Google TPU IronwoodGoogle
11
Inference-first TPU pods
- Version
- 7th-generation (Ironwood)
- Cost
- from $5.4/unit
- Model
- Gemini
AMD Instinct MI355X
AMD · #6 most active of 18 in Chips & Compute
288GB HBM3E accelerators with the ROCm software stack, used by several inference providers.
100
Trust score · Verified
2 data points checked · 2026-11-08
Current version
MI355X
Entry cost
Enterprise hardware, pricing not published
Changes / 30d
3
The short answers · verified November 3, 2026
- What is the latest version of AMD Instinct MI355X?
- The current shipped version of AMD Instinct MI355X is MI355X, as of November 3, 2026.
- How much does AMD Instinct MI355X cost?
- AMD Instinct MI355X has no published flat entry price. Enterprise hardware, pricing not published
- What AI model does AMD Instinct MI355X use?
- AMD Instinct MI355X runs primarily on CDNA.
Source: AMD — rocm.docs.amd.com · Reusable under CC BY 4.0 — cite Tomorrow
Capabilities
- 288GB HBM3E
- FP4/FP6 support
- ROCm 7
- open rack designs
- DPX compute partition mode
- CPX compute partition mode
- NPS2 memory partitioning
- GPU virtualization support
- SR-IOV support
- CDNA4 microarchitecture
- ROCm Core SDK support
- SR-IOV deployments
- GPU partitioning (DPX, CPX)
- ROCm support
- CDNA4 architecture
- GPU Partitioning
- DPX Compute Partition Mode
- CPX Compute Partition Mode
- NPS2 Memory Partitioning
- SR-IOV Deployments
- vLLM Support
- PyTorch Support
- JAX Support
- GPU computing
- AI inference
- HPC
- GPU partitioning
- GPU Virtualization
- DPX Compute Partition
- CPX Compute Partition
- CDNA4 Architecture
- ROCm Support
- SR-IOV virtualization
- GPU compute partition
- DPX compute partition
- CPX compute partition
- ROCm 7.14.0 support
- SR-IOV deployment
- AI inference optimization
- SR-IOV
- vLLM support
- PyTorch support
- JAX support
- GPU partitioning support
- SGLang support
- ROCm 7.14.0 compatibility
- GPU virtualization
- HPC workloads
- GPU compute partitioning
- ROCm Core SDK 7.14.0
- GPu Virtualization support
- CDNA4
- ROCm 10.0.0 Support
- GPU Virtualization Support
- AI Inference Acceleration
- Symmetric Memory Support
- ROCm 10.0.0 support
- Passthrough support
- Passthrough Ubuntu 22.04
- SR-IOV Support
- SGLang Support
- Symmetric Memory
- Triton Operator Tracing
- Passthrough Ubuntu
- HIP support
- ROCm Core SDK
- GPU Virtualization (SR-IOV)
- MI350 microarchitecture
- Passthrough Virtualization
- AI Inference
- HPC Workloads
- Passthrough Support
- CDNA4 Microarchitecture
- ROCm Core SDK 10.0.0 Support
- TensorFlow Support
- AI training
- GPU Computing
- ROCm Core SDK 10.0.0
- Passthrough Ubuntu support
- AI inference acceleration
- CDNA4 architecture support
- HIP runtime support
- KVM Passthrough support
- KVM Passthrough
- Virtualization (KVM Passthrough)
- ROCm SDK support
- KVM SR-IOV support
- ROCm Platform Support
- CDNA Architecture
- KVM SR-IOV
- CDNA architecture
- KVM SR-IOV Virtualization
- KVM SR-IOV virtualization
- HIP Runtime support
- KVM SR-IOV Support
- GPU virtualization (KVM SR-IOV)
- ROCm Core SDK 10.1.0
- ROCm SDK Support
Pricing
Open Source
Free open-source software platform for AMD Instinct GPUs
Free
Entry price over time
Geek mode
Models under the hood
- disclosed
CDNA
AMD
- disclosed
CDNA 4
AMD
- disclosed
ROCm
AMD
Context window
—
Public API
no
Multi-model routing
no
Stack signals
- ROCm
- CDNA4
- HIP
Shared model stack
Other tracked products running on the same foundation models — a quick read on how much of the catalog moves when one of these models changes.
- CDNAAMD · 1 product
- CDNA 4AMD · 1 product
- ROCmAMD · 1 product
green disclosed · amber inferred · grey unattributed. Aliases are folded into one model; provider concentration counts only models with an established vendor.
Reported scale
Reference data. Each figure is whatever the source actually said — weekly users, downloads, revenue run-rate — with its own definition and date. These are not comparable between tools and are never used to rank anything.
No public usage figure on record for this product.
Public market scorecard
AMD investor relationsAMD trades as NASDAQ: AMD. Figures below are as reported for Q2 2026 on Aug 4, 2026; market levels are the close on Aug 6, 2026 and are a reference marker, not a live quote.
Q2 2026 revenue
~$9.5B
+30% YoY
- Data center segment
- ~$4.5B
- MI series ramp
- Gross margin
- ~54%
Also in Chips & Compute
All Chips & ComputeCerebras WSE-3
Cerebras
9
Wafer-scale inference and training
- Version
- CS-4
- Cost
- Free trial credits available; developer tier has free promotional credits.
- Model
- Codex-Spark
Tenstorrent Blackhole
Tenstorrent
8
Open RISC-V AI hardware
- Version
- Blackhole
- Cost
- from $999/unit
- Model
- Blackhole Tensix
7
Cloud-native training silicon
- Version
- Trainium2
- Cost
- Pricing is not published on the page. Available via EC2 instances/UltraServers.
- Model
- Neuron compiler
NVIDIA Blackwell
NVIDIA
5
GB200/B200 rack-scale AI systems
- Version
- GB200 NVL72
- Cost
- Pricing is not published (enterprise/quote-only)
- Model
- Blackwell
Groq LPU
Groq
3
Deterministic low-latency inference
- Version
- GroqCloud API
- Cost
- Pricing not published on current page. Custom/usage rates apply.
- Model
- canopylabs/orpheus-arabic-saudi
Change history
- capability
- model
- capability
New capabilities: GPU virtualization (KVM SR-IOV)
95 tracked96 tracked · +GPU virtualization (KVM SR-IOV)
source - capability
- capability
New capabilities: KVM SR-IOV, CDNA architecture
89 tracked91 tracked · +KVM SR-IOV, CDNA architecture
source - capability
- capability
New capabilities: ROCm Platform Support, CDNA Architecture
87 tracked89 tracked · +ROCm Platform Support, CDNA Architecture
source - capability
- capability
- capability
- capability
- version
- version
- capability
New capabilities: ROCm 10.0.0 Support, GPU Virtualization Support, AI Inference Acceleration
52 tracked56 tracked · +ROCm 10.0.0 Support, GPU Virtualization Support, AI Inference Acceleration, Symmetric Memory Support
source - capability
New capabilities: Passthrough Ubuntu support, AI inference acceleration, CDNA4 architecture support
78 tracked81 tracked · +Passthrough Ubuntu support, AI inference acceleration, CDNA4 architecture support
source - capability
New capabilities: GPU Computing, ROCm Core SDK 10.0.0
76 tracked78 tracked · +GPU Computing, ROCm Core SDK 10.0.0
source - capability
- version
- capability
- version
- capability
- capability
New capabilities: ROCm Core SDK 10.0.0 Support
73 tracked74 tracked · +ROCm Core SDK 10.0.0 Support
source - capability
- capability
- capability
- capability
New capabilities: SR-IOV Support, SGLang Support, Symmetric Memory
59 tracked63 tracked · +SR-IOV Support, SGLang Support, Symmetric Memory, Triton Operator Tracing
source - capability
- version
- version
- capability
- capability
New capabilities: HIP support, ROCm Core SDK
64 tracked66 tracked · +HIP support, ROCm Core SDK
source - capability
New capabilities: GPU Virtualization (SR-IOV)
66 tracked67 tracked · +GPU Virtualization (SR-IOV)
source - version
- version
- version
- capability
- version
- capability
- capability
New capabilities: Passthrough Virtualization
68 tracked69 tracked · +Passthrough Virtualization
source - capability
New capabilities: Virtualization (KVM Passthrough)
84 tracked85 tracked · +Virtualization (KVM Passthrough)
source
AMD Instinct MI355X compared
Straight head-to-head pages against the busiest products in Chips & Compute.
How to cite this page
Free to cite and reuse under CC BY 4.0. Permalink: https://tomorrow.aliensquad.ai/tools/amd-instinct
APA
Tomorrow. (2026). AMD Instinct MI355X — version, pricing and model stack [Data set entry]. AlienSquad. Retrieved 2026-11-08, from https://tomorrow.aliensquad.ai/tools/amd-instinct
BibTeX
@misc{tomorrow-tools-amd-instinct,
author = {{Tomorrow}},
title = {AMD Instinct MI355X — version, pricing and model stack},
year = {2026},
publisher = {AlienSquad},
howpublished = {\url{https://tomorrow.aliensquad.ai/tools/amd-instinct}},
note = {Accessed: 2026-11-08}
}