“Découvrez la magie des TPU Google Cloud, y compris une vue exceptionnelle des centres de données où se déroule toute l’action. Nos clients utilisent des Cloud TPU pour exécuter certaines des charges de travail d’IA les plus importantes au monde, et cette puissance ne se résume pas à une simple puce. Dans cette vidéo, découvrez les composants du système TPU : mise en réseau de centres de données, commutateurs de circuits optiques, systèmes de refroidissement à eau, vérification de la sécurité biométrique, etc.”
Étiquette : cloud (Page 1 of 2)
“Amazon Bedrock provides you the flexibility to choose from a wide range of FMs built by leading AI startups and Amazon so you can find the model that is best suited for what you are trying to get done. With Bedrock’s serverless experience, you can get started quickly, privately customize FMs with your own data, and easily integrate and deploy them into your applications using the AWS tools and capabilities you are familiar with”
“La loi extraterritoriale américaine FISA -pour Foreign Intelligence Surveillance Act-, totalement ignorée par la communication gouvernementale et celle de Bleu et de S3ns, a pourtant été spécialement conçue pour les entreprises non-américaines. De plus, FISA s’applique au cloud, et permet aux services de renseignement américains de déposer des « backdoors » (portes dérobées) sur tout logiciel américain pour accéder aux données s’ils le souhaitent”
“We identified four best practices that reduce energy and carbon emissions significantly — we call these the “4Ms” — all of which are being used at Google today and are available to anyone using Google Cloud services.
- Model. Selecting efficient ML model architectures, such as sparse models, can advance ML quality while reducing computation by 3x–10x.
- Machine. Using processors and systems optimized for ML training, versus general-purpose processors, can improve performance and energy efficiency by 2x–5x. Mechanization. Computing in the Cloud rather than on premise reduces energy usage and therefore emissions by 1.4x–2x. Cloud-based data centers are new, custom-designed warehouses equipped for energy efficiency for 50,000 servers, resulting in very good power usage effectiveness (PUE). On-premise data centers are often older and smaller and thus cannot amortize the cost of new energy-efficient cooling and power distribution systems.
- Mechanization. Computing in the Cloud rather than on premise reduces energy usage and therefore emissions by 1.4x–2x. Cloud-based data centers are new, custom-designed warehouses equipped for energy efficiency for 50,000 servers, resulting in very good power usage effectiveness (PUE). On-premise data centers are often older and smaller and thus cannot amortize the cost of new energy-efficient cooling and power distribution systems.
- Map Optimization. Moreover, the cloud lets customers pick the location with the cleanest energy, further reducing the gross carbon footprint by 5x–10x. While one might worry that map optimization could lead to the greenest locations quickly reaching maximum capacity, user demand for efficient data centers will result in continued advancement in green data center design and deployment.
These four practices together can reduce energy by 100x and emissions by 1000x.”
Source : Google AI Blog: Good News About the Carbon Footprint of Machine Learning Training
“The Mandiant brand is synonymous with unmatched insights for organizations seeking to keep themselves secure in a constantly changing environment,” said Thomas Kurian, CEO, Google Cloud. “This is an opportunity to deliver an end-to-end security operations suite and extend one of the best consulting organizations in the world. Together we can make a profound impact in securing the cloud, accelerating the adoption of cloud computing, and ultimately make the world safer.”
“Google this evening began alerting Takeout users about the “technical issue.” From November 21-25, 2019, those that requested backups could have had videos in Google Photos “incorrectly exported to unrelated users’ archives.” In requesting a backup, some of your videos — but not pictures — might be visible to random users that were also downloading their data through Google Takeout.”
Source : Google Photos video backups sent to strangers last year – 9to5Google
Wow, this is cool: The duck-rabbit illusion apparently works on Google Cloud Vision. The system interprets it one way or the other, depending on the orientation of the image. pic.twitter.com/IoiftlRNrQ
— Steve Stewart-Williams (@SteveStuWill) October 22, 2019
“Python code to submit rotated images to the Cloud Vision API + R code for visualizing it. This repository was used to create this animation.”
“Le groupe vient de signer, discrètement, un accord avec Huawei qui s’occupera désormais du cloud des clients français de l’opérateur. Les services secrets, qui notaient la proximité entre la firme et Pékin, s’étranglent : les données des clients iront directement sur les serveurs chinois. L’opérateur nuance : cela ne concerne pas les entreprises, et, toujours selon le Canard, c’est à cause de Free, « si on veut rester compétitifs sans licencier, il faut bien qu’on noue des alliances avec les acteurs les moins chers du marché » se plaignait un cadre de l’opérateur.”
Source : Orange et Huawei : une amitié à géographie variable – Cyberguerre
“Research shows that 80 percent of business workloads have yet to move to the cloud, held back by the proprietary nature of today’s cloud market. This prevents portability of data and applications across multiple clouds, data security in a multi-cloud environment and consistent cloud management. IBM and Red Hat will be strongly positioned to address this issue and accelerate hybrid multi-cloud adoption. Together, they will help clients create cloud-native business applications faster, drive greater portability and security of data and applications across multiple public and private clouds, all with consistent cloud management.”
“Car makers are collecting massive amounts of data from the latest cars on the road. Now, they’re figuring out how to make money off it. With millions of cars rolling off dealer lots with built-in connectivity, auto companies are gaining access to unprecedented amounts of real-time data that allow them to track everything from where a car is located to how hard it is braking and whether or not the windshield wipers are on.
The data is generated by the car’s onboard sensors and computers, and then stored by the auto maker in cloud-based servers. Some new cars have as many as 100 built-in processors that generate data”.
Source : What Your Car Knows About You – WSJ