The 2024 annual work trends report released by Microsoft and LinkedIn shows that employers' preferences are shifting towards candidates with AI skills rather than those with extensive industry experience. 71% of corporate executives are inclined to hire job seekers with AI capabilities, but only 25% of companies plan to offer AI training, leading to 75% of knowledge workers learning AI tools on their own. With work pressure intensifying, 46% of employees consider resigning, and 45% worry that AI will replace their jobs. AI is reshaping the workplace, requiring professionals to actively learn and adapt.

Recently, MLPerf has added tests for training large language models and generating images based on text. Nvidia, Intel, and Google have all invested in large systems to participate in the new benchmark tests. Nvidia's 10,000 GPU system completed GPT-3 training in the shortest time. Intel's Gaudi 2 has significantly improved its 8-bit floating-point computing performance but still lags behind Nvidia's system. Nvidia and Microsoft's systems based on tens of thousands of GPUs are engaged in fierce competition for GPT-3 training. Training generative AI systems requires immense computing power, and companies are competing vigorously.