teleoperation. This approach showcases how massive online video datasets can be converted into scalable training data to overcome physical data collection bottlenecks.

2026.09.09
rice-robotpi-lab.github.io5 min readTLDR Hardware

ORION: A DIY 12-DOF QUADRUPED ROBOT DOG

Orion is an open-source, 12-degree-of-freedom quadruped robot designed, fabricated, and programmed from scratch using ROS 2, custom 3D-printed components, and serial bus servos. The repository provides complete mechanical CAD models, kinematic calculations for dynamic walking gaits, and step-by-step firmware integration guides. This project provides engineering teams and builders with an accessible reference architecture for prototyping low-cost legged locomotion platforms.

2026.09.09
github.com3 min readTLDR Hardware