All articles tagged with #gpu
A Steam Machine reportedly fails with a solid red LED about 20 minutes after setup and a software update, a sign Valve confirms points to a GPU failure. Because the GPU is soldered to the motherboard, repairs aren’t simple and typically require warranty repair or a motherboard replacement rather than swapping a graphics card.
A Steam Machine reportedly shows the Red Line of Death (RLOD) about 20 minutes after powering on, signaling a GPU failure. Steam’s support notes that RLOD patterns indicate different faults, with a mid-to-right red line pointing to a GPU issue. Because these machines use a soldered GPU, repairs typically require warranty service or motherboard-level work rather than a simple card swap, which could be slow given early batch shortages. The incident adds scrutiny to SteamOS hardware, especially as multi-platform support expands and prices for the Steam Machine remain a concern.
According to Mark Gurman, Apple will skip Pro and Max variants in the M6 line to accelerate the AI-focused M7, with the base M6 shipping in iPads and entry Macs while the M7 arrives in early 2027 and its Pro/Max by late 2027 and Ultra in 2028. The M6 will gain memory bandwidth, a faster CPU, a 12-core GPU option, upgraded video encoding/decoding, and an enhanced Neural Engine for AI tasks, while the M5 Ultra remains possible but postponed.
Allbirds has pivoted from footwear to AI infrastructure, selling its shoe business and rebranding as Smartbird with Nadia Carlsten as CEO to build a mid-market, single-tenant GPU platform for enterprises and sovereign AI. The company will source GPUs from multiple vendors and tailor infrastructure for specific customers, avoiding direct competition with hyperscalers.
Google inks a cloud deal with SpaceX to access its compute capacity, including 110,000 Nvidia GPUs, for about $920 million per month from Oct 2026 through Jun 2029, a contract that could exceed $30 billion. The agreement, tied to Google’s Gemini AI efforts and SpaceX’s expanding data-center infrastructure ahead of an IPO, includes termination rights if GPU delivery is insufficient and a 90-day notice window after Dec 31.
Nvidia has unveiled the RTX Spark, a new chip designed for Windows laptops and desktops to accelerate AI workloads and boost graphics performance, signaling a push toward on-device AI capabilities and more capable Windows PCs.
NVIDIA and AMD CEOs Jensen Huang and Lisa Su landed in Taipei ahead of Computex 2026, signaling a heavy AI and consumer-tech focus. NVIDIA is gearing up for a GTC-style Taipei event on June 1 with major AI reveals, while AMD is highlighting its 2nm Venice HPC production at TSMC and plans AI-centric and consumer-ready announcements, as both prepare for a jam-packed Computex schedule in Taipei’s Nangang district and key partner meetings.
An opinion piece arguing that Nvidia’s real moat isn’t hardware but CUDA—the software platform and libraries that lock AI workloads to Nvidia GPUs—elevating the company to a software powerhouse. CUDA’s ecosystem, PTX-level control, and a large team of software engineers create a practical edge that outperforms rivals like OpenCL, ROCm, and oneAPI in real-world AI work, akin to Apple’s ecosystem moat. While challengers exist (e.g., Modular), CUDA’s software lock-in makes Nvidia’s dominance durable beyond silicon specs.
Kaspersky revisited a 2024 study and found MD5 remains dangerously fast to crack with modern GPUs: about 60% of leaked real-world passwords could be cracked in under an hour and 48% in under a minute, highlighting the need to move to slower, purpose-built hash functions like bcrypt or Argon2 and to enable MFA or passkeys for extra protection.
ASUS’s ROG Equalizer cable for the RTX 5090 reduces overheating and improves power delivery by balancing load across all pins, lowering under-load temperatures from about 60°C to 50°C and keeping voltages near 12V with minimal variation. The tin-plated copper wires support up to around 17A per pin versus ~9A for standard cables, boosting stability and potentially overclocking performance. Priced at roughly $15, it offers a safer, more stable power solution for RTX 5090 users.
Allbirds unveiled a plan to exit its footwear business and environmental commitments to become an AI compute provider, including buying GPUs and offering GPU-as-a-service, selling footwear assets, and removing its public-benefit corporation status; the shift could impact its ESG reputation, even as the stock jumped about 582% after the announcement.
Allbirds announced a pivot from footwear to artificial intelligence, rebranding as NewBird AI and pursuing GPU-based AI compute infrastructure; the plan has sent its stock higher (intraday around 582%) while the company secures $50 million in funding and moves toward a sale to American Exchange Company, exiting its public-benefit status to become a conventional corporation and aiming to offer GPU-as-a-Service and AI-native cloud solutions—though the long-term viability of the strategy remains uncertain and the rally could be meme-driven.
Allbirds plans to pivot away from footwear by turning its assets into NewBird AI, a GPU‑as‑a‑Service and AI‑native cloud provider. After selling its name and assets to American Exchange for $39 million, the company aims to raise about $50 million to acquire high‑performance GPUs and deploy them under long‑term leases to AI customers. The move reflects the broader surge in demand for AI compute as GPU lead times grow and data-center capacity tightens, and NewBird AI’s stock jumped more than 600% intraday on the news.
Allbirds plans a dramatic pivot from shoes to AI compute, renaming to NewBird AI pending shareholder approval (vote May 18) and raising $50 million to acquire GPUs and related HPC infrastructure to become a GPU-as-a-Service AI-native cloud provider; the move also involves removing references to its environmental branding. The news sent the stock up roughly 400%, though some analysts warn the rally may be short-lived. Allbirds previously had been valued around $4 billion before selling most of its shoe business for about $39 million earlier this year, reflecting a broader trend of companies pivoting toward AI.
Allbirds plans a dramatic pivot from footwear to AI compute, renaming itself NewBird AI to build a GPU-as-a-Service platform. It will use a $50 million convertible financing facility and the sale of its IP to American Exchange Group to fund high-performance GPUs and an AI-native cloud business, with shareholder approval pending; investors have reacted positively, sending the stock up roughly 400% as the company shifts away from retail toward the AI‑compute boom.