---
format: "aidr-story-markdown/v1"
id: "8936c063704937dc9c5e8b501453d4f01614d83fa30d62fd87f662b24ef8cba9"
canonical_url: "https://aidr.today/8936c063?lang=en"
title: "8 High HBM Now Minimum Stack Height for AI Inference Products"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-26T02:53:32.000Z"
category: "Chips"
topics: ["nvidia","inference"]
source_urls: ["https://marketbrief.now/ai/8-high-hbm-now-minimum-stack-height-for-ai-inference-products-3a754c4f","https://x.com/jukan05/status/2114499959563370930","https://x.com/vikramskr/status/2114505341832658983","https://huggingnews.com/ai/8-high-hbm-now-minimum-stack-height-for-ai-inference-products-3a754c4f"]
summary: "Cloud service providers and AI laboratories have shifted to 8-high as the minimum stack height for HBM in low-end inference products, prioritizing bandwidth over cost efficiency, with Nvidia modifying Rubin chip specs and AMD's MI450 projected to use 12 HBM stacks."
---

# 8 High HBM Now Minimum Stack Height for AI Inference Products

> [Open the canonical story](<https://aidr.today/8936c063?lang=en>)

**Published:** 2026-10-26T02:53:32.000Z
**Category:** Chips
**Topics:** nvidia, inference

## Summary

Cloud service providers and AI laboratories have shifted to 8\-high as the minimum stack height for HBM in low\-end inference products, prioritizing bandwidth over cost efficiency, with Nvidia modifying Rubin chip specs and AMD's MI450 projected to use 12 HBM stacks\.

## Sources

- [Story source](<https://marketbrief.now/ai/8-high-hbm-now-minimum-stack-height-for-ai-inference-products-3a754c4f>)
- [Story source](<https://x.com/jukan05/status/2114499959563370930>)
- [Supporting source](<https://x.com/vikramskr/status/2114505341832658983>)
- [Story source](<https://huggingnews.com/ai/8-high-hbm-now-minimum-stack-height-for-ai-inference-products-3a754c4f>)