Under the fast-changing technology trend, no matter which way China's AI industry chooses to build its ecosystem or stay compatible with CUDA, the elimination and consolidation processes...
Nvidia's exclusive CUDA ecosystem has pushed competitors aside, forcing international hardware and software giants to form alliances to challenge the GPU giant's dominance. Server...
Nvidia announced at GPU Technology Conference (GTC) that it will launch the Blackwell architecture B100 and B200 GPUs in the second half of 2024, and will be manufactured by TSMC's...
Generative AI has driven huge demand for High-Performance Computing (HPC) chips and AI servers, turning Taiwan into the hardware distribution center for Generative AI hardware in...
Players in the memory industry are fiercely competing with each other in developing the fifth-generation High Bandwidth Memory (HBM) HBM3E due to the intense demand from AI GPUs....
Nvidia recently warned developers against using its CUDA on non-Nvidia GPUs, a move where the company tries to fend off other competitors. It also highlights another challenge in...
TSMC has been increasing its monthly production capacity for CoWoS (chip-on-wafer-on-substrate) advanced packaging, allowing it to increase AI GPU output for Nvidia and AMD, which...
Nvidia's many customers, including Open AI CEO Sam Altman, are trying to find ways to cut dependency on its AI chips. While specialists have pointed out that Altman's lack of understanding...
Although AI PC has become a hot topic the market has been following lately, AMD's AI GPUs, such as the MI300, were the center of attention during AMD's latest earnings call.
A significant ramp-up in CoWoS production capacity at TSMC is anticipated to propel server companies' AI server sales to new heights in 2024, according to industry sources.
