RED filed a lawsuit yesterday suing (PDF) Nikon for infringing on its video compression patents. PetaPixel reports: The lawsuit was filed in a southern California federal court today and asserts that the Japanese camera manufacturer and its United States subsidiaries have illegally infringed on seven patents that deal specifically with "a video camera that can be configured to highly compress video data in a visually lossless manner." In the filing, RED notes a type of compression that it says it has patented and is in use by Nikon in the Z9: "The camera can be configured to transform blue and red image data in a manner that enhances the compressibility of the data. The data can then be compressed and stored in this form. This allows a user to reconstruct the red and blue data to obtain the original raw data for a modified version of the original raw data that is visually lossless when demosaiced. Additionally, the data can be processed so the green image elements are demosaiced first, and then the red and blue elements are reconstructed based on values of the demosaiced green image elements." This compression comes thanks to a partnership with intoPIX's TicoRAW which was announced last December. [...] The TicoRAW feature has been in the news for months, but RED was likely waiting for it to be implemented into a competitor's camera before filing a lawsuit. RED's lawsuit says Nikon's infringement on its patent was "willful" and claims Nikon would have known about RED's patents. [...] RED then cites multiple lawsuits it has filed against Kinefinity, Sony, and Nokia over the years. RED is seeking damages or royalties for the infringement as well as an injunction to ban Nikon from further infringing.

Read more of this story at Slashdot.

"Quanta not patching vulnerable baseboard management controllers leaves data centers vulnerable," writes long-time Slashdot reader couchslug. "Pantsdown was disclosed in 2019..." Ars Technica reports: In January 2019, a researcher disclosed a devastating vulnerability in one of the most powerful and sensitive devices embedded into modern servers and workstations. With a severity rating of 9.8 out of 10, the vulnerability affected a wide range of baseboard management controllers (BMC) made by multiple manufacturers. These tiny computers soldered into the motherboard of servers allow cloud centers, and sometimes their customers, to streamline the remote management of vast fleets of computers. They enable administrators to remotely reinstall OSes, install and uninstall apps, and control just about every other aspect of the system -- even when it's turned off. Pantsdown, as the researcher dubbed the threat, allowed anyone who already had some access to the server an extraordinary opportunity. Exploiting the arbitrary read/write flaw, the hacker could become a super admin who persistently had the highest level of control for an entire data center. Over the next few months, multiple BMC vendors issued patches and advisories that told customers why patching the vulnerability was critical. Now, researchers from security firm Eclypsium reported a disturbing finding: for reasons that remain unanswered, a widely used BMC from data center solutions provider Quanta Cloud Technology, better known as QCT, remained unpatched against the vulnerability as recently as last month. As if QCT's inaction wasn't enough, the company's current posture also remains baffling. After Eclypsium privately reported its findings to QCT, the solutions company responded that it had finally fixed the vulnerability. But rather than publish an advisory and make a patch public -- as just about every company does when fixing a critical vulnerability -- it told Eclypsium it was providing updates privately on a customer-by-customer basis. As this post was about to go live, "CVE-2019-6260," the industry's designation to track the vulnerability, didn't appear on QCT's website. [...] "[T]hese types of attacks have remained possible on BMCs that were using firmware QCT provided as recently as last month," writes Ars' Dan Goodin in closing. "QCT's decision not to publish a patched version of its firmware or even an advisory, coupled with the radio silence with reporters asking legitimate questions, should be a red flag. Data centers or data center customers working with this company's BMCs should verify their firmware's integrity or contact QCT's support team for more information."

Read more of this story at Slashdot.

An anonymous reader shares an excerpt from an opinion piece, written by Macworld's Michael Simon: Apple killed its Touch Bar on the 14-inch and 16-inch MacBook Pro last year, but PC makers seem determined to prove the company wrong. First Dell introduced the XPS 13 Plus which sports a "new capacitive touch experience that allows you to switch between media and function keys easily." The laptop is available for purchase but back-ordered for weeks, and there haven't been any reviews so we don't know for sure how it will be received, but Dell's touch bar concept seems even less useful than Apple's: the buttons are static, they merely float above the actual keyboard, and they don't appear to add any functionality. Then Dell added a touch bar to the trackpad on the Latitude 9330. [...] Now there's a new PC touch bar, this time on the Voyager a1600, Corsair's first-ever gaming laptop. Corsair hasn't named or even officially announced the new feature -- it only appeared as a sneak peek -- but the company told The Verge that the strip features "10 easy-access customizable S-key shortcut buttons." [...] Corsair's Touch Bar doesn't replace the row of function keys but it is in an odd location -- on the hinge below the display. Even in pictures, it looks incredibly uncomfortable to reach. According to renders, you can still access the Touch Bar when the laptop is closed, which seems like an accident waiting to happen (not to mention a battery drain). But the biggest question I have is: why? No one shed a tear for the Touch Bar when it was killed. While it has its merits, it was never a proper pro-level feature and the implementation didn't evolve past the original idea. It was too skinny, lacked tactile feedback, required constant scrolling, and didn't actually save time. It looked nice, but even Apple didn't seem to know what to do with it. The MacBook Pro Touch Bar was one of Apple's most polarizing features and it never really caught on with developers. Maybe a niche use like gaming or video conferencing will have better results, but ultimately the Touch Bar, Apple's or otherwise, is a failed concept that should stay in the past.

Read more of this story at Slashdot.

Broadcom has signaled its $61 billion acquisition of VMware will involve a "rapid transition from perpetual licenses to subscriptions." The Register reports: That's according to Tom Krause, president of the Broadcom Software Group, on Thursday's Broadcom earnings call. He was asked how the semiconductor giant plans to deliver on its guidance that VMware will add approximately $8.5 billion of pro forma EBITDA to Broadcom within three years of the deal closing -- significant growth given VMware currently produces about $4.7 billion. And subscriptions was the answer. Krause also repeatedly said Broadcom intends to invest in VMware's key product portfolio and is pleased to be acquiring a sales organization and channel relationships that give it reach Broadcom does not currently enjoy. [...] Krause and Broadcom CEO Hock Tan both said Broadcom plans to nurture VMware's 300,000-plus customer base. The move to subscription-based licensing will apparently happen over the course of the next few years. [...] VMware may also experience slower growth in the short term due to the licensing shift. Krause said Broadcom is willing to live with lower margins for VMware than it expects from CA and Symantec, with R&D to benefit as a result. The software boss pledged ongoing investment and innovation for VMware's core infrastructure products, naming vSphere, VSAN, vRealize and NSX as the subjects of ongoing love and attention

Read more of this story at Slashdot.

Lavapipe as Mesa's software/CPU-based Vulkan implementation akin to LLVMpipe for OpenGL is now officially Vulkan 1.2 conformant...

New submitter Grokew writes: "GoodWill ransomware group propagates very unusual demands in exchange for the decryption key," reports CloudSEK. "The Robin Hood-like group is forcing its victims to donate to the poor and provides financial assistance to the patients in need." ["Once infected, the GoodWill ransomware worm encrypts documents, photos, videos, databases, and other important files and renders them inaccessible without the decryption key," reports CloudSEK.] In order for the victims to obtain the decryption keys, they must provide proof of donating to the homeless, sharing a meal with the less fortunate, and pay a debt of someone who can't afford it. [The decryption kit includes the main decryption tool, password file and a video tutorial on how to recover all important files. It's only given to infected users after the three activities are verified by the ransomware operators, who appear to be operating out of India.]

Read more of this story at Slashdot.

An anonymous reader quotes a report from Ars Technica: The IndyCar racing series is switching to an entirely renewable fuel next year. On Friday, ahead of Sunday's Indianapolis 500 race, IndyCar announced that starting next year, the race cars will be powered by a new, second-generation renewable ethanol race fuel developed by Shell. The manufacturing process for IndyCar's ethanol will be slightly less exotic than that seen in the low-carbon fuel that Formula 1 is considering for 2026. Rather than carbon capture and electrolysis, Shell will use sugarcane waste and other renewable feedstocks, which are hydrolyzed and fermented at a plant in Brazil operated by Raizen (Shell is a co-owner). Shell says that the switch "enables at least" 60 percent less carbon dioxide emissions than fossil fuel gasoline, although IndyCar currently runs on an E85 blend of gasoline rather than 100 percent fossil fuel or the 100 percent methanol that powered the sport for so many years. [...] Among other changes to help green the sport is the installation of a 150 kW DC fast charger at the Indianapolis Motor Speedway. And the roughly 5,000 tires that Firestone will transport to the track by Sunday will be hauled there from the tire maker's central Indiana warehouse by one of Penske's electric Freightliner eCascadia trucks.

Read more of this story at Slashdot.

In response to pesticides, many cockroach females have lost their taste for sweet stuff, which changes how they make the next generation of insects. From a report: When a male cockroach wants to mate with a female cockroach very much, he will scoot his butt toward her, open his wings and offer her a homemade meal -- sugars and fats squished out of his tergal gland. As the lovely lady nibbles, the male locks onto her with one penis while another penis delivers a sperm package. If everything goes smoothly, a roach's romp can last around 90 minutes. But increasingly, cockroach coitus is going really, weirdly wrong, and is contributing to roach populations in some places that are more difficult to vanquish with conventional pesticides. Back in 1993, scientists working at North Carolina State University discovered a trait in the German cockroach, a species that inhabits every continent except Antarctica. Specifically, these new cockroaches seemed to have no affection for a form of sugar called glucose, which was strange because -- as anyone who has ever battled against a cockroach infestation knows -- cockroaches normally cannot get enough of the sweet stuff. So, where did these new, health-conscious cockroaches come from? It seems we created them by accident, after decades of trying to kill their ancestors with sweet powders and liquids laced with poison. The cockroaches that craved sweets ate the poison and died, while cockroaches less keen on glucose avoided the death traps and survived long enough to breed, thus passing that trait down to the next cockroach generation. "When we think of evolution, we usually imagine wild animals, but actually, it's also happening with small animals living in our kitchens," said Ayako Wada-Katsumata, an entomologist at North Carolina State University. Dr. Wada-Katsumata and her colleagues have just introduced yet another wrinkle to the cockroach's story: According to a study published this month in the journal Communications Biology, the same trait that might help a female cockroach avoid sweet-tasting poison baits also makes her less likely to stick around and mate with normal cockroach males. This is because cockroach saliva is capable of rapidly breaking down complex sugars, like those found in the male's courtship offering, and turning them into simple sugars, such as glucose. So when one of these glucose-averse females takes a bite of the male's nuptial gift, it literally turns bitter in her mouth, and she bolts before he can complete the double barrel lock-and-pop maneuver. "Great!" you may be thinking. "The fewer cockroach hookups, the fewer infestations we'll have." Not so fast, said the researchers. "As to how this will affect the population, it's really complicated," said Dr. Wada-Katsumata. That's because, despite the hang-ups, glucose-averse cockroaches still find ways to do the deed.

Read more of this story at Slashdot.

Elon Musk has announced that Starlink, the satellite internet service launched by SpaceX, has been approved in Nigeria and Mozambique. From a report: This news is coming three days after Musk answered a tweet about the service launch in Africa. "Yes, first countries in Africa to be announced coming soon," he tweeted. "Starlink will serve everywhere on Earth that we're legally allowed to serve."

Read more of this story at Slashdot.

Some academic researchers in Russia are quietly working to prevent colleagues who have supported their country's invasion of Ukraine from being elected to the Russian Academy of Sciences this month. From a report: If they succeed, they will deny those who back the war a prized credential that confers prestige in Russian institutions of higher learning. Their campaign could also show that some acts of protest remain possible despite a government crackdown on dissent. The Russian Academy of Sciences is a nonprofit network of research institutes in a variety of disciplines across the Russian Federation. It has just under 1,900 members in Russia and nearly 450 nonvoting foreign members. The academy elects new members every three years. The upcoming poll, starting on Monday, is for 309 seats, including 92 for senior academicians and 217 for corresponding members. The competition is steep: More than 1,700 candidates have applied. This month, a group of Russian researchers started circulating a list of dozens of candidates who have publicly supported Russia's invasion of Ukraine by signing pro-war declarations or letters their universities or institutions released or by making such statements themselves. Hundreds of high-ranking officials at Russian universities, most of whom were administrators rather than prominent scientists, also signed a letter in support of the war in March. But many academic researchers have taken an antiwar stance. More than 8,000 Russian scientists and science journalists have signed an open letter opposing the invasion since it was first published in February. Three academic researchers -- who were not identified because they risk job loss, imprisonment and their safety by publicly opposing the war -- said in interviews that they helped create the list of those who supported the war to prevent them from being elected to the academy.

Read more of this story at Slashdot.

Inside of hard disk drives are platters which hold all your data; these are all manufactured by one company in Japan called Showa Denko which has announced it expects to "realize near-line HDD having storage capacity of more than 30TB" by the end of 2023. From a report: Deciphering that statement, we'd assume it will provide platters with a storage capacity of more than 3TB, sometime in 2023, to partners such as Toshiba, Seagate and Western Digital, who will then produce the hard disk drives, targeting hyperscalers and data centers operators. We'd expect some of them to end up in NAS and 3.5-inch external hard drives, but that won't be the main target markets, as performance is likely to be optimized for nearline usage. Showa Denko has now started shipment of the platters that will go into new 26TB Ultrastar DC HC670 UltraSMR hard disk drives announced by Western Digital only a few days ago. A 2.6TB platter -- which uses energy-assisted magnetic recording and shingled magnetic recording -- also marks an important milestone as it hits the symbolic 1TB/in^2 density. Showa Denko's announcement comes as a surprise as Toshiba recently suggested 30TB drives (rather than higher capacities) would not come until 2024. A 30TB model would comprise of 11 platters with 2.73TB capacities each, a slight improvement on the 2.6TB capacity that are on the way. Given the fact that 26TB HDDs have now been announced in the first half of 2022, there's a remote chance that we could see 30TB drives before the end of the year or (as the saying goes), depending on market conditions.

Read more of this story at Slashdot.

Russia's communications regulator Roskomnadzor said on Friday it had opened administrative cases against Alphabet's Google and six other foreign technology companies for alleged violations of personal data legislation. From a report: Moscow has clashed with Big Tech over content, censorship, data and local representation in a simmering dispute that has erupted into a full-on information battle since Russia sent tens of thousands of troops into Ukraine on Feb. 24. Russia fined Google 3 million roubles ($46,540) last year for not storing the personal data of Russian users in databases on Russian territory, and on Friday said it had opened a new case over what it called Google's repeated failure to comply with Russian legislation.

Read more of this story at Slashdot.

An anonymous reader shares a report: With the 2020 release of the Xbox Series X and PlayStation 5, we've started to see the era of console games that finally make full use of TVs capable of 4K resolutions (i.e., "Ultra HD" 3840Ã--2160 pixels) that have become increasingly popular in the marketplace. Now, though, at least one TV manufacturer is already planning to support 8K-capable consoles (i.e., 7680Ã--4320 resolution) that it thinks could launch in the next year or two. Polish gaming site PPL reports on a recent public presentation by Chinese TV and electronics maker TCL. Tucked away in a slide during that presentation is a road map for what TCL sees as "Gen 9.5" consoles coming in 2023 or '24. Those supposed consoles -- which the slide dubs the PS5 Pro and "New Xbox Series S/X" -- will be capable of pushing output at 8K resolution and up to 120 frames per second, according to TCL's slide.

Read more of this story at Slashdot.

An anonymous reader shares a report: With the 2020 release of the Xbox Series X and PlayStation 5, we've started to see the era of console games that finally make full use of TVs capable of 4K resolutions (i.e., "Ultra HD" 3840Ã--2160 pixels) that have become increasingly popular in the marketplace. Now, though, at least one TV manufacturer is already planning to support 8K-capable consoles (i.e., 7680Ã--4320 resolution) that it thinks could launch in the next year or two. Polish gaming site PPL reports on a recent public presentation by Chinese TV and electronics maker TCL. Tucked away in a slide during that presentation is a road map for what TCL sees as "Gen 9.5" consoles coming in 2023 or '24. Those supposed consoles -- which the slide dubs the PS5 Pro and "New Xbox Series S/X" -- will be capable of pushing output at 8K resolution and up to 120 frames per second, according to TCL's slide.

Read more of this story at Slashdot.

The HID subsystem changes were merged this week into the Linux 5.19 kernel...

SteamOS 3.2 débarque et avec lui une mise à jour des fonctionnalités prévues pour la console. Certaines sont plus intéressantes que d’autres avec par exemple une possibilité de limiter la fréquence d’affichage à 10, 20 ou 40 images par seconde qui est censée augmenter l’autonomie de la machine.

J’ai un peu de mal a voir qui va accepter un engin qui tournera à 10 images par seconde en échange des quelques minutes d’autonomie en plus. Un passionné de Point & Clic peut être ? Un fan de jeu de puzzle ? Cela servira probablement à certains utilisateurs et je suppose que si les équipes de Valve le proposent c’est qu’il doit y avoir une demande.

Mais l’élément le plus intéressant dans cette nouvelle mise à jour est l’arrivée d’un système de ventilation plus intelligent. Alors cela ne veut pas dire qu’un cerveau a poussé sur le ventilateur mais que les ingénieurs en charge du système ont décidé de créer des scénarios d’usage plus poussés pour piloter l’alimentation de celui-ci. Quand le Steam Deck est peu sollicité, le ventilateur sera moins alimenté ce qui limitera le bruit qu’il génère et évitera de vider inutilement la batterie.

Quand le processeur est plus sollicité, la ventilation suivra en montant dans les tours mais peut être de manière plus fine qu’auparavant. Ce qui évitera également une situation où la console basculera immédiatement dans un mode survie en poussant sa ventilation au maximum. Le système indique également une réaction plus rapide aux différents évènements. On imagine donc que le lancement d’un jeu aura un impact plus mesuré et qu’après les sollicitations de la phase de chargement, la ventilation sera rapidement plus calme. Les premiers retours sont assez intéressants et montrent une vraie évolution de ce poste. De quoi jouer tranquillement sans énerver trop vos voisins ? 

J’ai parfois l’impression de voir apparaitre avec le Steam Deck ce que j’ai constaté avec les consoles de salon de Nvidia, les Shield. Un excellent élève qui va continuer a travailler son projet après l’avoir rendu pour examen. Et c’est peut être ce qui fera la différence sur le long terme. Valve a les moyens et les ambitions de proposer un excellent materiel mais pourrait surtout proposer une solution logicielle de mieux en mieux adaptée au jeu et au materiel qui l’entoure.

Au vu des annonces récentes d’AMD et de ses puces Mendocino, il est possible qu’une nouvelle gamme de produits se développe dans le sillage des Steam Deck. Et quelle meilleure solution employer pour les piloter qu’un système conçu sur mesures pour elles ?

AMD Mendocino : des puces abordables sous Zen 2 et RDNA 2

 

Une mise à jour fait taire le Steam Deck © MiniMachines.net. 2022.

A top Federal Reserve official gave a stark warning to House lawmakers on Thursday: Move too slow in issuing a central bank digital currency and the dollar's global dominance could eventually be in jeopardy. From a report: "We shouldn't take the dollar's global status as the dominant payment currency for granted," Lael Brainard, the Fed's vice chair, said at a congressional hearing on central bank digital currencies (CBDCs). "If major foreign jurisdictions move to the issuance of their own digital currencies, it's important to think about whether the United States would continue to have the same kind of dominance without also issuing. I would hate for Congress to decide five years from now: 'You, Federal Reserve, you need to catch up. China's out there. The [European Central Bank] is out there.'" The Fed just wrapped a public comment period on its highly-anticipated report laying out the pros and cons of developing its own CBDC. This came amid the explosion in popularity of stablecoins, which aim to tie their value to a fiat currency (like the U.S. dollar).

Read more of this story at Slashdot.

Avec son FSR, AMD a pris une voie philosophique identique à celle de son meilleur ennemi NVIDIA : calculer l'image dans une définition inférieure pour gagner des images par seconde, et la restituer à la définition supérieure de votre écran par le biais d'algorithmes (celui de Lanczos principalement). La différence entre les deux est importante : les rouges font faire ces calculs aux unités classiques là où NVIDIA avec son DLSS le fait par ses tensor cores... [Tout lire]

Le Throttling est quelque chose d’assez simple, c’est ce qu’il se produit quand une puce chauffe trop. Pour se mettre en sécurité au delà d’une certaine température, le processeur baisse sa fréquence afin de baisser sa consommation et par la même occasion la chaleur qu’il dégage. 

Le Lenovo Yoga 9i sous Core i7-1260P

Le processeur propose sa fréquence maximale, monte en consommation, chauffe, atteint un pic, baisse sa fréquence, refroidit et, lorsqu’il a suffisamment baissé sa température peut éventuellement remonter en horloge pour entamer un nouveau cycle.

Problème, les nouveaux processeurs Intel Alder Lake-P sont conçus pour être intégrés dans des châssis très fins pour des machines ultramobiles : tablettes, ultraportables millimétriques et autres engins du genre. Des châssis tellement fins que le dispositif de refroidissement choisi peut impacter énormément le comportement des puces et c’est ce que tend à démontrer NotebookCheck.

Le Core i5-1240P devant le Core i7-1260P

Ainsi le test d’un Lenovo Yoga Slim 7 Pro 14 sous Intel Core i5-1240P offre des résultats surprenants parce que, au final, cette puce s’avère plus rapide que des processeurs i7-1260P et même i7-1270P intégrés dans d’autres machines. Le TDP nominal de toutes ces puces est de 28 watts avec une consommation maximale pouvant varier mais souvent presque 3 fois plus importantes. La limite théorique du Core i5-1240P est ainsi de 64 watts. Des chiffres, surtout le second, qui font peur quand on songe à des intégrations dans des tablettes ou dans des consoles mobiles. 

Et le résultat est très dépendant des capacités de refroidissement mises en œuvre par les constructeurs. Les processeurs cités sont tous construits de la même manière avec 4 coeurs P et 8 coeurs E. Suivant les modèles, seules les capacités graphique du chipset Xe et leurs fréquences peuvent réellement les différencier. Mais quand ces fréquences sont volontairement baissées par les processeurs eux même afin de ne pas monter trop haut en température, leurs performances théoriques peuvent s’avérer totalement inutiles.

Sur ce test Cinebench, on a  un processeur Core i7-1260P en rouge qui est toute le temps derrière un Core i5-1240P. Les engins ne sont pas les mêmes bien qu’ils soient signés tous les deux par Lenovo. Le Yoga 9i qui embarque le Core i7 est un 14″ hybride haut de gamme bénéficiant d’un double ventilateur avec caloduc. Malgré tout, on voit que sa performance décroit rapidement avant de se stabiliser sous celle  d’un Core i5-1240P intégré dans un Yoga Slim 7 Pro de même diagonale du même constructeur.  Le Core i7-1270P intégré dans un Thinpad X1 Yoga G7 en noir est également largement en dessous des deux autres références alors que cette puce est censée proposer plus de performances globales. 

Le ThinkPad X1 Yoga Gen 7 de Lenovo sous Core i7-1270P

Et cela n’est pas dû spécifiquement aux puces puisque le Core i5-1240P du Galaxy Book 2 Pro en vert, puce identique à celle qui est devant en Orange, se retrouve en dernière position du classement.

En bleu, une solution de 11e génération en Core i7-11370H intégrée dans un portable Surface Laptop Studio montre l’évolution de la gamme. Ces puces ne sont pas mauvaises, leurs performances sont tout à fait intéressantes face à la solution de génération précédente mais cela montre l’importance capitale de la construction de ces PC. Le fait qu’un unique constructeur comme Lenovo propose trois machines qui s’avèrent si différentes dans leur comportement et qui vont proposer plus de performances dans leurs déclinaisons moyenne gamme que haut de gamme est, je pense, suffisamment explicite.

Le Samsung Galaxy Book2 Pro est ici un très mauvais élève

La qualité d’assemblage, le système de ventilation et, tout bêtement, la place disponible pour effectuer le refroidissement de chaque puce seront primordiaux. Bien plus que la fiche technique pure et dure annonçant telle ou telle fréquence de processeur. Il vaut mieux un Core i5-1240P qui fonctionne réellement à 3.3 GHz dans la durée qu’un Core i7-1260P qui décroche de son 3.4 GHz quasi immédiatement faute  d’espace pour souffler.

Le site NotebookCheck présente d’autres tableaux et en particulier des comparaisons avec les solutions AMD en gammes Ryzen-U et Ryzen-HS. Là encore, les résultats sont intéressants bien qu’il faille évidemment prendre en compte les spécificités des machines testées et ne pas en faire une règle absolue. La consommation de chaque dispositif est à mesurer ainsi que l’espace de refroidissement de chaque engin. Les Ryzen 9 5900HS sont annoncés comme ayant un TDP de 35 watts de base, les Ryzen 7 6800U restent quant à eux dans une enveloppe allant de 15 à 28 Watts de TDP.

Vous l’aurez compris, il est impossible de déterminer ici si telle ou telle puce est plus performante ou moins performante qu’une autre en règle générale. La seule leçon à retenir ici est l’importance de regarder un futur achat sur l’ensemble de ses aspects. Il est parfois nécessaire de prendre du recul sur la fiche technique qui va classer les éléments de manière abrupte. Mais si un processeur sera toujours plus rapide qu’un autre dans un laboratoire, cela ne sera pas forcément le cas dans la vraie vie. Celle où les puces étouffent et baissent de fréquence jusqu’à proposer un résultat plus faible qu’une machine moins chère.

Peu d’écart de prix entre les deux versions de cette machine mais laquelle est la plus performante ?

L’examen d’un châssis, des solutions de ventilation, du comportement de l’engin au global ont toujours beaucoup de sens même si la fiche technique permet de faire un premier tri. Personnellement, j’ai toujours préféré acheter une machine plus silencieuse à une solution plus fine par exemple. Je sais qu’un engin plus épais aura souvent le gros avantage de mieux dissiper la chaleur des composants internes même si il est moins « esthétique » suivant les critères actuels. C’est un détail qui m’assure également de choisir un portable efficace et non pas un résultat de test de laboratoire sans rapport avec le réel.

La différence de performance du premier tableau entre un Samsung Galaxy Book 2 Pro 13 et Yoga Slim Pro 14 est suffisamment révélatrice du problème. Dans les deux cas, la puce est la même mais la première propose un niveau de performances totalement bridé alors que la seconde arrive à exprimer tout le potentiel voulu par Intel.

Des Core i5-1240P plus rapides que des Core i7-1260P ? © MiniMachines.net. 2022.