Road projects, golf, business updates: 6 trending Austin-area stories  Community Impact Newspaper

Trending Topics: Who has been the most surprising player this preseason?  NBA.com

SMU trending with 4-star defensive target: 'They're definitely leading now'  On3

Google’s top trending Halloween costumes revealed — and you may have never heard of them  New York Post

Trending stocks this week as tech selloff deepens, volatility surges  Seeking Alpha

1. Spotify’s ‘QR code verification’ cracks down on sharing and could ban you for a year  9to5Google
2. Spotify bans Family members who miss QR scan from other plans for 12 months  PPC Land
3. The End of Sharing Subscriptions with Friends? Spotify Has Devised a Clever Way to Expose Fake Families  Root-Nation.com
4. Spotify is cracking down on Family plans again, and the penalty can lock you out for a year  Digital Trends
5. Spotify has tightened its controls on shared subscriptions: violators will be blocked from accessing Premium Family and Duo for one year  Mezha

1. Apple Announces Oct. 28 Event for Touch-Screen MacBook Pro  Bloomberg.com
2. Apple announces product launch event for Oct. 28. Analysts expect touchscreen laptop  CNBC
3. Apple to host 'Pro and tell' event on October 28 in LA  Reuters
4. Apple’s first touchscreen MacBook is days away – with another OLED surprise  ZDNET
5. OLED iPad mini and touchscreen MacBook Pro: What to expect from 'Pro and Tell'  AppleInsider

1. iPhone Duo battery specs revealed: Good news for Galaxy Z Fold 8 and other Android rivals  Android Authority
2. Apple reveals iPhone Duo battery size, screen hardness details  Mashable
3. Here’s how iPhone Duo’s battery capacity compares to other iPhones  9to5Mac
4. iPhone Duo battery capacity lands in between iPhone 18 Pro & Pro Max  AppleInsider
5. iPhone Duo has a bigger battery than Galaxy Z Fold 8 and Pixel 11 Pro Fold  9to5Google

1. Google is linking your phone number to accounts you wanted to keep separate  Android Central
2. PSA: Google is silently linking your phone number across all signed-in accounts  Android Authority
3. Google quietly ties your phone number to all your signed-in accounts  Android Police
4. ‘I'm so sick of this:’ Google has started automatically linking your phone number to all your accounts, and Android users are calling it a privacy nightmare  Yahoo Tech
5. Google links phone numbers to all accounts on Android devices | turning off settings won't remove existing links | Inshorts  Inshorts

1. Android Auto could soon make commute part of your workday, whether you like it or not  Android Authority
2. Android Auto ‘Work Account’ support would allow for easy switching in work vehicles  9to5Google
3. All Users Can Now Download Android Auto 18.0 As "Cielo" Update Is Coming  autoevolution
4. Android Auto Could Soon Bring Work Apps to Your Car’s Screen  Android Headlines
5. Upcoming updates mean Android Auto will no longer lock your work life away  Android Police

Rockets vibrate, shake, and flex intensely as they climb into space, and spacecraft packed with delicate scientific instruments must survive those forces. Before any rocket leaves the ground, NASA engineers work to predict how those stresses will affect every part of a spacecraft.

To do that, engineers in the Dynamic Environments Group with NASA’s Launch Services Program Office at the agency’s Kennedy Space Center in Florida have streamlined how they ensure spacecraft and rockets can withstand harsh vibration, shock, stress, or other physical forces, strengthening mission safety in the process.

Using a commercial digital software tool in a new way, the engineers developed a standardized modeling workflow for finite element modeling to strengthen shock and vibration analysis and ensure hardware is prepared for the extreme environments of launch and ascent.

“This computer-based method helps predict how rockets and spacecraft behave under extreme conditions before hardware is built and helps identify weak spots and potential failure risks early so the team can address the findings,” said Brian Beaver, chief of the Flight Analysis Division with the Launch Services Program Office at NASA Kennedy.

The team recently used the software for the first time while certifying a commercial launch provider’s rocket selected for a NASA mission. The commercial partner provided vibration testing but not shock testing. The Dynamic Environments Group determined shock testing was needed and used a template the group created to run what-if comparisons, visualize structural behavior, and answer complex questions that would have required extensive manual modeling in the past. The team verified the company’s testing results for flight environments, and the analysis provided new insights addressing shock during transport.

“We are using a tool that’s existed for decades, but it’s not normally used by the dynamic environments disciplines,” said Alex Hardt, aerospace engineer with the Launch Services Program Office at NASA Kennedy. “But we’re changing that. We’re using it in a relatively new area of dynamics.”

Although rockets and spacecraft are designed to absorb shocks and vibrations, those forces can cause wear and tear or potentially flight failures, especially during launch and ascent.

“We define how severe the random vibration levels and shock levels are, and then we verify that all the parts going on future rockets have been tested to those levels, maximizing their longevity,” Hardt said.

Software time saver

For decades, engineers have estimated shock and vibration environments using rules of thumb or by scaling previous data to estimate how each rocket component or spacecraft equipment would be impacted once assembled and flown. The approach worked, but relied heavily on engineering judgment and simple approximations drawn from past missions.

Because most dynamic environments engineers use finite element modeling tools infrequently, they often need to re-learn complex workflows. The software replaces older tools and provides a standard template that guides engineers more quickly and thoroughly through vibration profiles and shock conditions.

Finite element analysis is a powerful simulation technique that breaks a complex structure into smaller components so computers can accurately predict how a spacecraft or rocket could respond to specific forces or launch environments. This is important because issues can be addressed on the ground; once the rocket leaves the launch pad, most problems cannot be repaired or corrected, making early identification essential.

“We’ve used it to get really valuable insight very quickly into verifying a provider’s system,” Hardt said. “It’s already provided some really deep insight into something that normally takes a lot of effort.”

Hardt says having the ability to visualize non-intuitive dynamics problems using finite element models also highlights the software’s valuable impact on learning, efficiency, and alignment with NASA’s people focused culture. The workflow can help the engineers share knowledge, refine assessments with additional flight data, and increase proficiency with modeling tools. This capability already has saved months by allowing the team to run analysis independently without having to request and wait for analysis reports or data, and by helping them identify risks that weren’t previously considered.

NASA’s Launch Services Program Office is responsible for launching rockets delivering spacecraft that observe the Earth, visit other planets and explore the universe, from weather satellites to space telescopes to Mars rovers. The team is constantly adopting new tools like this to give the agency the highest practical chance of success for each mission.

Research teams at NASA’s Glenn Research Center are involved in developing theory and software to predict the noise from aircraft engines. This software helps engineers to design and analyze new and improved aircraft propulsion systems. Copies of our software can be requested from NASA’s Software Catalog.

JeNo

Development of jet noise prediction codes is multifaceted. Since JeNo predicts jet noise using Lilley’s Acoustic Analogy and the mean flow solutions from a Reynolds-Averaged Navier-Stokes (RANS) code, accurate acoustic predictions hinge on accurate aerodynamic predictions. This is why acoustic source modeling and turbulence modeling, together with the development of advanced measurement systems, are the leading areas of research in jet noise prediction at NASA Glenn. RANS-based jet noise prediction schemes based on Lilley’s Acoustic Analogy offer relatively quick answers compared to those from alternative methods like Large-Eddy Simulations (LES) or Direct Numerical Simulation (DNS).

The JeNo jet noise prediction code calculates the far-field sound spectral density produced by axisymmetric jets at a user-specified observer location and frequency range. Currently, JeNo has been validated for subsonic and low supersonic, unheated axisymmetric jets. Copies of the JeNo code can be requested from NASA’s Software Catalog.

MGBK

Development of the MGBK code preceeded JeNo. Researchers at NASA GRC have used MGBK to predict trends in jet noise for complicated aircraft engine exhaust nozzles. MGBK was used to analyze the noise reduction benefits of chevrons–serrations on teh edge of turbofan nozzles–which are a familiar site on many commercial aircraft engines today. Copies of the MGBK codes can be requested from NASA’s Software Catalog.

V072

Predicated on simplistic models of the fan flow and geometry, V072 has been proven useful in studying rotor/stator interaction tone noise. When you need to design a low-noise fan stage, tone noise level predictions from V072 can help you make smart choices for such important design parameters as blade and vane counts, rotor-stator spacing, and stator vane sweep and lean. When you’re after quick noise estimates early in the design cycle, V072 may be your tool of choice. Copies of V072 can be requested from NASA’s Software Catalog.

TFaNS

Taking rotor/stator interaction noise prediction to the next level, not only can TFaNS calculate the tone noise levels generated by rotor/stator interaction it can also account for the blade row acoustic coupling effects such transmission and reflection and can propagate those tones to the far field. Like V072, TFaNS runs quickly and can be used in early design stages for studying trends. Copies of TFaNS can be requested from NASA’s Software Catalog

Linflux

When simple models are too simple, an analysis with Linflux gives you the power to study realistic, 3-D flows and blade geometries. Compared with V072 and TFaNS, Linflux requires significantly more computational resources since it models much more of the physics involved. Linflux predicts rotor/stator tone noise and, like all the software developed by the NASA Glenn Acoustics Branch, has been rigorously validated against experimental data. Copies of Linflux can be requested from NASA’s Software Catalog.

BFANS

Using estimates of turbulence intensity and characteristic length scale as input, BFANS can be used to calculate the broadband noise generated by turbulence impinging on the leading edges of the rotor blades and stator vanes. Calculations with BFANS, combined with experimental data from tests at facilities like NASA Glenn’s 9’x 15’ Low Speed Wind Tunnel, can be used to hunt down the noisiest components in a fan stage. Copies of BFANS can be requested from NASA’s Software Catalog.

RSI

RSI is a computer code for calculating the spectrum of broadband noise produced as a result of the interaction of the fan rotor wake turbulence with the fan exit guide vanes. The software computes the spectra of acoustic power upstream and downstream of the stator on a mode-by-mode basis at each frequency of interest given information about the incident turbulence characteristics. The target frequencies are arbitary and need not be harmonics of the blade passing frequency of the fan. Copies of RSI can be requested from NASA’s Software Catalog.

WOBBLE

Propeller noise can be strongly affected by any inflow that leads to unsteady blade loading. The WOBBLE code computes the tone noise associated with propellers with varying inflow angle and predicts the axial and circumferential directivity. Inflow angle is specified in terms of yaw, pitch, and roll angles of the aircraft. The acoustic model is exact and accounts for all unsteady sources such as thickness and loading effects from the blades. Copies of WOBBLE can be requested from NASA’s Software Catalog.

MPT

Multiple pure tone noise (aka “buzz saw” noise), created by rotor locked shocks present when the fan blade tip speed is supersonic, can be estimated with the MPT code. The MPT codes also use flow predictions from a Computational Fluid Dynamics (CFD) code as input. This tool can be used by engineers to study the real-world effects of blade-to-blade difference caused during fabrication (manufacturing variations) or during hot-running operation (blade setting angle and untwist variations). Copies of the MPT code can be requested from NASA’s Software Catalog.

ADPAC

Developed by the Allison Engine Company under contracts with Glenn Research Center, ADPAC solves tightly coupled internal/external flows through future-concept short-duct turbofan engines. ADPAC is one of our flexible validated CFD tools. Characteristics of the problems that can be modeled with ADPAC include: 2D/axisymmetric/3D/mixed, steady/unsteady, inviscid/viscous, rotating/non-rotating/mixed components, with bleeds, leakages, and injection. ADPAC was developed specifically for propulsion system aerodynamic analysis Copies of the ADPAC code can be requested from NASA’s Software Catalog.

Contact Us

Need to reach us for more information about our research, software, or data? Get in touch via the link below.

Research teams at the NASA Glenn Research Center perform experiments in world-class facilities such as the 9- by 15-Foot Low-Speed Wind Tunnel (9 x15 LSWT), the Aero-Acoustic Propulsion Laboratory (AAPL), the Small Hot Jet Acoustic Rig (SHJAR), the Nozzle Acoustic Test Rig (NATR), the DGEN AeroPropulsion Research Turbofan (DART), and the Acoustical Testing Laboratory (ATL).

9×15 Low-Speed Wind Tunnel (9×15 LSWT)

The 9- by 15-Foot Low Speed Wind Tunnel is specially designed to give researchers a controlled environment to study aircraft engine aerodynamics and acoustics at takeoff and landing conditions. This helps us to reduce noise in communities near airports. Major modifications were completed in 2019 to improve the flow and lower the background noise of the tunnel. A new 5-sensor traversing microphone rake helps measure sound produced by scale experimental propulsors. Read more about the capabilities of the 9 x 15 LSWT on our facilities page.

Aero-Acoustic Propulsion Laboratory (AAPL)

The Aero-Acoustic Propulsion Laboratory (AAPL) is a dual-purpose 65-foot radius geodesic dome used to reduce community noise and perform aeroacoustic tests of aircraft engines and engine components. The test area inside the dome can be configured as a hemi-anechoic or anechoic chamber. There are no echoes in the dome during tests because 17,000 custom-designed 2-ft thick fiberglass wedges are mounted to the inside surface of the dome walls and placed on the floor and around test rigs to absorb sound and prevent sound reflections. When AAPL was constructed in 1993, it was the first known free-standing geodesic dome to include materials to reduce transmission of sound through the walls and absorb sound within it. AAPL currently houses the Nozzle Acoustic Test Rig (NATR), Small Hot Jet Acoustic Rig (SHJAR), and the DGEN 380 Aeropropulsion Research Turbofan (DART). Read more about the capabilities of the AAPL on our facilities page.

Small Hot Jet Acoustic Rig (SHJAR)

SHJAR is one of three rigs installed inside the anechoic dome of the Aeroacoustic Propulsion Lab. It can be used to test single stream nozzles up to 5 cm in diameter, Mach 2.0 and 800 K. SHJAR is a key tool in connecting the dots between nozzle geometry, jet flow, and noise. Creative nozzle designs are inexpensively created for this rig, and their impact on flow and noise is measured with world-leading measurement techniques. Researchers at NASA Glenn have been using SHJAR to develop advanced flow measurement techniques like Rayleigh Scattering, Raman Scattering, digital Schleiren, and real-time Particle Image Velocimetry. Three-dimensional microphone arrays have been deployed to map the sources of the noise. These techniques hold the promise of unlocking some of the mysteries of how the turbulence caused by the nozzle creates noise. The data from this rig is used to validate the computer codes that are being developed to calculate jet noise. Better noise prediction codes will help engineers design quieter nozzles.

Among the many firsts in turbulent jet research produced on SHJAR, time-resolved measurements of turbulent temperature were recently made using Raman-based optical methods, achieving high-frequency measurements of thermodynamic temperature fluctuations needed to complete turbulence models used in heat transfer as well as jet noise. In 2006, researchers had used the first time-resolved particle image velocimetry to measured space-time correlations of turbulent velocities in hot jets, a flow statistic key to the production of noise. Time-resolved velocimetry was also used to map convection speeds of turbulence, another flow metric required by jet noise theory but previously unavailable in hot jets.

The SHJAR is not only a good place to develop advanced diagnostic techniques, it has been a good testbed for exploring canonical problems in jet noise such as how the installation of the jet on an aircraft changes the noise produced by the jet. Jet-surface interaction, where the plume of the jet interacts with the wings, flaps, and tail of an aircraft, was identified as a fundamentally different noise mechanism in SHJAR tests.

Most of these ground-breaking measurement techniques and practical noise insights first developed on the relatively simple SHJAR have been deployed on the high-fidelity Nozzle Acoustic Test Rig (NATR), co-located in the Aero-Acoustic Propulsion Lab. Read more about the capabilities of the SHJAR on our facilities page.

Nozzle Acoustic Test Rig (NATR)

The Nozzle Acoustic Test Rig (NATR), a free-jet wind tunnel, plays a key role in evaluating jet noise reduction concepts so that only the most promising concepts are selected for full-scale engine tests. Scale models of nozzles from both military and commercial aircraft engines can be tested in the NATR, which is one of three test rigs housed in the NASA Glenn Aero-Acoustic Propulsion Lab. The NATR can simulate take-off and landing flight conditions up to Mach 0.35. A jet engine simulator with up to three streams of air at a variety of pressures and temperatures simulate the flow of a turbofan engine to evaluate advanced nozzle concepts. The arc of microphones overhead simultaneously measures the far-field noise at all angles, allowing accurate translation of measurements to those of an actual aircraft flying overhead. The NATR enables engineers to evaluate a number of nozzle noise reduction concepts at a fraction of the cost and time required for a full-scale engine test.

Over its 30-year history, NATR has been the origin of many important jet noise breakthroughs. First built to support jet noise testing for the High-Speed Research program on commercial supersonic airliners, the facility has hosted many test programs led by NASA and industry. The chevron nozzles seen flying on today’s aircraft were first ‘discovered’ during a NASA-led test campaign in 1997. Over the years industry has used the rig to develop noise prediction methods for near-term designs. NASA has explored low-noise concepts for engines that are not yet available, such as engines with dual-fan systems. Recent tests have measured the noise likely to be heard from supersonic airliners, reducing the risk that these unique aircraft will not be allowed to fly from current airports. The tests also point to technologies to reduce their noise, allowing the aircraft designers to make the aircraft more efficient and environmentally acceptable.

The ability to test nozzle with flight is critical, as this has a strong impact on the noise produced by the relative velocity of the jet and the surrounding air, and it is a key feature allowing noise measurements to be directly transformed to what would be heard from a real airplane flying overhead. The large size of the anechoic space around the NATR is also critical to measuring the true acoustic far-field of the jet nozzle. The NATR in the Aero-Acoustic Propulsion Lab is one of the very few facilities in the world where these measurements are possible, and the only facility where both flow diagnostics and acoustics can be made on the same rig, a key capability in understanding the relationship between jet flow and noise. Read more about the capabilities of the NATR on our facilities page.

DGEN AeroPropulsion Research Turbofan (DART)

DGEN 380 is the primary component integrated into the NASA Glenn Research Center’s DGEN Aero-propulsion Research Turbofan (DART). It serves as a flexible, relevant, experimental aeroacoustic and aero-performance test bed, providing valuable information throughout NASA’s technology maturation process for a variety of multidisciplinary research that includes turbofan noise, engine controls, installation effects, and measurement technologies.

Recent tests were completed August 2019 to develop a better understanding of propulsion-noise sources to enable improved turbofan noise-prediction methods and noise-mitigation techniques. Simultaneous acoustic measurements were obtained using a circumferential infinite-tube-probe array at the core-nozzle exit in conjunction with sideline and farfield microphone arrays. The extensive data set will allow for the application of advanced source-separation and phased-array methods to elucidate not only the core-noise structure, but also the propagation characteristics of other propulsion noise sources. Read more about the capabilities of the DART on our facilities page.

Acoustical Testing Laboratory

The Acoustical Testing Laboratory at NASA GRC is used to support NASA’s human space exploration and aeronautics research missions. The ATL is a versatile test chamber that can be configured as an anechoic or hemi-anechoic chamber. Dense rubber material in the walls and ceiling of the ATL absorb or reflect sound from the outside. 34-inch deep fiberglass wedges that line the walls and floor are designed to absorb 99% of incident sound above 100 Hz which helps create an ideal environment for making acoustic measurements. A separate control room is attached for any noisy equipment or persons needed to perform the measurements.

The ATL has been used to quantify the noise emissions of experiments performed on the International Space Station, to measure noise sources and levels from electric motors, to calibrate microphone equipment, to measure the effectiveness of acoustic liners for aircraft engine models and to evaluate the over-the-wing engine placement for reducing takeoff and landing noise. Recent tests in the ATL have also included a scale model of a supersonic aircraft engine inlet and a spacecraft cabin ventilation fan.

To get a copies of NASA technical reports related to research in the Acoustical Testing Laboratory, please visit the NASA Technical Reports Server (NTRS).

Contact Us

Need to reach us for more information about our research, software, or data? Get in touch via the link below.

The four crew members in NASA’s simulated Mars mission have spent the past year living and working in isolation in a confined habitat at NASA’s Johnson Space Center in Houston. For the past 365 days, the Crew Health and Performance Exploration Analog (CHAPEA) Mission 2 crew members have operated as though they were on the Red Planet, growing crops, managing limited resources, navigating communication delays, and conducting suited extravehicular activities on a simulated Martian surface.

Crew members include Ross Elder, mission commander; Ellen Ellis, medical officer; Matthew Montgomery, science officer; and James Spicer, flight engineer. They entered their 1,700 square foot 3D-printed habitat Oct. 19, 2025. Leading up to the one-year mark of their 378‑day mission, science teams at NASA prepared to process health and performance data collected on the crew. This critical data can help NASA balance priorities and develop methods that counter spaceflight risks during future long-duration missions on Mars.

“We are extremely proud of the crew for their commitment and achievements over the last year on such a fast-paced, highly demanding mission,” said Grace Douglas, CHAPEA principal investigator. “The entire CHAPEA team is excited to congratulate the crew in person.” 

The crew will exit their habitat on Saturday, Oct. 31, to conclude their mission. They will have a chance to reconnect with their families, who they’ve only spoken with via email over the course of their mission with a Mars-like communication delay. The crew will then begin two weeks of post-mission evaluations and debriefs with NASA scientists.

These debriefs and evaluations will help scientists learn what it takes for crews to adapt to the routines and demands of daily life on long-duration missions, giving NASA insight into what future explorers may need to stay healthy, productive, and connected during missions far from Earth.

For example, data could shed light on how limited food and exercise systems will influence NASA’s ability to achieve mission objectives, including spacewalks on the surface of Mars. Some data also could help NASA foster autonomy and build decision-making support systems that best aid crew with communication delays lasting up to 45 minutes each way.

“This mission has been incredibly valuable in starting to teach us how to understand requirements for year-long missions, including medical and psychological care, via communication delays,” said Michael Stenger, chief scientist for NASA’s Human Research Program. “When combining this Mars-realistic mission analog with a suite of physiological, psychological, and performance measures in a rigorously controlled environment, we’ll be able to translate the knowledge from these experiences to Moon Base and beyond.”

To watch the CHAPEA Mission 2 egress ceremony on Oct. 31, visit:

https://www.youtube.com/live/2MONMgAmXNU

____

NASA’s Human Research Program 

NASA’s Human Research Program pursues methods and technologies to support safe, productive human space travel. Through science conducted in laboratories, ground-based analogs, commercial missions, the International Space Station and Artemis missions, the program scrutinizes how spaceflight affects human bodies and behaviors. Such research drives the program’s quest to innovate ways that keep astronauts healthy and mission ready as human space exploration expands to the Moon, Mars, and beyond. 

To measure airflow and sound, research teams at the NASA Glenn Research Center use special instrumentation like particle image velocimetry, hot wire anemometry, Raman spectroscopy, microphones in phased arrays or continuously rotating rakes. The global aviation noise reduction community relies on data from NASA GRC experiments to validate theory and noise prediction software.

Rotating Rake Mode Measurement System (RRMS)

Hear an annoying tone from your turbofan engine? Engineers might prescribe an advanced liner or specially tuned active noise control system to quiet it down.

To get it right, though, more information about the modal content of the tone is needed. This kind of data can’t be measured with a stationary microphone, which is why NASA Glenn engineers pioneered procedures using rotating microphones in the 1990s.

Rotating Microphone Rakes have been used by NASA Glenn researchers in a wide variety of tests ranging from low-speed proof of concept tests in the Advanced Noise Control Fan Rig and 9- by 15-Foot Low Noise Wind Tunnel to full-scale turbofan engine ground tests.

Microphone Phased Arrays

In a phased array system, microphones are placed in a 1D, 2D or 3D arrangement and when sound measurements from these microphones are properly processed, noise sources can be pinpointed.

NASA Glenn has an Array48 microphone system that has been used extensively to locate noise sources in scale-model jet flows, a counter-rotating open rotor fan model, and even a full-scale F404 military engine. Through NASA funding, a method has been developed to use microphones embedded in the fan duct of an engine to determine the location of noise sources on the rotating blades.

Hotwire, Hotfilm, and Laser Doppler Velocimetry

We know that annoying tone noises can be created by periodic disturbances, such as when the wake from a fan blade interacts with a downstream stator vane. We also know that broadband noise is generated when an aircraft engine ingests turbulent air. Therefore, we must be able to record detailed velocity measurements of the air passing through an experimental propulsor so we can identifying noise sources and work to mitigate them. We must use special instrumentation like hotwire probes and Laser Doppler Velocimeters (LDV) to measure the air velocity upstream, downstream, and between the rotor and the stator of an experimental propulsor. Plots of the velocity measurements can be made which allow us to visualize the disturbances that the rotor created. Hotwire and LDV data are routinely used to validate calculations used to estimate noise from turbofan engines.

Contact Us

Need to reach us for more information about our research, software, or data? Get in touch via the link below.

Mayor Zohran Mamdani said Monday night that New Yorkers must resist the narrative that an AI takeover is inevitable and demand regulations that protect the public and put power back in their hands.

In August, Nscale and Anthropic announced a $45 billion compute deal. The two companies are reportedly discussing yet another deal of the same magnitude.

AI labs have staked their reputations on model benchmarking — the supposedly neutral and accurate system of evaluations that are designed to measure whether one AI product is better than another. However, the methods behind these evaluations are complicated and still maturing, said Irene Solaiman, Hugging Face’s chief policy officer, during a discussion at TechCrunch […]

Microsoft wants enterprises to closely watch the agents that employees use with systems not controlled by AI labs like OpenAI and Anthropic.

Amazon wants to break Nvidia’s hold on PyTorch, the free toolkit used to build most AI apps, so that more developers run their apps on alternative chips like its own.