Search Results: HBM

In a strategic pivot, Samsung Electronics is reportedly reallocating up to 40% of its HBM3E production capacity to mainstream DRAM, driven by higher profit margins and unprecedented demand for memory chips from the artificial intelligence industry.

Transformers have hit a scaling wall: dense models keep getting bigger, but compute and latency won’t play along. Mixture-of-Experts (MoE) offers a surgical way out—massively increasing capacity while activating only a sliver of parameters per token. Here’s how it works under the hood, why systems teams care, and what you should know before wiring MoE into your own stack.

As hyperscalers pour billions into AI, the real choke point isn’t just GPUs—it’s the fabric that connects them. From Nvidia and Broadcom to Lightmatter, Celestial AI, and PsiQuantum, a new arms race in photonic and custom interconnects is quietly deciding who will own the next decade of AI infrastructure.

The OnePlus 15 launches globally on November 13th, boasting an unprecedented 7,300mAh battery and a 165Hz OLED display, challenging flagship norms. Powered by the Snapdragon 8 Elite Gen 5, it ditches the Hasselblad partnership for an untested in-house camera system, making it a high-stakes value proposition.

Oracle announces an unprecedented 18 zettaFLOPS AI infrastructure rollout using 800,000 Nvidia Blackwell and 50,000 AMD MI450X GPUs, marking one of history's largest compute deployments. The clusters will leverage Nvidia's Spectrum-X networking and AMD's open UALink architecture, with OpenAI positioned as a key beneficiary. This massive scaling raises critical questions about practical FP4 utilization and the staggering energy demands of next-gen AI.

China's Innosilicon unveils the Fenghua No. 3 GPU, boasting RISC-V architecture, alleged CUDA compatibility, and 112GB+ HBM memory for AI workloads like large language models. This 'all-function' GPU targets gaming, medical imaging, and scientific computing, signaling a bold step in China's semiconductor self-sufficiency drive—but technical validation remains critical.