Major hardware manufacturers are planning significant price increases for servers and PCs due to soaring memory component costs. According to channel sources, server prices could jump 15% while PCs see a 5% hike as demand from AI datacenters strains supply.
NVIDIA’s CFO, Colette Kress, recently addressed key industry concerns. She dismissed talk of an AI bubble, downplayed competition from custom ASIC chips, and confirmed the Vera Rubin AI platform is taped out and on schedule for the second half of 2026.
Intel’s CFO is expressing new confidence in the company’s foundry business. He says yields on the critical 18A node are improving predictably and that advanced packaging is attracting strategic customer conversations.
The open-source firmware ecosystem is seeing significant movement on two fronts. 9elements is now maintaining support for Intel’s older Xeon Scalable processors in Coreboot, and 3mdeb has ported its Dasharo firmware to a modern ASRock Rack motherboard.
According to The Verge, Valve has been quietly funding open-source projects since 2016 to make Windows PC games run on Arm chips. This tech, including the Fex emulator, is what powers the new Steam Frame VR headset and could enable Steam games on future phones and laptops. The goal is to eliminate t
Nvidia is pushing its latest DLSS 4 tech into new and old games this week, with Where Winds Meet leading the charge. They’re also running a flashy giveaway for a one-of-a-kind RTX 5090. It’s a classic feature update with a major hardware tease.
Avalue has unveiled the EMS-ARH, a fanless edge computing system powered by Intel’s Core Ultra processor. It delivers up to 99 TOPS of AI performance for discriminative tasks like classification and anomaly detection. The rugged box is designed for flexibility and harsh environments.
Samsung has finally unveiled the long-rumored Galaxy Z TriFold smartphone. It opens to a 10-inch tablet-like display and will launch in the US in the first quarter of 2026, likely costing over $2,400.
The openSUSE project is taking two significant steps forward for users. Its rolling release, Tumbleweed, is starting to get support for Intel’s latest AI-focused NPU hardware. Meanwhile, the upcoming stable Leap 16 release has been promised a full 24 months of maintenance and community support.
Support for AMD GPU managed memory has been officially merged into the upcoming GCC 16 compiler. This feature, part of the OpenMP offloading support, allows for a unified memory space between host and device. It aims to make programming for AMD’s ROCm platform more straightforward.
