Skip to content
MCS · Rigs · Compare

Side-by-side, with the numbers that matter.

Two rigs, one table. Higher numbers win in cobalt; TDP and price flip.

Rig A

The Stargate Project

OpenAI / Oracle / SoftBank / MGX · workstation

Change

Rig B

NVIDIA GB200 NVL72

NVIDIA · workstation

Change

0 of 38 fields comparable

⇄ swap sides
SpecThe Stargate ProjectNVIDIA GB200 NVL72
Abilene Trained ModelGPT-5.5—
Abilene Trained Model Version5.5—
Blackwell Gpus—72
Capacity Added Prior 90 Days GW MIN3—
Capacity Added Window Days90—
Capacity Update Date2026-04-29—
Compute Trays—18
Cooling—liquid
CPU Memory Bandwidth TB S—14
CPU Memory Lpddr5x TB—17
Custom Silicon AnnouncedBroadcom (10 GW)—
Fp16 Bf16 Pflops—360
FP32 TFLOPS—5,760
FP64 TFLOPS—2,880
FP8 FP6 Pflops—720
GPU Memory Bandwidth TB S—576
Grace Cpus—36
Int8 Pops—720
Nvfp4 Pflops Dense—720
Nvfp4 Pflops Sparse—1,440
Nvlink Generation—5
Nvswitch Trays—9
Oracle Partnership GW5—
PartnersOpenAI, SoftBank, Oracle, MGX—
Planned Capacity GW 20257—
Planned Chip Count Million2—
Primary Accelerator AnnouncedNVIDIA GB200 (initial Oracle racks)—
Primary SiteAbilene, Texas—
Rack Power KW Approx—120
Secondary Accelerator AnnouncedAMD Instinct (up to 6 GW)—
Secured Capacity Commitment Deadline Year2,029—
Secured Capacity Commitment GW10—
Status Checked2026-09-24—
Tf32 Pflops—180
Total ARM Cores—2,592
Total Hbm3e TB—13.4
Total Nvlink Bandwidth TB S—130
Total Planned Investment USD Billion400—