Tom’s Hardware Verdict The MSI MEG Ai1300P PCIE5 is an impressive ATX 3.0-compliant PSU that excels in power quality and thermal performance. However, its high price and increased fan noise under heavy loads may be drawbacks for some users. Great efficiency Amazing voltage filtering Very resilient to high temperatures Long warranty Compact for its power output Noisy under heavy load Limited fan size Loose 12V regulation Only one 12+4 pin GPU connector...
It seems that MSI is planning to bundle yellow-tipped 16-pin power adapters with its upcoming Nvidia GeForceRTX 50-series graphics cards, including the RTX 5070 Ti,RTX 5080, andRTX 5090. This move aims to address issues related to power delivery and overheating, which plagued some users of previous-generation adapters. According to a report fromVideoCardz,these adapters are designed to provide a visual guide, ensuring users fully insert the connectors into the GPU for safe operation....
Back in 2023,we reportedthat NASA had lost telemetry and thus alignment, steering controls, and usable research data from its historic Voyager 1 probe as it drifted deeper into space, leaving its fate uncertain— but today, a newNASA blog postgives us some more hope for the Voyager. The agency now has a full readout of the ship’s FDS memory, which could hopefully lead to a fix. NASA never actually lost communication with the device....
AI enthusiast青龍聖者has discovered two fascinating graphics cards in China: the GeForce RTX 4090D 48GB and GeForce RTX 4080 Super 32GB. The mysterious SKUs are clearly modified versions of theGeForce RTX 4090DandGeForce RTX 4080 Super, which contend with thebest graphics cards. As the model names already indicated, the GeForce RTX 4090D 48GB and GeForce RTX 4080 Super 32GB have double the memory of their regular versions. More VRAM is beneficial when dealing with AI workloads....
Nvidia has announced a competition for May The Fourth 2025, which should mean at least two PC enthusiasts win a freeGeForce RTX 5090by the end of the day. It is very easy to enter, asking you to only make public your allegiance with an attached #RTXON tag. We also guess it is worthwhile making your reply grab the attention of the Nvidia, in a good way. The ultimate #MayThe4thBeWithYou showdownWhat side of the galaxy are you choosing?...
Nvidia caught the imaginations of many technology enthusiasts on Tuesday with its show-closing, ‘just one more thing’- style reveal of theProject Digits AI supercomputer. But seasoned graphics chip designer Raja Koduri and disruptive AI server startup Tiny Corp were not enchanted by the small golden box or blinded by Jensen’s shimmering jacket. Both have publicly scorned Nvidia’s Grace Blackwell AI Supercomputer. At first glance, the purported provision of 1 PFLOPS of performance in such a tiny device might seem like a vision of the future, even if it is priced at $3,000....
TheGeForce RTX 4060 Ti 16GBand theRadeon RX 7700 XT, two of thebest graphics cards, are slowly dropping in price. This is good news for consumers who want a mid-range gaming graphics card and have been waiting for a deal. Competition is always good for the consumer. The GeForce RTX 4060 Ti 16GB and the Radeon RX 7700 XT are two mid-range graphics cards that have been competing for some time now, so prices have improved — somewhat....
Dell, one of the world’s largest server makers, has spilled the beans on Nvidia’s upcoming AI GPUs, codenamed Blackwell. Apparently, these processors will consume up to 1000 Watts, a 40% increase in power over the prior-gen, requiring Dell to use its engineering ingenuity to cool these GPUs down. Dell’s comments might also hint at some of the architectural peculiarities of Nvidia’s upcoming compute GPUs. “Obviously, any line of sight to changes that we are excited about what’s happening with the H200 and its performance improvement,” said Yvonne Mcgill, Dell’s chief financial officer....
Jensen Huang of Nvidia gave a sneak peek at what the trillion-dollar GPU company is planning to do with future iterations ofDeep Learning Super Sampling(DLSS). During a Q&A session atComputex2024 (reported byMore Than Moore), Huang answered a DLSS-related topic, saying that in the future, we will see generated textures and objects that will be created purely through AI. Huang also stated that AI NPCs will also be generated purely through DLSS....
Expanding its wings into the automotive ecosystem, Intelsharedits upcoming SDV (Software-Defined Vehicle) SoC designs at Auto Shanghai 2025 yesterday. Slated to be the industry’s first disaggregated design, the company presented its second-generation SDV SoC, internally codenamed Frisco Lake. A detailed investigation by3elife, a Chinese tech and news publication, purports these SoCs are derivatives of Intel’s Panther Lake design, with their successor allegedly based on Nova Lake. Software-defined vehicles are automobiles where a majority of the functionality is handled through software, rather than traditional physical, mechanical, or electronic components....
Memory specialist Patriot has launched a new PCIe 5.0 SSD to compete with thebest SSDson the market. The Patriot Viper PV573 is now official, but we’ve seen it before. We enjoyed ahands-onwith the device duringCES2024 back in January. Though several months have passed, the Viper PV573’s attractions remain the same: it will be available in 1TB, 2TB, and 4TB capacities, includes a chunky blower cooler for comfortable thermals under load, and delivers up to 14,000 MB/s sequential reads....
Phison’s booth at GTC 2024 held an unexpected surprise: The company demoed a single workstation with four GPUs using SSDs and DRAM to expand the effective memory space for AI workloads, allowing it to run a workload that typically requires 1.4TB of VRAM spread across 24 H100 GPUs. The company’s new aiDaptiv+ platform is designed to lower the barriers of AI LLM training by employing system DRAM and SSDs to augment the amount of GPU VRAM available for training, which Phison says will allow users to accomplish intense generative AI training workloads at a fraction of the cost of just using standard GPUs, albeit trading the lower cost of entry for reduced performance and thus longer training times....
Among the various RAID levels, RAID 0 and RAID 1 are the most basic ones. Understanding their data distribution is simple and the setup is comparatively less tricky. Like any other RAID type, their primary purpose is to provide better speed, redundancy, and fault tolerance. While RAID 0 is based on disk striping, RAID 1 follows the disk mirroring or shadowing principle. The former doesn’t provide redundancy and is completely performance-driven while the latter falls behind read/write speed but offers redundancy....
