GPU
Shop By
Showing 19–36 of 560 results
-
NVIDIA 900-9D3B6-00CV-AA0（B3220）P-Series BlueField – 3 NDR200Gb/s dual port QSFP112 InfiniBand/Ethernet DPU
Quote10 pcs(MOQ)**NVIDIA BlueField-3 DPU: Next-Generation Data Processing Powerhouse**
The NVIDIA BlueField-3 DPU (B3220) delivers revolutionary data center infrastructure acceleration with dual-port NDR200 InfiniBand/Ethernet connectivity via QSFP112 interfaces. This cutting-edge Data Processing Unit offloads critical networking, storage, and security tasks from CPUs, dramatically improving overall system performance and efficiency.
**Key Features:**
– Dual 200Gbps NDR InfiniBand/Ethernet ports with QSFP112 connectors
– Advanced Arm-based processing cores for infrastructure workloads
– Hardware-accelerated encryption, compression, and virtualization
– Programmable packet processing and network function virtualization
– Zero-trust security enforcement at the hardware level**Core Benefits:**
– **Performance Liberation**: Frees up CPU resources by handling data center infrastructure tasks independently
– **Ultra-Low Latency**: Hardware acceleration delivers microsecond-level response times for mission-critical applications
– **Scalable Security**: Built-in cryptographic acceleration and isolation capabilities protect against evolving threats
– **Future-Proof Connectivity**: NDR200 speeds ensure readiness for next-generation data center requirements
– **Operational Efficiency**: Reduces total cost of ownership through improved resource utilization and energy efficiency**Unique Value Proposition**: The only solution combining 200Gbps networking speeds with comprehensive infrastructure acceleration, enabling organizations to build secure, high-performance data centers that adapt to AI and cloud-native workloads while maintaining deterministic performance.
Bulk Order Discounts Available -
Gigabyte GeForce RTX 5090 GAMING OC 32G Key Features GPU
Quote1 Piece(MOQ)**Gigabyte GeForce RTX 5090 GAMING OC 32G**
Premium flagship graphics card delivering uncompromising 4K and 8K gaming performance with massive 32GB GDDR7 memory. Features advanced Ada Lovelace architecture with enhanced RT cores for superior ray tracing, DLSS 4 Frame Generation for maximum fps, and AV1 encoding for content creators. The GAMING OC variant includes factory overclocking, Gigabyte’s WINDFORCE cooling system with triple fans and robust heatsink design for optimal thermal management. Supports latest PCIe 5.0, DisplayPort 2.1, and HDMI 2.1a connectivity. Ideal for enthusiasts demanding ultimate gaming fidelity, professional 3D rendering, AI workloads, and future-proof performance. RGB Fusion 2.0 lighting and reinforced metal backplate complete the premium package.
Bulk Order Discounts Available -
NVIDIA H200 Tensor Core GPU,141 GB HBM3e, NVLink, PCIe 5.0, 700 W
Quote1 Piece(MOQ)The NVIDIA H200 Tensor Core GPU delivers unprecedented AI and HPC performance with 141GB of ultra-fast HBM3e memory—nearly double the capacity of its predecessor. This powerhouse accelerator features advanced Tensor Cores optimized for transformer models and generative AI workloads, enabling faster training and inference for large language models. With high-speed NVLink interconnect technology, multiple H200s can work together seamlessly for massive parallel processing. The PCIe 5.0 interface ensures maximum data throughput, while the 700W TDP provides exceptional performance-per-watt efficiency. Key advantages include dramatically reduced model loading times, support for larger AI models in memory, superior multi-GPU scaling, and industry-leading performance for enterprise AI applications, scientific computing, and data analytics workloads.
Bulk Order Discounts Available -
MSI GeForce RTX 5090 32G VENTUS 3X OC G5090-32V3C
Quote10 pcs(MOQ)The MSI GeForce RTX 5090 32G VENTUS 3X OC delivers flagship gaming performance with a massive 32GB GDDR6X memory buffer for handling 4K gaming, content creation, and AI workloads without compromise. Features MSI’s triple-fan VENTUS 3X cooling system with factory overclocking for enhanced performance straight out of the box. The card excels at ray tracing, DLSS 4 technology, and professional rendering tasks while maintaining efficient thermal management. Ideal for enthusiasts demanding maximum visual fidelity, creators working with large datasets, and gamers future-proofing their systems for next-generation titles. The generous VRAM capacity eliminates bottlenecks in memory-intensive applications, making it the ultimate choice for high-end gaming rigs and professional workstations.
Bulk Order Discounts Available -
NVIDIA L4 24G Tensor Core GPU
Quote1 Piece(MOQ)The NVIDIA L4 24G Tensor Core GPU delivers exceptional AI inference performance and versatility in a compact, energy-efficient design. Featuring 24GB of high-speed memory, advanced Tensor Cores, and NVIDIA’s latest architecture, it accelerates machine learning workloads, real-time AI applications, and video processing tasks with up to 120x faster inference than CPUs. Its low-profile form factor and optimized power consumption make it ideal for edge computing, data centers, and workstations where space and energy efficiency are critical. The L4 excels at computer vision, natural language processing, recommendation systems, and video analytics while supporting multiple AI frameworks and offering seamless integration with existing infrastructure. This GPU represents the optimal balance of performance, efficiency, and affordability for organizations deploying AI at scale.
Bulk Order Discounts Available -
NVIDIA A100 80G SXM4
Quote1 Piece(MOQ)The NVIDIA A100 80GB SXM4 is a flagship data center GPU designed for AI, machine learning, and high-performance computing workloads. Built on the Ampere architecture with 54 billion transistors, it delivers exceptional performance with 624 tensor TFLOPS for AI training and 1,248 TFLOPS for inference. The 80GB high-bandwidth memory (HBM2e) with 2 TB/s memory bandwidth enables processing of massive datasets and complex models without memory constraints. Key features include Multi-Instance GPU (MIG) technology that partitions the GPU into up to seven isolated instances for optimal resource utilization, third-generation NVLink for ultra-fast GPU-to-GPU communication, and advanced sparsity support for accelerated inference. The SXM4 form factor provides superior power efficiency and thermal management in server environments. This GPU accelerates breakthrough research, reduces training times from weeks to hours, and enables real-time AI inference for mission-critical applications across industries including healthcare, finance, and autonomous vehicles.
Bulk Order Discounts Available -
NVIDIA T1000 8GB GDDR6 PCIe3.0 900-5G172-2270-000 Professional Cards
Quote1 pcs(MOQ)The NVIDIA T1000 8GB is a professional-grade graphics card engineered for creative and technical workloads. Featuring 8GB of high-speed GDDR6 memory and PCIe 3.0 connectivity, it delivers exceptional performance for CAD design, 3D modeling, video editing, and visualization tasks. Built on NVIDIA’s Turing architecture, the T1000 provides hardware-accelerated ray tracing, AI-enhanced workflows, and certified driver support for professional applications like AutoCAD, SolidWorks, and Adobe Creative Suite. Its low-profile, energy-efficient design makes it ideal for workstations with space and power constraints while maintaining ISV certification for mission-critical professional software. The card offers superior multi-display support, enabling seamless productivity across multiple 4K monitors, making it the perfect solution for professionals demanding reliable performance, certified stability, and exceptional visual quality in compact workstation environments.
Bulk Order Discounts Available -
NVIDIA A10 Tensor Core GPU
Quote1 Piece(MOQ)The NVIDIA A10 Tensor Core GPU is a versatile data center accelerator designed for AI inference, graphics workloads, and virtual workstations. Built on the Ampere architecture with 24GB GDDR6 memory, it delivers exceptional performance for machine learning inference, real-time ray tracing, and professional visualization. Key features include third-generation RT Cores for accelerated ray tracing, third-generation Tensor Cores for AI workloads, and support for multiple precision formats (FP32, FP16, INT8, INT4). The A10 excels in edge AI deployments, cloud gaming, virtual desktop infrastructure (VDI), and mixed workloads requiring both compute and graphics capabilities. Its single-slot, power-efficient design (150W) makes it ideal for dense server deployments while providing enterprise-grade reliability and comprehensive software ecosystem support through NVIDIA’s AI Enterprise suite.
Bulk Order Discounts Available -
MSI GeForce RTX 5090 32G GAMING TRIO OC G5090-32GTC
Quote10 pcs(MOQ)**MSI GeForce RTX 5090 32G GAMING TRIO OC** delivers ultimate 4K and 8K gaming performance with 32GB GDDR6X memory and factory overclocking. Features MSI’s triple-fan TORX 4.0 cooling system for whisper-quiet operation under extreme loads, robust dual BIOS design for enhanced reliability, and premium build quality with reinforced backplate. RGB Mystic Light synchronization adds customizable aesthetics while advanced thermal design ensures sustained peak performance. Ideal for enthusiast gamers, content creators, and professionals demanding uncompromising graphics power for ray tracing, AI workloads, and future-proof gaming at maximum settings.
Bulk Order Discounts Available -
NVIDIA Quadro T1000 8G – Professional Graphics Card 900-5G172-2570-000
Quote1 pcs(MOQ)The NVIDIA Quadro T1000 8GB is a professional-grade graphics card engineered for creative professionals and technical workstations. Featuring 8GB of GDDR6 memory and 896 CUDA cores, it delivers exceptional performance for CAD modeling, 3D rendering, video editing, and data visualization. The card supports four 5K displays simultaneously through Mini DisplayPort outputs, enabling expansive multi-monitor workflows. Its compact single-slot design and low 50W power consumption make it ideal for small form factor workstations without sacrificing performance. Built on NVIDIA’s Turing architecture, it provides real-time ray tracing capabilities and AI-enhanced workflows through RT and Tensor cores. The T1000 includes certified drivers for professional applications like SolidWorks, AutoCAD, Maya, and Premiere Pro, ensuring reliable performance and stability for mission-critical projects. This card strikes an optimal balance between affordability and professional-grade capabilities, making advanced visual computing accessible to a broader range of professionals.
Bulk Order Discounts Available -
NVIDIA L20 48GB GPU
Quote10 pcs(MOQ)The NVIDIA L20 48GB GPU is a high-performance accelerator designed for AI inference and lightweight training workloads. Featuring 48GB of high-bandwidth memory, it delivers exceptional performance for large language models, computer vision, and recommendation systems while maintaining energy efficiency. Key advantages include superior memory capacity for handling complex AI models, optimized architecture for inference acceleration, and cost-effective deployment for enterprise AI applications. The L20 bridges the gap between consumer and data center GPUs, offering professional-grade AI capabilities with improved price-performance ratio, making it ideal for organizations seeking powerful AI processing without the premium cost of flagship data center solutions.
Bulk Order Discounts Available -
NVIDIA Ampere A16, PCIe, 250W, 64GB Passive, Double Wide, Full Height GPU
Quote10 pcs(MOQ)The NVIDIA Ampere A16 is a powerful enterprise-grade GPU designed for virtualized workloads and data center applications. Featuring 64GB of high-speed memory and 250W power efficiency, this PCIe card delivers exceptional performance for AI inference, virtual desktop infrastructure (VDI), and cloud gaming. Its passive cooling design ensures silent operation while the double-wide, full-height form factor maximizes computational density. Built on the advanced Ampere architecture, the A16 provides superior multi-user support with hardware-accelerated encoding/decoding, making it ideal for organizations requiring scalable graphics virtualization, remote workstations, and GPU-accelerated computing in noise-sensitive environments. The massive 64GB memory capacity enables handling of complex datasets and multiple concurrent users without performance degradation.
Bulk Order Discounts Available -
NVIDIA RTX A4000 16GB PCI Express 4.0 x16 Video Card VCNRTXA4000-PB
Quote1 Piece(MOQ)The NVIDIA RTX A4000 is a professional-grade graphics card designed for demanding creative and technical workflows. Featuring 16GB of GDDR6 memory and PCIe 4.0 x16 connectivity, it delivers exceptional performance for 3D rendering, CAD design, video editing, and AI workloads. Built on NVIDIA’s Ampere architecture with RT cores for real-time ray tracing and Tensor cores for AI acceleration, the A4000 enables photorealistic visualizations and machine learning tasks. Its compact single-slot design fits into space-constrained workstations while providing enterprise-level reliability, ECC memory support, and optimized drivers for professional applications like Autodesk, Adobe Creative Suite, and Blender. The card offers superior price-to-performance ratio compared to higher-tier Quadro models while maintaining professional certification and support standards essential for mission-critical work environments.
Bulk Order Discounts Available -
PNY VCNRTXA5000-PB RTX A5000 24GB 384-bit GDDR6 PCI Express 4.0 Workstation Video Card
Quote1 Piece(MOQ)The PNY RTX A5000 is a professional-grade workstation graphics card featuring 24GB of high-speed GDDR6 memory with a 384-bit interface, delivering exceptional performance for demanding creative and technical workflows. Built on NVIDIA’s Ampere architecture with RT cores and Tensor cores, it accelerates ray tracing, AI workloads, and complex 3D rendering tasks. The PCIe 4.0 interface ensures maximum bandwidth for data-intensive applications. Key benefits include massive memory capacity for handling large datasets and models, professional driver support with ISV certifications for reliability in mission-critical environments, and versatility across CAD, content creation, scientific computing, and machine learning applications. This workstation-class GPU offers enterprise-level performance and stability that consumer cards cannot match.
Bulk Order Discounts Available -
Quadro A2 16G
Quote10 pcs(MOQ)The Quadro A2 16G is a professional-grade graphics card designed for demanding creative and technical workflows. Featuring 16GB of high-speed memory, it delivers exceptional performance for 3D modeling, CAD design, video editing, and AI workloads. Key advantages include certified drivers for professional applications ensuring stability and reliability, support for multiple 4K displays, real-time ray tracing capabilities, and optimized performance for industry-standard software like AutoCAD, Maya, and Premiere Pro. Its compact form factor makes it ideal for workstations with space constraints while maintaining enterprise-level performance. The card excels in professional visualization, content creation, and computational tasks where precision, stability, and certified compatibility are critical requirements that consumer graphics cards cannot match.
Bulk Order Discounts Available -
Nvidia H100 80GB PCIe Core GPUs
Quote10 pcs(MOQ)The Nvidia H100 80GB PCIe is a flagship data center GPU engineered for AI training, inference, and high-performance computing workloads. Built on the advanced Hopper architecture with 4th-gen Tensor Cores, it delivers up to 9x faster AI training and 30x faster inference compared to previous generations. The massive 80GB HBM3 memory provides exceptional capacity for large language models and complex datasets, while 3TB/s memory bandwidth ensures rapid data processing. Key features include support for FP8 precision for enhanced AI performance, PCIe 5.0 connectivity for broad server compatibility, and advanced security with confidential computing capabilities. The H100 excels in transformer-based AI models, scientific simulations, and enterprise AI applications, offering unmatched performance density and energy efficiency for organizations seeking to accelerate their most demanding computational workloads.
Bulk Order Discounts Available -
NVIDIA GPU L40S 48GB Deep learningrendering GPU
Quote10 pcs(MOQ)**NVIDIA L40S 48GB: Professional AI & Visual Computing Powerhouse**
The NVIDIA L40S delivers exceptional performance for AI training, inference, and professional visualization with 48GB of ultra-fast GDDR6 memory. Built on the Ada Lovelace architecture, it features 18,176 CUDA cores and advanced RT cores for accelerated ray tracing and rendering workflows.
**Key Features:**
– 48GB GDDR6 memory for handling massive datasets and complex models
– 142 RT cores for real-time ray tracing and photorealistic rendering
– 568 Tensor cores with Transformer Engine for AI acceleration
– PCIe Gen4 interface with 300W power efficiency
– Multi-instance GPU (MIG) support for workload isolation**Benefits:**
– Trains large language models and neural networks faster than previous generations
– Enables real-time collaboration on complex 3D scenes and simulations
– Reduces time-to-insight for data scientists and researchers
– Supports multiple concurrent AI workloads with enterprise-grade reliability**Unique Selling Points:**
– Dual-purpose design optimizing both AI compute and professional graphics
– Industry-leading memory capacity for uncompromised model complexity
– Enterprise-ready with comprehensive software ecosystem including NVIDIA Omniverse, CUDA, and AI frameworksIdeal for data centers, research institutions, and creative studios requiring maximum performance for AI development and high-end visual computing.
Bulk Order Discounts Available -
PNY VCNRTXA6000-BLK RTX A6000 48GB 384-bit GDDR6 PCI Express 4.0 x16 Workstation Video Card – OEM
Quote1 Piece(MOQ)The PNY RTX A6000 is a professional-grade workstation graphics card featuring 48GB of GDDR6 memory with a 384-bit interface, delivering exceptional performance for demanding creative and technical workflows. Built on NVIDIA’s Ampere architecture with PCI Express 4.0 x16 connectivity, it provides massive memory capacity for handling complex 3D models, large datasets, and high-resolution content creation. Key benefits include real-time ray tracing, AI-accelerated workflows, support for multiple 8K displays, and certified drivers for professional applications like CAD, 3D rendering, and scientific computing. The card’s substantial VRAM makes it ideal for professionals working with massive textures, simulations, and memory-intensive tasks that would overwhelm consumer graphics cards, while its workstation-class reliability ensures consistent performance in mission-critical environments.
Bulk Order Discounts Available
Bulk Order Discounts Available
