Showing 25–36 of 46 systems
SKU: 676547
~166 TB HBM3e (576× 288 GB) · 288× NVIDIA Grace (ARM) · Grace LPDDR5X coherent memory · 8 PB parallel NVMe · Spine-leaf InfiniBand (NDR/XDR)
GB300 NVL72 SuperPOD
Fits Frontier multi-rack
SKU: 876046
1,128 GB HBM3e (8× 141 GB) · 2× AMD EPYC 9005 · 2 TB DDR5 ECC · 30 TB NVMe · InfiniBand NDR 400G
Cluster-ready training building block
Fits 180B+
SKU: 504910
1,128 GB HBM3e (8× 141 GB) · 2× Intel Xeon 6 · 2 TB DDR5 ECC · 60 TB NVMe · 8× InfiniBand NDR 400G
Large-model training node
Fits 405B
SKU: 696670
6,912 GB HBM3e (36× 192 GB) · 18× NVIDIA Grace (ARM) · Grace LPDDR5X coherent memory · 240 TB NVMe · NVLink + InfiniBand NDR 400G
Unified NVLink-domain training rack
Fits Trillion-scale
SKU: 438718
13,824 GB HBM3e (72× 192 GB) · 36× NVIDIA Grace (ARM) · Grace LPDDR5X coherent memory · 480 TB NVMe · NVLink domain + InfiniBand spine
72-GPU unified NVLink training rack
Fits Trillion-scale
SKU: 584582
20,736 GB HBM3e (72× 288 GB) · 36× NVIDIA Grace (ARM) · Grace LPDDR5X coherent memory · 480 TB NVMe · NVLink domain + InfiniBand spine
Blackwell Ultra unified NVLink rack
Fits Trillion-scale frontier
SKU: 300072
20 TB · 36× Grace · 17 TB LPDDR5X · Configurable · 130 TB/s NVLink
Training & inference
Fits Trillion-parameter
₹42 Cr+ GST · Made to order · Ships in 12 weeks
Add to cart
SKU: 645587
100 TB usable (NVMe) · 80 GB/s read · 24× 7.68 TB NVMe · 2× 200G InfiniBand/Ethernet
Fits AI training data & checkpoints
SKU: 620866
192 GB GDDR7 (2× 96 GB) · 2× Intel Xeon 6 · 512 GB DDR5 ECC · 8 TB NVMe · 2× 25 GbE
Production inference & multi-model serving
Fits 70B
SKU: 298000
250 TB usable (NVMe) · 160 GB/s read · 24× 15.36 TB NVMe · 4× 200G InfiniBand/Ethernet
Fits AI training data & checkpoints
SKU: 778492
384 GB GDDR7 (4× 96 GB) · 2× Intel Xeon 6 · 1 TB DDR5 ECC · 16 TB NVMe · 2× 25 GbE
Inference at scale & fine-tuning
Fits 70B+
SKU: 469979
50 TB usable (NVMe) · 40 GB/s read · 12× 7.68 TB NVMe · 2× 100G
Fits AI training data & checkpoints
Showing 25–36 of 46 systems 