AMD publishes no four-bit integer figure for this accelerator. Its four-bit entry is MXFP4, a floating-point format, which is shown here as FP4. An INT4 figure that had been recorded was a copy of this card's INT8 number and has been removed.

AMD Instinct MI350X OAM

Name: AMD Instinct MI350X OAM
Brand: AMD

CDNA 4 OAM 2025 3nm/6nm Spec confidence: Official

FP8 (dense)

4,614

TFLOPS

FP32

144

TFLOPS

VRAM

288 GB

Bandwidth

8.0 TB/s

memory

TDP

1.0 kW

144.2 TFLOPS/kW

Download datasheet Rental pricing from $6.16/hr →

Performance

Peak theoretical throughput by precision type

FP6

9,228 TFLOPS

sparse not published

INT8

8-bit integer

4,614 TOPS

9,227 with sparsity

FP8

8-bit floating point

4,614 TFLOPS

9,227 with sparsity

FP4

4-bit floating point

9,228 TFLOPS

sparse not published

FP16

16-bit floating point

2,310 TFLOPS

4,619 with sparsity

BF16

Brain Float 16

2,310 TFLOPS

4,619 with sparsity

FP32

32-bit floating point

144 TFLOPS

FP64

64-bit floating point

72 TFLOPS

Dense Added with sparsity

4.614 TFLOPS/W FP8 at 1000 W

2.310 TFLOPS/W FP16 at 1000 W

0.144 TFLOPS/W FP32 at 1000 W

The Instinct MI350X OAM in the GPU landscape

Peak FP16 TFLOPS (dense) against TDP, single-GPU parts tracked by Flopper

Higher and further left is better: more half-precision throughput for less power.

Specifications

Architecture

CDNA 4

Form Factor

OAM

Launch Year

2025

Process Node

3nm/6nm

Memory

288 GB

Bandwidth

8,000 GB/s

TDP

1.0 kW

Max Power

~1.1 kW

Spec Confidence

Official

Full Specifications

Memory
VRAM	288 GB
Bandwidth	8.0 TB/s
Power & Thermal
TDP	1.0 kW
Enterprise Features
Sparsity	Yes
General
Form Factor	OAM
Architecture	CDNA 4
Launch Year	2025

Datasheet & Resources

Flopper Datasheet

AMD Instinct MI350X OAM specifications, generated from our database. Printable.

View PDF

Vendor product page

AMD documentation for the Instinct MI350X OAM

View

AMD Instinct Documentation

ROCm guides, architecture details

Browse

Data Provenance

Every figure traced to a source

Primary Source

Document: --
Publisher: AMD
Published: --

Data Quality

Spec confidence: Official
Clock basis: Boost
Precisions tracked: 8
Normalization: All values in TFLOPS

Rent this GPU

$6.16 / GPU-hr

Lowest of 2 live listings across 1 provider

See all rental prices →

Price history and every provider on the pricing page. Some links are affiliate links.

Similar GPUs

AMD Instinct MI355X OAM

OAM · 2025

FP32: 157 TFLOPS

Compare vs Instinct MI355X

Frequently Asked Questions

How many TFLOPS does the AMD Instinct MI350X have?

The AMD Instinct MI350X delivers 144.2 TFLOPS for FP32 operations, 2309.6 TFLOPS for FP16, and 4614 TFLOPS for FP8 precision.

What is the power consumption of the AMD Instinct MI350X?

The AMD Instinct MI350X has a TDP (Thermal Design Power) rating of 1000 watts.

How much memory does the AMD Instinct MI350X have?

The AMD Instinct MI350X is equipped with 288 GB of VRAM with 8000 GB/s memory bandwidth.

What architecture is the AMD Instinct MI350X based on?

The AMD Instinct MI350X is based on the CDNA 4 architecture, launched in 2025.

Stay Updated on GPU Releases

Get notified when new GPUs are added or specifications are updated.

Loading verification...

No spam, unsubscribe anytime.

Back to GPUs