Sarvam AI

Sarvam AI is an Indian artificial intelligence company headquartered in Bengaluru, Karnataka. Founded in 2023, the company develops large language models (LLMs) and multimodal AI systems with a focus on Indian languages and region-specific use cases. The company has received venture capital backing and has participated in government-supported AI initiatives, including India's sovereign large language model programme under the IndiaAI Mission. == History == Sarvam AI was founded in August 2023 by Vivek Raghavan and Pratyush Kumar, who were previously associated with AI4Bharat at the Indian Institute of Technology Madras. In December 2023, the company announced a combined seed and Series A funding round of approximately US$41 million. The round was led by Lightspeed Venture Partners, with participation from Peak XV Partners and Khosla Ventures. In April 2025, the Ministry of Electronics and Information Technology (MeitY) selected Sarvam AI as one of the companies to develop an indigenous foundational model under the IndiaAI Mission. As part of the initiative, the company received access to government-supported computing infrastructure, including GPUs allocated for model training over a specified period. In February 2026, Sarvam AI introduced two large language models at the AI Impact Summit held at Bharat Mandapam, New Delhi. == Products and technology == Sarvam AI develops language models trained on datasets that include multiple Indian languages and code-mixed text. The company uses mixture-of-experts (MoE) architectures in some of its models. === Foundational language models === On 18 February 2026, the company announced the release of two foundational models: Sarvam-30B – A 30-billion parameter model based on a mixture-of-experts design. According to company disclosures reported by the media, the model activates approximately 1 billion parameters per token and supports a 32,000-token context window. Sarvam-105B – A 105-billion parameter model activating approximately 9 billion parameters per token, with a 128,000-token context window. The model is positioned for complex reasoning and enterprise applications. On 20th February 2026, the company released a beta version of the Sarvam-105B model which is named Indus. It is available on the Apple App Store, Google Play Store and the web. === Speech and vision systems === Sarvam AI has also developed multimodal systems including speech-to-text and vision-language models. Its speech model, referred to as Saaras V3 in company materials, supports multiple Indian languages. The company has also introduced a vision-language model known as Sarvam Vision, intended for document understanding and optical character recognition (OCR) in Indian scripts. === Devices === 'Sarvam Kaze' is an indigenous AI-powered wearable glass that listens, understands, and captures what users see the world through their eyes in real time. The device supports more than 10 Indian languages, enabling voice-based interaction and potentially real-time translation. The company plans to launch the device in May 2026. == Startup support == In March 2026, Sarvam AI launched the Sarvam Startup Program, an initiative providing selected early-stage companies with 6–12 months of API credits scaled to their needs, priority engineering support, and access to production infrastructure for developing multilingual AI applications in areas such as speech, translation, and large language models. == Open-source release == In February 2026, Sarvam AI announced and open-sourced two large language models: Sarvam 30B (30 billion parameters) and Sarvam 105B (105 billion parameters, using a Mixture-of-Experts architecture with 10.3 billion active parameters). Both models were trained from scratch on datasets focused on Indian languages and support advanced reasoning, multilingual tasks, mathematics, and coding. The models are hosted on Hugging Face under the Apache License and are intended for enterprise and developer applications in Indian languages. The models were subsequently released as open source under the Apache License 2.0, with model weights made available on Hugging Face (sarvamai/sarvam-30b and sarvamai/sarvam-105b) and AIKosh in early March 2026. == Government and institutional collaborations == In 2025, Sarvam AI was selected to contribute to India's sovereign AI model initiative under the IndiaAI Mission. The initiative aims to support domestic AI infrastructure and model development. In March 2025, the Unique Identification Authority of India (UIDAI) announced a collaboration with Sarvam AI to integrate AI-based voice interactions and multilingual support into Aadhaar-related services. Sarvam AI has also worked with AI4Bharat and academic institutions on language datasets and speech research projects. == Industry participation == Sarvam AI presented its foundational models at the India AI Impact Summit 2026 in New Delhi. The company has also been listed among Indian members of the AI Alliance, a consortium focused on open-source artificial intelligence initiatives. == List of models ==

Saliency map

In computer vision, a saliency map is an image that highlights either the region on which people's eyes focus first or the most relevant regions for machine learning models. The goal of a saliency map is to reflect the degree of importance of a pixel to the human visual system or an otherwise opaque ML model. For example, in this image, a person first looks at the fort and light clouds, so they should be highlighted on the saliency map. == Application == === Overview === Saliency maps have applications in a variety of different problems. Some general applications: ==== Human eye ==== Image and video compression: The human eye focuses only on a small region of interest in the frame. Therefore, it is not necessary to compress the entire frame with uniform quality. According to the authors, using a salience map reduces the final size of the video with the same visual perception. Image and video quality assessment: The main task for an image or video quality metric is a high correlation with user opinions. Differences in salient regions are given more importance and thus contribute more to the quality score. Image retargeting: It aims at resizing an image by expanding or shrinking the noninformative regions. Therefore, retargeting algorithms rely on the availability of saliency maps that accurately estimate all the salient image details. Object detection and recognition: Instead of applying a computationally complex algorithm to the whole image, we can use it to the most salient regions of an image most likely to contain an object. the primary visual cortex (V1) appears to be responsible for the saliency map, according to the V1 Saliency Hypothesis. ==== Explainable artificial intelligence ==== Saliency maps are a prominent tool in explainable artificial intelligence, providing visual explanations of the decision-making process of machine learning models, particularly deep neural networks. These maps highlight the regions in input data that are most influential on the model's output, effectively indicating where the model is "looking" when making a prediction. In image classification tasks, for example, saliency maps can identify pixels or regions that contribute most to a specific class decision. Developed for convolutional neural networks, saliency mapping techniques range from simply taking the gradient of the class score with respect to the input data to more complex algorithms, such as integrated gradients and class activation mapping. In transformer architecture, attention mechanisms led to analogous saliency maps, such as attention maps, attention rollouts, and class-discriminative attention maps. === Saliency as a segmentation problem === Saliency estimation may be viewed as an instance of image segmentation. In computer vision, image segmentation is the process of partitioning a digital image into multiple segments (sets of pixels, also known as superpixels). The goal of segmentation is to simplify and/or change the representation of an image into something that is more meaningful and easier to analyze. Image segmentation is typically used to locate objects and boundaries (lines, curves, etc.) in images. More precisely, image segmentation is the process of assigning a label to every pixel in an image such that pixels with the same label share certain characteristics. == Algorithms == === Overview === There are three forms of classic saliency estimation algorithms implemented in OpenCV: Static saliency: Relies on image features and statistics to localize the regions of interest of an image. Motion saliency: Relies on motion in a video, detected by optical flow. Objects that move are considered salient. Objectness: Objectness reflects how likely an image window covers an object. These algorithms generate a set of bounding boxes of where an object may lie in an image. In addition to classic approaches, neural-network-based are also popular. There are examples of neural networks for motion saliency estimation: TASED-Net: It consists of two building blocks. First, the encoder network extracts low-resolution spatiotemporal features, and then the following prediction network decodes the spatially encoded features while aggregating all the temporal information. STRA-Net: It emphasizes two essential issues. First, spatiotemporal features integrated via appearance and optical flow coupling, and then multi-scale saliency learned via attention mechanism. STAViS: It combines spatiotemporal visual and auditory information. This approach employs a single network that learns to localize sound sources and to fuse the two saliencies to obtain a final saliency map. There's a new static saliency in the literature with name visual distortion sensitivity. It is based on the idea that the true edges, i.e. object contours, are more salient than the other complex textured regions. It detects edges in a different way from the classic edge detection algorithms. It uses a fairly small threshold for the gradient magnitudes to consider the mere presence of the gradients. So, it obtains 4 binary maps for vertical, horizontal and two diagonal directions. The morphological closing and opening are applied to the binary images to close the small gaps. To clear the blob-like shapes, it utilizes the distance transform. After all, the connected pixel groups are individual edges (or contours). A threshold of size of connected pixel set is used to determine whether an image block contains a perceivable edge (salient region) or not. === Example implementation === First, we should calculate the distance of each pixel to the rest of pixels in the same frame: S A L S ( I k ) = ∑ i = 1 N | I k − I i | {\displaystyle \mathrm {SALS} (I_{k})=\sum _{i=1}^{N}|I_{k}-I_{i}|} I i {\displaystyle I_{i}} is the value of pixel i {\displaystyle i} , in the range of [0,255]. The following equation is the expanded form of this equation. SALS(Ik) = |Ik - I1| + |Ik - I2| + ... + |Ik - IN| Where N is the total number of pixels in the current frame. Then we can further restructure our formula. We put the value that has same I together. SALS(Ik) = Σ Fn × |Ik - In| Where Fn is the frequency of In. And the value of n belongs to [0,255]. The frequencies is expressed in the form of histogram, and the computational time of histogram is ⁠ O ( N ) {\displaystyle O(N)} ⁠ time complexity. ==== Time complexity ==== This saliency map algorithm has ⁠ O ( N ) {\displaystyle O(N)} ⁠ time complexity. Since the computational time of histogram is ⁠ O ( N ) {\displaystyle O(N)} ⁠ time complexity which N is the number of pixel's number of a frame. Besides, the minus part and multiply part of this equation need 256 times operation. Consequently, the time complexity of this algorithm is ⁠ O ( N + 256 ) {\displaystyle O(N+256)} ⁠ which equals to ⁠ O ( N ) {\displaystyle O(N)} ⁠. ==== Pseudocode ==== All of the following code is pseudo MATLAB code. First, read data from video sequences. After we read data, we do superpixel process to each frame. Spnum1 and Spnum2 represent the pixel number of current frame and previous pixel. Then we calculate the color distance of each pixel, this process we call it contract function. After this two process, we will get a saliency map, and then store all of these maps into a new FileFolder. ==== Difference in algorithms ==== The major difference between function one and two is the difference of contract function. If spnum1 and spnum2 both represent the current frame's pixel number, then this contract function is for the first saliency function. If spnum1 is the current frame's pixel number and spnum2 represent the previous frame's pixel number, then this contract function is for second saliency function. If we use the second contract function which using the pixel of the same frame to get center distance to get a saliency map, then we apply this saliency function to each frame and use current frame's saliency map minus previous frame's saliency map to get a new image which is the new saliency result of the third saliency function. == Datasets == The saliency dataset usually contains human eye movements on some image sequences. It is valuable for new saliency algorithm creation or benchmarking the existing one. The most valuable dataset parameters are spatial resolution, size, and eye-tracking equipment. Here is part of the large datasets table from MIT/Tübingen Saliency Benchmark datasets, for example. To collect a saliency dataset, image or video sequences and eye-tracking equipment must be prepared, and observers must be invited. Observers must have normal or corrected to normal vision and must be at the same distance from the screen. At the beginning of each recording session, the eye-tracker recalibrates. To do this, the observer fixates their gaze on the screen center. The session is then started, and saliency data are collected by showing sequences and recording eye gazes. The eye-tracking device is a high-speed camera, capable of recording eye movements at least 250 fr

DVD

DVD (digital video disc or digital versatile disc) is a digital optical disc data storage format. It was invented and developed in 1995 and first released on November 1, 1996, in Japan. The medium can store any kind of digital data and has been widely used to store video programs (watched using DVD players), software and other computer files. DVDs offer significantly higher storage capacity than compact discs (CD) while having the same dimensions. A standard single-layer DVD can store up to 4.7 GB of data, a dual-layer DVD up to 8.5 GB. Dual-layer, double-sided DVDs can store up to a maximum of 17.08 GB. Prerecorded DVDs are mass-produced using molding machines that physically stamp data onto the DVD. Such discs are a form of DVD-ROM because data can only be read and not written or erased. Blank recordable DVD discs (DVD-R and DVD+R) can be recorded once using a DVD recorder and then function as a DVD-ROM. Rewritable DVDs (DVD-RW, DVD+RW, and DVD-RAM) can be recorded and erased many times. DVDs are used in DVD-Video consumer digital video format and less commonly in DVD-Audio consumer digital audio format, as well as for authoring DVD discs written in a special AVCHD format to hold high definition material (often in conjunction with AVCHD format camcorders). DVDs containing other types of information may be referred to as DVD data discs. == Etymology == The Oxford English Dictionary comments that, "In 1995, rival manufacturers of the product initially named digital video disc agreed that, in order to emphasize the flexibility of the format for multimedia applications, the preferred abbreviation DVD would be understood to denote digital versatile disc." The OED also states that in 1995, "The companies said the official name of the format will simply be DVD. Toshiba had been using the name 'digital video disc', but that was switched to 'digital versatile disc' after computer companies complained that it left out their applications." "Digital versatile disc" is the explanation provided in a DVD Forum Primer from 2000 and in the DVD Forum's mission statement, which the purpose is to promote broad acceptance of DVD products on technology, across entertainment, and other industries. Because DVDs became highly popular for the distribution of movies in the 2000s, the term DVD became popularly used in English as a noun to describe specifically a full-length movie released on the format; for example the phrase "to watch a DVD" describes watching a movie on DVD. == History == === Development and launch === Released in 1987, CD Video used analog video encoding on optical discs matching the established standard 120 mm (4.7 in) size of audio CDs. Video CD (VCD) became one of the first formats for distributing digitally encoded films in this format, in 1993. In the same year, two new optical disc storage formats were being developed. One was the Multimedia Compact Disc (MMCD), backed by Philips and Sony (developers of the CD and CD-i), and the other was the Super Density (SD) disc, supported by Toshiba, Time Warner, Matsushita Electric, Hitachi, Mitsubishi Electric, Pioneer, Thomson, and JVC. By the time of the press launches for both formats in January 1995, the MMCD nomenclature had been dropped, and Philips and Sony were referring to their format as Digital Video Disc (DVD). On May 3, 1995, an ad hoc industry technical group formed from five computer companies (IBM, Apple, Compaq, Hewlett-Packard, and Microsoft) issued a press release stating that they would only accept a single format. The group voted to boycott both formats unless the two camps agreed on a single, converged standard. They recruited Lou Gerstner, president of IBM, to pressure the executives of the warring factions. In one significant compromise, the MMCD and SD groups agreed to adopt proposal SD 9, which specified that both layers of the dual-layered disc be read from the same side—instead of proposal SD 10, which would have created a two-sided disc that users would have to turn over. Philips/Sony strongly insisted on the source code, EFMPlus, that Kees Schouhamer Immink had designed for the MMCD, because it makes it possible to apply the existing CD servo technology. Its drawback was a loss from 5 to 4.7 Gigabytes of capacity. As a result, the DVD specification provided a storage capacity of 4.7 GB (4.38 GiB) for a single-layered, single-sided disc and 8.5 GB (7.92 GiB) for a dual-layered, single-sided disc. The DVD specification ended up similar to Toshiba and Matsushita's Super Density Disc, except for the dual-layer option. MMCD was single-sided and optionally dual-layer, whereas SD was two half-thickness, single-layer discs which were pressed separately and then glued together to form a double-sided disc. Philips and Sony decided that it was in their best interests to end the format war, and on September 15, 1995 agreed to unify with companies backing the Super Density Disc to release a single format, with technologies from both. After other compromises between MMCD and SD, the group of computer companies won the day, and a single format was agreed upon. The computer companies also collaborated with the Optical Storage Technology Association (OSTA) on the use of their implementation of the ISO-13346 file system (known as Universal Disk Format) for use on the new DVDs. The format's details were finalized on December 8, 1995. In November 1995, Samsung announced it would start mass-producing DVDs by September 1996. The format launched on November 1, 1996, in Japan, mostly with music video releases. The first major releases from Warner Home Video arrived on December 20, 1996, with four titles being available. The format's release in the U.S. was delayed multiple times, from August 1996, to October 1996, November 1996, before finally settling on early 1997. Players began to be produced domestically that winter, with March 24, 1997, as the U.S. launch date of the format proper in seven test markets. Approximately 32 titles were available on launch day, mainly from the Warner Bros., MGM, and New Line libraries, with the notable inclusion of the 1996 film Twister. However, the launch was planned for the following day (March 25), leading to a distribution change with retailers and studios to prevent similar violations of breaking the street date. The nationwide rollout for the format happened on August 22, 1997. DTS announced in late 1997 that they would be coming onto the format. The sound system company revealed details in a November 1997 online interview, and clarified it would release discs in early 1998. However, this date would be pushed back several times before finally releasing their first titles at the 1999 Consumer Electronics Show. In 2001, blank DVD recordable discs cost the equivalent of $27.34 US dollars in 2022. === Adoption === Movie and home entertainment distributors adopted the DVD format to replace the ubiquitous VHS tape as the primary consumer video distribution format. Immediately following the formal adoption of a unified standard for DVD, two of the four leading video game console companies (Sega and The 3DO Company) said they already had plans to design a gaming console with DVDs as the source medium. Sony stated at the time that they had no plans to use DVD in their gaming systems, despite being one of the developers of the DVD format and eventually the first company to actually release a DVD-based console. Game consoles such as the PlayStation 2, Xbox, and Xbox 360 use DVDs as their source medium for games and other software. Contemporary games for Windows were also distributed on DVD. Early DVDs were mastered using DLT tape, but using DVD-R DL or +R DL eventually became common. TV DVD combos, combining a standard definition CRT TV or an HD flat panel TV with a DVD mechanism under the CRT or on the back of the flat panel, and VCR/DVD combos were also available for purchase. For consumers, DVD soon overtook VHS as the favored choice for home movie releases. In 2001, DVD players outsold VCRs for the first time in the United States. At that time, one in four American households owned a DVD player. By 2007, about 80% of Americans owned a DVD player, a figure that had surpassed VCRs; it was also higher than personal computers or cable television. == Specifications == The DVD specifications created and updated by the DVD Forum are published as so-called DVD Books (e.g. DVD-ROM Book, DVD-Audio Book, DVD-Video Book, DVD-R Book, DVD-RW Book, DVD-RAM Book, DVD-AR (Audio Recording) Book, DVD-VR (Video Recording) Book, etc.). DVD discs are made up of two discs; normally one is blank, and the other contains data. Each disc is 0.6 mm thick, and they are glued together to form a DVD disc. The gluing process must be done carefully to make the disc as flat as possible to avoid both birefringence and "disc tilt", which is when the disc is not perfectly flat, preventing it from being read. Some specifications for mechanical, physical and optical characteristics of DV

DVD

News ticker

A news ticker (sometimes called a crawler, crawl, slide, zipper, ticker tape, or chyron) is a horizontal or vertical (depending on the language's writing system) text-based display either in the form of a graphic that typically resides in the lower third of the screen space on a television station or network (usually during news programming) or as a long, thin scoreboard-style display seen around the facades of some offices or public buildings dedicated to presenting headlines or minor pieces of news. It is an evolution of the paper strips tapes, a continuous paper print-out of stock quotes from a printing telegraph which was mainly used to transmit companies' share price information over telegraph lines before the advance of technology in the 1960s. News tickers have been used in Europe in countries such as United Kingdom, Germany and Ireland for some years; they are also used in several Asian countries and Australia. In the United States, tickers were long used on a special event basis by broadcast television stations to disseminate weather warnings, school closings, and election results. Sports telecasts occasionally used a ticker to update other contests in progress before the expansion of cable news networks and the internet for news content. In addition, some ticker displays are used to relay continuous business and financial information. Most tickers are traditionally displayed in the form of scrolling text running from right to left across the screen or building display (or in the opposite direction for right-to-left writing systems such as Arabic script and Hebrew), allowing for headlines of varying degrees of detail; some used by television broadcasters, however, display stories in a static manner (allowing for the seamless switching of each story individually programmed for display) or utilize a "flipping" effect (in which each individual headline is shown for a few seconds before transitioning to the next, instead of scrolling across the screen, usually resulting in a relatively quicker run through of all of the information programmed into the ticker). Since the growth in usage of the World Wide Web, some news tickers have syndicated news stories posted largely on websites of broadcasters or by other independent news agencies. == Current uses == === Television === The presentation of headlines or other information in a news ticker has become a common element of many different news networks. The use of the ticker has differed on a number of channels: News networks and local newscasts commonly use a setup in which news headlines are scrolled across an area near the bottom of the screen, though some variations have formed, such as showing one headline at a time with a scrolling or "flipper" effect. Financial news channels use two or more tickers displaying company shares prices and business headlines. Networks with a focus on sports often use a slightly different system, where scores and statuses of ongoing and finished games are displayed one by one, along with minor sports highlights, statistics and sports news headlines. They are typically divided into categories devoted to specific leagues and events (with college basketball and football usually focusing on the top 25 ranked teams on the AP Poll, occasionally supplemented by sections for specific conferences). Some programs, including news-based programs emphasizing viewer interactivity, or special events, may also use tickers to display messages and reactions from viewers and others that relate to the program. These comments are often sourced from social networking services such as Facebook and Twitter, typically curating comments from a specific page or hashtag. Due to their current prevalence, they have been occasionally been made targets of pranks and vandalism. In one such example, News 14 Carolina allowed viewers to submit relevant information such as school closings or traffic delays via telephone or the Internet that would be incorporated into the ticker; the system was exploited in February 2004 to display humorous and crude messages, including the infamous "All your base are belong to us". Occasionally messages intended for training accidentally end up being put on the live ticker as happened on BBC News in 2022 when "Weather rain everywhere" and "Manchester United are rubbish" appeared on the live news ticker. Some businesses and organizations have utilized tickers intended for relaying weather-related closings as a surreptitious source for free guerrilla marketing, proclaiming they were open rather than closed and giving their phone number if possible, allowing them to 'advertise' on a television station all day for free. Since then, many stations have required pre-registration of businesses or organizations with an authorized representative and a signed affidavit on company letterhead affirming their authenticity, along with filtering out unfamiliar businesses and organizations, before being able to display their closing announcements. Stations also confirm all closings involving school districts with authorized officials to prevent situations in which students either show up to canceled classes in dangerous conditions, or do not attend school due to an erroneous, prank-submitted, or false listing. === On personal computers === Various applications have been developed over time to install news tickers on personal computer desktops using RSS feeds from news organizations, which are displayed in a fashion similar to those used by television channels but enable the user to access to underlying news stories, a feature not offered by traditional television channels. The Bloomberg Terminal and other financial information-tracking programs and devices also utilize tickers. A ticker may also be used as an unobtrusive method by businesses in order to deliver important information to their staff. The ticker can be set to reappear, stay on screen, or be put into a retractable mode (where a small tab is left visible on-screen). In the United Kingdom, broadcasters have stopped using this technology as other forms of communications have become available and increased in popularity. BBC News and Sky News discontinued their respective desktop tickers in March 2011 and 2012 to focus on other products, such as smartphone applications, to deliver updated information on breaking news and sport stories. === News tickers on buildings === Since the advent of the telegraph, newspapers commonly used their buildings to share the latest headlines. At first simple chalkboard signs were used for bulletins, but limelight illumination, electric lights, magic lantern projections, and other novel techniques were later employed. The method of using electric lights to spell out moving letters was invented by Frank C. Reilly (August 20, 1888 – April 10, 1947) and patented in 1923. Reilly called his invention the Motograph News Bulletin. In 1928, The New York Times installed a Motograph News Bulletin to display news headlines on the sides of Times Tower. The display was 388 feet (118 m) long, 5 feet (1.5 m) high, and employed over 14,800 light bulbs. Popularly known as the "Zipper", the sign remained in use until the building was sold in 1961. The sign was darkened during World War II to comply with wartime lighting restrictions. The Motograph operated until 1994 and was replaced by an electronic version in 1995, which was in turn removed in 2017 due to the replacement of all individual screens on the front of One Times Square with a 350 foot (110 m)-tall LED billboard in 2018. Ticker displays appear today on the exterior of the News Corp Building, which houses the headquarters for Fox News Channel/News Corp in the west extension of Manhattan's Rockefeller Center, as well as one that displays delayed stock market data that is located in Times Square. NASDAQ itself features a large display screen on the facade of the NASDAQ MarketSite building in Times Square. The Reuters buildings at Canary Wharf and in Toronto have news and stock tickers; the latter type features market data for the New York Stock Exchange, NASDAQ and London Stock Exchange, while the Toronto building's ticker also includes quotes from the Toronto Stock Exchange. A red-LED ticker was added to the perimeter of 10 Rockefeller Center in 1994, as the building was being renovated to accommodate the studios for NBC's Today. Placed at the juncture of the first and second floors, the ticker is visible to spectators in Rockefeller Plaza and passersby on West 49th Street and updates continuously, even at times when Today is not being produced and broadcast. As of 2015, the ticker strip is only a small part of a large two-floor LCD video display that is placed within the window of the studio showing promotional information. The Martin Place Headquarters of Seven News, the news division of Australian television broadcaster Seven Network, also incorporates a ticker that wraps around the building. == In popular culture == The use of new

FoundationDB

FoundationDB is a free and open-source multi-model distributed NoSQL database owned by Apple Inc. with a shared-nothing architecture. The product was designed around a "core" database, with additional features supplied in "layers." The core database exposes an ordered key–value store with transactions. The transactions are able to read or write multiple keys stored on any machine in the cluster while fully supporting ACID properties. Transactions are used to implement a variety of data models via layers. The FoundationDB Alpha program began in January 2012 and concluded on March 4, 2013, with their public Beta release. Their 1.0 version was released for general availability on August 20, 2013. On March 24, 2015, it was reported that Apple has acquired the company. A notice on the FoundationDB web site indicated that the company has "evolved" its mission and would no longer offer downloads of the software. On April 19, 2018, Apple open sourced the software, releasing it under the Apache 2.0 license. == Main features == The main features of FoundationDB include the following: Ordered key–value store In addition to supporting standard key-based reads and writes, the ordering property enables range reads that can efficiently scan large swaths of data. Transactions Transaction processing employs multiversion concurrency control for reads and optimistic concurrency for writes. Transactions can span multiple keys stored on multiple machines. ACID properties FoundationDB guarantees serializable isolation and strong durability via redundant storage on disk before transactions are considered committed. Layers Layers map new data models, APIs, and query languages to the FoundationDB core. They employ FoundationDB's ability to update multiple data elements in a single transaction, ensuring consistency. An example is their SQL layer. Commodity clusters FoundationDB is designed for deployment on distributed clusters of commodity hardware running Linux. Replication FoundationDB stores each piece of data on multiple machines according to a configurable replication factor. Triple replication is the recommended mode for clusters of 5 or more machines. Scalability FoundationDB is designed to support horizontal scaling through the addition of machines to a cluster while automatically handling data replication and partitioning. Systems supported FoundationDB supports packages for Linux, Windows, and macOS. The Linux version supports production clusters, while the Windows and macOS versions support local operation for development purposes. Configurations on Amazon EC2 are also supported. Programming language bindings FoundationDB supports language bindings for Python, Go, Ruby, Node.js, Java, PHP, and C, all of which are made available with the product. == Design limitations == The design of FoundationDB results in several limitations: Long transactions FoundationDB does not support transactions running over five seconds. Large transactions Transaction size cannot exceed 10 MB of total written keys and values. Large keys and values Keys cannot exceed 10 kB in size. Values cannot exceed 100 kB in size. == History == FoundationDB, headquartered in Vienna, Virginia, was started in 2009 by Nick Lavezzo, Dave Rosenthal, and Dave Scherer, drawing on their experience in executive and technology roles at their previous company, Visual Sciences. In March 2015 the FoundationDB Community site was updated to state that the company had changed directions and would no longer be offering downloads of its product. The company was acquired by Apple Inc., which was confirmed March 25, 2015. On April 19, 2018, Apple open sourced the software, releasing it under the Apache 2.0 license.

GlTF

glTF (Graphics Library Transmission Format or GL Transmission Format and formerly known as WebGL Transmissions Format or WebGL TF) is a standard file format for three-dimensional scenes and models. A glTF file uses one of two possible file extensions: .gltf (JSON/ASCII) or .glb (binary). Both .gltf and .glb files may reference external binary and texture resources. Alternatively, both formats may be self-contained by directly embedding binary data buffers (as base64-encoded strings in .gltf files or as raw byte arrays in .glb files). An open standard developed and maintained by the Khronos Group, it supports 3D model geometry, appearance, scene graph hierarchy, and animation. It is intended to be a streamlined, interoperable format for the delivery of 3D assets, while minimizing file size and runtime processing by apps. As such, its creators have described it as the "JPEG of 3D". == Overview == The glTF format stores data primarily in JSON. The JSON may also contain blobs of binary data known as buffers, and refer to external files, for storing mesh data, images, etc. The binary .glb format also contains JSON text, but serialized with binary chunk headers to allow blobs to be directly appended to the file. The fundamental building blocks of a glTF scene are nodes. Nodes are organized into a hierarchy, such that a node may have other nodes defined as children. Nodes may have transforms relative to their parent. Nodes may refer to resources, such as meshes, skins, and cameras. Meshes may refer to materials, which refer to textures, which refer to images. Scenes are defined using an array of root nodes. Most of the top-level glTF properties use a flat hierarchy for storage. Nodes are saved in an array and are referred to by index, including by other nodes. A glTF scene refers to its root nodes by index. Furthermore, nodes refer to meshes by index, which refer to materials by index, which refer to textures by index, which refer to images by index. All glTF data structures support being extended using a JSON property, allowing arbitrary JSON data to be added. == Releases == === glTF 1.0 === Members of the COLLADA working group conceived the file format in 2012. At SIGGRAPH 2012, Khronos presented a demo of glTF, which was then called WebGL Transmissions Format (WebGL TF). On October 19, 2015, Khronos released the glTF 1.0 specification. ==== Adoption of glTF 1.0 ==== At SIGGRAPH 2016, Oculus announced their adoption of glTF citing the similarities to their ovrscene format. In October 2016, Microsoft joined the 3D Formats working group at Khronos to collaborate on glTF. === glTF 2.0 === The second version, glTF 2.0, was released in June 2017, and is a complete overhaul of the file format from version 1.0, with most tools adopting the 2.0 version. Based on a proposal by Fraunhofer originally presented at SIGGRAPH 2016, physically based rendering (PBR) was added, replacing WebGL shaders used in glTF 1.0. glTF 2.0 added the GLB binary format into the base specification. Other upgrades include sparse accessors and morph targets for techniques such as facial animation, and schema tweaks and breaking changes for corner cases or performance such as replacing top-level glTF object properties with arrays for faster index-based access. There is ongoing work towards import and export in Unity and an integrated multi-engine viewer and validator. ==== Adoption of glTF 2.0 ==== On March 3, 2017, Microsoft announced that they would be using glTF 2.0 as the 3D asset format across their product line, including Paint 3D, 3D Viewer, Remix 3D, Babylon.js, and Microsoft Office. Sketchfab also announced support for glTF 2.0. The glTF and GLB formats are used on and supported by companies including DGG, UX3D, Sketchfab, Facebook, Microsoft, Meta, Google, Adobe, Box, TurboSquid, Unreal Engine, Unity, and Qt Quick 3D. The format has been noted as an important standard for augmented reality, integrating with modeling software such as Autodesk Maya, Autodesk 3ds Max, and Poly. In February 2020, the Smithsonian Institution launched their Open Access Initiative, releasing approximately 2.8 million 2D images and 3D models into the public domain, using glTF for the 3D models. In July 2022, glTF 2.0 was released as the ISO/IEC 12113:2022 International Standard. Khronos stated they would make regular submissions to bring updates and new widely adopted glTF functionality into refreshed versions of ISO/IEC 12113 to ensure that there is no long-term divergence between the ISO/IEC and Khronos specifications. The open-source game engine Godot supports importing glTF 2.0 files since version 3.0 and export since version 4.0. === Extensions === The glTF format can be extended with arbitrary JSON to add new data and functionality. Extensions can be placed on any part of a glTF, including nodes, animations, materials, textures, and on the entire document. Khronos keeps a non-comprehensive registry of glTF extensions on GitHub, including all official Khronos extensions and a few third-party extensions. PBR extensions model the physical appearance of real-world objects, allowing developers to create realistic 3D assets that have the correct appearance. As new PBR extensions are released, they continue to expand PBR capabilities within the glTF framework, allowing a wider range of scenes and objects to be realistically rendered as 3D assets. The KTX 2.0 extension for universal texture compression enables 3D models in the glTF format to be highly compressed and to use natively supported texture formats, reducing file size and boosting rendering speed. Draco is a glTF extension for mesh compression, to compress and decompress 3D meshes, to help reduce the size of 3D files. It compresses vertex attributes, normals, colors, and texture coordinates. Various glTF extensions for game engine interoperability have been developed by OMI group. This includes extensions for physics shapes, physics bodies, physics joints, audio playback, seats, spawn points, and more. The VRM consortium has developed glTF extensions for advanced humanoid 3D avatars including dynamic spring bones and toon materials. == Derivative formats == 3D Tiles, an OGC Community Standard, builds on glTF to add a spatial data structure, metadata, and declarative styling for streaming massive heterogeneous 3D geospatial datasets. VRM, a model format for VR, is built on the .glb format. It is a 3D humanoid avatar specification and file format. == Software ecosystem == Khronos maintains the glTF Sample Viewer for viewing glTF assets. Khronos also maintains the glTF Validator for validating if 3D models conform to the glTF specification. Khronos maintains a glTF Compressor tool to interactively optimize and fine-tune compression settings for glTF assets using KTX 2.0 textures. glTF loaders are in open-source WebGL engines including PlayCanvas, Three.js, Babylon.js, Cesium, PEX, xeogl, and A-Frame. The Godot game engine supports and recommends the glTF format, with both import and export support. Open-source glTF converters are available from COLLADA, FBX, and OBJ. Assimp can import and export glTF. glTF files can also be directly exported from a variety of 3D editors, such as Blender, Unity (using the glTFast importer/exporter), Freecad, Vectary, Autodesk 3ds Max (natively or using Verge3D exporter), Autodesk Maya (using babylon.js exporter), Autodesk Inventor, Modo, Houdini, Paint 3D, Godot, and Substance Painter. Open-source glTF utility libraries are available for programming languages including JavaScript, Node.js, C++, C#, Python, Haskell, Java, Go, Rust, Haxe, Ada, and TypeScript. Khronos keeps a list of these libraries and other related applications on their ecosystem site. The Khronos 3D Commerce Working Group released Asset Creation Guidelines in 2020 outlining best practices for use of the glTF file format in 3D Commerce. In 2025, the Working Group launched Asset Creation Guidelines 2.0, a continuously updated resource with additional guidance for geometry, mesh optimization, UV maps, textures, materials/PBR performance, and web optimization. The Khronos PBR Neutral Tone Mappers specification is a tone mapper designed to faithfully reproduce an object's base color, hue, and saturation when using PBR rendering under grayscale lighting, supporting brand- and product-accurate color representation. Khronos maintains the glTF Asset Auditor to allow retailers and advertising technology platforms to validate 3D assets against either a default Audit Profile modelled on the 2020 3D Commerce Asset Creation Guidelines or a custom profile defined by the target application.

Saliency map

DVD

DVD

News ticker

FoundationDB

GlTF

Related Articles