Instinct MI100

AMDdatacenter
Overview
ArchitectureCDNA 1 (2020)
Launch date2020-11-16
Process nodeTSMC 7nm FinFET
GPU target (gfx)gfx908
Memory
VRAM32 GB HBM2
Memory bandwidth1,200 GB/s
ECCYes
Compute — vector
FP6411.5 TFLOPS
FP3223.1 TFLOPS
Compute — matrix / tensor
BF1692.3 TFLOPS
FP16184.6 TFLOPS
INT892.3 TOPS
INT492.3 TOPS
Cores & clocks
Compute Units120
Shader cores7,680
Peak clock1,502 MHz
Board & system
TDP300 W
Form factorPCIe Add-in Card
CoolingPassive
PCIePCIe 4.0 x16, PCIe 3.0 x16
InterconnectInfinity Fabric — 92 GB/s
Multi-instance supportN/A

Compatible ROCm versions

4.5.04.5.25.0.05.0.15.0.25.1.05.1.15.1.35.2.05.2.15.2.35.3.05.3.25.3.35.4.05.4.15.4.25.4.35.5.05.5.15.6.05.6.15.7.05.7.16.0.06.0.26.1.06.1.16.1.26.1.56.2.06.2.16.2.26.2.46.3.06.3.16.3.26.3.36.4.06.4.16.4.26.4.37.0.07.0.17.0.27.1.07.1.17.2.07.2.17.2.27.2.37.2.47.14.010.0.0
fp8TFLOPS omitted — CDNA 1 predates FP8 matrix support (added in CDNA 3 with MI300). No transistor count or Matrix Cores count is published for this generation. Unlike MI250X/MI300-series, bfloat16 (92.3 TFLOPs) is HALF of fp16TFLOPS (184.6) on this part rather than equal to it — AMD's page lists them as genuinely different rates here; int8TOPS/int4TOPS (92.3 each) match the bf16 rate, not the fp16 rate. No separate FP64/FP32 matrix figures are published — only the single "Peak Double/Single Precision (FP64/FP32) Performance" rows, used directly (FP32 Matrix and FP32 vector figures happen to be identical on this page, 23.1, so only one field is populated). No msrpUSD — AMD doesn't publish Instinct list prices. multiInstanceSupport set to N/A — no SR-IOV/partitioning listed (introduced starting with MI300-series). Source: AMD's official product/spec page (amd.com/en/products/accelerators/instinct/mi100.html, "Expand All" spec accordion), pulled 2026-09-01.

← All GPUs · Compare GPUs →