Rack & Pod Scale

AI GPU Platforms

Enterprise AI runs on systems that combine many GPUs into a single product. Compare the rack-scale and pod-scale configurations that power today's frontier AI models. See their average FLOPS, power draw, and bandwidth side by side, and understand the machines behind world-class AI training and inference.

Compare Systems

Filters

Vendor

Min GPU Count

Max GPU Count

Available Systems

#						View
1	Huawei Atlas 950 SuperPoD pod	1024× Ascend 950DTDa Vinci v3	2.00 EF FP4	1.00 EF	—	View →
2	NVIDIA Groq 3 LPX rack	256× Groq 3 LPUGroq LPU	307 PF FP8	307 PF	—	View →
3	AMD Helios rack	72× Instinct MI455X OAMCDNA 5	2,899 PF FP4	1,450 PF	—	View →
4	Google TPU 8i Pod pod	1024× TPU 8iTPU 8	10.3 EF FP4	—	—	View →
5	Google TPU 8t Superpod pod	9600× TPU 8tTPU 8	121 EF FP4	—	—	View →
6	NVIDIA Vera Rubin NVL72 rack	72× Rubin SXMRubin	2,520 PF NVFP4	1,260 PF	—	View →
7	Huawei Atlas 900 A3 SuperPoD pod	384× Ascend 910CDa Vinci v2 (dual-die)	614 POPS INT8	—	—	View →
8	NVIDIA DGX GB300 NVL72 rack	72× GB300 NVL72 GPU 288GBBlackwell Ultra	1,080 PF FP4	360 PF	142.0 kW	View →
9	NVIDIA DGX SuperPOD GB200 pod	2304× GB200 NVL72 GPU 186GBBlackwell	23.0 EF FP4	11.5 EF	3,840.0 kW	View →
10	NVIDIA GB200 NVL32 rack	32× GB200 NVL72 GPU 186GBBlackwell	320 PF FP4	160 PF	53.3 kW	View →
11	NVIDIA GB200 NVL36	36× GB200 NVL72 GPU 186GBBlackwell	360 PF FP4	180 PF	60.0 kW	View →
12	NVIDIA GB200 NVL576	576× GB200 NVL72 GPU 186GBBlackwell	5.76 EF FP4	2.88 EF	960.0 kW	View →
13	NVIDIA GB200 NVL72 rack	72× GB200 NVL72 GPU 186GBBlackwell	720 PF FP4	360 PF	120.0 kW	View →
14	NVIDIA GB300 NVL72 rack	72× GB300 NVL72 GPU 288GBBlackwell Ultra	—	—	142.0 kW	View →
15	Google TPU7x (Ironwood) Pod pod	9216× TPU v7 192GBTPU	42.5 EF FP8	42.5 EF	—	View →
16	AWS Trn3 UltraServer rack	144× Trainium3NeuronCore-v4	362 PF FP8	362 PF	—	View →
17	Cerebras CS-3 Inference Cluster (16-node) cluster	16× WSE-3Wafer Scale Engine 3	—	—	475.0 kW	View →
18	Cerebras CS-3 Inference Cluster (32-node) cluster	32× WSE-3Wafer Scale Engine 3	—	—	945.0 kW	View →
19	Cerebras CS-3 Inference Cluster (64-node) cluster	64× WSE-3Wafer Scale Engine 3	—	—	1,860.0 kW	View →
20	Cerebras CS-3 Inference Cluster (88-node) cluster	88× WSE-3Wafer Scale Engine 3	—	—	2,540.0 kW	View →
21	AWS EC2 trn2.48xlarge rack	16× Trainium2NeuronCore-v3	20.8 PF FP8	20.8 PF	—	View →
22	Google TPU v6e Pod pod	256× TPU v6e 32GBTPU	470 POPS INT8	—	—	View →
23	AWS Trn2 UltraServer rack	64× Trainium2NeuronCore-v3	83.2 PF FP8	83.2 PF	—	View →
24	Google TPU v5e Pod pod	256× TPU v5e 16GBTPU	101 POPS INT8	—	—	View →
25	Google TPU v5p Pod pod	8960× TPU v5p 95GBTPU v5p	4.11 EF BF16	4.11 EF	—	View →
26	Google TPU v4 Pod pod	4096× TPU v4 32GBTPU	1.13 EF BF16	—	—	View →
27	Inventec Artemis II Rack rack	72× GB300 NVL72 GPU 288GBBlackwell Ultra	—	—	136.0 kW	View →
28	Inventec Artemis III Rack rack	72× Rubin SXMRubin	—	—	228.0 kW	View →
29	Inventec Artemis Rack rack	72× GB200 NVL72 GPU 186GBBlackwell	—	—	125.0 kW	View →
30	EF Enflame Cloud Blazer POD P2010 rack	32× Yunsui T20GCU / DTU 2.0	—	—	—	View →
31	EF Enflame Cloud Blazer POD P2110 rack	64× Yunsui i20GCU / DTU 2.5	—	—	—	View →
32	d-Matrix Corsair SquadRack rack	64× CorsairCorsair	614 POPS INT4	—	—	View →
33	Tenstorrent Galaxy Blackhole rack	32× Blackhole p150aBlackhole	23.0 PF FP8	23.0 PF	12.0 kW	View →
34	Lenovo Lenovo NVIDIA GB300 NVL72 rack	72× GB300 NVL72 GPU 288GBBlackwell Ultra	—	—	135.0 kW	View →
35	HPE NVIDIA GB300 NVL72 by HPE rack	72× GB300 NVL72 GPU 288GBBlackwell Ultra	—	—	132.0 kW	View →
36	Rebellions RebelPOD (32-node) pod	256× Rebel100REBEL	524 PF FP8	524 PF	224.0 kW	View →
37	Rebellions RebelRack rack	32× Rebel100REBEL	65.5 PF FP8	65.5 PF	28.0 kW	View →
38	SambaNova SambaRack SN40L-16 rack	16× SN40LRDU	10.2 PF BF16	—	10.0 kW	View →
39	MetaX 曦云C500X 光链超节点 (Optical Supernode) cluster	64× C500X	—	—	—	View →
40	MetaX 曦云C550 3D Mesh 超节点 cluster	64× C550	—	—	—	View →
41	MetaX 曦云C550 Shanghai Cube 液冷整机柜 rack	128× C550	—	—	—	View →

per page

1/1

Rent These Systems

Cloud providers offering managed instances with these GPU systems.

Provider	Configuration	GPU	Price/GPU-hr	Checked
CoreWeave	4 GPUs	GB200 NVL72 GPU 186GB	$10.50	2h ago	View →
Oracle Cloud (OCI)	1 GPU	GB200 NVL72 GPU 186GB	$16.00	2h ago	View →
DataCrunch	1 GPU Spot	GB300 NVL72 GPU 288GB	$4.31	2h ago	View →
DataCrunch	1 GPU	GB300 NVL72 GPU 288GB	$8.62	2h ago	View →
Oracle Cloud (OCI)	1 GPU	GB300 NVL72 GPU 288GB	$18.00	2h ago	View →

AI GPU Platforms

Filters

Available Systems

Rent These Systems

Stay Updated on GPU Specs