---
schema_version: '1.0'
id: news-20260826-3360e6
title: NVIDIA「Groq 3 LPX」が量産開始、毎秒3,400トークンでAIエージェントの応答を高速化 | 財経新聞
url: https://www.zaikei.co.jp/article/20260826/867467.html
url_hash: 3360e62179c0347fe06ed9351bd5d46405b231f877023ddf2e6dc1eb8f83732b
canonical_url: https://www.zaikei.co.jp/article/20260826/867467.html
source: www.zaikei.co.jp
category: news/tech
category_raw: technology
region: null
tags:
- technology
- ai
- semiconductor
- gpu
- inference_accelerator
lang: null
published_at: '2026-08-26T14:41:59Z'
fetched_at: '2026-08-26T16:47:48.830577Z'
updated_at: '2026-08-26T16:52:00Z'
status: published
content_hash: null
content_changed_at: null
license_note: link-only
summary: null
summary_source: null
summary_en: null
summary_omitted_reason: license_restricted
redistribution: false
keywords_en:
- AI
- artificial intelligence
- gpu
- GPU HBM Memory
- Groq
- inference_accelerator
- LPX
- Nordic Semiconductor
- NVIDIA
- semiconductor
- technology
entities_en:
- name: AI
  type: concept
- name: GPU HBM Memory
  type: artifact
- name: Groq
  type: organization
- name: Nordic Semiconductor
  type: organization
- name: NVIDIA
  type: organization
- name: Technology
  type: UNKNOWN
topics_en:
- artificial intelligence
- technology
numeric_facts: []
entities:
- name: Technology
  type: UNKNOWN
- name: AI
  type: concept
- name: Nordic Semiconductor
  type: organization
- name: GPU HBM Memory
  type: artifact
- name: NVIDIA
  type: organization
- name: Groq
  type: organization
key_facts: []
related: []
related_auto:
- name: 法律業界
  type: location
  weight: 4.0
- name: デザイン
  type: concept
  weight: 2.0
- name: ライティング
  type: concept
  weight: 2.0
- name: Windows
  type: location
  weight: 1.0
- name: Infrastructure Maintenance
  type: UNKNOWN
  weight: 1.0
computed_at: '2026-08-26T16:51:52Z'
---

# NVIDIA「Groq 3 LPX」が量産開始、毎秒3,400トークンでAIエージェントの応答を高速化 | 財経新聞

## TL;DR
（要約なし・原文はリンク先を参照 / No summary available; see the source link.）

## Key Points
- technology / ai / semiconductor / gpu / inference_accelerator

## Details
(本文なし。リンク先参照)

## Source
元記事: [NVIDIA「Groq 3 LPX」が量産開始、毎秒3,400トークンでAIエージェントの応答を高速化 | 財経新聞](https://www.zaikei.co.jp/article/20260826/867467.html) — published 2026-08-26T14:41:59Z