Category
Graphics cards compared by architecture, memory, power, benchmarks, and workload fit.
Latest models
Models with known release dates, newest first.
Browse and compare
275 items
Select two items to compare.
PNY's flagship Blackwell-based GPU featuring 32GB of ultra-fast GDDR7 memory on a 512-bit interface. Designed for 8K gaming and intensive AI compute, it utilizes the VERTO triple-fan cooling system and EPIC-X RGB ligh...
7 listed key facts
Enthusiast-grade graphics card built on the Blackwell architecture, featuring ultra-fast GDDR7 memory and a military-grade TUF cooling solution for extreme stability and performance in 4K gaming.
7 listed key facts
A high-end RDNA 4 graphics card based on the Navi 48 XTX core, featuring 16GB of GDDR6 memory and advanced AI/ray tracing accelerators, designed to compete in the high-performance 4K gaming segment.
7 listed key facts
A high-performance graphics card featuring the NVIDIA Blackwell architecture, equipped with 6,144 CUDA cores and 12GB of high-speed GDDR7 memory. The Ventus 3X model utilizes a triple-fan cooling system with TORX Fan...
7 listed key facts
The first data center GPU based on the CDNA architecture, featuring 120 Compute Units and Matrix Core technology to deliver high performance in HPC and AI training workloads.
6 listed key facts
A high-performance data center GPU accelerator designed for HPC and AI training. It features 64GB of HBM2e memory and lacks display outputs or DirectX support, resulting in a 0 score for 3DMark gaming benchmarks.
6 listed key facts
A server-grade accelerator designed for deep learning, high-performance computing (HPC), and virtualization. Based on the Vega 10 architecture, it features high-bandwidth HBM2 memory and is optimized for large-scale A...
6 listed key facts
The AMD Instinct MI250 is a professional-grade data center accelerator based on the CDNA 2 architecture, specifically designed for exascale HPC and AI workloads. It features a massive 128GB HBM2e memory capacity with...
6 listed key facts
A flagship data center accelerator based on the CDNA 2 architecture, featuring a multi-chip module (MCM) design. It is optimized for exascale HPC and high-end AI workloads, providing industry-leading FP64 and FP32 per...
6 listed key facts
The world's first data center APU, integrating 24 Zen 4 CPU cores and CDNA 3 GPU compute units with 128GB of unified HBM3 memory. As a pure HPC/AI accelerator, it lacks a hardware rasterizer and DirectX support, resul...
5 listed key facts
AMD's flagship generative AI accelerator based on the CDNA 3 chiplet architecture. It features a massive 192GB HBM3 memory capacity and 5.3TB/s bandwidth, designed to challenge the NVIDIA H100 in large language model...
6 listed key facts
The AMD Instinct MI325X is a high-performance data center accelerator designed for generative AI and large-scale model training. It features 256GB of HBM3e memory with 6TB/s bandwidth, delivering a significant boost i...
6 listed key facts
Choose MI350P for Linux/ROCm AI or compute servers needing 144 GB of dedicated HBM3E on a PCIe add-in card. Passive cooling and a 600 W maximum board-power specification make chassis airflow and power delivery central...
4 listed key facts
The AMD Instinct MI350X is a flagship data center accelerator based on the CDNA 4 architecture, specifically engineered for large-scale AI training and generative AI inference. Built on TSMC's 3nm process, it features...
6 listed key facts
The AMD Instinct MI355X is a flagship AI accelerator built on the CDNA 4 architecture. It features 288GB of HBM3e memory and a massive 8TB/s bandwidth. Designed for large-scale AI training and inference, it delivers u...
6 listed key facts
Plan MI430X-based systems for scientific simulation and converged HPC/AI projects that prioritize FP64 arithmetic and 432 GB HBM4 capacity. Match the accelerator to the complete scientific software stack and cluster i...
3 listed key facts
Choose MI455X within an engineered rack-scale AI system for large-model training or inference that benefits from 432 GB HBM4 per accelerator. Its EAM module, direct liquid cooling and scale-up links demand a compatibl...
4 listed key facts
The AMD Instinct MI50 is a professional-grade data center accelerator and the first GPU built on a 7nm process. Based on the Vega 20 architecture, it is designed for high-performance computing (HPC) and deep learning,...
6 listed key facts
The world's first 7nm datacenter GPU, designed for high-performance computing (HPC) and AI workloads. It features a massive 1TB/s memory bandwidth and specialized hardware for high-precision scientific calculations.
6 listed key facts
A flagship mobile integrated graphics solution based on the RDNA 4 architecture. Designed for high-end APUs, it delivers performance comparable to mid-range discrete mobile GPUs, featuring enhanced ray tracing hardwar...
7 listed key facts
An integrated graphics solution found in entry-level Ryzen 7020 'Mendocino' series mobile processors. Based on the RDNA 2 architecture with 2 Compute Units (128 shaders), it is designed for basic productivity, web bro...
6 listed key facts
An entry-level integrated graphics solution based on the RDNA 2 architecture (Mendocino), featuring 2 Compute Units and 128 stream processors. It is designed for basic productivity and light multimedia tasks in budget...
6 listed key facts
An integrated graphics solution found in Ryzen 5 6000 series mobile processors, featuring 6 Compute Units (384 shaders) based on the RDNA 2 architecture.
6 listed key facts
High-performance integrated graphics processor (iGPU) featuring 12 Compute Units (768 shaders) based on the RDNA 2 architecture, typically found in Ryzen 6000 and 7000 series mobile APUs.
6 listed key facts
Compare
Daily
