AMD Instinct MI455X GPU launches with 432GB of HBM4 memory and next-gen CDNA 5 architecture

Q: What are the key hardware specifications of the AMD Instinct MI455X (process node, transistor count, clock speed, HBM4 capacity, and peak memory bandwidth)?

The AMD Instinct MI455X is built on a 2nm process and contains 320 billion transistors. It has a maximum clock speed of 2.4 GHz, 432 GB of HBM4 memory, and a peak memory bandwidth of 23.3 TB/sec.

Q: How is the MI455X configured differently from the MI355X in terms of XCDs, Shader Engines, and Work Group Processors?

Both GPUs use eight XCD chiplets, but whereas the MI355X had 32 active compute units per XCD, the MI455X maintains the same total number of Work Group Processors while reorganizing each XCD into two Shader Engines with 16 WGPs apiece. The MI455X’s WGPs are newly defined in CDNA 5, with each WGP made up of four 32-thread SIMD execution units and four scalar execution units, improving math and tensor throughput.

Q: How much faster is the MI455X in math throughput (MXFP8/MXFP4) and tensor/vector workloads compared to the MI355X?

The MI455X delivers up to 4X higher peak math throughput in MXFP8/MXFP4 operations compared to the MI355X. For tensor and vector workloads it is up to 2X faster.

Q: What token throughput improvement does AMD claim for the MI455X versus the MI355X, and how does that affect token cost?

AMD says the MI455X delivers about 34 times the token throughput of the MI355X. This higher throughput "in turn ... leads to a significant reduction in token cost," with no specific cost figures provided.

AMD Instinct MI455X GPU - 2nm process, 320 billion transistors, 2.4 GHz clock, 432GB of high-speed HBM4 memory, and 23.3 TB/sec peak memory bandwidth.

VIEW GALLERY - 4

Kosta Andreadis

Senior Editor

Published Jul 23, 2026 1:44 PM CDT

1 minute & 45 seconds read time

TL;DR: AMD launched the Instinct MI455X GPU built on 2nm with 320 billion transistors, 2.4 GHz clocks, 432 GB HBM4 and 23.3 TB/s bandwidth. CDNA 5 uses eight XCD chiplets with redesigned WGPs, boosting MXFP throughput up to 4× and token throughput 34× versus the MI355X.

Voice: Kosta AndreadisSpeed

0:00 / 3:51

Coinciding with AMD's Advancing AI 2026 event, the company has launched its latest AMD Instinct data center GPUs purpose-built for the AI era, with the Instinct MI400 Series led by the flagship AMD Instinct MI455X GPU. Built on a 2nm process, it sports 320 billion transistors, a maximum clock speed of 2.4 GHz, and 432 GB of high-speed HBM4 memory with a peak memory bandwidth of 23.3 TB/sec.

AMD Instinct MI455X GPU launches with 432GB of HBM4 memory and next-gen CDNA 5 architecture 1

VIEW GALLERY - 4 IMAGES

According to the company, the AMD Instinct MI455X GPU is its first GPU to be designed and built for rack-scale AI deployments, which is why it serves as a key part of the new AMD Helios rack, which features 72 Instinct MI455X GPUs for 31TB of HBM4 memory with an impressive 1.7 PB/s memory bandwidth and up to 2.9 exaflops of compute performance.

And with the AMD Instinct MI455X being built for Helios, this means the GPU has been designed with scale-up and scale-out capability with shared-memory capacity, communication bandwidth, and more connectivity compared to previous AMD Instinct GPU generations.

AMD Instinct MI455X GPU launches with 432GB of HBM4 memory and next-gen CDNA 5 architecture 2

Powering the AMD Instinct MI455X GPU is the company's latest CDNA 5 architecture, which introduces several improvements to accelerate and expand AI performance. Like its predecessor, the AMD Instinct MI355X, the new MI455X features eight Accelerator Complex Dies, or XCD, chiplets. However, where previously you had 32 Active Compute Units per XCD, the AMD Instinct MI455X GPU has the same number of Work Group Processors (WGPs), with each XCD divided into two Shader Engines (Ses) with 16 WGPs each.

The new WGP design comes from CDNA 5, which optimizes performance and latency. Each WGP itself is comprised of four 32-thread SIMD (single-instruction, multiple-data) execution Units, and four scalar execution units. With parallel processing, peak throughput in math operations per clock (MXFP8 and MXFP4) has increased by up to 4X compared to the AMD Instinct MI355X, and up to 2X when looking at tensor and vector data.

AMD Instinct MI455X GPU launches with 432GB of HBM4 memory and next-gen CDNA 5 architecture 3

And when it comes to token throughput measured in tokens per second, the AMD Instinct MI455X GPU is reportedly an impressive 34 times faster than the Instinct MI355X. This in turn, leads to a significant reduction in token cost.

Frequently Asked Questions

TweakBot answers common questions about this news using TweakTown's own coverage from this page and related content from our archive. Tap a question to reveal the answer, or type your own below.

Question #1

What are the key hardware specifications of the AMD Instinct MI455X (process node, transistor count, clock speed, HBM4 capacity, and peak memory bandwidth)?

Question #2

How is the MI455X configured differently from the MI355X in terms of XCDs, Shader Engines, and Work Group Processors?

Question #3

How much faster is the MI455X in math throughput (MXFP8/MXFP4) and tensor/vector workloads compared to the MI355X?

Question #4

What token throughput improvement does AMD claim for the MI455X versus the MI355X, and how does that affect token cost?

Have a question not listed here? Ask below and TweakBot will answer it.

In addition to the flagship AMD Instinct MI455X GPU, AMD has also launched the Instinct MI430X GPUs for sovereign AI and HPC, which can deliver up to 288 TFLOPS of hardware-based FP64 performance.