Vue normale

Reçu — 20 mai 2026 Actualités numériques

Intuit To Lay Off Over 3,000 Employees To Refocus On AI

Par : BeauHD
20 mai 2026 à 23:00
Intuit is reportedly cutting about 3,000 jobs, or 17% of its workforce, as it restructures around AI and simplifies its corporate organization. TechCrunch reports: The layoffs come during a bad year for the tech workforce. The tech industry has already cut more than 100,000 jobs this year, per Statista, and is on track to outpace both 2024 and 2025 if the layoff trend continues. Companies such as Amazon, Block, Cisco, Cloudflare, Meta, Microsoft, and Oracle have let go of thousands of employees each, all of them citing a need to refocus expenditures around AI projects as a reason to cut jobs and restructure their organizations. [...] Intuit, however, hasn't been perceived as a beneficiary of the AI boom, with its shares consistently underperforming in the broader S&P 500 over the past 12 months. The company has been caught up in the broader current of worries that traditional software-as-a-service firms will not be able to keep up or compete, as new and upcoming AI products and services threaten to change how software is developed and how it is used. In its fiscal second quarter ended January, Intuit reported revenue of $4.65 billion, a 17% increase, and net profit of $693 million, a 48% improvement compared to a year earlier. The company expects revenue to increase by about 10% in the third quarter, for which it will report results later today.

Read more of this story at Slashdot.

Google Publishes Exploit Code Threatening Millions of Chromium Users

Par : BeauHD
20 mai 2026 à 22:00
An anonymous reader quotes a report from Ars Technica: Google on Wednesday published exploit code for an unfixed vulnerability in its Chromium browser codebase that threatens millions of people using Chrome, Microsoft Edge, and virtually all other Chromium-based browsers. The proof-of-concept code exploits the Browser Fetch programming interface, a standard that allows long videos and other large files to be downloaded in the background. An attacker can use the exploit to create a connection for monitoring some aspects of a user's browser usage and as a proxy for viewing sites and launching denial-of-service attacks. Depending on the browser, the connections either reopen or remain open even after it or the device running it has rebooted. The unfixed vulnerability can be exploited by any website a user visits. In effect, a compromise amounts to a limited backdoor that makes a device part of a limited botnet. The capabilities are limited to the same things a browser can do, such as visit malicious sites, provide anonymous proxy browsing by others, enable proxied DDoS attacks, and monitor user activity. Nonetheless, the exploit could allow an attacker to wrangle thousands, possibly millions, of devices into a network. Once a separate vulnerability becomes available, the attacker could use it to then compromise all those devices. "The dangerous part here is that you can just have a lot of different browsers together that you can in the future run something on that you figure out," said Lyra Rebane, the independent researcher who discovered the vulnerability and privately reported it to Google in late 2022 in an interview. He said using the exploit code Google prematurely published would be "pretty easy," although scaling it to wrangle large numbers of devices into a single network would require more work. In the thread of Rebane's disclosure to Google, two developers said in separate responses that it was a "serious vulnerability." Its severity was rated S1, the second-highest classification. Since its reporting 29 months ago, the vulnerability remained unknown except to Chromium developers. Then on Wednesday morning, it was published to the Chromium bug tracker. Rebane initially assumed the vulnerability was finally fixed. Shortly thereafter, he learned that, in fact, it remained unpatched. While Google removed the post, it remains available on archival sites, along with the exploit code. Google representatives didn't immediately respond to an email asking how and why it published the vulnerability and if or when a fix would become available. The exploit works by abusing Chromium's Browser Fetch API to open a service worker that remains persistently active. A malicious website can trigger it through JavaScript, creating a connection that can be used "for monitoring some aspects of a user's browser usage and as a proxy for viewing sites and launching denial-of-service attacks," reports Ars. Depending on the browser, those connections "either reopen or remain open even after it or the device running it has rebooted," effectively turning the device into part of a "limited botnet."

Read more of this story at Slashdot.

AMDGPU HDMI 2.1 FRL To Be Initially Disabled-By-Default

20 mai 2026 à 21:34
One of the most exciting developments in recent times for the open-source AMDGPU kernel graphics driver is HDMI 2.1 FRL support for the AMDGPU driver along with Display Stream Compression support as they work toward providing full HDMI 2.1 support for this open-source AMD Radeon driver. The details how AMD managed to pull this feat off given prior resistance from the HDMI Forum remains to be confirmed, but it's moving ahead and out today is the latest iteration of the HDMI 2.1 FRL+DSC patches...

RHEL 10.2 Released With New AI Command Line Assistance

Par : BeauHD
20 mai 2026 à 21:00
Red Hat has released RHEL 10.2 and 9.8 with new AI-assisted command-line tools. The releases also add updated developer toolchains such as Go 1.26, LLVM 21, Rust 1.92, Python 3.14, and PHP 8.4. Phoronix reports: Red Hat Enterprise Linux has introduced the goose command for power users. Goose is an optional CLI AI assistance with model context protocol (MCP) integration. There is also improved visual output via color output enhancements. As for their rationale with the new AI integration: "The business value: Faster problem resolution, and a quicker path for new administrators to become proficient. This translates into higher developer productivity and accelerated project timelines."

Read more of this story at Slashdot.

GitHub's Internal Repos Breached Via Employee's Use of Malicious VS Code Extension

Par : BeauHD
20 mai 2026 à 20:00
Longtime Slashdot reader Himmy32 writes: GitHub has announced on X that their internal repositories have been breached through a compromised VS Code Extension on an employee's workstation. Bleeping Computer reported that the attack is linked to TeamPCP who have been in the news for a recent campaign affecting Checkmarx, Trivy, SAP, TanStack, and Bitwarden. The group appears to be attempting to sell the stolen code on cybercrime forums. "Yesterday we detected and contained a compromise of an employee device involving a poisoned VS Code extension. We removed the malicious extension version, isolated the endpoint, and began incident response immediately," the company said. "Our current assessment is that the activity involved exfiltration of GitHub-internal repositories only. The attacker's current claims of ~3,800 repositories are directionally consistent with our investigation so far." Although the investigation remains ongoing, GitHub says it has "no evidence of impact to customer information stored outside of GitHub's internal repositories." The company has also not said whether it's in contact with the hackers or if it's received a ransom demand.

Read more of this story at Slashdot.

The Very Exciting Cache Aware Scheduling Looks Like It Will Land For Linux 7.2

20 mai 2026 à 19:44
As a very exciting development ahead of the Linux 7.2 kernel merge window opening in about one month's time, it looks like the long-awaited Cache Aware Scheduling support will finally be merged! CONFIG_SCHED_CACHE has made it into a TIP branch with all the Cache Aware Scheduling code for helping with Linux performance on modern CPUs sporting multiple last level caches...

Forza Horizon 6 vs Forza Horizon 5 : comparaison directe de la qualité graphique

Sur H&Co, nous vous avons proposé pas mal de contenu au sujet du jeu Forza Horizon 6 ces derniers jours. Il y a avant tout évidemment le Performance Test de notre Thibaut national, mais nous avons également relayé celui de VCG pour continuer à vous forger un avis aussi complet que possible. Pour...

Anna's Archive Hit With Global Domain Takedown Order

Par : BeauHD
20 mai 2026 à 19:00
An anonymous reader quotes a report from TorrentFreak: A coalition of thirteen major publishers has won a massive $19.5 million default judgment against shadow library Anna's Archive. A New York federal judge fully approved the publishers' requests, issuing a broad permanent injunction that orders more than twenty specific global registries, hosts, and service providers to immediately disable the site's remaining domains. [...] At first glance, the damages award is the headline figure. Judge Rakoff granted the maximum statutory damages of $150,000 for each of the 130 "Works in Suit." This brings the final damages bill amount to a staggering $19,500,000. However, as with the $322 million judgment won by the music industry against Anna's Archive in the related Spotify case, it's highly unlikely that this money will be recouped. For now, the operators of Anna's Archive remain strictly anonymous, which doesn't help either. The default judgment (PDF) addresses this and requires the operators to unmask their identities and provide a sworn statement with valid contact information to the court within 10 days. However, since the operators have previously stated they hide their identities to avoid "decades of prison time," it is safe to assume that the operators will simply ignore this request. The true power of this default judgment lies in the permanent injunction. Anna's Archive is known to evade enforcement and change domain names when needed, so the injunction targets the technical intermediaries that keep the site online. Specifically, the injunction orders "all domain name registries and registrars of record" to permanently disable access to Anna's Archive's domains and prevent their transfer to anyone other than the publishers or the music industry plaintiffs in the related case. In addition to domain name services, the order also extends to international hosting providers, who are also ordered to stop working with the site. Leaving no room for interpretation, the order specifically names more than twenty companies and organizations. This includes familiar names like Cloudflare, Njalla, and DDOS-Guard, as well as the domain name registries of the site's current active domains [...]. The names include some intermediaries that were already listed in the Spotify default judgment, as well as new ones.

Read more of this story at Slashdot.

Seagate Sparks Memory Sell-Off As CEO Says It Would 'Take Too Long' To Build New Factories

Par : BeauHD
20 mai 2026 à 18:00
Seagate CEO Dave Mosley said Monday that building new memory chip factories or adding capacity would "take too long" to keep up with AI-driven storage demand. "If we took the teams off and started building new factories or bringing up new machines, that would just take too long. You would end up with more capacity, but then you'd slow the rate of growth on that technology," Mosely said. CNBC reports: Memory chip stocks have soared in recent months as a flood of AI investing has sent demand soaring, with the chips a key part of the AI buildout in data centers. Chip production cycles stretch over many quarters for a single unit, and investors are increasingly wary of how long the leading memory makers can capture demand. CME Group is launching a new futures market for semiconductors, enabling more traders to lock in prices and hedge against the rising prices of computing power. At Monday's conference, Mosely also addressed the "very long lead times" and maintaining predictability with its clients. "We know what's coming out a year from now," he said. "And we've basically gone to the customers and said, 'Look, if you want to plan this really well, which it should be for your data centers, we know what's coming out. You can buy this stuff up to a certain period.' And so we want to keep that four or five quarters of visibility very, very solid for what's being built. But the demand is significantly higher than that."

Read more of this story at Slashdot.

La réédition du Ryzen 7 5800X3D est disponible... enfin en Inde pour le moment

Le 16 avril 2026, nous vous partagions une photo mettant en évidence le retour du célèbre Ryzen 7 5800X3D au catalogue d'AMD. Le tout premier processeur doté de la fameuse 3D V-Cache faisant un retour triomphant pour les possesseurs de configurations AM4 souhaitant la faire durer encore quelques ann...

Yearslong Fight Over Users' Right To Tweak Smart TV Software Heads To Trial

Par : BeauHD
20 mai 2026 à 17:00
A long-running lawsuit over Vizio's Linux-based smart TV software is headed to trial in August, with the Software Freedom Conservancy arguing that GPL rules require Vizio to release complete source code owners could use to modify, maintain, or strip ads and tracking from their TVs. Ars Technica reports: The outcome could reverberate across the industry. Because many of today's popular smart TV operating systems are Linux-based, the case may help determine how much control many owners have over their sets. Access to the full code would allow users to make meaningful changes to how their TVs work, including limiting ads or deactivating automatic content recognition. [...] The Software Freedom Conservancy argues it has the right to Vizio OS's source code because it owns several Vizio TVs and because the operating system is based on Ubuntu, a Linux distribution. (SFC employees bought seven Vizio TVs from 2018 to 2021 after getting complaints about Vizio not sharing its TVs' source code, according to the complaint.) In general, the Linux kernel is provided under the terms of GPLv2, as noted by kernel.org, which is run by the Linux Kernel Organization. SFC's lawsuit alleges that Vizio breached GPLv2 and LGPLv2.1 by failing to make available the complete source code for Vizio OS. The case is currently in the Orange County Superior Court of the State of California. The lawsuit targets Vizio specifically, but the impact could extend to other Linux-based smart TV OSes such as LG's webOS, Samsung's Tizen, and Roku's Roku OS. "We expect all companies who distribute Linux and other software using right-to-repair agreements like the GPL in their products would comply with these agreements," Denver Gingerich, the director of compliance at SFC, told Ars. [...] SFC expects a ruling within three to six months of the conclusion of the trial, which is currently scheduled for August 10.

Read more of this story at Slashdot.

Regional Winners of Prestigious Literary Prize Suspected of Using Chatbots

Par : BeauHD
20 mai 2026 à 16:00
The 2026 Commonwealth Short Story Prize is facing backlash after several winning entries were accused of being AI-generated, with one Caribbean winner's story flagged as fully AI-written by a detector that WIRED says it independently confirmed. From the report: Each year, the Commonwealth Foundation, a nongovernmental organization in London, awards its short story prize to one writer in each of five regions: Africa, Asia, Canada and Europe, the Caribbean, and the Pacific. One overall winner is then selected from that short list. Regional winners take home [about $3,350], while the top winner, to be announced next month, claims [about $6,700]. On May 12, the respected UK literary magazine Granta published the top five 2026 entries -- all previously unpublished, per the rules of the contest -- on its website. (It has hosted the winning submissions for the prize since 2012.) Within days, however, one entry aroused suspicion. "The Serpent in the Grove," a story by Jamir Nazir of Trinidad and Tobago, which had taken honors for the Caribbean region, struck a few people as bearing the stylistic tells of AI-generated text. "Well, this is a first: a ChatGPT-generated story won a prestigious literary prize," wrote researcher and entrepreneur Nabeel S. Qureshi, a former visiting scholar of AI at the Mercatus Center at George Mason University, in a post on X on Monday. "'Not X, not Y, but Z' sentences everywhere, the 'hums' trope, and plenty of other obvious markers of AI writing. A major milestone for AI, at any rate..." "They say the grove still hums at noon," Nazir's mysterious and atmospheric tale begins. In his screenshot of the opening paragraphs, Quereshi highlighted the second line as what he considered to be a signature example of AI syntax: "Not the bees' neat industry or the clean rasp of cutlass on vine, but a belly sound -- as if the earth swallows a shout and holds it there." As the literary community undertook a closer read of Nazir's story, many criticized its language and metaphors as nonsensical, wondering how the Commonwealth judges could have seen any merit to them. Others shared screenshots showing that the AI-detection tool Pangram flagged "The Serpent in the Grove" as 100 percent AI-generated, a result that WIRED independently confirmed. (While no AI-detection software is perfect, third-party analysis has consistently determined Pangram to be the most accurate, with a near-zero rate of false positives.) [...] Besides Nazir, two more winning authors have drawn allegations of using AI in their work. Pangram finds that "The Bastion's Shadow," by Maltese writer John Edward DeMicoli, winner for the Canada and Europe region, is fully AI-generated; it scans "Mehendi Nights," by Indian writer Sharon Aruparayil, winner for the Asia region, as partly AI-generated. Neither DeMicoli nor Aruparayil immediately returned requests for comment when reached through their respective social media accounts. The other two short-listed stories, by Holly Ann Miller of New Zealand and Lisa-Anne Julien of South Africa, deliver "fully human-written" results from Pangram. Wired also reports that one of the judges for the prize has been "accused of using AI to craft her descriptive blurb that accompanied the listing of 'The Serpent in the Grove' as a regional winner.'" Pangram labels the text as "AI-assisted."

Read more of this story at Slashdot.

Initial Benchmarks Of The SpacemiT K3 RVA23 RISC-V CPU With The K3 Pico-ITX

20 mai 2026 à 15:43
One of the RISC-V SoCs we have been most looking forward to this year is the SpacemiT K3 that features the X100 RISC-V cores that are RVA23 compliant and among the first readily available RVA23 RISC-V platform for running on the likes of Ubuntu 26.04 LTS. In this article is a preview of some very early benchmarks of the SpacemiT K3 with the new Pico-ITX single board computer offering.

Les AMD Gorgon Halo alias Ryzen AI Max 400 Series sur le point d'arriver, avec jusqu'à 192 Go de LPDDR5X ?

Le 23 janvier 2026, nous vous présentions une rumeur venue du site HKEPC qui expliquait qu'AMD serait en train de préparer l'arrivée des successeurs des très convoités processeurs Strix Halo. Leur petit nom de code : Gorgon Halo, et qu'il s'agirait de légers refresh plus qu'autre chose. Nous sommes...

Bon Plan : Batterie UGREEN Nexode 55W 20000 mAh à 34.49€

20 mai 2026 à 15:09

Nexode 55W

Batterie UGREEN Nexode 55W – 20000 mAh – certifications CE et CCC – recharge Rapide – Câble USB Type-C plat tressé intégré – Power Delivery 45W – 15.2 x 7.1 x 3.0 cm – 470g –  Puissance max par port : ports USB Type-C 1 (câblé) et USB Type-C 2 : 55W. Port USB Type-A 3 : 22.5W. Charge possible pendant l’usage. Dispo à 34.49€.

Voir l’offre sur Amazon

Bon Plan : Batterie UGREEN Nexode 55W 20000 mAh à 34.49€ © MiniMachines.net. 2026

Le Mini PC AMD Ryzen AI Halo arrive pour répondre au DGX Spark de NVIDIA, ou pas...

20 mai 2026 à 14:26

Après le DGX Spark développé par NVIDIA et MediaTek, voilà qu'AMD prépare à son tour une machine compacte dédiée à l'IA locale. Son petit nom : Ryzen AI Halo Mini PC. Une plateforme attendue pour juin 2026, pensée autant pour les développeurs que pour les utilisateurs avancés qui veulent faire tourner des modèles IA directement en local, sous Windows comme sous Linux. Au cœur de cette machine, on retrouve un Ryzen AI Max+ 395, alias Strix Halo. AMD mise donc sur une très grosse configuration APU avec 16 cœurs Zen 5 et 32 threads, une fréquence Boost pouvant atteindre 5.1 GHz, 64 Mo de cache L3, mais aussi une partie graphique RDNA 3.5 dotée de 40 CU. À cela s'ajoute un NPU XDNA 2 capable de dépasser les 50 TOPS pour les traitements IA. Et surtout, la plateforme pourra embarquer jusqu'à 128 Go de mémoire unifiée, un élément clé pour manipuler des modèles lourds localement. […]

Lire la suite

Notre très CHER Herman Miller se lance dans le bureau Gaming, de 1095 à 1825 dollars...

20 mai 2026 à 14:03

Chez Herman Miller, on connaît déjà les fauteuils hors de prix qui font rêver les gamers, streamers et travailleurs du quotidien. Cette fois, la marque revient avec un nouveau produit, à savoir un bureau gaming motorisé baptisé COYL, un modèle réglable en hauteur qui veut clairement jouer dans la cour des grands avec un traitement PREMIUM. Et autant le dire tout de suite : à partir de 1095 dollars, il faudra aimer autant l'ergonomie que vider son compte bancaire. Un bureau motorisé qui mise beaucoup sur le design et le câble management Le nom COYL ne sort pas de nulle part. Herman Miller met énormément en avant le fameux câble d'alimentation spiralé qui accompagne le bureau et qui devient presque un élément de design à part entière. Et il est vrai que cela apporte une touche assez originale et unique à ce bureau. Le bureau profite évidemment d'un système de réglage électrique de la hauteur, mais ici pas de boutons classiques. La marque opte pour une molette rotative accompagnée d'un petit affichage LED afin de sélectionner précisément la hauteur souhaitée. Simple, propre et plutôt classe il faut l'avouer. […]

Lire la suite

Une RTX 5080 passive : c'est possible avec ce mod complétement improbable !

20 mai 2026 à 14:03

Vous ne rêvez pas un système complet, orienté gaming peut être totalement passif, c'est le concept du mod improbable de Billet Labs et jugez par vous-même, la configurait-on se compose d'un processeur AMD Ryzen 7 98000X3D, de 32 Go de mémoire DDR5, d'une carte mère Gigabyte AORUS Pro B850, d'un SSD de 2 To et d'une petite carte graphique ZOTAC Gaming RTX 5080 Solid Core. En ne s'attardant que sur le CPU et GPU, nous sommes déjà à des TDP de 120 Watts et 360 Watts, en pic... Du cuivre et des radiateurs, on a chaud ! Pour atteindre son objectif, Billet Labs a accumulé pas moins de trois radiateurs, empilés les uns sur les autres, le plus grand est au format 200 × 400 mm, le moyen a un format de 140 × 280 mm, tandis que le plus petit est au format 120 × 240 mm. La carte mère, l'alimentation et la carte graphique sont posés sur une large plaque en aluminium, divers éléments en aluminium ont été spécifiquement moulés afin de refroidir au mieux les composants. Lors des tests en conditions réelles, ils s'avèrent que le concept promet des températures un poil élevés, en jeux, sous Clair Obscur Expedtion 33, le CPU peut atteindre les 73 degrés, et le GPU les 71 degrés, dans Cyberpunk 2077, le CPU atteint un pic de température de 81 degrés et le GPU un pic de 74 degrés, la conclusion est sans appel, le concept, bien qu'intéressant, ne permet pas d'aboutir à une configuration utilisable au quotidien ! […]

Lire la suite

be quiet! Dark Rock 6, un simple tour 135 mm qui voit grand

20 mai 2026 à 13:40

Terminons notre découverte de la nouvelle génération Dark Rock de be quiet! avec le Dark Rock 6, un radiateur simple tour en 135 mm qui représente l'entrée de gamme. Mais à 89.99 U+20AC tout de même. La marque allemande revoit le positionnement tarifaire de son modèle phare et en profite pour revoir bien d'autres choses. Le style, bien entendu, mais aussi les spécifications puisqu'on passe d'un format 120 mm à un format 135 mm en n'ajoutant qu'un petit millimètre à la hauteur du radiateur. Un Silent Wings silencieux puisqu'il ne tourne pas Comme pour le Dark Rock Pro 6, on retrouve un interrupteur bien placé qui va permettre de changer la courbe PWM du ventilateur : classique en mode Performance et avec un mode 0 rpm sous 40 % du cycle PWM en mode Quiet. Un petit détail qui change tout si on ne veut pas utiliser le logiciel de la carte mère, et ça marche bien, très bien même. […]

Lire la suite
❌