{"title":"Towers","description":"\u003cp\u003eTower desktops and AI workstations with room for discrete graphics cards and expansion.\u003c\/p\u003e","products":[{"product_id":"msi-edgexpert-ms-c931-8211-ai-supercomputer-based-on-nvidia-dgx-spark-platform-for-deep-learning","title":"MSI EdgeXpert MS-C931 – AI Supercomputer Based on NVIDIA DGX Spark Platform for Deep Learning","description":"\u003cp\u003eThe MSI EdgeXpert MS-C931 is a desktop-class AI supercomputer built on the NVIDIA DGX Spark platform. It suits developers, researchers and data scientists who want data-center-grade AI capability on their own desk instead of metered by the hour in the cloud. It pairs the NVIDIA GB10 Grace Blackwell Superchip with 128GB of unified memory in a 1.19-litre chassis, and MSI adds a curated software ecosystem.\u003c\/p\u003e\u003ch2\u003eWho it's for\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eDevelopers and researchers fine-tuning, quantizing and serving models up to roughly 200B parameters\u003c\/li\u003e\n\u003cli\u003eTeams building legal AI, machine vision, offline voice translation or agentic AI products with the bundled software\u003c\/li\u003e\n\u003cli\u003eTeams with data-residency or IP concerns keeping datasets and model weights on-premises\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eKey features\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA GB10 Grace Blackwell Superchip:\u003c\/strong\u003e a 20-core Arm CPU (10x Cortex-X925 + 10x Cortex-A725) fused to a Blackwell-architecture GPU over NVIDIA NVLink-C2C, delivering 1000 AI TOPS \/ 1 petaFLOP of FP4 (sparse) performance.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e128GB unified LPDDR5x memory at 273GB\/s:\u003c\/strong\u003e a single coherent 256-bit memory pool shared between CPU and GPU, holding models up to roughly 200B parameters (and up to 405B across two stacked units via ConnectX-7).\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e1TB or 4TB self-encrypting NVMe SSD:\u003c\/strong\u003e hardware self-encryption at either tier.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA ConnectX-7 Smart NIC (200Gbps):\u003c\/strong\u003e data-center-class networking for clustering two units, plus 10GbE, Wi-Fi 7 and Bluetooth 5.4.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA DGX OS pre-installed:\u003c\/strong\u003e the full NVIDIA AI stack ships ready to run, and work is container-portable to other DGX Spark-class systems or DGX Cloud.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eCurated third-party AI software:\u003c\/strong\u003e MSI ships access to AIPLUX Legal AI Suite, Memorence Operagents (machine vision and operational AI), Ubestream offline voice translation and Galene.AI Elettra’s agentic AI platform.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eCompact and quiet:\u003c\/strong\u003e a 151 x 151 x 52mm, 1.2kg chassis with smart fan control, 4x USB 3.2 Type-C, HDMI 2.1a and 3x DisplayPort 1.4a.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eEnterprise reliability:\u003c\/strong\u003e TPM onboard, a 0–35°C operating temperature range, and FCC\/CE\/UKCA\/VCCI\/BSMI certification.\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eSpecifications\u003c\/h2\u003e\u003ctable\u003e\n\u003cthead\u003e\u003ctr\u003e\n\u003cth\u003eSpec\u003c\/th\u003e\n\u003cth\u003eDetail\u003c\/th\u003e\n\u003c\/tr\u003e\u003c\/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003eModel Number\u003c\/td\u003e\n\u003ctd\u003eEdgeXpert MS-C931\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eArchitecture\u003c\/td\u003e\n\u003ctd\u003eNVIDIA Grace Blackwell GB10 Superchip\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCPU\u003c\/td\u003e\n\u003ctd\u003e20-core Arm (10x Cortex-X925 + 10x Cortex-A725)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCPU-GPU Interconnect\u003c\/td\u003e\n\u003ctd\u003eNVIDIA NVLink-C2C\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eTensor Performance\u003c\/td\u003e\n\u003ctd\u003e1 petaFLOP (1000 AI TOPS), FP4, Sparse\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eSystem Memory\u003c\/td\u003e\n\u003ctd\u003e128GB LPDDR5x coherent unified system memory\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMemory Interface \/ Bandwidth\u003c\/td\u003e\n\u003ctd\u003e256-bit \/ 273GB\/s\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage\u003c\/td\u003e\n\u003ctd\u003e1TB or 4TB NVMe M.2 SSD with self-encryption\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eUSB\u003c\/td\u003e\n\u003ctd\u003e4x USB 3.2 Type-C\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eEthernet\u003c\/td\u003e\n\u003ctd\u003e1x RJ-45 (10GbE)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eNetwork Interface Card\u003c\/td\u003e\n\u003ctd\u003eNVIDIA ConnectX-7 Smart NIC (200Gbps)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eDisplay Connectors\u003c\/td\u003e\n\u003ctd\u003e1x HDMI 2.1a; 3x DisplayPort 1.4a via USB-C\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eVideo Encode\/Decode\u003c\/td\u003e\n\u003ctd\u003e1x NVENC \/ 1x NVDEC\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eTPM \/ Smart Fan Control\u003c\/td\u003e\n\u003ctd\u003eYes \/ Yes\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating System\u003c\/td\u003e\n\u003ctd\u003eNVIDIA DGX OS\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMax Model Size\u003c\/td\u003e\n\u003ctd\u003eUp to 200B parameters (405B via dual-unit ConnectX stacking)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eDimensions\u003c\/td\u003e\n\u003ctd\u003e151 x 151 x 52mm (1.19L)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eWeight\u003c\/td\u003e\n\u003ctd\u003e1.2kg\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating Temperature \/ Humidity\u003c\/td\u003e\n\u003ctd\u003e0–35°C \/ 10–90% non-condensing\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage Temperature \/ Humidity\u003c\/td\u003e\n\u003ctd\u003e0–60°C \/ 10–90% non-condensing\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePower Supply\u003c\/td\u003e\n\u003ctd\u003e~240W via USB-C with external adapter\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCertifications\u003c\/td\u003e\n\u003ctd\u003eFCC Class B, CE, UKCA, VCCI, BSMI\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eBundled AI Software\u003c\/td\u003e\n\u003ctd\u003eAIPLUX Legal AI Suite, Memorence Operagents, Ubestream Voice Translation, Galene.AI Elettra\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eIn the Box\u003c\/td\u003e\n\u003ctd\u003eEdgeXpert MS-C931 system, external ~240W USB-C power adapter, power cord, quick start guide\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\u003ch2\u003eGood to know\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eThe 1 petaFLOP \/ 1000 TOPS figure is a peak FP4-sparse number. Real dense-precision throughput on a given model will be lower.\u003c\/li\u003e\n\u003cli\u003eTraining a foundation model from scratch is still cluster or cloud territory. This machine is built for fine-tuning, quantization, inference serving and applied-AI development.\u003c\/li\u003e\n\u003cli\u003eA 70B model in FP4\/INT4 quantization, or a 30B model at higher precision with a long context window, runs without memory workarounds.\u003c\/li\u003e\n\u003cli\u003eChoose 1TB for a handful of active checkpoints and datasets, or 4TB for multiple quantized variants, larger training corpora or a growing library of fine-tuned checkpoints.\u003c\/li\u003e\n\u003c\/ul\u003e\u003cp\u003ePick the MS-C931 over a bare DGX Spark unit if the bundled software overlaps with what you are building.\u003c\/p\u003e","brand":"NVIDIA","offers":[{"title":"4 TB","offer_id":58232536269182,"sku":"MSI-EDGEXPERT-MSC932","price":6499.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1022\/4468\/0062\/files\/product_175686100704b1a75fc1f20b2de1b0354880526c9a_f33d2699-cf99-4e2c-97ab-a8f795e8b410.webp?v=1792094004"},{"product_id":"acer-veriton-ai-mini-workstation-personal-computer-gn100-ud11-series","title":"Acer Veriton AI Mini Workstation Personal Computer GN100-UD11 Series","description":"\u003cp\u003eThe Acer Veriton GN100 AI Mini Workstation brings server-grade AI compute to your desk instead of a data center. It is built around the NVIDIA GB10 Grace Blackwell Superchip, the same silicon at the heart of NVIDIA’s own DGX Spark platform, and delivers up to 1 petaFLOP of FP4 AI performance in a sealed chassis you can hold in one hand. It suits AI developers, researchers and businesses that need to run and fine-tune large models locally without cloud API costs, egress fees or data leaving the building.\u003c\/p\u003e\u003ch2\u003eWho it's for\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eAI developers running and fine-tuning models with 200B+ parameters locally\u003c\/li\u003e\n\u003cli\u003eLegal, healthcare, financial services and R\u0026amp;D teams working with proprietary models, regulated data or client-confidential information\u003c\/li\u003e\n\u003cli\u003eTeams running NVIDIA NIM microservices locally, pre-configured through DGX OS\u003c\/li\u003e\n\u003cli\u003eGroups that have outgrown a single high-end GPU workstation but are not ready to commission a server room\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eKey features\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA GB10 Grace Blackwell Superchip:\u003c\/strong\u003e a unified 20-core Arm CPU (10 Cortex-X925 performance cores + 10 Cortex-A725 efficiency cores) fused with a Blackwell-generation GPU carrying 6,144 CUDA cores and fifth-generation Tensor Cores.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eUp to 1 PFLOPS of FP4 AI performance (with sparsity):\u003c\/strong\u003e enough throughput to run and fine-tune models with 200B+ parameters locally.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e128GB unified LPDDR5x memory at 273GB\/s:\u003c\/strong\u003e shared between CPU and GPU, removing the CPU-GPU data transfer bottleneck that limits typical PC-plus-GPU setups.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e4TB self-encrypting NVMe SSD:\u003c\/strong\u003e secure local storage for models, datasets and containers, backed by an integrated Kensington lock and a tamper-resistant 1.2kg sealed chassis.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDual 200Gbps NVIDIA ConnectX-7 ports:\u003c\/strong\u003e direct-attach a second GN100 for combined 405B-parameter inference, or link up to four units through a RoCE 200GbE switch for a small AI cluster.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA DGX OS pre-installed:\u003c\/strong\u003e a Linux-based platform tuned for the NVIDIA AI stack, with CUDA, PyTorch, NIM microservices and the NeMo Framework ready to run out of the box.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eEveryday connectivity:\u003c\/strong\u003e Wi-Fi 7, four USB 3.2 Type-C ports, HDMI 2.1b output and a dedicated Gigabit LAN port alongside the high-speed ConnectX-7 links.\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eSpecifications\u003c\/h2\u003e\u003ctable\u003e\n\u003cthead\u003e\u003ctr\u003e\n\u003cth\u003eSpec\u003c\/th\u003e\n\u003cth\u003eDetail\u003c\/th\u003e\n\u003c\/tr\u003e\u003c\/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003eProcessor\u003c\/td\u003e\n\u003ctd\u003eNVIDIA GB10 Grace Blackwell Superchip, 20-core Arm CPU (10x Cortex-X925 + 10x Cortex-A725)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eGPU\u003c\/td\u003e\n\u003ctd\u003eBlackwell-generation GPU, 6,144 CUDA cores, 5th-gen Tensor Cores\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eAI Performance\u003c\/td\u003e\n\u003ctd\u003eUp to 1 PFLOPS FP4 (with sparsity)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMemory\u003c\/td\u003e\n\u003ctd\u003e128GB unified LPDDR5x, 273GB\/s bandwidth (shared CPU\/GPU)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage\u003c\/td\u003e\n\u003ctd\u003e4TB self-encrypting NVMe PCIe Gen5 SSD\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eNetworking\u003c\/td\u003e\n\u003ctd\u003e2x NVIDIA ConnectX-7 (200Gbps each), Gigabit LAN, Wi-Fi 7\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePorts\u003c\/td\u003e\n\u003ctd\u003e4x USB 3.2 Type-C, HDMI 2.1b\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating System\u003c\/td\u003e\n\u003ctd\u003eNVIDIA DGX OS (Linux), pre-loaded with CUDA, PyTorch, NIM microservices, NeMo Framework\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eSecurity\u003c\/td\u003e\n\u003ctd\u003eSelf-encrypting SSD, integrated Kensington lock slot, tamper-resistant sealed enclosure\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eForm Factor\u003c\/td\u003e\n\u003ctd\u003eMini workstation, approximately 1.2kg\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eClustering\u003c\/td\u003e\n\u003ctd\u003eDirect-connect two units for 405B-parameter class inference, or up to four units via a RoCE 200GbE switch\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\u003ch2\u003eGood to know\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eInference and fine-tuning run entirely on-device: nothing touches a third-party cloud, nothing is billed per token and nothing depends on an outside provider’s availability.\u003c\/li\u003e\n\u003cli\u003eBecause CPU and GPU share the same 128GB pool, model size is not capped by discrete VRAM, so you can load large open-weight models whole and iterate on fine-tuning jobs.\u003c\/li\u003e\n\u003c\/ul\u003e\u003cp\u003eChoose the GN100 when you want a GB10 system that scales from one unit to two directly connected units, or to a four-node cluster over a RoCE switch, rather than a single-box-only setup.\u003c\/p\u003e","brand":"Acer","offers":[{"title":"Default Title","offer_id":58232537153918,"sku":"ACER-VERITON-GN100","price":1049.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1022\/4468\/0062\/files\/61ILf-VLN-L._SL1500.jpg?v=1792094023"},{"product_id":"lenovo-thinkstation-pgx-gb10-1t-linux-workstation-30kl0002us","title":"Lenovo ThinkStation PGX GB10 1T LINUX Workstation (30KL0002US)","description":"\u003cp\u003eThe Lenovo ThinkStation PGX is Lenovo’s first workstation built around the NVIDIA GB10 Grace Blackwell Superchip. It gives ML engineers, researchers and applied-AI teams data-center-class model development power on their own desk, backed by Lenovo’s workstation-grade build and enterprise support. If you rent cloud GPU instances just to fine-tune or evaluate open-weight models, it moves that work into a 1.13-litre box beside your keyboard.\u003c\/p\u003e\u003ch2\u003eWho it's for\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eML engineers running LoRA\/QLoRA fine-tunes and evaluation harness runs on hardware they own\u003c\/li\u003e\n\u003cli\u003eApplied-AI teams prototyping RAG pipelines, agentic workflows, and computer-vision or multimodal models\u003c\/li\u003e\n\u003cli\u003eOrganizations handling data that cannot leave the premises\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eKey features\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA GB10 Grace Blackwell Superchip:\u003c\/strong\u003e a 20-core Arm CPU (10x Cortex-X925 performance cores + 10x Cortex-A725 efficiency cores) coupled to a Blackwell-architecture GPU with 5th Gen Tensor Cores and 4th Gen RT Cores over NVLink-C2C, rated for 1000 TOPS \/ 1 petaFLOP of FP4 AI performance with sparsity.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e128GB unified LPDDR5x-8533 memory at 273GB\/s:\u003c\/strong\u003e CPU and GPU share one coherent pool with no PCIe copy penalty, holding models up to roughly 200 billion parameters on a single unit without offloading to disk or splitting across GPUs.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDual 200Gb\/s QSFP ports (NVIDIA ConnectX-7):\u003c\/strong\u003e link two PGX units to scale past a single unit’s memory ceiling.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eSelf-encrypting NVMe storage:\u003c\/strong\u003e 1TB, 2TB or 4TB PCIe Gen4 TLC Opal SSD options, so you can size local storage to your dataset and checkpoint volume.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA DGX OS with the full AI stack pre-installed:\u003c\/strong\u003e CUDA 13, the complete NVIDIA AI software stack, the GB10 Dashboard and NVIDIA AI Workbench, using the same container stack as DGX systems, so a workload moves to a cluster or DGX Cloud unchanged.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eEnterprise-grade connectivity:\u003c\/strong\u003e 10 Gigabit Ethernet, Wi-Fi 7, Bluetooth 5.3, three USB4 ports carrying DisplayPort 2.1, and HDMI 2.1a.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e1.13-litre chassis, from 1.2kg:\u003c\/strong\u003e 150 x 150 x 50.5mm, on a single 240W USB-C adapter.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eWorkstation-grade security:\u003c\/strong\u003e AES SED NVMe, TPM 2.0, UEFI Secure Boot, an NVLink-C2C hardware enclave and NVIDIA Firmware Recovery.\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eSpecifications\u003c\/h2\u003e\u003ctable\u003e\n\u003cthead\u003e\u003ctr\u003e\n\u003cth\u003eSpec\u003c\/th\u003e\n\u003cth\u003eDetail\u003c\/th\u003e\n\u003c\/tr\u003e\u003c\/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003ePart Number (MTM)\u003c\/td\u003e\n\u003ctd\u003e30KL0002US\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eModel\u003c\/td\u003e\n\u003ctd\u003eThinkStation PGX GB10 1T LINUX\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eSuperchip\u003c\/td\u003e\n\u003ctd\u003eNVIDIA GB10 Grace Blackwell Superchip\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCPU\u003c\/td\u003e\n\u003ctd\u003eNVIDIA Grace 20-core Arm CPU (10x Cortex-X925 + 10x Cortex-A725)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eGPU\u003c\/td\u003e\n\u003ctd\u003eNVIDIA Blackwell architecture GPU with CUDA Cores, 5th Gen Tensor Cores, 4th Gen RT Cores\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eAI Performance\u003c\/td\u003e\n\u003ctd\u003e1000 TOPS \/ 1 petaFLOP (FP4, with sparsity)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMemory\u003c\/td\u003e\n\u003ctd\u003e128GB LPDDR5x-8533MT\/s unified system memory, 273GB\/s bandwidth\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage\u003c\/td\u003e\n\u003ctd\u003e1TB \/ 2TB \/ 4TB SSD M.2 2242 PCIe Gen4 TLC Opal (self-encrypting)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating System\u003c\/td\u003e\n\u003ctd\u003eNVIDIA DGX OS (Ubuntu Linux Pro base)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePre-installed Software\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\u003eMax Model Size\u003c\/td\u003e\n\u003ctd\u003eUp to 200B parameters (single unit)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eRear I\/O\u003c\/td\u003e\n\u003ctd\u003e1x USB-C power in (PD 3.1); 3x USB-C USB4 (20Gb\/s) with DisplayPort 2.1; 1x HDMI 2.1a with multichannel audio; 1x RJ-45 10GbE; 2x QSFP (NVIDIA ConnectX-7, 200Gb\/s)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eWireless\u003c\/td\u003e\n\u003ctd\u003eWi-Fi 7 (802.11be) + Bluetooth 5.3 LE\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePower Supply\u003c\/td\u003e\n\u003ctd\u003e240W USB-C power adapter\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eForm Factor\u003c\/td\u003e\n\u003ctd\u003eSmall form factor desktop, 1.13 litre, 150 x 150 x 50.5mm\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eWeight\u003c\/td\u003e\n\u003ctd\u003eFrom 1.2kg\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eSecurity\u003c\/td\u003e\n\u003ctd\u003eSelf-encrypting NVMe (AES SED), TPM 2.0, NVLink-C2C enclave, NVIDIA Firmware Recovery, AMI setup password, UEFI Secure Boot\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eWarranty\u003c\/td\u003e\n\u003ctd\u003e1-year limited onsite service\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eSustainability\u003c\/td\u003e\n\u003ctd\u003ePackaging with 90% recycled and\/or FSC-certified content; RoHS compliant\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\u003ch2\u003eGood to know\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eThe 1 petaFLOP figure is a peak FP4-with-sparsity number. Real dense-precision throughput on a given model will be lower.\u003c\/li\u003e\n\u003cli\u003eThis is a fine-tuning, inference and prototyping machine, not a foundation-model training rig. Pre-training multi-billion-parameter models from random initialization belongs on a cluster.\u003c\/li\u003e\n\u003cli\u003eA 70B-parameter model in FP4\/INT4 quantization, or a 30B model at higher precision with generous context length, loads and runs without memory juggling.\u003c\/li\u003e\n\u003cli\u003ePick 1TB, 2TB or 4TB by how many checkpoints, quantized variants and datasets you keep local.\u003c\/li\u003e\n\u003c\/ul\u003e\u003cp\u003eChoose the ThinkStation PGX if you want Lenovo’s 1-year limited onsite service and a choice of 1TB, 2TB or 4TB storage.\u003c\/p\u003e","brand":"Lenovo","offers":[{"title":"1TB SSD","offer_id":58232537284990,"sku":"30KL0002US","price":5849.0,"currency_code":"USD","in_stock":true},{"title":"2TB SSD","offer_id":58232537317758,"sku":"30KL000EUS","price":6749.0,"currency_code":"USD","in_stock":true},{"title":"4TB SSD","offer_id":58232537350526,"sku":"30KL000BUS","price":7879.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1022\/4468\/0062\/files\/27515107407_ThinkStationPGX_Front_202507310437231759166660807_179aba0a-4a17-48a0-8817-845619962d3d.png?v=1792094025"},{"product_id":"asus-ascent-gx10-ai-supercomputer-nvidia-gb10-superchip","title":"ASUS Ascent GX10 AI Supercomputer (NVIDIA GB10 Superchip)","description":"\u003cp\u003eThe ASUS Ascent GX10 puts the NVIDIA GB10 Grace Blackwell Superchip in a stackable, toolless desktop chassis. It offers DGX Spark-class local compute for ML engineers and small AI teams, with 128GB of unified memory. It sits on a desk alone or stacks with a second unit when a project outgrows one machine's memory.\u003c\/p\u003e\u003ch2\u003eWho it's for\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eFine-tuning, inference and prototyping with 70B-parameter models in FP4\/INT4 quantization\u003c\/li\u003e\n\u003cli\u003eRunning 30B models at higher precision with a long context window\u003c\/li\u003e\n\u003cli\u003eAgentic AI development with OpenClaw and NVIDIA NemoClaw\u003c\/li\u003e\n\u003cli\u003eTeams pooling two units' unified memory over ConnectX-7 200GbE\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eKey features\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA GB10 Superchip:\u003c\/strong\u003e An Arm v9.2-A 20-core CPU (10x Cortex-X925 + 10x Cortex-A725) joined to an integrated Blackwell GPU with 5th-gen Tensor Cores and 4th-gen RT Cores over NVLink-C2C, which ASUS states gives roughly 5x the bandwidth of PCIe Gen 5, rated up to 1 petaFLOP (1000 AI TOPS) of FP4 performance.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e128GB unified LPDDR5x at 273GB\/s:\u003c\/strong\u003e One coherent memory pool at 8533MT\/s shared by CPU and GPU, so large language models fit that would exceed the VRAM of most discrete workstation GPUs.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003e1TB, 2TB or 4TB NVMe storage:\u003c\/strong\u003e A user-upgradeable M.2 2242 SSD reached through the base grille.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDual-port NVIDIA ConnectX-7 (200GbE):\u003c\/strong\u003e Networking for clustering multiple GX10 units, plus a dedicated 10GbE RJ-45 port.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eToolless, stackable chassis:\u003c\/strong\u003e The power switch is on the front rather than the rear panel, which helps when units are stacked or rack-mounted.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eDual-fan cooling:\u003c\/strong\u003e Active cooling engineered to stay quiet under sustained AI workloads, with improved thermal headroom versus comparably sized systems according to ASUS.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eNVIDIA DGX OS:\u003c\/strong\u003e Ubuntu-based, with CUDA, PyTorch and TensorFlow pre-installed, and support for agentic AI frameworks including OpenClaw and NVIDIA NemoClaw.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eConnectivity:\u003c\/strong\u003e 3x USB-C 3.2 Gen2x2 (20Gbps) carrying DisplayPort 2.1, 1x USB-C with 180W PD 3.1 EPR power delivery, HDMI 2.1, Wi-Fi 7 and Bluetooth 5.4.\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eSpecifications\u003c\/h2\u003e\u003ctable\u003e\n\u003cthead\u003e\u003ctr\u003e\n\u003cth\u003eSpec\u003c\/th\u003e\n\u003cth\u003eDetail\u003c\/th\u003e\n\u003c\/tr\u003e\u003c\/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003eSuperchip\u003c\/td\u003e\n\u003ctd\u003eNVIDIA GB10 Grace Blackwell\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCPU\u003c\/td\u003e\n\u003ctd\u003eArm v9.2-A, 20-core (10x Cortex-X925 + 10x Cortex-A725)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eGPU\u003c\/td\u003e\n\u003ctd\u003eIntegrated NVIDIA Blackwell, 5th Gen Tensor Cores, 4th Gen RT Cores\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCPU-GPU interconnect\u003c\/td\u003e\n\u003ctd\u003eNVLink-C2C\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eAI performance\u003c\/td\u003e\n\u003ctd\u003eUp to 1 petaFLOP \/ 1000 AI TOPS (FP4, with sparsity)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMemory\u003c\/td\u003e\n\u003ctd\u003e128GB LPDDR5x, 8533MT\/s, 273GB\/s bandwidth\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage\u003c\/td\u003e\n\u003ctd\u003e1TB \/ 2TB (PCIe Gen4) or 4TB (PCIe Gen5) NVMe SSD, user-upgradeable M.2 2242\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eNetworking\u003c\/td\u003e\n\u003ctd\u003eDual-port NVIDIA ConnectX-7 (200GbE), 10 Gigabit Ethernet (RJ-45)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eWireless\u003c\/td\u003e\n\u003ctd\u003eWi-Fi 7, Bluetooth 5.4\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePorts\u003c\/td\u003e\n\u003ctd\u003e3x USB-C 3.2 Gen2x2 (20Gbps) with DisplayPort 2.1, 1x USB-C with 180W PD 3.1 EPR, 1x HDMI 2.1\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating system\u003c\/td\u003e\n\u003ctd\u003eNVIDIA DGX OS (Ubuntu-based), CUDA\/PyTorch\/TensorFlow pre-installed\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eAgentic AI support\u003c\/td\u003e\n\u003ctd\u003eOpenClaw, NVIDIA NemoClaw compatible\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePower supply\u003c\/td\u003e\n\u003ctd\u003e240W external USB-C adapter\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eDimensions\u003c\/td\u003e\n\u003ctd\u003e150 x 150 x 51mm\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eWeight\u003c\/td\u003e\n\u003ctd\u003e1.48kg\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eCooling\u003c\/td\u003e\n\u003ctd\u003eDual-fan active cooling\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eChassis\u003c\/td\u003e\n\u003ctd\u003eToolless, stackable, front-mounted power switch\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\u003ch2\u003eGood to know\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eThe 1 petaFLOP \/ 1000 TOPS figure is a peak FP4 number with sparsity, so real dense-precision throughput on a given model will be lower.\u003c\/li\u003e\n\u003cli\u003eThis is a fine-tuning, inference and prototyping platform, not a from-scratch foundation-model trainer. The 140W GB10 package is cooled by the dual-fan design.\u003c\/li\u003e\n\u003cli\u003eThe standard DGX OS and CUDA stack means workloads developed here are portable to other DGX Spark-class systems or DGX Cloud.\u003c\/li\u003e\n\u003c\/ul\u003e\u003cp\u003eChoose the GX10 if you want a GB10 system with a front power switch, an upgradeable M.2 slot and a stackable design. It runs Arm, so it suits teams working in the CUDA ecosystem.\u003c\/p\u003e","brand":"ASUS","offers":[{"title":"1TB PCIe NVMe SSD","offer_id":58232540168574,"sku":"ASUS-ASCENT-GX10-1TB","price":5899.0,"currency_code":"USD","in_stock":true},{"title":"2TB PCIe NVMe SSD","offer_id":58232540201342,"sku":"ASUS-ASCENT-GX10-2TB","price":6499.0,"currency_code":"USD","in_stock":true},{"title":"4TB PCIe NVMe SSD","offer_id":58232540234110,"sku":"ASUS-ASCENT-GX10-4TB","price":7899.0,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1022\/4468\/0062\/files\/new-asus-ascent-gx10-1_53208dfd-244d-4433-8f44-0aeb03890e20.jpg?v=1792094099"},{"product_id":"hp-z2-mini-g1a-workstation-amd-ryzen-ai-max-pro","title":"HP Z2 Mini G1a Workstation (AMD Ryzen AI MAX PRO)","description":"\u003cp\u003eThe HP Z2 Mini G1a is a compact, workstation-certified desktop built around AMD's Ryzen AI Max PRO processor family. HP positions it for CAD, 3D design, rendering and local AI\/LLM inference, with ISV certifications, MIL-STD durability testing and an enterprise warranty. It suits engineers, designers and ML practitioners who want workstation reliability in a footprint small enough to mount behind a monitor.\u003c\/p\u003e\u003ch2\u003eWho it's for\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eCAD, 3D design, rendering and content creation with ISV-certified applications\u003c\/li\u003e\n\u003cli\u003eLocal LLM inference and light fine-tuning on up to 128GB of unified memory\u003c\/li\u003e\n\u003cli\u003eTeams already running HP Z-series workstations who want the same procurement and support\u003c\/li\u003e\n\u003cli\u003eShared deployments, with roughly five units per 4U of rack space\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eKey features\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eAMD Ryzen AI MAX PRO options:\u003c\/strong\u003e From the Ryzen AI MAX PRO 385 (8 cores\/16 threads, up to 5.0GHz, Radeon 8050S) to the Ryzen AI MAX+ PRO 395 (16 cores\/32 threads, up to 5.1GHz, 64MB L3 cache, Radeon 8060S).\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eUp to 128GB unified LPDDR5x:\u003c\/strong\u003e One memory pool shared by CPU and GPU, with up to 96GB assignable to the integrated GPU through AMD Variable Graphics Memory.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eUp to 50 TOPS NPU:\u003c\/strong\u003e Dedicated AI acceleration on top of CPU and Radeon graphics compute.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eStorage to 8TB:\u003c\/strong\u003e Dual NVMe modules with RAID support.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eConnectivity:\u003c\/strong\u003e 1x USB4-C (40Gbps\/Thunderbolt 4), 1x USB-C (10Gbps), 2x USB-A (10Gbps), 2x Mini DisplayPort 2.1 UHBR10 driving up to four 4K displays, 2.5GbE, Wi-Fi 7, Bluetooth 5.4, and two Flex I\/O slots with ten additional I\/O module options.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eFlexible mounting:\u003c\/strong\u003e On the desk, behind a monitor, or under a desk with an optional VESA mount accessory.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eEnterprise build:\u003c\/strong\u003e MIL-STD 810H durability tested, a high-airflow 3D mesh chassis, and 60% recycled plastics \/ 20% recycled steel.\u003c\/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eWindows 11 Pro, WSL2 or Linux:\u003c\/strong\u003e Configure the OS to suit your development environment.\u003c\/li\u003e\n\u003c\/ul\u003e\u003ch2\u003eSpecifications\u003c\/h2\u003e\u003ctable\u003e\n\u003cthead\u003e\u003ctr\u003e\n\u003cth\u003eSpec\u003c\/th\u003e\n\u003cth\u003eDetail\u003c\/th\u003e\n\u003c\/tr\u003e\u003c\/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd\u003eProcessor (as configured)\u003c\/td\u003e\n\u003ctd\u003eAMD Ryzen AI MAX PRO 385 (8C\/16T, up to 5.0GHz, 32MB cache, Radeon 8050S) or AMD Ryzen AI MAX+ PRO 395 (16C\/32T, up to 5.1GHz, 64MB cache, Radeon 8060S)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eNPU\u003c\/td\u003e\n\u003ctd\u003eAMD XDNA 2, up to 50 TOPS\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMemory\u003c\/td\u003e\n\u003ctd\u003eUp to 128GB unified LPDDR5x, up to 96GB allocatable to GPU via AMD Variable Graphics Memory\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eStorage\u003c\/td\u003e\n\u003ctd\u003eUp to 8TB via dual NVMe modules with RAID support\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eDisplay support\u003c\/td\u003e\n\u003ctd\u003eUp to 4x 4K displays\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePorts\u003c\/td\u003e\n\u003ctd\u003e1x USB4-C (40Gbps\/Thunderbolt 4), 1x USB-C (10Gbps), 2x USB-A (10Gbps), 2x Mini DisplayPort 2.1 UHBR10, 2x Flex I\/O expansion slots, UAJ headset jack\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eNetworking\u003c\/td\u003e\n\u003ctd\u003e2.5GbE (with HP DASH manageability), Wi-Fi 7, Bluetooth 5.4\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eMounting\u003c\/td\u003e\n\u003ctd\u003eDesk, behind-monitor, or VESA wall\/arm mount (accessory)\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003ePower supply\u003c\/td\u003e\n\u003ctd\u003e300W internal PSU\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eDurability\u003c\/td\u003e\n\u003ctd\u003eMIL-STD 810H tested\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eSustainability\u003c\/td\u003e\n\u003ctd\u003e60% recycled plastics, 20% recycled steel, 100% sustainably sourced outer packaging, EPEAT Gold Climate+ certified\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eOperating system\u003c\/td\u003e\n\u003ctd\u003eWindows 11 Pro, WSL2, or Linux\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003ctr\u003e\n\u003ctd\u003eConfigurations available\u003c\/td\u003e\n\u003ctd\u003eAI MAX PRO 385 \/ 32GB \/ 1TB SSD; AI MAX PRO 385 \/ 64GB \/ 1TB SSD; AI MAX PRO 395 \/ 64GB \/ 1TB SSD; AI MAX PRO 395 \/ 128GB \/ 1TB SSD\u003c\/td\u003e\n\u003c\/tr\u003e\n\u003c\/tbody\u003e\n\u003c\/table\u003e\u003ch2\u003eGood to know\u003c\/h2\u003e\u003cul\u003e\n\u003cli\u003eReviewers have tested the top configuration running 120-billion-parameter class models without a discrete GPU.\u003c\/li\u003e\n\u003cli\u003eIt is not a training rig for large models from scratch, and ROCm maturity still trails CUDA's depth for the newest inference frameworks.\u003c\/li\u003e\n\u003c\/ul\u003e\u003cp\u003eChoose the Ryzen AI MAX PRO 385 configuration for CAD, office work and light AI. Step up to the AI MAX+ PRO 395 with 128GB when local LLM inference and larger models drive the purchase.\u003c\/p\u003e","brand":"HP","offers":[{"title":"AI MAX PRO 385 \/ 32GB \/ 1TB SSD","offer_id":58232540266878,"sku":"HP-Z2MINI-G1A-385-32-1TB","price":3523.99,"currency_code":"USD","in_stock":true},{"title":"AI MAX PRO 385 \/ 64GB \/ 1TB SSD","offer_id":58232540299646,"sku":"HP-Z2MINI-G1A-385-64-1TB","price":4002.38,"currency_code":"USD","in_stock":true},{"title":"AI MAX PRO 395 \/ 64GB \/ 1TB SSD","offer_id":58232540332414,"sku":"HP-Z2MINI-G1A-395-64-1TB","price":3980.29,"currency_code":"USD","in_stock":true},{"title":"AI MAX PRO 395 \/ 128GB \/ 1TB SSD","offer_id":58232540365182,"sku":"HP-Z2MINI-G1A-395-128-1TB","price":4406.83,"currency_code":"USD","in_stock":true}],"thumbnail_url":"\/\/cdn.shopify.com\/s\/files\/1\/1022\/4468\/0062\/files\/new-hp-z2-mini-g1a-1_0cf9cfd9-4610-43cb-8816-629a92edcec4.jpg?v=1792094100"}],"url":"https:\/\/computahardware.net\/collections\/towers.oembed","provider":"ComputaHardware","version":"1.0","type":"link"}