{"product_id":"lenovo-thinkstation-pgx-nvidia-gb10-128-gb-1-tb-ssd","title":"Lenovo ThinkStation PGX – NVIDIA GB10, 128 GB, 1 TB SSD","description":"\u003cp\u003e\u003cstrong\u003eLenovo's first workstation built only for AI development.\u003c\/strong\u003e The ThinkStation PGX runs on the NVIDIA GB10 Grace Blackwell Superchip with 128 GB of unified memory, and arrives with NVIDIA DGX OS and the NVIDIA AI software stack already installed. This configuration (30KL000CBF) has a 1 TB self-encrypting NVMe SSD.\u003c\/p\u003e\u003cul\u003e\n\u003cli\u003eUp to \u003cstrong\u003e1 petaFLOP\u003c\/strong\u003e of AI performance (FP4, with sparsity), 1,000 TOPS\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e128 GB\u003c\/strong\u003e LPDDR5x unified memory at 273 GB\/s for models of up to \u003cstrong\u003e200 billion parameters\u003c\/strong\u003e\n\u003c\/li\u003e\n\u003cli\u003eLink two PGX units over ConnectX-7 for models of up to \u003cstrong\u003e405 billion parameters\u003c\/strong\u003e\n\u003c\/li\u003e\n\u003cli\u003e1 TB M.2 NVMe SSD with Opal 2.0 self-encryption\u003c\/li\u003e\n\u003cli\u003e1.13 litres, 150 × 150 × 50.5 mm, from 1.2 kg\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch3\u003eA personal AI development machine\u003c\/h3\u003e\u003cp\u003eLenovo designed the PGX for the whole AI workflow, from classic statistics and machine learning to computer vision, generative AI and large language models, robotics and agentic AI. It gives developers a local, sandboxed place to prototype, fine-tune and run inference, and it sits next to an existing workstation to add compute rather than replace it. Lenovo positions it for AI developers, data scientists, researchers and students who want local inference without the complexity and running cost of cloud resources.\u003c\/p\u003e\u003ch3\u003eReady to work out of the box\u003c\/h3\u003e\u003cp\u003eThe PGX ships with NVIDIA DGX OS (Ubuntu Linux with the NVIDIA Base OS), the NVIDIA AI software stack, CUDA 13, the GB10 Dashboard and NVIDIA AI Workbench, plus familiar tools such as PyTorch and Jupyter Notebooks. Because it runs the same NVIDIA AI platform as data-centre and cloud systems, Lenovo notes that models built on the desk can move to accelerated cloud or data-centre infrastructure with virtually no code changes.\u003c\/p\u003e\u003ch3\u003eKeep models and data on-premises\u003c\/h3\u003e\u003cp\u003ePrototyping on a local machine frees cluster and cloud capacity for training and production, and keeps intellectual property in-house. Security features include a self-encrypting NVMe drive (AES SED), TPM 2.0, an NVLink-C2C enclave, NVIDIA firmware recovery, a setup password and UEFI Secure Boot.\u003c\/p\u003e\u003ch3\u003eScale to two or three systems\u003c\/h3\u003e\u003cp\u003eTwo 200 Gbps QSFP ports on the built-in NVIDIA ConnectX-7 NIC link PGX units directly: two systems handle models of up to 405 billion parameters, and Lenovo supports direct connections between two or three PGX systems. More than three systems need a compatible QSFP switch. Lenovo's option cable for this is the ThinkStation PGX QSFP Link Cable (4X91U42988), sold separately.\u003c\/p\u003e\u003ch3\u003eSpecifications\u003c\/h3\u003e\u003ctable\u003e\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eManufacturer part no.\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e30KL000CBF\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eEAN\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e0199275392009\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eProcessor\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eNVIDIA GB10 Grace Blackwell Superchip – 20-core Arm (10× Cortex-X925 + 10× Cortex-A725)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eGPU\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eNVIDIA Blackwell architecture (integrated in GB10): Blackwell CUDA cores, 5th-gen Tensor Cores, 4th-gen RT Cores, NVENC encoder, NVDEC decoder\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eAI performance\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eUp to 1 PFLOP (FP4, with sparsity), 1,000 TOPS\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eMemory\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e128 GB LPDDR5x-8533, 256-bit, unified system memory, 273 GB\/s\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eStorage\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e1 TB M.2 2242 PCIe 4.0 x4 NVMe SSD, Opal 2.0 self-encryption\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eRear ports\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e2× QSFP 200 Gbps (NVIDIA ConnectX-7)\u003cbr\u003e1× USB-C 20 Gbps with USB PD 3.1 power input and DisplayPort 2.1\u003cbr\u003e3× USB-C 20 Gbps with DisplayPort 2.1\u003cbr\u003e1× HDMI 2.1a (multichannel audio)\u003cbr\u003e1× RJ-45 10 GbE\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eClustering\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eDirect connection of two or three ThinkStation PGX systems over ConnectX-7; more than three need a compatible QSFP switch. Lenovo option cable: ThinkStation PGX QSFP Link Cable (4X91U42988), sold separately\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eWireless\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eWi-Fi 7, Bluetooth 5.3\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eOperating system\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eNVIDIA DGX OS (Ubuntu Linux with NVIDIA Base OS)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003ePre-installed software\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eNVIDIA AI software stack, CUDA 13, GB10 Dashboard, NVIDIA AI Workbench\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eSecurity\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eSelf-encrypting NVMe (AES SED), TPM 2.0, NVLink-C2C enclave, NVIDIA firmware recovery, setup password, UEFI Secure Boot\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003ePower\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e240 W USB-C power adapter (included)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eVolume\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e1.13 litres\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eDimensions (W × D × H)\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003e150 × 150 × 50.5 mm\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eWeight\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eFrom 1.2 kg\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eColour\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eBlack\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003ePackaging\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eCarton 90% recycled and\/or FSC-certified content; cushion 90% recycled EPE\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eCertifications\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eRoHS compliant\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003e\u003cstrong\u003eIn the box\u003c\/strong\u003e\u003c\/td\u003e\n\u003ctd\u003eThinkStation PGX, 240 W power adapter with power cord, USB-C to USB-A adapters; QSFP link cable sold separately\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\u003c\/table\u003e\u003cp\u003e\u003cem\u003eSources: Lenovo ThinkStation PGX Product Guide (Lenovo Press LP2321), Lenovo PSREF for 30KL000CBF, the ThinkStation PGX Hardware Maintenance Manual and Lenovo's ThinkStation PGX product page. Images are illustrative and show the ThinkStation PGX chassis.\u003c\/em\u003e\u003c\/p\u003e","brand":"Lenovo","offers":[{"title":"Default Title","offer_id":58755699081599,"sku":"30KL000CBF","price":5773.07,"currency_code":"EUR","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1037\/5207\/3599\/files\/lenovo-thinkstation-pgx-1.png?v=1791275334","url":"https:\/\/aistorelabs.com\/products\/lenovo-thinkstation-pgx-nvidia-gb10-128-gb-1-tb-ssd","provider":"AI StoreLabs","version":"1.0","type":"link"}