Tools And Hardware – Car Link Mall Wed, 25 Aug 2021 05:51:28 +0000 en-US hourly 1 Tools And Hardware – Car Link Mall 32 32 Orolia adds a new GNSS simulator to the BroadSim product line Wed, 25 Aug 2021 05:00:50 +0000

BroadSim Solo enables the creation of advanced GNSS scenarios on the engineer’s desktop

Orolia Defense & Security released the latest addition to its GNSS simulator family – BroadSim Solo – at the Institute of Navigation Joint Navigation Conference (JNC) taking place this week in Covington, Kentucky.

The Solo joins the BroadSim line of Skydel-powered GNSS simulators, which includes models suitable for hardware-in-the-loop and multi-element / Controlled Reception Pattern Antenna (CRPA) testing.

Live demonstrations of BroadSim Solo are taking place in the JNC exhibition hall at Orolia Defense & Security booth # 117 until August 26. Orolia provides M-code solutions for resilient positioning, navigation and synchronization.

BroadSim Solo shares the same Skydel simulation engine that runs on a standard BroadSim, BroadSim Anechoic, and BroadSim Wavefront. It supports the advanced scenario-building features and benefits provided by a software-defined architecture, such as high dynamics, iteration update rate of 1000Hz, and ultra-low latency of 5ms. .

Photo: Orolia

Almost any civilian GNSS signal can be generated through Solo’s single RF output (one frequency band at a time), as well as jamming or impersonation signals and the AES GPS M code. AES is an encryption method; rather than using MNSA encryption, it is possible to use AES for testing purposes only.

BroadSim Solo’s compact form factor is designed to eliminate clutter, fitting comfortably into a typical office or workstation. In addition, Solo addresses the ongoing challenge engineers face in terms of laboratory capacity and availability.

“Creating complex test cases can be a tedious process, especially when emulating challenging environments,” said Tim Erbes, director of engineering for Orolia Defense & Security. “The ability to script on your desktop frees up a lot of lab time and space to run these important simulations. In addition, scenario creation is no longer limited to one person attached to a system. Imagine a team of engineers, each with a BroadSim Solo, simultaneously building tests. Have a whole fleet of BroadSim Solos? It changes the game.

BroadSim Solo with the Skydel simulation engine offers an intuitive user interface, a full API supporting Python, C ++ and C #, as well as automation tools and custom plugins that will speed up development cycles, increase performance and ultimately boost innovation.

“In an effort to improve the customer experience and extend the reach of advanced GNSS simulators, we wanted to offer an affordable solution with all of the same basic features as our most advanced BroadSim systems,” said Tyler Hohman, product manager for Orolia Defense & Security. “This gives our customers the opportunity to put more simulators in the hands of engineers and scientists without sacrificing capabilities. Our hope is that customers will find value in a scalable simulation ecosystem based on their needs. “

Photo: Orolia

Photo: Orolia

Source link

Cutlery and hand tool market trends involve the manufacture of metal additives Mon, 23 Aug 2021 09:09:00 +0000

Cutlery and Hand Tools Market Report 2021: Impact of COVID-19 and Recovery to 2030

The Business Research Company Global Cutlery and Hand Tools Market Report 2021: Impact of COVID-19 and Recovery to 2030

LONDON, GREATER LONDON, UK, 23 Aug 2021 / – Metal additive manufacturing is an emerging trend in the cutlery and hand tool market. The demand for metal additive manufacturing is increasing around the world, especially in countries like Germany, Japan, China and India. Metal additive manufacturing is a 3D printing technology used to make the final product by stacking layers of material and then polishing to a seamless look. In 2016, according to the KPMG report, 26% of global metallurgical company executives said they had already introduced 3D printing technologies for metal additive manufacturing, and 27% of them said they planned to invest more in the future. Increased investment in metal additive manufacturing technology further leads to process improvements and lower production costs in related manufacturing industries.

The cutlery and hand tools market comprises the sales of cutlery and hand tools by entities (organizations, sole proprietorships, and partnerships) that are engaged in one or more of the following fields: kitchen of metal (except those produced by molding (eg stoves) or stamped without other manufacture), utensils and / or cutlery and cutlery of non-precious and precious metal plated; manufacture of saw blades of all types (including those for power sawing machines); and the manufacture of non-motorized hand and cutting tools.

The major players in the global cutlery and hand tools market are Snap-on Inc, Stanley Black & Decker Inc, Brüder Mannesmann AG, Acme United Corporation.

Learn more about the Global Cutlery and Hand Tools Market report:

The global cutlery and hand tools market is expected to grow from $ 1.40 trillion in 2020 to $ 1.54 trillion in 2021 at a compound annual growth rate (CAGR) of 10.5%. The growth is mainly due to companies reorganizing their operations and recovering from the impact of COVID-19, which previously led to restrictive containment measures involving social distancing, remote working and the closure of business activities which resulted in operational challenges. The market is expected to reach $ 2.01 trillion in 2025 at a CAGR of 7%.

The cutlery and hand tools market is segmented into kitchen utensils, utensils, cutlery and metal cutlery; saw blade and hand tool, and segmented by application in household, commercial.

Global Cutlery and Hand Tools Market Report 2021: Impact and Recovery of COVID-19 through 2030 is part of a series of new reports from The Business Research Company that provides an overview of the cutlery market and Cutlery and Hand Tools Market size and growth for the Overall Market, Cutlery and Hand Tools market segments, and geographies, market trends cutlery and hand tools, cutlery and hand tools market drivers, restraints, revenues, profiles and market shares of major competitors.

Request for Sample of Global Cutlery and Hand Tools Market Report:

Here is a list of similar reports from the Business Research Company:
Global Metal Products Market Report 2021 – By Type (Forged & Stamped Goods, Cutlery & Hand Tools, Architectural & Structural Metals, Boiler, Tank & Shipping Container, Hardware, Spring & Wire Products, Workshops machining, turned product and screw, nut, And bolts, coated, etched and heat treated metal products, metal valves, other fabricated metal products), by end use (construction, manufacturing), COVID-19 impact and recovery

Global Artificial Intelligence Market Report 2021: Growth and Evolution of COVID-19 to 2030

Global Big Data and Analytics Market Report 2021: Growth and Change of COVID-19 through 2030

Want to learn more about The Business Research Company?
Learn more about us at

Or get a quick overview of our services here:

The Business Research Company is a market research and intelligence company that excels in business, market and consumer research. It has more than 200 research professionals in offices in India, UK and US, as well as a network of trained researchers around the world. It has consultants who specialize in a wide range of industries including manufacturing, healthcare, financial services and technology.

Call us now for personal assistance with your purchase:
Europe: +44 207 1930 708
Asia: +91 88972 63534
Americas: +1 315 623 0293

The commercial research company
Follow us on LinkedIn:
Follow us on Twitter:
Check out our blog:

Olivier Guirdham
The commercial research company
+44 20 7193 0708
Visit us on social networks:

Source link

Tracing the growth of the Solana ecosystem Sun, 22 Aug 2021 12:29:31 +0000

Key points to remember

  • Solana is a high-speed Layer 1 blockchain offering fast, low-cost transactions.
  • The project has been described as one of Ethereum’s strongest competitors.
  • Solana had a great year, with an increase in the value of SOL and rapid development of its DeFi ecosystem.

Share this article

We explain how Solana and its rapidly growing ecosystem have established a place at the forefront of the cryptocurrency space.

A new competitor of Ethereum

In early June, Solana made headlines after closing a $ 314 million private sales round led by Andreessen Horowitz and Polychain.

Funding has come thanks to the rapidly growing ecosystem that is developing on Solana and the growing status of one of the main competitors of Ethereum, the most widely used public blockchain.

In the past, the huge demand for Ethereum block space has caused network congestion, resulting in very high transaction fees.

This congestion has created opportunities for Layer 2 solutions, side chains and new Layer 1 networks that aim to create scalable dApps beyond Ethereum. Solana is one of those Layer 1 networks.

The project was founded in 2017 amid the ICO mania when its team raised more than $ 25 million in private and public rounds. The mainnet beta was finally released in March 2020.

Solana has been recognized for its 400ms block time and high throughput of 50,000 transactions per second, thousands of times greater than Bitcoin and the current version of Ethereum, both of which rely on Proof-of-Work consensus. (Ethereum plans to switch to Proof-of-Stake in the future).

With a focus on scalability for consumer adoption, Solana can theoretically scale up to 700,000 transactions per second, as noted in the white paper.

How does Solana achieve scalability?

Solana’s architecture explains how the network achieves such high scalability. Blockchain The sea level runtime enables horizontal parallel processing of transactions. This means that Solana can continue to scale with improvements to the Validator’s GPU, which keeps fees low as transactions scale.

According to the founder and CEO of Solana, Anatoly Yakovenko, the level of scalability promised by Solana is proportionately linked to the computer hardware. Essentially, the network can run tens of thousands of smart contract transactions in parallel, using as many GPU cores as are available to validators.

The main disadvantage of Solana is that specialized hardware that can cost thousands of dollars is required to run a validator.

Along with other features such as proof of history and the Tower BFT consensus algorithm, an optimized version for proof of history of BFT, the goal of the project is to have a distributed system capable of scaling transactions in proportion to network bandwidth.

In addition, Solana allows transactions to scale alongside network bandwidth. This means that it can scale as network usage increases without relying on partitioning or Layer 2 solutions.

There are over 900 validators on Solana today. While Ethereum is still the most decentralized smart contract network, Solana is more decentralized than many other Layer 1 chains, including Avalanche, Binance Smart Chain, and Fantom.

The Solana ecosystem

Many new projects have chosen to rely on Solana to benefit from its high speed and ultra-low transaction fees.

Taking advantage of Solana’s low-cost and instantaneous block purpose, high-efficiency blockchain, the growing DeFi ecosystem now consists of dozens of dApps.

The ecosystem includes decentralized exchanges (HydraSwap, Orca), automated market makers (Raydium, Popsicle Finance), yield aggregators (SolFarm, Solyard), stable coin exchange platforms (Mercurial Finance, Saber), portfolios (Solflare, Phantom, Solong), NFT marketplaces (Solanart, Sollectify), derivatives (Parrot, Mango Markets) and games (SOLife, Sollamas, SolPunks).

Numerous infrastructure-based projects such as data analysis tools, block explorers, oracles, and launch pads have also been built in the past six months.

Like Ethereum, Solana’s biggest area of ​​growth has been decentralized finance. Solana’s fast block times and low transaction fees have proven to be attractive for on0chain trading protocols. For DeFi traders, the real-time block finality enables accurate account margin values ​​and real-time profit and loss calculations.

Another big contributor to Solana’s DeFi boom was Sam Bankman-Fried, the CEO of FTX exchange and one of the network’s biggest supporters. In August 2020, Bankman-Fried announced the launch of Serum, a fast decentralized, non-custodial exchange. The serum has become a great catalyst for Solana’s rapid growth.

Bankman-Fried’s confidence in Solana was enough to bring massive levels of liquidity to Serum by integrating some of the top market makers, including Alameda Research (which he founded) and Jump Trading. Alameda Research has also invested in many emerging projects in the ecosystem.

While functioning as a decentralized exchange native to Solana, Serum offers a trading experience similar to centralized exchanges using a limited order book executed on the network.

An order book allows features such as limit orders and instant profit and loss updates for more control and precision in trading. In addition, any other project on Solana can be linked to the liquidity of Serum’s on-chain order book. Traders can place limited buy and sell orders, which can be matched through Serum. Various types of commercial and financial projects are now integrated into Serum’s order book.

Over the past year, Serum has become the core infrastructure that powers several Solana projects, including Radium, an automated market maker that has some similarities to projects like Uniswap. In return, these projects help to increase the volume of Serum trade.

Solana’s fast blocking time enables high fidelity Oracle data through projects like Pyth Network. This makes it possible to share precise information between the various stakeholders and to adjust it on the chain in real time.

Solana hosts many popular stablecoins to provide significant liquidity and scalability to support orderbook-based DEXs. Stablecoins are considered to be one of the fundamental building blocks of DeFi. Most recently, USDC’s supply to Solana exceeded $ 1 billion.

Besides stablecoins, there are many Ethereum native DeFi projects that have deployed their code to the network or are looking for ways to expand in the future. Aave, Ethereum’s main lending market, hinted that it would launch into Solana through Neon Labs earlier this month.

With the infrastructure in place, new projects built on Solana also benefit from what is known as the “Solana summer”. A new derivatives trading app on Solana, Mango Markets, recently raised over $ 60 million in a fundraising round. Over $ 500 million has been deposited on Mango Markets for a chance to qualify for the public sale, showing the extent of interest in Solana’s nascent DeFi ecosystem.

The NFT mania is also making its way into the Solana ecosystem.

Just a few days ago, a collection of 10,000 Degenerate Ape Academy NFT avatars sold out within eight minutes of launching on Solanart. The NFT sale came as Solana’s SOL token climbed to a historic price of $ 63. It has since hit $ 81, now ranking among the top 10 cryptocurrencies by market cap.

Smart contracts and interoperability

Solana does not support Solidity, the programming language used in Ethereum. This means that it lacks EVM compatibility, which could make it difficult to compete with Ethereum’s network effect.

Instead, Solana uses Rust for development. Rust is one of the most popular languages ​​in the developer community.

Projects like Neon Labs are also working to offer EVM compatibility on Solana by porting Solidity smart contracts to run on the network.

Additionally, an inter-chain bridge called a Wormhole allows the transit of assets from Ethereum, although such solutions come with compromises in security. Millions of dollars have been lost in bridge attacks this year.

Solana’s most promising value proposition is to deliver low latency blocking times and the highest bandwidth of any blockchain. With web-scale performance, Solana offers one of the best user experiences of any Layer 1 blockchain, which can lead to wider adoption. With over $ 300 million in freshly raised capital, Solana is also in a good position to accelerate the development of many other native projects.

Share this article

Source link

Worldwide Smart Buildings AI & Machine Learning in Smart Wed, 11 Aug 2021 09:33:35 +0000

Dublin, August 11, 2021 (GLOBE NEWSWIRE) – The report “Smart Buildings AI & Machine Learning in Smart Commercial Buildings” has been added to offer.

This report is a new study from 2021 that performs an independent assessment of the artificial intelligence and machine learning technologies market and their application in intelligent commercial buildings from 2020 to 2025.

This assessment of current and projected revenue is based on a comprehensive bottom-up model that assesses the AI ​​and machine learning offerings of a total of 255 companies, ranging from the world’s largest tech companies to niche startups based on a total. from 17 different countries around the globe.

The information presented in this report is based on our expertise in the associated areas of IoT, Big Data, cybersecurity and application specific markets such as physical security, video analytics, analytics. occupancy and intelligent lighting.

Why do you need this report?

  • Go through the hype to understand what AI IS (and what is NOT) and how it is applied to systems in the built environment. There are a growing number of use cases and case studies of commercially available solutions that deliver tangible value to building owners and occupants, which we discuss in this report.
  • Understand the solutions that have grown in importance in the wake of the COVID-19 pandemic and how they harness the ability of AI technology to detect, monitor and track the actions, behaviors and movements of individuals.
  • Discover the market opportunity. The report estimates that the AI ​​and machine learning in intelligent commercial buildings market generated total revenues of $ 1.11 billion in 2020, and we forecast it will grow by 24.3% CAGR until 2025, almost tripling in value to reach approx. $ 3.3 billion by 2025.
  • Understand the competitive landscape. Over 45% and 43% of companies respectively provide solutions related to the security and access control and space / occupancy and movement of people markets. The market for AI-based energy management and sustainability services is also growing rapidly and attracting a good number of new entrants to the market, with 32% of providers now offering services for this space.

Across its 265 pages and 42 charts and tables, the report filters out all the key facts and draws conclusions, so you can understand exactly how AI technology will be applied to commercial buildings and why;

  • While the organic sales growth of new AI-based solutions will be responsible for the majority of the expected growth in the market, it will also be due, in part, to AI taking over a still proportion. larger of existing building systems already in operation. . AI devices will increasingly replace older generations of edge devices, and AI-based analytics will replace some more traditional forms of software analytics sold today.
  • The majority of hardware revenue comes from edge devices, especially the different types of AI-enabled cameras. Other market analyzes devoted to the broader AI solutions market show a much higher weighting of software-generated revenue, but the relative importance of computer vision solutions to the smart building market means that revenues hardware is a strong market share and will continue to do so. so go ahead. We estimate that hardware revenue currently represents 35.5% of the market.
  • The global AI industry is attracting significant investment and this trend also applies to those offering solutions for the smart building market. 120 of the 255 companies on our list have received some form of declared equity investment. Of these, 111 received $ 1 million or more in reported funding, with the median total amount of funding received being $ 12 million for all of the companies on our list.
  • For AI & ML startups involved in the smart buildings market, our analysis of the total funding received since 2010 shows that Chinese companies receive the largest total amount of funding, at over $ 6.3 billion, compared to 3.9 billion. billion dollars for American companies. These two countries are by far the most important in terms of private investment in AI.

Who should buy this report?

The information in this report will be useful to anyone engaged in the management, operation and investment in intelligent commercial buildings (and their advisers) across the world. In particular, those who wish to understand exactly how AI and machine learning technologies impact commercial real estate will find it particularly useful.

Main topics covered:

1. Scope and methodology

2. An introduction to AI for smart buildings
2.1 The fundamentals of AI
2.2 IoT and Big Data
2.3 Cloud
2.4 AI hardware
2.5 Tools for AI development

3. Applications and use cases
3.1 Analysis of use cases
3.2 Security and access control
3.3 Space, occupation and movement of people
3.4 Energy management and sustainability
3.5 Predictive maintenance and FDD
3.6 Experience, comfort and productivity
3.7 Commitment, feeling and behavior
3.8 Emergency notice
3.9 Air quality and environmental analysis
3.10 Lighting
3.11 Water management
3.12 Fire safety
3.13 Elevators and escalators
3.14 Cybersecurity and device management
3.15 Digital twin and AI platforms

4. Vertical market application and use cases
4.1 Retail trade
4.2 Hospitality
4.3 Health care
4.4 Education
4.5 Airports
4.6 Data centers

5. Impact assessment of COVID-19
5.1 AI adoption and investment impacts
5.2 Impacts of smart buildings
5.3 Cyber ​​security implications
5.4 Impacts specific to vertical markets
5.5 COVID-specific applications and use cases

6. Market dynamics
6.1 Development trends
6.2 adoption trends
6.3 Maturity of the solution
6.4 The future of AI for smart buildings
6.5 Market drivers
6.6 Challenges and obstacles
6.7 Governance and ethics

7. Market size and growth prospects
7.1 Global Growth Forecast
7.2 Market Forecast by Hardware and Software
7.3 Market Forecast by Use Case
7.4 Market Forecast by Vertical
7.5 Market Forecast by Region
7.6 Regional growth indicators
7.7 Market Forecast – The Americas
7.8 Market Forecast – EMEA
7.9 Market Forecast – Asia Pacific

8. The competitive landscape
8.1 Geographical distribution of AI providers
8.2 Mapping of ecosystems
8.3 Investment trends
8.4 Strategic partnerships and alliances
8.5 Merger and acquisition activity

Companies mentioned

  • 3Divi
  • 6th Energy Technologies
  • 720 degrees
  • 75F
  • Accent
  • ACIC
  • Actuate
  • Agent VI
  • Aifi
  • Aislelabs
  • Aitek
  • Alcatraz AI
  • AllGoVision
  • amadeus
  • Amazon (Go)
  • Ambarella
  • AnyVision
  • Aquaséca
  • Aquicore
  • Arcarithm
  • Arloid Automation
  • ARM
  • Aruba Networks
  • Arup
  • AskPorter
  • Athena Security
  • Avigilon
  • Cloud Axiom
  • Axis Communications
  • AxxonSoft
  • Ayonix
  • BeeBryte
  • Bentley Systems
  • in the blink of an eye
  • BlockDox
  • BlueWave-ai
  • Bosch
  • Rock AI
  • Artificial intelligence
  • BrainChip
  • Brief camera
  • BuildingIQ
  • Cambricon Technologies
  • Carbon Lighthouse
  • CBRE
  • Brain Systems
  • Cisco
  • Cityzenith
  • Clock analysis
  • CloudMinds
  • Cloudwalk technology
  • Cognitec
  • Competent
  • cohesion
  • CopperTree
  • Dabbel
  • Dahua
  • Deep shine
  • Deep camera
  • Defense
  • Demand logic
  • Density
  • Device42
  • Digital barriers
  • Distech Controls (Acuity)
  • Eagle Eye Networks
  • Energy efficient outlook
  • Ecolibrium
  • IA element
  • Energetic
  • Envizi
  • EQuota Energy
  • Ethera
  • Everbridge
  • EVolution Networks
  • Evolution of technology
  • FaceFirst
  • Facet
  • Flir Systems
  • foghorn
  • Foobot
  • Fujitec
  • Gemalto
  • Genetec
  • Geovision
  • Google
  • Gorilla Technology Group
  • IA GoSpace
  • Graphcore
  • Green wave technologies
  • Gridium
  • Hanvon
  • Hanwha Techwin
  • Harman International
  • Hella Aglaia Mobile Vision GmbH
  • Helvar
  • Herta
  • HID Global
  • HiKVision
  • Hitachi
  • Honyewell
  • Huawei Technologies
  • IBM
  • ice
  • IDIS
  • Igor
  • Picture
  • IndigoVision
  • Infinova
  • Infogrid
  • Innovation
  • Inpixon
  • Insiteo
  • Intelligence
  • intelGlas
  • IntelliVision
  • iOmniscient
  • Ipsos Retail performance
  • Ipsotek
  • Irisys
  • IronYun
  • ISS
  • Johnson control
  • Kairos AR
  • Ketos
  • Kloudspot
  • Kognition
  • Koné
  • Leaftech
  • Leanheat
  • Logical buildings
  • Mapped
  • Matterport
  • MCloud Technologies
  • MCS Solutions (Spacewell)
  • Measurable
  • Megvii
  • Microsoft
  • Milestone systems
  • spirit tree
  • Mist (Juniper)
  • Mobotix
  • NEC Company
  • Neurotechnology
  • NiroVision
  • Nlyte software
  • Novion
  • Nozomi Networks
  • NUUO
  • Octo (Heads Upp)
  • Omnilert
  • On the semiconductor
  • OpenAI
  • OpenSensors
  • Otis
  • Panasonic
  • Passive logic
  • petasense
  • Pivot3
  • Point taking
  • Normative data
  • Qualcomm
  • Qualvision
  • Quividi
  • R&B Technology Group
  • ReconaSense
  • Resonai
  • Next Business
  • Rhombus Systems
  • Samsung
  • SAP
  • Schindler
  • Schneider Electric
  • Scylla
  • SecuriThings
  • SenseTime
  • Sensor flow
  • Sensormatic (Shoppertrak)
  • AI shapes
  • Shayp
  • Shenzhen TVT Digital Technology
  • Shepherds networks
  • Shield IOT
  • Siemens
  • Sightcorp
  • SkyFoundry
  • Smart Spaces operating system
  • Smart spaces
  • Intelligent space AI
  • SpaceIQ
  • Springboard
  • Standard cognition
  • Station A
  • Rod
  • StoneLock
  • Sunbird
  • Sunell
  • Supreme
  • Automation of switches
  • Technicians
  • Terminus Technologies
  • Thought Thread
  • Thyssenkrupp
  • Tiandy
  • Retail business
  • Trigo
  • UbiqiSense
  • uhoo
  • Ultimate
  • Umbo computer vision
  • Univue
  • V-count
  • Vaak
  • Vantiq
  • Vert-de-Gris Technologies
  • Vergesense
  • Verint
  • Verkada
  • Vertiv
  • Vigilant
  • Vintra
  • Virdi
  • VisionLabs
  • Vivotek Inc.
  • Volan technology
  • Vyntelligence
  • To win
  • Xjera Laboratories
  • Xovis
  • XY direction
  • Yardi
  • Yitu technology
  • Zebra Technologies
  • Zero Eyes
  • Zippin
  • ZKTEco
  • ZTE

For more information on this report, visit

Source link 

]]> The challenge – and the opportunity – to be a niche AI ​​cloud Mon, 09 Aug 2021 15:56:21 +0000

In this age of hyperscaler and cloud builder titans, seven of which account for roughly half of the world’s purchased IT infrastructure, it’s important to remember the importance of niches and the vital role that other system makers, of other system vendors, and other system tenants all play in the IT ecosystem. Never has a niche player been so important, and never has it been so difficult to be.

And yet, this is the path that Lambda – which is no longer called Lambda Labs because it is now a fast growing company, not an experience in building AI systems and building it. operating an AI cloud – chose herself. As niche players ourselves here at The next platform, of course we honor this choice, which is not an easy path. If you can’t be the greatest or you can’t be the first, then the only choice is to try to be the best. And really, there is no trial, as Master Yoda rightly points out. There is what to do or not to do.

If you had to pick a niche, this one is pretty good. Machine learning training models are growing exponentially, so are the data sets they depend on, and the performance of the systems that provide training is not keeping up. Which means customers have to buy more and more machines even as this equipment becomes more powerful thanks to Moore’s Law advancements in parallel computing, memory bandwidth and I / O bandwidth. Whenever software infrastructure grows faster than hardware infrastructure – the proliferation of batch processing on mainframes in the 1970s, the relational database revolution of the 1980s, the commercial Internet boom of the late 1990s and early 2000s, the explosion of data analytics in the 2000s, and the AI ​​revolution of the late 2010s and early 2020s – this is the perfect time to play a niche with specialist hardware and software and engineering expertise to keep everything buzzing.

We did an in-depth profile of Lambda in December 2020, when we spoke to Michael Balaban, the company’s co-founder and CTO, and in May of this year we looked at some price / performance metrics that Lambda released Cloud GPU Lambda Instances based on Nvidia A6000 GPU Accelerators versus Nvidia A100 GPU Instances running in the Amazon Web Services Cloud. Lambda’s argument was that the Chevy Truck’s GPU is good enough for many AI training workloads and superior to the Cadillac model in some cases. At this point, Lambda doesn’t care about the inference, and there’s no reason for it to. The company wants to build AI coaching Infrastructure. Full stop. Inference is supposed to run on internal infrastructure, and it can range from CPUs and GPUs to FPGAs and custom ASICs, and Stephen Balaban, co-founder (along with his older brother) and CEO of Lambda, is not interested in selling inference systems. At least not yet, but that can – and we think it will likely change – change. But it’s important for startups to stay very focused. You shouldn’t trust those who don’t, in fact, because both money and time are running out.

Lambda wants to ride this wave of AI training with not only specialized hardware, but also creating its own AI cloud built on its own hardware and its own software stack, obviously called Lambda Stack, which is being developed. by its own software engineers. Lambda recently secured $ 15 million in a Series A funding round plus a $ 9.5 million credit facility, giving it the funds to support its own explosive growth. Series A was led by 1517, Gradient Ventures, Bloomberg Beta, Razer, and Georges Harik – most of whom were angel investors when Lambda started nine years ago – and the credit facility came from Silicon Valley Bank.

We took the opportunity to discuss with Stephen Balaban, the CEO, the state of the company and what it sees happening in what is still a still very nascent AI training market.

“Unlike other clouds and other system vendors, we are only focusing on this particular use case, which is deep learning training,” Balaban said. The next platform. “Our product base spans from laptops to desktops, servers, clusters and the cloud, and we are vertically integrated on those devices with our own Lambda stack. But there’s another side to that, and customers need to ask themselves if they really need the gold-plated data center service experience of having Amazon Web Services as their operations team to manage the infrastructure. , because it is very expensive, as you can imagine.

This is because Lambda creates the kind of cloud that you would probably like to build on your own, if you had the skills to do it. It is designed not to use the most general GPU compute engines, as public clouds must do given the diversity of their workloads, but rather those GPUs that have enough parallel computing, enough memory capacity. and sufficient memory bandwidth at the lowest price. to reduce the total cost of ownership. When workloads run smoothly, you need to lower your total cost of ownership in a world where models and data grow faster than capacity grows. Public clouds have to massively overprovision their general-purpose machines, then sell you the idea that you should run your spiky workloads there, and then charge you a hefty premium for the privilege of doing so. It’s better and cheaper, says Balaban, to run your AI training on your own hardware (made by Lambda of course) and then process the bursts on the Lamba cloud, which is cheaper than AWS or Microsoft. Azure.

So far, this niche game has worked well, and it’s the one Lambda must have found because as a pioneer in AI software, it couldn’t afford to run its AI applications on AWS. without going bankrupt because of the instantaneous and explosive popularity of the AI ​​tools it has brought to the web. The Balaban brothers learned the hard way that sometimes success is more difficult than failure, and that’s how a niche computer hardware company and niche cloud were formed.

What Lambda is doing clearly resonates with organizations trying to master AI training and put it into production. In 2017, after being in business as an AI application maker and local cloud builder to support them, Lambda had its first full year of selling AI training systems and pulled around 3 million dollars in revenue from this material. Two years later, it’s grown to $ 30 million, and in 2021, two years later, it’s on track to make $ 60 million.

“We have found that there is a huge demand – and a growing demand – for deep learning training systems that just work,” Balaban says, and that funding for Series A is to strengthen the teams at hardware and software engineering and sales teams to really see how important this addressable market for Chevy systems is compared to the Cadillac systems that big cloud builders need to design – and bill for – because they need to support a variety of clients and workloads on their devices, where Lambda simply doesn’t.

The software is going to be a key differentiator, and the Lambda stack, which is designed to run on Debian Linux and includes Nvidia drivers and libraries such as CUDA and cuDNN as well as TensorFlow, Keras, PyTorch, Caffe, Caffe 2 and Training Frameworks. in machine learning from Theano. Along with fundraising, Lambda will expand the software that runs on its cloud and make it much more user-friendly than these frameworks (many of them developed by hyperscalers and cloud builders who seem to like Byzantine and bizarre structures like a matter of pride) are when they are released into the wild on GitHub. Ultimately, this refined AI training stack will be available for Lambda customers to deploy to their laptops and desktops, internal clusters, and the Lambda GPU cloud.

And this is the secret of the niche. The experience will be the same for Lambda customers no matter where they build their AI models. They won’t even know the difference. The market will tell Lambda how valuable such an experience is, but we can infer it from Apple’s actual experience with its music business, right?

Source link

5 essential video editing tips everyone should know Fri, 06 Aug 2021 18:46:53 +0000

If you’re a content creator of any kind, you’re probably looking for some new video editing tips. Capturing footage for your next project is only one piece of the puzzle, after all, and even the most elegant 4K scenes can fall flat in the hands of an inexperienced and poorly equipped video editor.

Plus, in this eclectic era of ever-changing technology, there are a myriad of methods to achieve the same video effects with similar software; Ultimately, refining your own artistic process is all about amassing the right tools.