AI 模型和平台
NVIDIA 详细情况哈萨克斯坦 GPU 部署和机器人计划
在 Google 上的首选来源中添加 Unite.AI
NVIDIA 称哈萨克泰莱科姆是哈萨克斯坦最大的电信连，已部署400台NVIDIA H200和480台NVIDIA B300 GPU，计划明年在另外两个地点各部署1000多台GPU. 这些数字出现于2026年10月22日的一篇文章中，副总裁为NVIDIA Rev Lebaredian，并在公司的GTC柏林新闻序列中发表。
部署和计划部署的GPU能力
哈萨克电信公司(Hakhistelecom)是全国最大的电信公司，明年，公司计划在数据中心谷地部署1000多个GPU，同时在埃基巴图兹地区部署1000多个NVIDIA Vera Rubin GPU. NVIDIA认为，这种基础设施旨在服务于企业，开发商和机构可以大规模使用的各种工作量。
Lebaredian的帖子证明哈萨克斯坦的人工智能生态系统正在从基础设施规划转向部署，该帖子指出，更多的本地能力为开发者和公司提供了培训微调并服务能力日益增强的模型的空间，成功的应用反过来又创造了对更计算的需求。
阿斯塔纳机器人毒物勘探
在哈萨克斯坦，阿斯塔纳是第一个部署城市。 公司将评估将Pony.ai的第七代机器人轴引入InDrive应用的可行性，由NVIDIA AI基础设施支持，该基础设施建立在NVIDIA Hyperion传感器上，并用NVIDIA Halos安全系统计算平台。
Pony.ai的机器人平台有文献记载的计算历史。 在2026年4月25日的公告中，自主驾驶技术公司称与NVIDIA合作开发了升级后的4级自主驾驶域控制器，该控制器基于NVIDIA DRIVE Hyperion平台，由NVIDIA DRIVE AGX Thor与NVIDIA NVLink供电。 采用灵活的单芯片和多芯片配置，公司称，平台可以链接两个DRIVE 托尔芯片用于4000FP4 TFLOPS的综合最大计算性能。 Pony.ai表示，自2017年起与NVIDIA合作，发布专有汽车级计算单元于2022年由NVIDIA DRIE Orin芯片提供动力，并于2025年缩放为它所描述的世界上首个四级域控制器，其质量生产的第七代机器人Axi拥有4个DRIE Orin X芯片。
2025年AI工厂和2026年6月协议
哈萨克泰莱科姆的主权AI建设在本周宣布前开始。 2025年5月29日，名为中亚首家商业主权AI工厂的宣布 发布公司，由NVIDIA H200 GPU提供动力，估计总性能为56.4 petaflops(FP16)，计划于2025年夏季上线。 平台将作为云服务向哈萨克斯坦全国的政府机构、私营企业、创业企业和大学提供，Bagdat Mussin 称董事会主席，当时该平台旨在作为哈萨克斯坦数字主权的基础。
2026年6月15日，奥尔扎斯·别克捷诺夫总理会见了勒巴雷迪安和火鸟联合创始人拉兹米格·霍瓦希米安和亚历山大·叶赛扬。 据总理办公室称，会议之后签署了关于人工情报和数字基础设施合作协定，总额达100亿美元。 其中包括人工情报和数字开发部与火鸟部之间的战略合作协定，为人工智能生态系统发展建立框架，在阿莱姆。艾建立哈萨克斯坦火鸟实验室，以及哈萨克电信公司与火鸟公司之间的捆绑定期表，概述数据中心谷项目内合作的技术和组织参数。
政府发布将数据中心谷描述为埃基巴斯图兹的旗舰项目，已有300兆瓦的电力容量，设计允许分阶段扩建至1GW. Bektenov指出，2026年被宣布为哈萨克斯坦数字化和人工智能年，该国在该周通过了数字化卡扎克斯坦国家数字转型和AI开发战略。
副总理兼人工智能和数字发展部长扎斯兰·马迪耶夫·称计划投资100亿美元，将使哈萨克斯坦能够建立一个由包括NVIDIA GB300和Vera Rubin在内的10万GPU芯片供电的大规模计算集群，估计该项目将至少产生30亿美元的年度出口收入。 火鸟CEOHovaghimian 称计划2027年的火鸟发布将让哈萨克斯坦成为AI基础设施世界前十强国家之一。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
AI Models & Platforms
NVIDIA Details Kazakhstan GPU Deployments and Robotaxi Plans
Add Unite.AI to your preferred sources on Google
NVIDIA said Kazakhtelecom, Kazakhstan’s largest telecommunications company, has deployed 400 NVIDIA H200 and 480 NVIDIA B300 GPUs in Astana and Almaty, with plans to deploy more than 1,000 additional GPUs at each of two further sites next year. The figures appeared in an October 22, 2026 post bylined to NVIDIA Vice President Rev Lebaredian and published in the company’s GTC Berlin news sequence.
Deployed and Planned GPU Capacity
The post described Kazakhtelecom as the nation’s largest telecommunications company and stated it has already deployed 400 H200 and 480 B300 GPUs across the two cities. Next year, the company plans to deploy more than 1,000 GPUs in the Data Center Valley, along with more than 1,000 NVIDIA Vera Rubin GPUs in the Ekibastuz region. According to NVIDIA, that infrastructure is intended to serve diverse workloads that enterprises, developers and institutions can use at scale.
Lebaredian’s post framed the announcements as evidence that Kazakhstan’s AI ecosystem is moving from infrastructure planning to deployment, spanning what NVIDIA describes as the full AI stack: infrastructure, models and applications. More local capacity, the post stated, gives developers and companies room to train, fine-tune and serve increasingly capable models, with successful applications in turn creating demand for more compute.
Robotaxi Exploration in Astana
The post also disclosed that inDrive and Pony.ai are collaborating to explore robotaxi services in Kazakhstan, with Astana as the first deployment city. The companies will assess the feasibility of bringing Pony.ai’s seventh-generation robotaxis to the inDrive app, supported by NVIDIA AI infrastructure built on the NVIDIA Hyperion sensor and compute platform with the NVIDIA Halos safety system.
Pony.ai’s robotaxi platform has a documented compute history. In an April 25, 2026 announcement, the autonomous-driving technology company said it had developed an upgraded Level 4 autonomous-driving domain controller in collaboration with NVIDIA, built on the NVIDIA DRIVE Hyperion platform and powered by NVIDIA DRIVE AGX Thor with NVIDIA NVLink. With flexible single-chip and multi-chip configurations, the company said, the platform can link two DRIVE Thor chips for a combined maximum computing performance of 4,000 FP4 TFLOPS. Pony.ai stated it has partnered with NVIDIA since 2017, launched proprietary automotive-grade computing units powered by NVIDIA DRIVE Orin chips in 2022, and scaled in 2025 to what it described as the world’s first Level 4 domain controller featuring four DRIVE Orin X chips for its mass-produced seventh-generation Robotaxi.
The 2025 AI Factory and the June 2026 Agreements
Kazakhtelecom’s sovereign AI buildout began before this week’s announcement. On May 29, 2025, the company announced the launch of what it called Central Asia’s first commercial sovereign AI factory, powered by NVIDIA H200 GPUs with an estimated total performance of 56.4 petaflops (FP16) and scheduled to go live in summer 2025. Kazakhtelecom said the platform would be delivered as a cloud service to government institutions, private businesses, startups and universities throughout Kazakhstan, with full access to the NVIDIA AI Enterprise software suite. Chairman of the Board Bagdat Mussin said at the time that the platform was intended to serve as the foundation for Kazakhstan’s digital sovereignty.
On June 15, 2026, Prime Minister Olzhas Bektenov met Lebaredian and Firebird co-founders Razmig Hovaghimian and Alexander Yesayan. According to the Prime Minister’s office, agreements on artificial intelligence and digital infrastructure cooperation totaling $10 billion were signed following the meeting. They included a Strategic Cooperation Agreement between the Ministry of Artificial Intelligence and Digital Development and Firebird, establishing a framework for AI ecosystem development and the creation of Firebird Labs Kazakhstan at Alem.ai, and a Binding Term Sheet between Kazakhtelecom and Firebird outlining the technical and organizational parameters of cooperation within the Data Center Valley project.
The government release described the Data Center Valley as a flagship project in Ekibastuz, with 300 MW of power capacity already available and a design allowing phased expansion to 1 GW. Bektenov noted that 2026 has been declared Kazakhstan’s Year of Digitalization and Artificial Intelligence and that the country had adopted the Digital Qazaqstan national strategy for digital transformation and AI development that week.
Deputy Prime Minister and Minister of Artificial Intelligence and Digital Development Zhaslan Madiyev said the planned $10 billion investment would enable Kazakhstan to establish a large-scale computing cluster powered by 100,000 GPU chips, including NVIDIA GB300 and Vera Rubin, and estimated the project would generate at least $3 billion in annual export revenue. Firebird CEO Hovaghimian said a launch planned for 2027 would place Kazakhstan among the world’s top ten countries in AI infrastructure.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.