NVIDIA AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GB10 desktop AI development platform
CPU / platform NVIDIA GB10; 20-core Arm CPU
GPU / graphics Integrated Blackwell GPU
Memory 128 GB LPDDR5x unified memory
Storage Up to 4 TB NVMe M.2 SSD
Form factor / display 150 × 150 × 50.5 mm mini desktop Suitable for Local LLM inference, model testing and AI development
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA AI PCs / workstations
Official NVIDIA DGX Station platform reference image; actual OEM chassis and configuration are confirmed in the quotation. 1 images GB300 deskside AI workstation
CPU / platform GB300; Grace 72-core Neoverse V2
GPU / graphics Blackwell Ultra; 252 GB HBM3e
Memory 496 GB CPU memory; 748 GB total coherent memory
Networking ConnectX-8; up to 800 Gb/s Ethernet
Form factor / display Deskside AI supercomputer platform Suitable for Large-model development, fine-tuning and shared AI compute
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Compact GB10 AI development workstation
CPU / platform NVIDIA GB10; 20-core Arm CPU
GPU / graphics Integrated NVIDIA Blackwell GPU
Memory 128 GB LPDDR5x unified memory
Storage 1 TB or 4 TB self-encrypting NVMe SSD
Form factor / display 1.13 L; 150 × 150 × 50.5 mm Suitable for Desktop AI model development and local inference
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Tower workstation; select GPUs for the application
CPU / platform Up to dual 4th Gen Intel Xeon Scalable
GPU / graphics Up to 4 RTX PRO 6000 Blackwell Max-Q GPUs, configuration dependent
Memory Up to 2 TB DDR5 ECC RDIMM
Storage M.2 NVMe SSD and SATA storage options
Form factor / display Tower workstation with rack-mount options Suitable for Multi-GPU local AI, engineering simulation and large 3D rendering
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Mobile workstation; RTX PRO configuration enquiries
CPU / platform Up to Core Ultra 9 285HX; 24 cores
GPU / graphics Up to RTX PRO 5000 Blackwell Laptop; 24 GB GDDR7 ECC
Memory Up to 192 GB non-ECC / 128 GB ECC DDR5
Storage Up to 12 TB; 3 M.2 SSD slots
Form factor / display 16-inch; optional 3.2K Tandem OLED touch Suitable for Mobile AI development, CAD / BIM and professional rendering
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Mobile workstation; match software and memory needs
CPU / platform Up to Core Ultra 9 285H; 16 cores
GPU / graphics Up to RTX PRO 2000 Blackwell Laptop; 8 GB GDDR7
Memory Up to 64 GB LPCAMM2 LPDDR5x, upgradeable
Storage Up to 8 TB; 2 M.2 SSD slots
Form factor / display 16-inch mobile workstation; IPS / OLED display options Suitable for Mobile CAD, content creation and lighter AI development
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Business AI laptop; regional configuration to confirm
CPU / platform Up to Intel Core Ultra X7 Series 3
GPU / graphics Up to Intel Arc 12Xe integrated graphics
Memory Up to 64 GB LPDDR5x
Storage Up to 2 TB M.2 PCIe Gen5 SSD
Form factor / display 14-inch; WUXGA IPS / 2.8K OLED options Suitable for Mobile business, AI-assisted productivity and video collaboration
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Business AI laptop; quotation by complete MTM
CPU / platform Intel Core Ultra 5 / 7 / X7 Series 3
GPU / graphics Intel Graphics / Arc B370 / B390, processor dependent
Memory Up to 64 GB LPCAMM2 LPDDR5x
Storage Up to 2 TB M.2 SSD; PCIe 5.0 slot
Form factor / display 14-inch business laptop; optional 5G / smart card Suitable for Enterprise daily work, AI productivity and mobile deployment
Enquiries open · Sourced on request
Product page · images & specifications →
Dell AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Compact local AI development platform
CPU / platform NVIDIA GB10; 20-core Arm CPU
GPU / graphics Integrated NVIDIA Blackwell GPU
Memory 128 GB LPDDR5x unified memory
Storage Up to 4 TB NVMe SSD
Form factor / display 150 × 150 × 51 mm compact desktop Suitable for Local AI development, RAG and edge inference
Enquiries open · Sourced on request
Product page · images & specifications →
Dell AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Deskside AI development and compute platform
CPU / platform NVIDIA GB300 Grace Blackwell Ultra
GPU / graphics Blackwell Ultra; up to 252 GB GPU memory
Memory Up to 748 GB coherent memory (CPU + GPU)
Storage Up to 16 TB SSD
Networking ConnectX-8; 2 × 400 GbE QSFP Suitable for Large language models, research labs and enterprise AI teams
Enquiries open · Sourced on request
Product page · images & specifications →
Dell AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Tower workstation; GPU selection by requirements
CPU / platform Up to Intel Core Ultra 9 285K
GPU / graphics Up to RTX PRO 6000 Blackwell; 96 GB GDDR7, configuration dependent
Memory Up to 128 GB DDR5; ECC options
Storage Up to 12 TB SSD / 24 TB HDD, configuration dependent
Form factor / display 32 L tower workstation; 4 PCIe slots Suitable for AI / data analysis, CAD and expandable desktop compute
Enquiries open · Sourced on request
Product page · images & specifications →
Dell AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. 14-inch business AI laptop
CPU / platform Intel Core Ultra 5 / 7 Series 2 U or V options
GPU / graphics Intel integrated graphics / Arc, processor dependent
Memory V-series options: 16 / 32 GB LPDDR5x
Storage Up to 2 TB M.2 NVMe SSD
Form factor / display 14-inch; FHD+ / QHD+ and touch options Suitable for Business AI productivity, collaboration and mobile work
Enquiries open · Sourced on request
Product page · images & specifications →
Dell AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. 14-inch mobile workstation
CPU / platform Intel Core Ultra H Series 2; configuration dependent
GPU / graphics Up to RTX PRO 2000 Blackwell; 8 GB GDDR7
Memory Up to 64 GB LPDDR5x, onboard
Storage Up to 2 TB NVMe SSD
Form factor / display 14-inch; optional 2880 × 1800 OLED touch Suitable for Mobile design, professional applications and AI-assisted creation
Enquiries open · Sourced on request
Product page · images & specifications →
HP AI PCs / workstations
HP manufacturer ZGX Nano G1n image (US appearance reference); confirm the regional version, ports and accessories in the quotation. Compact local AI development platform
CPU / platform NVIDIA GB10; 20-core Grace Arm CPU
GPU / graphics Integrated NVIDIA Blackwell GPU
Memory 128 GB coherent unified memory
Storage 2 TB or 4 TB self-encrypting NVMe M.2 SSD
OS NVIDIA AI software stack and HP Z Toolkit Suitable for Local model prototyping, fine-tuning and inference
Enquiries open · Sourced on request
Product page · images & specifications →
HP AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Ryzen AI Max PRO mini workstation
CPU / platform Up to AMD Ryzen AI Max+ PRO 395
GPU / graphics AMD Radeon integrated graphics; CPU dependent
Memory Up to 128 GB unified memory; up to 96 GB allocatable to GPU
Storage Up to 8 TB; dual NVMe modules
Form factor / display Mini workstation; up to 50 NPU TOPS Suitable for Local LLMs, 3D design and multi-application work
Enquiries open · Sourced on request
Product page · images & specifications →
HP AI PCs / workstations
HP manufacturer Japanese-version appearance reference; confirm keyboard, region, ports and configuration at quotation. AMD business AI laptop
CPU / platform Ryzen AI 7 PRO 360 / AI 9 HX PRO 375 options
GPU / graphics AMD Radeon 880M / 890M, CPU dependent
Memory Up to 64 GB LPDDR5x, onboard
Storage Up to 1 TB PCIe Gen4 NVMe SSD, configuration dependent
Form factor / display 14-inch business AI laptop; up to 55 NPU TOPS Suitable for AI-assisted office work, collaboration and local productivity
Enquiries open · Sourced on request
Product page · images & specifications →
HP AI PCs / workstations
HP manufacturer Japanese-version appearance reference; confirm keyboard, region, ports and configuration at quotation. Intel business AI laptop
CPU / platform Example configuration: Intel Core Ultra 7 258V
GPU / graphics Intel Arc integrated graphics
Memory Example configuration: 32 GB LPDDR5x, onboard
Storage Example configuration: 1 TB PCIe NVMe SSD
Form factor / display 14-inch; WUXGA IPS touch and Sure View options Suitable for Mobile business, video collaboration and AI productivity
Enquiries open · Sourced on request
Product page · images & specifications →
ASUS AI PCs / workstations
Manufacturer model reference; confirm accessories, ports, regional version and delivered configuration at quotation. GB10 desktop AI development platform
CPU / platform NVIDIA GB10; 20-core Arm CPU
GPU / graphics Integrated NVIDIA Blackwell GPU
Memory 128 GB LPDDR5x unified memory
Storage 1 / 2 / 4 TB M.2 NVMe SSD options
Form factor / display 150 × 150 × 51 mm; 10 GbE + ConnectX-7 Suitable for Compact desktop AI development, local inference and model testing
Enquiries open · Sourced on request
Product page · images & specifications →
ASUS AI PCs / workstations
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GB300 deskside AI platform
CPU / platform NVIDIA GB300 Grace Blackwell Ultra
GPU / graphics Integrated Blackwell Ultra GPU
Memory 748 GB coherent memory (CPU + GPU)
Networking NVIDIA ConnectX-8 SuperNIC
Form factor / display Deskside AI platform based on DGX Station Suitable for Large-model training / inference, research and AI agent development
Enquiries open · Sourced on request
Product page · images & specifications →
MSI AI PCs / workstations
Manufacturer model reference; confirm accessories, ports, regional version and delivered configuration at quotation. Desktop AI platform; memory variant to confirm
CPU / platform NVIDIA GB10; 20-core Arm CPU
GPU / graphics Integrated NVIDIA Blackwell GPU
Memory 64 GB or 128 GB LPDDR5x unified memory
Networking ConnectX-7, up to 200 GbE
OS NVIDIA DGX OS / Ubuntu Linux Suitable for Local AI agents, edge inference and data science
Enquiries open · Sourced on request
Product page · images & specifications →
Apple AI PCs / workstations
Apple 2026 M5 Max/M5 Ultra Mac Studio manufacturer image; displays and software in workflow illustrations are not included by implication. Apple Silicon configurations for local AI and creation
CPU / platform Apple M5 Max / M5 Ultra configuration options
GPU / graphics Up to 40-core GPU with M5 Max; up to 80-core with M5 Ultra
Memory Up to 128 GB unified memory with M5 Max; up to 512 GB with M5 Ultra
Storage Up to 8 TB SSD with M5 Max; up to 16 TB with M5 Ultra
Form factor / display Compact desktop; Thunderbolt 5 and 10 Gb Ethernet Suitable for macOS creation and MLX / Metal-compatible local AI
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
PNY example of this GPU model, showing the PNY standard cooling design. Confirm the final brand, board, memory variant and configuration at quotation. GeForce RTX 50 series; board and memory variant to select
GPU architecture NVIDIA Blackwell
VRAM 32 GB GDDR7
Memory interface 512-bit
Interface PCIe Gen 5
Reference graphics power 575 W Suitable for Local AI inference, image generation and creation; choose VRAM for the model size.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
PNY example of this GPU model, showing the PNY standard cooling design. Confirm the final brand, board, memory variant and configuration at quotation. GeForce RTX 50 series; board and memory variant to select
GPU architecture NVIDIA Blackwell
VRAM 16 GB GDDR7
Memory interface 256-bit
Interface PCIe Gen 5
Reference graphics power 360 W Suitable for Local AI inference, image generation and creation; choose VRAM for the model size.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
PNY example of this GPU model, showing the PNY standard cooling design. Confirm the final brand, board, memory variant and configuration at quotation. GeForce RTX 50 series; board and memory variant to select
GPU architecture NVIDIA Blackwell
VRAM 16 GB GDDR7
Memory interface 256-bit
Interface PCIe Gen 5
Reference graphics power 300 W Suitable for Local AI inference, image generation and creation; choose VRAM for the model size.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
PNY example of this GPU model, showing the PNY standard cooling design. Confirm the final brand, board, memory variant and configuration at quotation. GeForce RTX 50 series; board and memory variant to select
GPU architecture NVIDIA Blackwell
VRAM 12 GB GDDR7
Memory interface 192-bit
Interface PCIe Gen 5
Reference graphics power 250 W Suitable for Local AI inference, image generation and creation; choose VRAM for the model size.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
PNY example of this GPU model, showing the PNY standard cooling design. Confirm the final brand, board, memory variant and configuration at quotation. GeForce RTX 50 series; board and memory variant to select
GPU architecture NVIDIA Blackwell
VRAM 16 GB / 8 GB GDDR7
Memory interface 128-bit
Interface PCIe Gen 5
Reference graphics power 180 W Suitable for Local AI inference, image generation and creation; choose VRAM for the model size.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Official PNY reference image of this NVIDIA RTX PRO board; confirm bracket, supplied accessories and configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 96 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 600 W
Form factor / cooling Full-height dual-slot; double-flow-through cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Manufacturer model reference; confirm accessories, ports, regional version and delivered configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 96 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 300 W
Form factor / cooling Full-height dual-slot; active cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Official PNY reference image of this NVIDIA RTX PRO board; confirm bracket, supplied accessories and configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 48 GB / 72 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 300 W
Form factor / cooling Full-height dual-slot; active cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Official PNY reference image of this NVIDIA RTX PRO board; confirm bracket, supplied accessories and configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 32 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 200 W
Form factor / cooling Full-height dual-slot; active cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Official PNY reference image of this NVIDIA RTX PRO board; confirm bracket, supplied accessories and configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 24 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 145 W
Form factor / cooling Full-height single-slot; active cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Official PNY reference image of this NVIDIA RTX PRO board; confirm bracket, supplied accessories and configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 24 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 70 W
Form factor / cooling Low-profile dual-slot; active cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Official PNY reference image of this NVIDIA RTX PRO board; confirm bracket, supplied accessories and configuration at quotation. Professional workstation GPU; match chassis and software
GPU architecture NVIDIA Blackwell
VRAM 16 GB GDDR7 ECC
Interface PCIe Gen 5
Maximum power 70 W
Form factor / cooling Low-profile dual-slot; active cooling Suitable for Professional AI development, CAD/3D visualization and creative workstations; confirm software and chassis compatibility.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
NVIDIA manufacturer Server Edition reference: air/liquid cooling options; confirm the system-specific cooling configuration at quotation. 2 manufacturer images Server GPU; configure with a suitable system platform
GPU architecture NVIDIA Blackwell
GPU memory 96 GB GDDR7 ECC
Interface PCIe Gen 5
Configurable maximum power 600 W
Platform / cooling Server only; air-cooled dual-slot or liquid-cooled single-slot Suitable for Enterprise AI inference, visual computing and multi-workload servers.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Manufacturer reference image; supplied version confirmed in quotation. 1 images Data-centre PCIe GPU; cooling and system compatibility to check
GPU architecture NVIDIA Ada Lovelace
GPU memory 48 GB GDDR6 ECC
Interface PCIe Gen 4 x16
Maximum power 350 W
Form factor / cooling Dual-slot passive; requires server airflow Suitable for Data-centre AI inference, rendering and visual computing.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Data-centre PCIe GPU; cooling and system compatibility to check
GPU architecture NVIDIA Hopper
GPU memory 141 GB HBM3e
Memory bandwidth 4.8 TB/s
Interface / form factor PCIe Gen 5; dual-slot air-cooled
Configurable maximum power 600 W Suitable for Enterprise LLM inference and HPC; match NVLink and server configuration.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
Lenovo-integrated HGX B200 eight-GPU liquid-cooled board; not a desktop graphics card. 1 images HGX/SXM platform enquiry; quotation as a system configuration
GPU architecture NVIDIA Blackwell
GPU memory 180 GB HBM3e/GPU; 1.44 TB across 8 GPUs
Platform form factor 8 × SXM GPU baseboard
GPU interconnect Fifth-generation NVLink; NVSwitch fabric
Deployment Compatible system; verify facility power and cooling Suitable for Large-model training, inference and multi-GPU AI compute platform.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPUs / AI accelerators
NVIDIA HGX B300 eight-GPU baseboard; configure as a compatible system. 1 images HGX/SXM platform enquiry; quotation as a system configuration
GPU architecture NVIDIA Blackwell Ultra
GPU memory 288 GB HBM3e/GPU; 2.30 TB across 8 GPUs
Platform form factor 8 × SXM GPU baseboard
GPU interconnect Fifth-generation NVLink; NVSwitch fabric
Deployment Compatible system; verify facility power and cooling Suitable for Large-model training, inference and multi-GPU AI compute platform.
Enquiries open · Sourced on request
Product page · images & specifications →
AMD GPUs / AI accelerators
ASUS Turbo Radeon AI PRO R9700 32GB manufacturer reference; confirm supplied brand and cooling version at quotation. Workstation AI GPU; software environment to check
GPU architecture AMD RDNA 4
VRAM 32 GB GDDR6
Interface PCIe 5.0 x16
Board power 300 W
Form factor / cooling Dual-slot; active cooling Suitable for Local AI inference, development and memory-intensive workloads; verify ROCm and application support.
Enquiries open · Sourced on request
Product page · images & specifications →
AMD GPUs / AI accelerators
Manufacturer Instinct MI325X OAM module; requires compatible system and cooling. 1 images Data-centre accelerator; enquire with a compatible system platform
GPU architecture AMD CDNA 3
GPU memory 256 GB HBM3E ECC
Memory bandwidth 6 TB/s
Module / interface OAM・PCIe 5.0 x16
Typical board power 1000 W Suitable for Data-centre AI training, inference and HPC; select with ROCm and compatible system requirements.
Enquiries open · Sourced on request
Product page · images & specifications →
AMD GPUs / AI accelerators
AMD MI350X module/platform manufacturer reference; a multi-GPU platform image does not mean a complete platform is included per accelerator. Data-centre accelerator; enquire with a compatible system platform
GPU architecture AMD CDNA 4
GPU memory 288 GB HBM3E ECC
Memory bandwidth 8 TB/s
Module / interface OAM・PCIe 5.0 x16
Typical board power 1000 W Suitable for Data-centre AI training, inference and HPC; select with ROCm and compatible system requirements.
Enquiries open · Sourced on request
Product page · images & specifications →
AMD GPUs / AI accelerators
Manufacturer MI350-series OAM module illustration; MI355X cooling confirmed with system. 1 images Data-centre accelerator; enquire with a compatible system platform
GPU architecture AMD CDNA 4
GPU memory 288 GB HBM3E ECC
Memory bandwidth 8 TB/s
Module / interface OAM・PCIe 5.0 x16
Typical board power 1400 W Suitable for Data-centre AI training, inference and HPC; select with ROCm and compatible system requirements.
Enquiries open · Sourced on request
Product page · images & specifications →
Intel GPUs / AI accelerators
Intel Arc Pro B50 reference design; supplied version confirmed in quotation. 1 images Arc Pro professional graphics; driver and application support to check
GPU architecture Intel Xe2
VRAM 16 GB GDDR6
Memory bandwidth 224 GB/s
Interface PCIe 5.0 x8
Board power 70 W Suitable for Local AI inference and professional design; verify Intel software and driver support.
Enquiries open · Sourced on request
Product page · images & specifications →
Intel GPUs / AI accelerators
Intel B60 manufacturer reference rendering; supplied brand, form factor and cooling confirmed in quotation. 1 images Arc Pro professional graphics; driver and application support to check
GPU architecture Intel Xe2
VRAM 24 GB GDDR6
Memory bandwidth 456 GB/s
Interface PCIe 5.0 x8
Board power 120–200 W (by board design) Suitable for Local AI inference and professional design; verify Intel software and driver support.
Enquiries open · Sourced on request
Product page · images & specifications →
Intel GPUs / AI accelerators
ASUS UGen Intel Arc Pro B70 32GB manufacturer reference; confirm supplied brand, ports and configuration at quotation. Arc Pro professional graphics; driver and application support to check
GPU architecture Intel Xe2
VRAM 32 GB GDDR6
Memory bandwidth 608 GB/s
Interface PCIe 5.0 x16
Board power 160–290 W (by board design) Suitable for Local AI inference and professional design; verify Intel software and driver support.
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPU servers
Dell PowerEdge XE9680 OEM system reference images. Confirm the actual OEM, GPU variant and complete configuration at quotation. 8-GPU Hopper server solution; OEM configuration selected for the workload
GPU platform 8 × NVIDIA H200 SXM; Hopper
GPU memory 141 GB HBM3e per GPU; 1,128 GB across 8 GPUs
Memory bandwidth 4.8 TB/s per GPU
GPU interconnect NVLink 900 GB/s per GPU; NVSwitch interconnect
System configuration CPU, system RAM, NVMe, networking and cooling selected by OEM system Suitable for LLM inference, fine-tuning and HPC
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPU servers
Lenovo ThinkSystem SR680a V3 with B200 OEM system reference images. Confirm the actual OEM, GPU variant and complete configuration at quotation. 8-GPU Blackwell server solution; OEM configuration selected for the workload
GPU platform 8 × NVIDIA B200 SXM; Blackwell
GPU memory 180 GB HBM3e per GPU; 1.44 TB across 8 GPUs
Memory bandwidth Up to 8 TB/s per GPU
GPU interconnect NVLink 1.8 TB/s per GPU; NVSwitch interconnect
System configuration CPU, system RAM, NVMe, networking and cooling selected by OEM system Suitable for Multi-GPU training, fine-tuning and enterprise AI compute
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPU servers
Lenovo ThinkSystem SR680a V4 OEM system reference images. Confirm the actual OEM, GPU variant and complete configuration at quotation. 8-GPU Blackwell Ultra server solution; OEM configuration selected for the workload
GPU platform 8 × NVIDIA B300 SXM; Blackwell Ultra
GPU memory 288 GB HBM3e per GPU; 2.30 TB across 8 GPUs
Memory bandwidth Up to 8 TB/s per GPU
GPU interconnect NVLink 1.8 TB/s per GPU; NVSwitch interconnect
System configuration CPU, system RAM, NVMe, networking and cooling selected by OEM system Suitable for Large-model and multimodal inference, multi-GPU training
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPU servers
Example H3C UniServer R5500 G5 chassis. Final OEM, chassis and configuration are confirmed at quotation. 1 images 8-GPU Ampere server solution; 80 GB per GPU reference configuration
GPU platform 8 × NVIDIA A800; HGX / Ampere
GPU memory 80 GB per GPU; 640 GB across 8 GPUs
Memory bandwidth Up to 2 TB/s per GPU
GPU interconnect NVLink 400 GB/s per GPU; NVSwitch interconnect
System configuration CPU, system RAM, NVMe, networking and cooling selected by OEM system Suitable for Multi-GPU training, model fine-tuning and HPC
Enquiries open · Sourced on request
Product page · images & specifications →
NVIDIA GPU servers
Dell PowerEdge XE9680 OEM system reference images. Confirm the actual OEM, GPU variant and complete configuration at quotation. 8-GPU Hopper server solution; 80 GB per GPU SXM5 reference configuration
GPU platform 8 × NVIDIA H800 SXM5; Hopper
GPU memory 80 GB per GPU; 640 GB across 8 GPUs
GPU power 700 W per GPU reference configuration
GPU interconnect NVLink 400 GB/s per GPU; NVSwitch interconnect
System configuration CPU, system RAM, NVMe, networking and cooling selected by OEM system Suitable for Large-model training, fine-tuning and inference
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor Rack GPU server
CPU support 2 × Intel Xeon 6
GPU configuration support Configurations with 8 × NVIDIA HGX B200 or B300
Memory support 32 DDR5 DIMM slots; up to 4 TB
Cooling Air cooled Suitable for Enterprise AI training, fine-tuning and inference; select GPUs for the workload.
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor 3 OU node; ORv3 rack infrastructure
CPU support 2 × Intel Xeon 6
GPU configuration support Configurations with 8 × NVIDIA HGX B200 or B300
Memory support 32 DDR5 DIMM slots
Cooling Liquid-cooled CPUs, GPUs and NVLink switches; facility water infrastructure required Suitable for Enterprise AI training, fine-tuning and inference; select GPUs for the workload.
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor Rack GPU server
CPU support 2 × AMD EPYC 9005
GPU configuration support Configurations with 8 × AMD Instinct MI355X or NVIDIA HGX B300
Memory support 24 DDR5 DIMM slots; up to 6 TB
Cooling Air cooled Suitable for Enterprise AI training, fine-tuning and inference; select GPUs for the workload.
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor 3 OU node; ORv3 rack infrastructure
CPU support 2 × AMD EPYC 9005
GPU configuration support Configurations with 8 × AMD Instinct MI355X or NVIDIA HGX B300
Memory support 24 DDR5 DIMM slots; up to 6 TB
Cooling Direct-to-chip liquid cooling for CPUs and GPUs; facility water infrastructure required Suitable for Enterprise AI training, fine-tuning and inference; select GPUs for the workload.
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor 4U rack
CPU support 2 × Intel Xeon 6
GPU configuration support Up to 8 dual-slot PCIe Gen5 GPUs, up to 600 W each; approved configurations apply
Memory support 32 DDR5 DIMM slots; up to 4 TB
Cooling Air cooled Suitable for Enterprise AI training, fine-tuning and inference; select GPUs for the workload.
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor 4U rack
CPU support 2 × AMD EPYC 9005
GPU configuration support Up to 8 dual-slot or 16 single-slot PCIe Gen5 GPUs; approved configurations apply
Memory support 24 DDR5 DIMM slots
Cooling Air cooled Suitable for Enterprise AI training, fine-tuning and inference; select GPUs for the workload.
Enquiries open · Sourced on request
Product page · images & specifications →
Dell GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. GPU server; configure GPU, network and facility requirements
Form factor 6U rack
CPU support 2 × 4th/5th Gen Intel Xeon Scalable
GPU configuration support Configurations with 8 × NVIDIA HGX H100 or H200; other accelerator platforms also available
Memory support 32 DDR5 DIMM slots; up to 4 TB
Cooling Air cooled Suitable for Multi-GPU AI training, inference and HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. 72-GPU rack-scale AI solution; enquire against facility, power and liquid-cooling requirements
Rack architecture 48U MGX rack; 18 compute trays and 9 NVLink switch trays
GPU platform 72 NVIDIA Blackwell Ultra GPUs; 4 per compute tray
GPU memory 288 GB HBM3e per GPU; approximately 20.7 TB across 72 GPUs
Processors 36 NVIDIA Grace CPUs; 2 per compute tray
GPU interconnect 5th-generation NVLink; 72-GPU domain within one rack
Cooling requirements Liquid-cooled GPUs, CPUs and NVSwitch with air cooling; CDU and facility water must be checked
Rack power Manufacturer reference: 135 kW TDP, up to 155 kW peak; verify facility power for the configuration Suitable for LLM training, multimodal reasoning and large-scale AI inference; assessed as a rack-scale project
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. AI/GPU server; system configuration and lead time to confirm
Form factor 8U rack
CPU support 2 × Intel Xeon 6 6700 series with P-cores
GPU configuration support NVIDIA HGX B300 8-GPU platform
Memory support 32 DDR5 DIMM slots; up to 4 TB
Cooling Air cooled Suitable for Generative AI, large-model training and scientific computing.
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. AI/GPU server; system configuration and lead time to confirm
Form factor 8U rack
CPU support 2 × 5th Gen Intel Xeon Scalable
GPU configuration support NVIDIA HGX B200 8-GPU platform
Memory support 32 DDR5 DIMM slots; up to 4 TB
Cooling Air cooled Suitable for Generative AI training, fine-tuning and HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. AI/GPU server; system configuration and lead time to confirm
Form factor 3U rack
CPU support Up to 2 × AMD EPYC 9004/9005
GPU configuration support Up to 8 PCIe GPUs, or an NVIDIA HGX H200 4-GPU configuration
Memory support 24 DDR5 DIMM slots; up to 6 TB
Cooling Air cooling for PCIe configurations; Neptune liquid-to-air hybrid cooling for HGX H200 Suitable for Scalable AI inference, training, visual computing and HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
Lenovo GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. AI/GPU server; system configuration and lead time to confirm
Form factor 5U rack
CPU support 2 × 5th Gen Intel Xeon Scalable
GPU configuration support Configurations with 8 × NVIDIA H100, H200 or B200; platform version applies
Memory support 32 DDR5 DIMM slots; up to 4 TB
Cooling Liquid cooled; facility water and CDU infrastructure must be checked Suitable for Dense multi-GPU AI training and HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
HPE GPU servers
Manufacturer front, rear or internal reference with port callouts; confirm shown options and delivered configuration at quotation. Enterprise GPU compute platform; configure for the workload
Form factor 4U dual-socket rack
CPU support Up to 2 × Intel Xeon 6
GPU configuration support Supports up to 8 × NVIDIA RTX PRO 6000 Blackwell Server Edition; specified kits required
Memory support 32 DDR5 DIMM slots
Cooling Air cooled Suitable for Enterprise AI inference, fine-tuning and visual computing.
Enquiries open · Sourced on request
Product page · images & specifications →
HPE GPU servers
Manufacturer front, rear or internal reference with port callouts; confirm shown options and delivered configuration at quotation. Enterprise GPU compute platform; configure for the workload
Form factor 5U single-node rack
CPU support 2 × 4th Gen Xeon (H100) or 5th Gen Xeon (H200)
GPU configuration support Configurations with 8 × NVIDIA H100 or H200 SXM5; PCIe GPUs not supported
Memory support Up to 32 DDR5 DIMMs
Cooling Air-cooled or direct-liquid-cooled configurations available Suitable for Large-model training, deep learning and HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
HPE GPU servers
Manufacturer front, rear or internal reference with port callouts; confirm shown options and delivered configuration at quotation. Enterprise GPU compute platform; configure for the workload
Form factor 2U dual-socket rack
CPU support Up to 2 × AMD EPYC 9004/9005
GPU configuration support GPU CTO chassis: up to 4 dual-slot or 8 single-slot GPUs; thermal/power limits depend on configuration
Memory support DDR5 RDIMM, up to 6400 MT/s; depends on CPU/memory configuration
Cooling Air cooled Suitable for Enterprise compute, AI inference and GPU workloads.
Enquiries open · Sourced on request
Product page · images & specifications →
Supermicro GPU servers
Manufacturer front-panel illustration; final configuration is confirmed at quotation. 1 images HGX B300 GPU system; facility requirements to check
Form factor 8U rack
CPU support 2 × Intel Xeon 6700 series with P-cores
GPU configuration support NVIDIA HGX B300 8-GPU platform
Memory support 32 DDR5 RDIMM slots; up to 8 TB
Cooling Air cooled Suitable for Large-model training, generative AI and HPC clusters.
Enquiries open · Sourced on request
Product page · images & specifications →
Supermicro GPU servers
Manufacturer front, rear or internal reference with port callouts; confirm shown options and delivered configuration at quotation. HGX B300 GPU system; facility requirements to check
Form factor 8U rack
CPU support 2 × AMD EPYC 9005/9004; approved configurations apply
GPU configuration support NVIDIA HGX B300 8-GPU platform
Memory support 24 DDR5 RDIMM slots; up to 6 TB
Cooling Air cooled Suitable for Large-model training, generative AI and HPC clusters.
Enquiries open · Sourced on request
Product page · images & specifications →
ASUS GPU servers
Manufacturer model / family reference image; confirm accessories, ports and the delivered configuration at quotation. Multi-GPU server platform; sourcing by configuration
Form factor 4U rack
CPU support 2 × AMD EPYC 9005/9004
GPU configuration support Up to 8 dual-slot PCIe Gen5 GPUs, up to 600 W each; compatibility list applies
Storage support 8 hot-swap 2.5-inch drive bays with NVMe support
Cooling Air cooled Suitable for Enterprise AI inference, fine-tuning and multi-GPU HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
GIGABYTE GPU servers
Official datasheet product image (front and rear); configuration confirmed by quotation. 1 images Multi-GPU server platform; sourcing by configuration
Form factor 4U rack
CPU support 2 × AMD EPYC 9005/9004
GPU configuration support Up to 8 dual-slot PCIe Gen5 GPUs; H200 NVL/RTX PRO 6000 support is conditional
Memory support 48 DDR5 RDIMM slots
Storage support 12 hot-swap 2.5-inch NVMe/SATA/SAS drive bays; SAS requires a controller Suitable for AI inference, visual computing and multi-GPU HPC.
Enquiries open · Sourced on request
Product page · images & specifications →
Cisco GPU servers
Official reference image; appearance and accessories depend on the quoted configuration. 1 images Multi-GPU server platform; sourcing by configuration
Form factor 4RU rack
CPU support 2 × 5th Gen AMD EPYC 9005
GPU configuration support Configurations with 2/4/6/8 approved PCIe GPUs, including H200 NVL, L40S and RTX PRO 6000
Memory support 32 DDR5 DIMM slots; up to 8 TB
Cooling Air cooled Suitable for Enterprise AI inference, RAG and generative AI fine-tuning.
Enquiries open · Sourced on request
Product page · images & specifications →
No matching models. Try another keyword or send your requirements directly.
