Intel Gaudi 3 128GB vs Intel Gaudi 3 HL-338

The Gaudi 3 128GB and the Gaudi 3 HL-338 deliver near-identical FP8 throughput (1,678 vs 1,678 TFLOPS dense).

Gaudi 3 128GB: Gaudi 3, 2024

Memory

128 vs 128 GB

Same capacity

Bandwidth

3.7 TB/s vs 3.7 TB/s

Even

TDP

900 W vs 600 W

Gaudi 3 HL-338 draws 300 W less

Download comparison sheet Printable version →

Intel Gaudi 3 128GB

Full specs →

What is the Intel Gaudi 3 128GB?

The Gaudi 3 128GB is an ASIC from Intel on the Gaudi 3 architecture, an accelerator built for AI workloads. It is an OAM module, the open accelerator format used in multi-GPU servers. Gaudi 3 128GB full specifications →

Can I rent the Gaudi 3 128GB in the cloud?

Flopper is not tracking any live rental listings for the Gaudi 3 128GB yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the Gaudi 3 128GB compare with other accelerators?

This page sets the Gaudi 3 128GB against the Gaudi 3 HL-338. Both are Intel parts; the Gaudi 3 128GB is an OAM module while the Gaudi 3 HL-338 is a PCIe card. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the Gaudi 3 128GB figures come from?

Every Gaudi 3 128GB figure on this page traces to Intel's own datasheet or specification table. How Flopper verifies specifications →

Intel Gaudi 3 HL-338

Full specs →

What is the Intel Gaudi 3 HL-338?

The Gaudi 3 HL-338 is an ASIC from Intel on the Gaudi 3 architecture, an accelerator built for AI workloads. It is a PCIe add-in card, so it fits standard servers and workstations. Gaudi 3 HL-338 full specifications →

Can I rent the Gaudi 3 HL-338 in the cloud?

Flopper is not tracking any live rental listings for the Gaudi 3 HL-338 yet. The provider directory lists the clouds that rent comparable accelerators on demand. GPU cloud providers →

How does the Gaudi 3 HL-338 compare with other accelerators?

This page sets the Gaudi 3 HL-338 against the Gaudi 3 128GB. Both are Intel parts; the Gaudi 3 HL-338 is a PCIe card while the Gaudi 3 128GB is an OAM module. The compare tool builds the same comparison for any two parts in the catalogue. All GPU comparisons →

Where do the Gaudi 3 HL-338 figures come from?

Every Gaudi 3 HL-338 figure on this page traces to Intel's own datasheet or specification table. How Flopper verifies specifications →

Specifications

Gaudi 3 128GB

Gaudi 3 HL-338

Architecture

Gaudi 3Gaudi 3

Launch Year

2024—

Form Factor

OAMPCIe

Memory

128 GB128 GB

Memory Bandwidth

3.7 TB/s3.7 TB/s

TDP

900 W600 W

Process Node

5nm5nm

Spec

Gaudi 3 128GB

Gaudi 3 HL-338

Architecture Gaudi 3Gaudi 3

Launch Year 2024—

Form Factor OAMPCIe

Memory 128 GB128 GB

Memory Bandwidth 3.7 TB/s3.7 TB/s

TDP 900 W600 W

Process Node 5nm5nm

Performance (TFLOPS)

Gaudi 3 128GB

Gaudi 3 HL-338

229 TFLOPS 229 TFLOPS

459 TFLOPS 459 TFLOPS

1,678 TFLOPS 1,678 TFLOPS

459 TFLOPS 459 TFLOPS

1,678 TFLOPS 1,678 TFLOPS

FP6

Precision

Gaudi 3 128GB

Gaudi 3 HL-338

229 TFLOPS 229 TFLOPS

459 TFLOPS 459 TFLOPS

1,678 TFLOPS 1,678 TFLOPS

459 TFLOPS 459 TFLOPS

1,678 TFLOPS 1,678 TFLOPS

FP6

Figures are dense. Where a vendor publishes a figure it sits under the dense one and never decides a winner. Hover any precision for its definition, or read the GPU glossary.

FLOPS by Precision

What actually differs

Both carry 128 GB of memory. Bandwidth is 3.7 TB/s against 3.7 TB/s, which sets the ceiling on decode speed.

Power is 900 W against 600 W, and at FP32 the Gaudi 3 HL-338 does more work per watt. Weight the FP8 and FP16 rows for LLM work and the FP64 row for HPC.

New to the terms? The glossary explains FP32, TF32, BF16, FP16, FP8, Sparsity and TFLOPS.

Frequently asked questions

Is the Gaudi 3 128GB faster than the Gaudi 3 HL-338?

They are close on paper: both deliver about 1,678 TFLOPS of dense FP8 throughput. Memory, bandwidth, and power are the deciding differences.

Which should you choose, the Gaudi 3 128GB or the Gaudi 3 HL-338?

On raw FP8 throughput the Gaudi 3 128GB and the Gaudi 3 HL-338 are level, so the decision rests on memory, power and price. The Gaudi 3 HL-338 draws 300 W less power, which matters in an air-cooled or power-capped rack. Packaging differs too: the Gaudi 3 128GB is an OAM module while the Gaudi 3 HL-338 is a PCIe card, so check what your servers accept.

Which has more memory, the Gaudi 3 128GB or the Gaudi 3 HL-338?

Both GPUs carry 128 GB of memory. Memory bandwidth is 3.7 TB/s for the Gaudi 3 128GB and 3.7 TB/s for the Gaudi 3 HL-338.

How much power do the Gaudi 3 128GB and the Gaudi 3 HL-338 draw?

The Gaudi 3 128GB is rated at 900 W TDP and the Gaudi 3 HL-338 at 600 W. On FP32 throughput per watt, the Gaudi 3 HL-338 is the more efficient part.

Get Comparison Updates

New GPUs added weekly. Be the first to see how they compare.

Open in Full Compare Tool → Gaudi 3 128GB full specs Gaudi 3 HL-338 full specs Estimate cluster cost in the calculator →

Intel Gaudi 3 128GB vs Intel Gaudi 3 HL-338

Intel Gaudi 3 128GB

What is the Intel Gaudi 3 128GB?

Can I rent the Gaudi 3 128GB in the cloud?

How does the Gaudi 3 128GB compare with other accelerators?

Where do the Gaudi 3 128GB figures come from?

Intel Gaudi 3 HL-338

What is the Intel Gaudi 3 HL-338?

Can I rent the Gaudi 3 HL-338 in the cloud?

How does the Gaudi 3 HL-338 compare with other accelerators?

Where do the Gaudi 3 HL-338 figures come from?

Specifications

Performance (TFLOPS)

FLOPS by Precision

What actually differs

Frequently asked questions

Is the Gaudi 3 128GB faster than the Gaudi 3 HL-338?

Which should you choose, the Gaudi 3 128GB or the Gaudi 3 HL-338?

Which has more memory, the Gaudi 3 128GB or the Gaudi 3 HL-338?

How much power do the Gaudi 3 128GB and the Gaudi 3 HL-338 draw?

Get Comparison Updates

Stay Updated on GPU Specs