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

Groq Language Processing Unit

Groq · accelerator

Change

Rig B

The Stargate Project

OpenAI / Oracle / SoftBank / MGX · workstation

Change

0 of 30 fields comparable

⇄ swap sides
SpecGroq Language Processing UnitThe Stargate Project
Abilene Trained Model—GPT-5.5
Abilene Trained Model Version—5.5
Capacity Added Prior 90 Days GW MIN—3
Capacity Added Window Days—90
Capacity Update Date—2026-04-29
Chip TO Chip Links11—
Custom Silicon Announced—Broadcom (10 GW)
DOJ Probe Reported2026-09-10—
Form factorPCIe Gen4 x16—
Groq3 LPX Full Production2026-08-24—
Groqcard Sales Ended Reported2024-04-05—
Nvidia Consideration AT Closing USD Billion13—
Nvidia Consideration Deferred USD Billion4—
Nvidia License Agreement2025-12-24—
ON Chip Sram MB230—
ON DIE Memory Bandwidth TB S80—
Oracle Partnership GW—5
Partners—OpenAI, SoftBank, Oracle, MGX
Peak Fp16 Tflops188—
Peak Int8 Tops750—
Planned Capacity GW 2025—7
Planned Chip Count Million—2
Primary Accelerator Announced—NVIDIA GB200 (initial Oracle racks)
Primary Site—Abilene, Texas
Process Node14nm—
Secondary Accelerator Announced—AMD Instinct (up to 6 GW)
Secured Capacity Commitment Deadline Year—2,029
Secured Capacity Commitment GW—10
Status Checked2026-09-242026-09-24
Total Planned Investment USD Billion—400