SpaceX 算力扩张目标或需超两百万块 Nvidia Rubin GPU

Aioga Editorial Team

AI资讯 / 行业动态

返回 AI资讯

SpaceX 算力扩张目标或需超两百万块 Nvidia Rubin GPU

The Decoder：AI News（RSS）Aioga 编辑团队2026-08-05T14:15:50.000Z热度 67

SpaceX 计划到 2027 年底将算力提升五倍以上，总功率超过两吉瓦，全部基于 Nvidia 芯片。现有 Colossus 集群约 54 万块 GPU，若全部采用 Vera...

行业动态The Decoder：AI News（RSS）

今日 AI 情报摘要

SpaceX 计划到 2027 年底将算力提升五倍以上，总功率超过两吉瓦，全部基于 Nvidia 芯片。

现有 Colossus 集群约 54 万块 GPU，若全部采用 Vera Rubin 平台，新目标所需 GPU 可能超过两百万块。 SpaceX 的 AI 业务（含 xAI 和 X）Q2 营收 25.6 亿美元，但运营亏损仍达 12.6 亿美元。

中文正文 · AI 翻译

根据埃隆·马斯克的说法，SpaceX希望在2027年底前将其计算能力增加五倍以上，使年终计算能力超过两吉瓦。整个建设计划将全部使用英伟达芯片。

马斯克在SpaceX第二季度财报电话会议上分享了这一目标。他说，到2027年底，容量应接近十吉瓦而不是五吉瓦。根据公司季度报告，目前容量为1.4吉瓦。SpaceX计划基于英伟达即将推出的Vera Rubin平台进行建设，马斯克称这是现有最好的架构。现有的Colossus集群运行在英伟达的H100、GB200和GB300系统上。

SpaceX的IPO文件显示，目前和计划中的Colossus建设大约包括100,000个H100、110,000个GB200和110,000个GB300处理器，另外计划至少增加220,000个GB300单元。总计约540,000个GPU。新的容量目标将使计算超过一百万个Rubin GPU。如果整个扩展都使用Vera Rubin，数量可能超过两百万。SpaceX尚未说明建设中有多少实际会使用Rubin。

马斯克的人工智能公司xAI于2026年2月通过一笔大部分以股票融资的交易并入SpaceX。交易中，SpaceX的估值为1万亿美元，xAI为2,500亿美元，总合计为1.25万亿美元。长期来看，马斯克希望在轨道上运行数据中心。

SpaceX的人工智能部门，包括xAI和X，在2026年第二季度实现了25.6亿美元收入，但运营亏损为12.6亿美元。在第一季度，该部门在仅有8.18亿美元收入的情况下亏损24.7亿美元。2026年上半年，该部门的运营亏损总计37.3亿美元。

大部分收入增长并非来自Grok，而是来自新的云合同，这些合同出租Colossus计算能力。一位未具名的AI客户约占该收入中的15.2亿美元。该客户很可能是Anthropic：https://the-decoder.com/anthropic-taps-spacexs-colossus-1-data-center-for-220000-gpus-to-power-claude/。

竞争对手将他们的计算分布在更多的供应商上。Anthropic 使用 AWS Trainium、Nvidia GPU 和 Google 的 TPU 来训练和运行 Claude：https://the-decoder.com/google-moves-billions-in-anthropic-chip-risk-off-its-balance-sheet/。该公司已公开宣布使用 Trainium 和 Graviton 系统在 AWS 上增加最多五千兆瓦的基础设施：https://the-decoder.com/amazon-pours-33b-into-anthropic-which-promises-to-spend-100b-right-back-on-aws/，以及大约 3.5 千兆瓦的额外 TPU 容量。OpenAI 已宣布六千兆瓦的 AMD：https://the-decoder.com/amd-signs-a-long-term-deal-to-supply-openai-with-multiple-generations-of-instinct-gpus/ 容量、至少十千兆瓦的 Nvidia：https://the-decoder.com/openai-and-nvidia-announce-10-gigawatt-partnership-for-ai-infrastructure/ 容量，以及大约两千兆瓦的 AWS Trainium 容量：https://the-decoder.com/openais-up-to-110-billion-raise-lines-up-almost-exactly-with-the-111-billion-it-just-added-to-its-cash-burn-forecast/。

保持对 AI 的了解。清晰、有用，无废话。

关注 The Decoder 以获取 AI 新闻、背景故事和专家分析。

解码器：https://the-decoder.com/

SpaceX wants to grow its compute capacity more than fivefold by the end of 2027, finishing the year above two gigawatts of power, according to Elon Musk. The entire buildout will run on Nvidia chips.

Musk shared the target during SpaceX's Q2 earnings call. By the end of 2027, capacity should be closer to ten gigawatts than five, he said. The company currently sits at 1.4 gigawatts, according to its quarterly report. SpaceX plans to build on Nvidia's upcoming Vera Rubin platform, which Musk called the best architecture available. The existing Colossus clusters run on Nvidia's H100, GB200, and GB300 systems.

SpaceX's IPO filings put the current and planned Colossus buildout at roughly 100,000 H100, 110,000 GB200, and 110,000 GB300 processors, plus at least 220,000 more planned GB300 units. That's about 540,000 GPUs total. The new capacity target would push the math well past one million Rubin GPUs. If the entire expansion used Vera Rubin, the number could exceed two million. SpaceX hasn't said how much of the buildout will actually be Rubin.

Musk's AI company xAI merged into SpaceX in February 2026 through a deal funded mostly with stock. The transaction valued SpaceX at $1 trillion and xAI at $250 billion, for a combined $1.25 trillion. Long-term, Musk wants to run data centers in orbit.

SpaceX's AI segment, which includes xAI and X, posted $2.56 billion in revenue during Q2 2026 alongside an operating loss of $1.26 billion. In Q1, the segment had lost $2.47 billion on just $818 million in revenue. For the first half of 2026, operating losses in the segment totaled $3.73 billion.

Most of the revenue jump didn't come from Grok. It came from new cloud contracts that lease out Colossus compute capacity. A single unnamed AI customer accounted for roughly $1.52 billion of that revenue. That customer is likely Anthropic：https://the-decoder.com/anthropic-taps-spacexs-colossus-1-data-center-for-220000-gpus-to-power-claude/.

Competitors spread their compute across more providers. Anthropic uses AWS Trainium, Nvidia GPUs, and Google's TPUs：https://the-decoder.com/google-moves-billions-in-anthropic-chip-risk-off-its-balance-sheet/ to train and run Claude. The company has publicly announced up to five gigawatts of additional AWS infrastructure with Trainium and Graviton systems：https://the-decoder.com/amazon-pours-33b-into-anthropic-which-promises-to-spend-100b-right-back-on-aws/, plus about 3.5 gigawatts of extra TPU capacity. OpenAI has announced six gigawatts of AMD：https://the-decoder.com/amd-signs-a-long-term-deal-to-supply-openai-with-multiple-generations-of-instinct-gpus/ capacity, at least ten gigawatts of Nvidia：https://the-decoder.com/openai-and-nvidia-announce-10-gigawatt-partnership-for-ai-infrastructure/ capacity, and around two gigawatts of AWS Trainium capacity：https://the-decoder.com/openais-up-to-110-billion-raise-lines-up-almost-exactly-with-the-111-billion-it-just-added-to-its-cash-burn-forecast/.

Stay in the loop on AI. Clear, useful, no fluff.

Follow The Decoder for AI news, background stories and expert analyses.

The Decoder：https://the-decoder.com/

情报判断

Aioga 编辑摘要

SpaceX拟在2027年底前将算力容量扩大五倍以上，功率超过2吉瓦并采用Nvidia芯片。按现有约54万块GPU及Vera Rubin方案估算，若扩建全部使用该平台，需求可能超过200万块，但具体采用比例尚未披露。

背景分析

材料称SpaceX当前容量为1.4吉瓦，Colossus现有及规划配置约含10万块H100、11万块GB200、11万块GB300和至少22万块拟增GB300。其AI业务包含xAI和X，第二季度营收25.6亿美元、运营亏损12.6亿美元。

Aioga 观点

Aioga判断，超过200万块Rubin GPU是基于“扩建全部采用Vera Rubin”的情景推算，并非SpaceX已公布的采购量。更确定的信息是算力扩张方向、功率目标和Nvidia技术路线，芯片数量仍存在较大不确定性。

影响与后续

若相关扩建按计划推进，可能带来规模可观的Nvidia芯片需求，并提高SpaceX在AI训练及算力租赁领域的基础设施投入。值得关注的是，AI业务营收增长与持续运营亏损并存，扩张所需投入及利用效率仍是关键变量。后续应重点核验SpaceX对2027年底功率目标、Vera Rubin实际采用比例、GPU采购与部署节奏的进一步披露，同时跟踪云计算合同收入是否持续、客户集中度是否变化，以及AI业务运营亏损能否收窄。