特斯拉CEO埃隆·马斯克近日在社交平台X上透露，特斯拉公司在全球英伟达H100芯片储备量上仅次于Meta公司，对一份将特斯拉与xAI排名低估的报告表达了不满。据用户“The Technology Brother”在X平台上发布的数据，Meta已囤积了多达35万块H100 GPU，高居榜首。马斯克指出，如果计算准确，特斯拉应排在第二位，而xAI则应位列第三。
马斯克的言论暗示，特斯拉目前可能持有3万至35万颗H100 GPU，这一数量显著超过了原先报告中提及的1万颗。相比之下，xAI的储备量估计在2.6万到3万颗之间。英伟达的H100芯片因其在人工智能和高性能计算领域的应用而备受追捧，因此，科技巨头们对这类芯片的大规模囤积反映出对先进计算能力的激烈竞争。
特斯拉和Meta的这种大规模囤积行为，不仅凸显了AI和自动驾驶等前沿技术对高性能芯片的依赖，也揭示了科技公司在全球芯片短缺背景下，为确保技术研发和生产能力所采取的策略。随着技术的不断发展，芯片储备已成为衡量公司竞争力的重要指标，而马斯克的言论无疑将引发业界对芯片分配和使用透明度的讨论。
英语如下：
News Title: “Musk Reveals: Tesla Could Become World’s Second-Largest H100 GPU Military-Armed Company, Challenging Meta’s Chip Hoarding Dominance”
Keywords: Elon Musk, Tesla, H100 GPU
News Content: Tesla CEO Elon Musk recently disclosed on the social platform X that Tesla ranks second globally in NVIDIA H100 chip reserves, trailing only Meta, and expressed dissatisfaction with a report that underestimated Tesla’s position in the xAI rankings. According to data posted by user “The Technology Brother” on X, Meta has amassed as many as 350,000 H100 GPUs, topping the list. Musk pointed out that, if the numbers are accurate, Tesla should be in second place, with xAI following in third.
Musk’s comments imply that Tesla may currently hold between 30,000 and 350,000 H100 GPUs, a figure significantly higher than the originally reported 10,000. In comparison, xAI’s estimated reserves are believed to be between 26,000 and 30,000. NVIDIA’s H100 chip is highly sought after due to its applications in artificial intelligence and high-performance computing, indicating a fierce competition among tech giants for advanced computing power.
This large-scale hoarding by Tesla and Meta underscores the reliance of cutting-edge technologies like AI and autonomous driving on high-performance chips and the strategies tech companies adopt to secure research and development capabilities amidst a global chip shortage. As technology continues to evolve, chip reserves have become a crucial indicator of a company’s competitiveness, and Musk’s remarks are certain to spark industry discussions on chip allocation transparency and usage.
【来源】https://www.ithome.com/0/760/681.htm
Views: 1