GPU & AI Accelerators

AMD Radeon RX 9050 — Entry-Level RDNA 4 Graphics Card Intel Arc Pro B65 — Professional Workstation GPU · Xe2 Battlemage Lisuan LX MAX — Domestic AI GPU Accelerator · LX Series NVIDIA A100 80GB PCIe NVIDIA A100 80GB SXM4 AMD Instinct MI210 64GB — AI Accelerator AMD Instinct MI250X 128GB — AI Accelerator AMD Instinct MI325X — 256GB HBM3E · 1307 FP16 TFLOPS · CDNA 3 AMD Instinct MI355X — 288GB HBM3E · 10.1 PFLOPs MXFP4 · CDNA 4 AMD Instinct MI400 — 432GB HBM4 · 40 PFLOPS FP4 · CDNA 5 · TSMC N2 AMD Instinct MI300X 192GB — AI Accelerator MI300X-OAM-192GB — AMD Instinct MI300X Accelerator 192GB AMD Radeon Instinct MI50 32GB HBM2 — Datacenter GPU Accelerator NVIDIA B200 SXM6 180GB — GPU Accelerator NVIDIA H100 NVL 94GB NVIDIA H100 PCIe 80GB — GPU Accelerator NVIDIA H100 SXM5 80GB — GPU Accelerator NVIDIA H200 SXM5 141GB — GPU Accelerator Hailo-10H M.2 — Edge AI Accelerator for Generative AI Intel Gaudi 2 HL-225 — AI Accelerator Intel Gaudi 3 — 128GB HBM2e · 1,835 TFLOPS FP8 Intel Gaudi 3 HL-338 PCIe Accelerator — 1,835 TFLOPS BF16 / 128GB HBM2e NVIDIA L20 48GB NVIDIA L40S 48GB — GPU Accelerator NVIDIA A10 24GB — Ampere Graphics & Inference GPU NVIDIA A100 80GB PCIe — TCS-A100-80GB-PCIE NVIDIA A40 48GB — TCS-A40-48GB NVIDIA A800 80GB PCIe — GPU Accelerator NVIDIA B200 180GB — 4,500 TFLOPS FP8 · 8.0 TB/s Bandwidth · 1,000W TDP NVIDIA B200 Blackwell Ultra — TCS-B200-192GB-SXM NVIDIA B200 PCIe 192GB — GPU Accelerator NVIDIA B300 — Next-Gen Blackwell Ultra GPU NVIDIA B300 Blackwell Ultra — 288GB HBM3e · 15 PFLOPS FP4 · 8 TB/s · Data Center GPU NVIDIA B300 NVL8 — 8-Way Blackwell Ultra HGX, 2.3TB HBM3e NVIDIA GB10 Grace Blackwell Superchip — TCS-GB10-128GB · GPU + CPU NVIDIA GH200 Grace Hopper Superchip — 480GB LPDDR5X + 144GB HBM3e NVIDIA H100 80GB PCIe — TCS-H100-80GB-PCIE NVIDIA H100 80GB SXM5 — GPU Accelerator NVIDIA H100 CN SXM5 80GB — Special Export Variant | QS Compute NVIDIA H100 PCIe 80GB — Hopper GPU NVIDIA H20 96GB — China-Market Hopper GPU NVIDIA H200 141GB — 900-2H200-0000-000 NVIDIA H200 NVL 141GB — Multi-GPU AI Accelerator NVIDIA H200 NVL 141GB — 3,958 TFLOPS FP8 · 4.8 TB/s Memory Bandwidth · 700W TDP 900-21010-2740-000 — NVIDIA H200 NVL 141GB PCIe GPU NVIDIA L4 Tensor Core GPU — 30.3 TFLOPS NVIDIA L4 24GB — 242 TFLOPS INT8 · Single-Slot LP · 72W TDP 900-2G193-0000-000 — NVIDIA L4 Tensor Core GPU 24GB NVIDIA L40 48GB — Data Center GPU NVIDIA L40 48GB — GPU Accelerator NVIDIA L40 CN 48GB — GPU Accelerator (China Market) NVIDIA L40S 48GB — GPU Accelerator NVIDIA L40S CN 48GB — Export-Compliant GPU NVIDIA L40S HGX — 8-GPU Baseboard NVIDIA L40S HGX 8-Way — 8× L40S 48GB Baseboard Server GPU NVIDIA RTX 4000 Ada SFF — VCNRTX4000ADA-SFF NVIDIA GeForce RTX 5090 — 32GB GDDR7 Blackwell GPU NVIDIA RTX 6000 Ada 48GB — TCS-RTX6000ADA-48GB NVIDIA RTX A4500 20GB — Professional GPU NVIDIA RTX A5000 24GB — Ampere Professional GPU NVIDIA RTX A6000 48GB — GPU Accelerator NVIDIA RTX PRO 2000 Blackwell — 16GB GDDR7 · 4,352 CUDA · 545 AI TOPS · 70W SFF NVIDIA RTX PRO 4000 Blackwell SFF — 24GB GDDR7 · 8960 CUDA · 70W Low-Profile NVIDIA RTX PRO 4500 Blackwell — 32GB GDDR7 · 811 TFLOPS FP8 · PCIe 5.0 NVIDIA RTX PRO 5000 Blackwell 72GB — 72GB GDDR7 / 2,142 AI TOPS NVIDIA RTX PRO 5000 Blackwell — 48GB GDDR7 · 2064 AI TOPS · PCIe 5.0 NVIDIA RTX PRO 5500 Blackwell — 84GB GDDR7 Workstation GPU · 1,398 GB/s NVIDIA RTX PRO 6000 Blackwell Server Edition Liquid — 96GB GDDR7 · Liquid-Cooled NVIDIA RTX PRO 6000 Blackwell — Workstation GPU NVIDIA RTX PRO 6000 Blackwell — TCS-RTXPRO6K-96GB NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition — TCS-RTXPRO6K-MQ-96GB NVIDIA RTX PRO 6000 Blackwell Server Edition — TCS-RTXPRO6K-SE-96GB NVIDIA Rubin GPU — 288GB HBM4 · 50 PFLOPS FP4 · Vera Rubin Platform NVIDIA T4 16GB — Turing Inference GPU 900-2G183-0000-000 NVIDIA RTX 4000 Ada Generation NVIDIA RTX 5000 Ada Generation — 65.3 TFLOPS NVIDIA RTX A6000 NVIDIA RTX PRO 6000 Blackwell 96GB — GPU Accelerator NVIDIA RTX 5090 32GB — GPU Accelerator NVIDIA RTX 6000 Ada 48GB — GPU Accelerator NVIDIA B100 Tensor Core GPU — 3.5 PFLOPS FP4 AMD Instinct MI300A APU — 5.3 TB/s NVIDIA RTX A2000 12GB — 8.0 TFLOPS Intel Arc Pro A60 12GB — 12 GB GDDR6 AMD Radeon PRO W7900 48GB — 61.3 TFLOPS Axelera Metis AIPU — Edge AI Accelerators (M.2 & PCIe, 214–856 TOPS) NVIDIA H200 NVL 141GB HBM3e PCIe — 900-21010-0040-000 NVIDIA H200 NVL 141GB HBM3e PCIe (Lenovo OEM) — 900-21010-2740-030 NVIDIA L20 48GB — 900-2G133-0080-000 NVIDIA L20 48GB — 900-2G133-00A0-000 NVIDIA RTX PRO 6000 Blackwell (Workstation) — 900-2G153-0000-000 NVIDIA RTX PRO 6000 Blackwell (Active Cooler) — 900-5G144-2200-000-01 NVIDIA RTX PRO 6000 Blackwell Max-Q — 900-5G153-2500-000 NVIDIA RTX Spark Superchip — 1 PFLOPS FP4 · Grace + Blackwell · 128GB Unified Qualcomm Cloud AI 200 / AI 250 — Datacenter Inference Accelerators NVIDIA RTX PRO 4500 Blackwell Workstation 900-5G147-2250-000-01 — 32GB GDDR7 NVIDIA RTX PRO 4500 Blackwell Server Edition 900-5G147-0150-000 — 32GB GDDR7 NVIDIA RTX PRO 6000 Blackwell Workstation 900-5G144-2500-000 — 96GB GDDR7 NVIDIA RTX PRO 6000 Blackwell 900-5G144-0000-000 — 96GB GDDR7 Workstation NVIDIA RTX 5880 Ada Generation — 48GB GDDR6 ECC Workstation GPU AMD Instinct MI350P — 144GB HBM3E PCIe AI Accelerator NVIDIA RTX PRO 5000 Blackwell 900-5G153-2250-000-01 — 48GB GDDR7 OEM NVIDIA RTX PRO 5000 Blackwell 900-5G153-2550-000 — 48GB GDDR7 Workstation NVIDIA RTX PRO 5000 72GB Blackwell 900-5G153-2270-000-01 — OEM NVIDIA RTX 5080 16GB NVIDIA RTX 5070 Ti 16GB AMD Radeon PRO W7800 32GB Intel Arc Pro B60 24GB AMD Radeon RX 9070 XT 16GB NVIDIA RTX 5070 16GB NVIDIA RTX 5060 Ti 16GB NVIDIA RTX 5060 8GB AMD Radeon RX 9070 16GB AMD Radeon RX 9060 XT 16GB NVIDIA RTX 5090D 32GB NVIDIA A2 16GB NVIDIA A30 24GB NVIDIA A16 64GB NVIDIA RTX A4000 16GB Intel Crescent Island — Xe3P Data Center GPU · 160GB LPDDR5X NVIDIA Groq 3 LPU — Language Processing Unit Inference Accelerator NVIDIA A100 80GB PCIe — 900-21001-0020-000 NVIDIA H100 NVL 94GB — 900-21010-0020-000 NVIDIA H100 NVL 94GB — 900-21010-0120-030 NVIDIA H100 PCIe 80GB — 900-21010-0000-000 NVIDIA RTX 4000 Ada 20GB — 900-5G190-2570-000 NVIDIA RTX 4000 Ada 20GB — 900-5G190-2270-000 NVIDIA RTX A4000 16GB — 900-5G190-2200-000 AMD Instinct MI250 — 128GB HBM2e CDNA 2 Accelerator NVIDIA RTX A2000 6GB — 6GB GDDR6 Workstation GPU NVIDIA RTX 5090 FE 900-1G144-2530-000 — 32GB GDDR7 Blackwell GPU NVIDIA RTX 5080 FE 900-1G144-2545-000 — 16GB GDDR7 Blackwell GPU PNY RTX 5090 OC VCG509032TFXPB1-O — 32GB GDDR7 Triple-Fan ADLINK EGX-MXM Blackwell — Embedded MXM GPU Modules (RTX PRO Blackwell) d-Matrix Corsair & Raptor — Digital In-Memory Compute (DIMC) AI Inference Accelerator NVIDIA Space-1 Vera Rubin Module — Space-Grade AI Compute for Orbital Data Centers NVIDIA GB300 Grace Blackwell Ultra Superchip — 288GB HBM3e · Grace CPU · Blackwell Ultra NVIDIA GB300 NVL72 — 72× B300 GPU + 36× Grace · Rack-Scale Blackwell Ultra NVIDIA GB200 Grace Blackwell Superchip — 384GB HBM3e · 2× B200 + Grace CPU NVIDIA GB200 NVL72 — 72× B200 + 36× Grace · Rack-Scale Grace Blackwell AMD Instinct MI350X — 288GB HBM3e · CDNA 4 · AI Accelerator Khadas Mind Graphics 2 — NVIDIA RTX 5060 Ti 16GB USB4 eGPU Dock Intel Arc B580 12GB — Battlemage BMG-G21 GPU Intel Arc B570 10GB — Battlemage BMG-G21 GPU NVIDIA RTX 5060 Ti 8GB — Blackwell GB206 GPU AMD Radeon RX 9060 XT 8GB — Navi 44 GPU NVIDIA RTX 4090 24GB — Ada Lovelace AD102 GPU NVIDIA L40S 48GB — 900-2G133-2780-030 GPU Accelerator AMD Instinct MI455X — 432GB HBM4 · CDNA 5 · AI Accelerator AMD Instinct MI450X — 432GB HBM4 · CDNA 5 · AI Accelerator AMD Instinct MI430X — 432GB HBM4 · 288 TFLOPS FP64 · HPC Accelerator AMD Instinct MI440X — 432GB HBM4 · CDNA 5 Entry Accelerator AMD Radeon RX 9060 XT LP — 16GB GDDR6 · Navi 44 · Low-Profile GPU NVIDIA L40S 48GB GPU — 900-2G133-0080-000 Blaize Xplorer X1600E / X600M — GSP Edge AI Inference Accelerators Cambricon MLU370-X8 — 256 TOPS Cloud AI Accelerator Card Tensordyne Napier — 3nm Log-Math AI Inference Accelerator, TDN72 Pod & Rack Hailo-8 Century — 52-208 TOPS PCIe Edge AI Inference Card (0802S-0808S) Euclyd CRAFTWERK — 32 PFLOPS FP4 Agentic AI SiP · 1TB UBM · 8,000 TB/s Axelera Europa AIPU — 629 TOPS INT8 · 64GB · PCIe Gen4 Moore Threads MTT S4000 — 48GB MUSA AI GPU · 200 TFLOPS FP16 · KUAE Cluster Concurrent Technologies Bragi / Condor — Rugged 3U OpenVPX GPGPU Cards Tenstorrent Blackhole p150 — 120 Tensix Cores · 32GB GDDR6 · 300W PCIe Gen5 AI Accelerator Qualcomm Cloud AI 100 Ultra — 128GB LPDDR4x · 870 TOPS INT8 · 150W Inference Accelerator SambaNova SN40L — 638 BF16 TFLOPS Dataflow RDU, 3-Tier Memory Graphcore C600 IPU-PROCESSOR Card — 1,472 IPU-Cores · 900 MB SRAM · 185W FuriosaAI RNGD — 512 TFLOPS FP8 · 48GB HBM3 · 180W LLM Inference Accelerator Rebellions REBEL (RBLN-CR13) — 144GB HBM3E · 2,048 TFLOPS FP8 · 350W Biren BR100 / BR104 — 2 PFLOPS INT8 · 64GB HBM2E · OAM & PCIe Enflame CloudBlazer Series — S60 / T20 / T21 / L600 AI Accelerator Cards MetaX Xiyun C500 — 560 TOPS INT8 · 64GB HBM2e · MXMACA Stack Iluvatar CoreX TianGai 100 — 7nm GPGPU · 147 TFLOPS FP16 · 32GB HBM2 Biren BR106 / BR110 / BR166 — Production GPGPU Family · 2.5D Chiplet · Birensupa NVIDIA L4 24GB — 900-2G193-0000-000 PCIe 4.0 Inference, Video & Graphics GPU Microsoft Maia 200 — Azure AI Inference Accelerator · 216GB HBM3e · FP4 10.15 PFLOPS Untether AI speedAI 240 Slim — 75W Low-Profile PCIe Inference Accelerator OpenAI Jalapeño — Broadcom LLM Inference ASIC Google Ironwood TPU v7 — 192GB HBM3E Accelerator Google TPU 8t & TPU 8i — 8th-Gen Training & Inference TPUs AWS Trainium3 — 2.5 PFLOP MXFP8 Accelerator Meta MTIA 300 / 400 / 450 / 500 — AI Accelerator Family Etched Sohu — Transformer-Only Inference ASIC AMD Instinct MI350P PCIe — 144GB HBM3E AI Accelerator (4.6 PFLOPS) Skymizer HTX301 — 384GB PCIe AI Accelerator for 700B LLM Inference 900-5G153-2200-000 — RTX PRO 6000 Blackwell Max-Q Workstation Edition 96GB GDDR7 900-2G153-0000-000 — RTX PRO 6000 Blackwell Server Edition 96GB GDDR7 Intel Gaudi 3 AI Accelerator HL-325L NVIDIA L40 48GB GPU — 900-2G133-0010-000 Intel Arc Pro B70 — 32GB Workstation GPU, 367 INT8 TOPS Axelera Embedded 110m / 113m M.2 Accelerators — EMBEDDED-110M / 113M Taalas HC1 — Hardwired Llama-3.1 8B Accelerator DEEPX DX-H1 Quattro — 100 TOPS INT8 PCIe AI Inference Card DEEPX DX-H1 V-NPU — 50 TOPS AI Inference with Dual Video Codec Axelera Edge 130p — 214 TOPS Metis AIPU PCIe Accelerator Axelera Server 150p — 856 TOPS Quad-Metis PCIe Accelerator Tenstorrent Blackhole p100a — 120 Tensix Cores, 28 GB GDDR6 7STARLAKE SK901-AD5000 — 3U OpenVPX GPGPU, NVIDIA 5000 Ada