Intel Gaudi 3 vs AMD Instinct MI300A Full Specs

Summary

16,384 Shaders 438MHz	14,592 Shaders 2.1GHz
128GB HBM2e3.7TB/s	128GB HBM35.33TB/s
Peak·FP32·FP16 -	Peak·FP32·FP16 122.6 TFLOPS
Form Factor OAM Module	Form Factor APU SH5 Socket
TDP 900W	TDP 560W

Highlights

Peak OPS

Peak OPS FP32 FP16 Tensor FP8-16 FP8-16 Tensor Sparse Tensor FP8-32 FP8-32 Tensor Sparse Tensor FP16-16 FP16-16 Tensor Sparse Tensor FP16-32 FP16-32 Tensor Sparse Tensor FP32 FP64 Tensor FP64 Tensor BF16 BF16 Tensor Sparse Tensor TF32 Tensor INT8 Tensor INT4 Memory Bandwidth TDP

Gaudi 31.84 PFLOPSFP8-32 Tensor

x1

Instinct MI300A3.92 POPSINT4 Tensor Sparse

x2.14

Show Top Score

Theoretical Performance

Clock Speed438MHz UC·Boost·OC	Clock Speed2100MHz UC·Base·Boost·OC
Peak OPS 1.84 PFLOPSFP8-32 Tensor	Peak OPS 3.92 POPSINT4 Tensor Sparse
Tensor FP8-16 - FP8-16 Tensor Sparse - Tensor FP8-32 1.84 PFLOPS FP8-32 Tensor Sparse -	Tensor FP8-16 1.96 PFLOPS FP8-16 Tensor Sparse 3.92 PFLOPS Tensor FP8-32 1.96 PFLOPS FP8-32 Tensor Sparse 3.92 PFLOPS
Tensor FP16-16 - FP16-16 Tensor Sparse - Tensor FP16-32 459.3 TFLOPS FP16-32 Tensor Sparse -	Tensor FP16-16 980.6 TFLOPS FP16-16 Tensor Sparse 1.96 PFLOPS Tensor FP16-32 980.6 TFLOPS FP16-32 Tensor Sparse 1.96 PFLOPS
Tensor BF16 1.84 PFLOPS BF16 Tensor Sparse -	Tensor BF16 980.6 TFLOPS BF16 Tensor Sparse 1.96 PFLOPS
Tensor TF32 459.3 TFLOPS	Tensor TF32 490.3 TFLOPS
FP32 - Tensor FP32 229.6 TFLOPS	FP32 122.6 TFLOPS Tensor FP32 122.6 TFLOPS
-	FP64 61.29 TFLOPS Tensor FP64 122.6 TFLOPS
Tensor INT4 -	Tensor INT4 1.96 POPS
-	Tensor INT8 1.96 POPS
Texture Rate -	Texture Rate 1.92 TTexel/s

Core Config

Shaders 16,384 Shaders	Shaders 14,592 Shaders
TMUs -	TMUs 912 TMUs
Tensor Cores 64 T-Cores	Tensor Cores 912 T-Cores
CUs 8 CUs	CUs 228 CUs

Clocks

Base Clock -	Base Clock 1GHz
Boost Clock 438MHz	Boost Clock 2.1GHz

Cache

L2 Cache -	L2 Cache 16.4MB shared
L3 Cache -	L3 Cache 192MB shared

Memory

128GB HBM2e	128GB HBM3
Memory Bus 8192-bit	Memory Bus 8192-bit
Memory Speed 3.6GT/s	Memory Speed 5.2GT/s
Memory Bandwidth 3.7TB/s	Memory Bandwidth 5.33TB/s
ECC Yes	ECC Yes
ESRAM 96MBSize 12800GB/sBandwidth	ESRAM -

Power & Thermals

TDP 900W

TDP 560W

Encoder

-	Encoder Model 2x VCN 2.6

Decoder

Decoder Model 14x Gaudi

Decoder Model 2x VCN 2.6

Card

Form Factor OAM Module	Form Factor APU SH5 Socket
Cooling Passive	Cooling Passive

Chip

Manufacturer Intel	Manufacturer AMD
Chip Designer Intel	Chip Designer AMD
Architecture Gaudi 3	Architecture CDNA 3
Family Gaudi	Family Instinct
Branding	Branding
Codename Gaudi 3	Codename Aqua Vanjaram
Chip Variant -	Chip Variant Aqua Vanjaram XT
Market Segment Server	Market Segment Server
Release Date Apr 9, 2024	Release Date Dec 6, 2023

Fabrication

Foundry TSMC	Foundry TSMC
Other Foundries -	Other Foundries TSMCActive Interposer
Fabrication Node N5 -	Fabrication Node N5 N6Active Interposer
Die Size -	Die Size 1110mm²
Transistor Count -	Transistor Count 109.5 Billion
Transistor Density -	Transistor Density 98.65 MTr/mm²

Images

No images available

No images available

Competitors

IntelGaudi 3 vs GPU AMDInstinct MI250X IntelGaudi 3 vs GPU AMDInstinct MI250 IntelGaudi 3 vs GPU AMDInstinct MI300X Show 12 more

AMDInstinct MI300A vs GPU AMDInstinct MI300X AMDInstinct MI300A vs GPU AMDInstinct MI250X AMDInstinct MI300A vs GPU AMDInstinct MI325X Show 12 more