Computer Engineering News, Latest Updates and Recent Announcements on Techminis

A naukri.com initiative

New

Home

Computer Engineering

Netflixtechblog

315

Image Credit: Netflixtechblog

Introducing Configurable Metaflow

Metaflow has introduced a new feature, Configs, to complement the existing Metaflow constructs of artifacts and Parameters, by allowing you to configure all aspects of the flow, decorators in particular, prior to any run starting.
The new feature allows a flow to be configurable using configuration files, so variants can be defined without changing the code. Configurations can be used more widely in flow code and can be used to set defaults for parameters.
Metaflow Configs are all stored as artifacts by Metaflow but differ in when they are persisted. Configuration parameters are resolved and persisted at the start of a run, while Metaflow Configs are resolved and persisted when the flow is deployed.
Configurations can be read from a pleasantly human-readable configuration file and specifications can be made for a triggering ‘@schedule’ and ‘@resource’ requirements.
Metaflow has always supported sweeping over parameter grids easily using foreaches, but altering the flow itself, for instance, to change @resources or @pypi/@conda dependencies for every experiment, hasn’t been easily possible. Configuration managers like Hydra allow for orchestration of experiments over multiple configurations or sweeping over parameter spaces.
Metaflow has partnered with Netflix’s Metaboost, an internal Netflix CLI tool that helps ML practitioners manage, develop and execute their cross-platform projects, to provide a single interface to three internal platforms at Netflix that manage ETL/Workflows (Maestro), Machine Learning Pipelines (Metaflow) and Data Warehouse Tables (Kragle).
Metaboost uses a configuration system that combines GIT-based parameters, global configurations and arbitrarily bound configuration files for use during execution against internal Netflix platforms.
Metaflow has made it incredibly simple to integrate this configuration system with Configs. Users can simply add a mix-in class to their FlowSpec and reference the configuration values in steps or decorators.
Configs allow for reproducible, consistent, low-boilerplate and easily configurable experiments and robust production deployments.
Configs turn into dictionary artifacts, they get versioned and persisted automatically as artifacts. You can access Configs of any past runs easily through the Client API.

Read Full Article

18 Likes

Physicsworld

854

Image Credit: Physicsworld

Hartmut Neven talks about Google Quantum AI’s breakthrough in quantum error correction

Hartmut Neven and colleagues at Google Quantum AI have been awarded half of the Physics World 2024 Breakthrough of the Year.
They have implemented quantum error correction below the surface code threshold in a superconducting chip.
Google's new Willow quantum processor integrates 105 superconducting physical qubits.
Google aims to create a processor with 100 or 1000 logical qubits by 2030.

Read Full Article

7 Likes

Physicsworld

340

Image Credit: Physicsworld

Mikhail Lukin and Dolev Bluvstein explain how they used trapped atoms to create 48 logical qubits

Mikhail Lukin, Dolev Bluvstein and colleagues receive half of the Physics World 2024 Breakthrough of the Year award.
They demonstrated quantum error correction on an atomic processor with 48 logical qubits.
In the podcast, Lukin and Bluvstein discuss the role of error correction in developing practical quantum computers.
They explain how atoms were moved in their quantum processor to create logical qubits and perform quantum computations.

Read Full Article

20 Likes

Fyfluiddynamics

282

Image Credit: Fyfluiddynamics

Inside the Squirting Cucumber

The squirting cucumber, despite being only 5 cm long, can spray its seeds up to 10 meters away.
The plant prepares for launch by moving water from the fruit into the stem, reorienting the cucumber at a 45-degree angle.
The squirting cucumber releases its seeds in an explosive burst, propelling them at speeds up to 20 m/s.
With its rotation and multiple fruits in different orientations, the plant spreads its seeds in an even ring around itself.

Read Full Article

17 Likes

COSMOS

377

Image Credit: COSMOS

Watch swarms of ant-like microrobots lift heavy objects and more

Nanoengineers have designed tiny magnetic robots that swarm together like ants to pick up objects many times their size or even form rafts to float on water.
The robots assemble through magnetic attractions and respond to and be guided by magnetic fields.
They can climb obstacles, transport cargo, and unclog tubes resembling blocked blood vessels.
However, they require higher levels of autonomy for real-world applications in complex or confined spaces.

Read Full Article

22 Likes

Knowridge

246

Image Credit: Knowridge

Dead stars reveal secrets of extreme gravity and dark matter

Scientists have confirmed a long-predicted effect in dying stars called white dwarfs: hotter white dwarfs are slightly puffier than cooler ones, even when they have the same mass.
Studying over 26,000 white dwarfs, this finding could provide insights into extreme gravity and potentially unveil the mysteries of dark matter.
White dwarfs are the leftover cores of stars that have burned through their nuclear fuel and are incredibly dense.
Understanding white dwarfs better could contribute to the hunt for dark matter and help refine theories about massive star evolution.

Read Full Article

14 Likes

Knowridge

Image Credit: Knowridge

Biological computers could use far less energy than current technology by working more slowly

Redesigning computers may be the key to high-speed computing that doesn't rapidly consume energy.
The root of the issue lies in the Landauer limit, which states that all computational tasks must expend 10-21 joules of energy.
To complete a computational task at the Landauer limit, it must be carried out infinitely slowly.
Operating at such a low energy level would necessarily eliminate concerns regarding electricity consumed by computing.
To achieve a biocomputing solution, it may be necessary to adopt a system that relies on biological motor proteins.
In a biocomputing network, a single computational task is split among thousands of tiny machines called biofilaments.
Each biofilament might be as much as a million times slower than a transistor, but it exactly matches the energy output the task demands.
Small biocomputers have been built by researchers, but scaling the technology remains an obstacle.
Experts suggest integrating with current technology and precisely controlling individual biofilaments could solve these problems.
Alternatively, neuromorphic computing uses novel computer hardware to emulate highly complex interconnections between human brain cells.

Read Full Article

5 Likes

Physicsworld

Image Credit: Physicsworld

New day dawns for quantum computing in the UK

The UK National Quantum Computing Centre (NQCC) located on the Harwell Campus in Oxfordshire has opened.
The building is designed to foster innovation and collaboration and has state-of-the-art experimental facilities.
Scientists and engineers from commercial quantum computing companies will work alongside NQCC technical teams to create a supply chain.
The hope is that a quantum cluster will be formed in and around the Harwell Campus in Oxfordshire.
The NQCC’s mandate is to build a community of hardware providers, software developers and end users to explore the value of quantum computing.
Applications engineers from the NQCC are working with Unisys and the University of Newcastle to explore how today’s quantum computers could be used to optimize the loading of cargo onto aircraft.
Further, the NQCC provides clear, trusted and impartial guidance to government, businesses and the public.
The National Quantum Technologies Programme has seen over £1bn of investment since 2014, helping to establish a collaborative culture.
Scientists and companies hope to translate scientific insights into innovative technologies, paving the way to deliver quantum computing at scale.
The NQCC strives to provide thought leadership on ethical considerations around quantum computing.

Read Full Article

2 Likes

Physicsworld

201

Image Credit: Physicsworld

Magnetically controlled microbots are small enough to diffract visible light

Researchers at Cornell University have developed magnetically controlled microbots that are small enough to diffract visible light.
The microbots utilize nanoscale hinges made using atomic layer deposition (ALD) and nanomagnet arrays for magnetic actuation.
The ability to control light at the microscale offers new possibilities for imaging and probing the microscopic world.
Potential applications include endoscopic imaging, high-resolution microscopy, and sensing magnetic fields and current in integrated circuits.

Read Full Article

12 Likes

Medium

Image Credit: Medium

Temporal Manipulation of Water Molecules Using Quantum Entanglement: A Framework for Retroactive…

This paper explores the theoretical and practical foundations for using quantum entanglement to encode and transmit information, specifically focusing on manipulating water molecules in the past.
The paper investigates the feasibility of retroactively encoding information into water molecules in historical timelines using quantum entanglement.
This research aims to determine if vibrational patterns induced in water molecules in the present can manifest as discernible effects, such as text or symbols, on water surfaces in the past.
This paper discusses the challenges, including decoherence, no-signaling constraints, and environmental interference, alongside proposals for overcoming these limitations.
The implications for quantum communication, causality, and historical interaction are profound, positioning this work at the cutting edge of speculative physics.
This paper outlines the theoretical and experimental groundwork for retroactive manipulation of water molecules using quantum entanglement.
The paper proposes that temporal entanglement can facilitate retroactive encoding of vibrational information into water molecules.
The paper discusses the physical and mathematical framework for encoding vibrational information into water molecules.
The paper concludes by outlining the scientific impact and ethical considerations of this research.
This research has profound implications for communication, causality, and the understanding of time itself.

Read Full Article

3 Likes

Physicsworld

302

Image Credit: Physicsworld

Opening doors with outreach: using your physics skills to engage, inspire and break down barriers

Outreach is vital for nurturing the next generation of physicists, promoting public understanding of science and building a skilled physics community.
Accessing shared resources and training will make starting out much easier.
The Physics Mentoring Project mentors school students to encourage them to continue studying physics, providing interactive training and online support.
Long-term strategic planning, target audience identification, and resource considerations are important for successful outreach.
Evaluation of outreach programs should be built in from the start. It doesn't always mean surveys and quantitative data.
Outreach helps to break down barriers to physics and reach those who may not normally engage with the subject.
External funding opportunities are available for engagement activities such as STFC's Spark awards and AI Enigneering's Ingenious grant scheme.
The Ogden Trust supports university physics outreach through funding, the Outreach Officer Network, and annual Outreach Awards.
The Taking a Strategic Approach to Outreach guide and annual leadership training course equip academics and teaching staff with skills to plan and deliver effective outreach.
Examples of successful outreach programs include the Physics Mentoring Project, Orbyts, and My Place, My Science.

Read Full Article

18 Likes

Physicsworld

265

Image Credit: Physicsworld

AI-powered tool detects residual tumour during brain surgery

FastGlioma is an artificial intelligence-based diagnostic system that can detect residual brain tumour that’s often missed during surgery. The developers of the new FastGlioma tool, at the University of Michigan and the University of California, San Francisco (UCSF), explain that it can predict if and the extent to which glioma remains in the brain while the surgical procedure is underway.
FastGlioma combines rapid, easy-to-use stimulated Raman histology (SRH) optical imaging with open-source visual foundation models (artificial intelligence models trained on massive, diverse datasets that can be adapted for a wide range of tasks) to perform a 10 s analysis of fresh tissue specimens in operating room suites.
The researchers also developed a rapid visualization strategy, called few-shot visualizations. By comparing feature similarity between the support set and the tissue sample being analysed, FastGlioma creates both a tumour-infiltration score and infiltration heat maps.
FastGlioma could detect and quantify the degree of tumour infiltration with an average accuracy of 92.1%. The tool maintained accurate tumour-infiltration scores despite significant cytological and histoarchitectural differences related to tumour grade, molecular genetics, treatment effect or WHO subtypes.
Another benefit is that the analytic speed of FastGlioma provides a rapid and scalable alternative to conventional intraoperative pathology methods.
FastGlioma significantly outperformed conventional methods, with only a 3.8% tumour miss rate, compared with a 24% miss rate.
The researchers also note that FastGlioma can accurately detect residual tumour for several non-glioma brain tumours.
Future research will focus on applying a similar workflow to other human cancers, including lung, prostate, head-and-neck and breast cancer.
FastGlioma represents the transformative potential of medical foundation models to unlock the role of artificial intelligence in care of patients with cancer.
In a prospective multicentre clinical study, principal investigators trained and validated FastGlioma to detect microscopic tumour infiltration in an international cohort of patients.

Read Full Article

15 Likes

Hackaday

339

Image Credit: Hackaday

Pulling Backward to Go Forward: The Brennan Torpedo Explained

The Brennan torpedo, invented in 1877 by Louis Brennan, was one of the first guided torpedoes.
It had no internal power source and operated by pulling a cable backward to propel the torpedo forward.
The torpedo had a pulley-type assembly with counter-rotating props and could be steered by adjusting the speed of the cables.
The Brennan torpedo was in service for over a decade before being replaced by torpedoes with internal power systems.

Read Full Article

20 Likes

Netflixtechblog

Image Credit: Netflixtechblog

Part 1: A Survey of Analytics Engineering Work at Netflix

This article highlights some of the analytics engineering work done at Netflix to empower the streaming giant to produce and deliver high-quality analytical insights and optimize member experience.
One area of focus is DataJunction, an open-source tool developed by Netflix that simplifies the metric creation process, bringing metric definitions to a central store that other teams can access. The tool also performs the dimensional join and filtering automatically.
Netflix is also democratizing analytics by developing LORE, a chatbot that uses language learning models to provide users with contextual insights and gain insights that were locked in dashboard products.
The company is leveraging Foundational Platform Data (FPD) to provide a centralized data layer for all platform data with a consistent model and standardized processing methodology. Cloud Efficiency Analytics (CEA) is built on top of FPD and provides time series efficiency metrics across various business use cases.
The Analytics Engineering is a key contributor to building Netflix's deep data culture. Netflix annually holds the Analytics Summit, where the community comes together to discuss analytical deliverables, build relationships, and share knowledge.
Netflix's analytics engineering work is essential to understanding what content will excite members and bring them back for more. Analytics Engineers build end-to-end analytical systems that translate business challenges into actionable insights.
The DataJunction tool heavily leverages the dimensional join feature to allow the Netflix Experimentation Platform to compare average streaming hours in different cells and reliably track data across different dashboards.
LORE, the chatbot developed by Netflix, allows internal users and business users to query contextualized data that previously required sophisticated knowledge of the underlying data models.
Foundational Platform Data provides a centralized data layer for all platform data and a standardized data processing methodology, and Cloud Efficiency Analytics offers an analytics data layer that provides insights into resource usage and costs across various granularities.
Analytics Engineering is a critical component of Netflix's culture and contributes to delivering optimized member experiences. Netflix promotes its Analytics Engineering conferences as a way to share knowledge, review progress, and discuss analytical deliverables.

Read Full Article

2 Likes

Netflixtechblog

350

Image Credit: Netflixtechblog

Cloud Efficiency at Netflix

Netflix uses Amazon Web Services (AWS) for their cloud infrastructure needs.
Netflix's ecosystem enables engineering teams to run applications and services at scale.
The democratization of highly curated data sources is crucial for the efficiency of Netflix’s services and teams.
Foundational Platform Data (FPD) and Cloud Efficiency Analytics (CEA) are the two components used by Netflix to provide a scalable analytic solution.
The idea is to enable engineering partners to understand their resource usage effectively and efficiently, promoting cost consciousness decision-making.
CEA provides time series efficiency metrics across various business use cases after consuming inventory, ownership, and usage data.
Data correctness, trust, and SLAs are crucial while ensuring data completeness; Netflix continuously iterates audits and receives feedback to refine this process.
FPD establishes data contracts with producers to ensure data quality and reliability; these contracts allow Netflix to leverage a common data model for ownership and usage.
Moving towards proactive approaches via predictive analytics and machine learning is the long-term goal for Netflix.
Netflix plans to cover nearly complete cost insight coverage in the upcoming year and extend FPD to other areas of the business such as security and availability.

Read Full Article

20 Likes

For uninterrupted reading, download the app

Computer Engineering

Discover more