In July 2026 Groq patented compiler technology for running large deep-learning models across multiple tensor streaming processors in AI computing systems.
US patent US12688075
Abstract
A graph partitioning compiler partitions an AI program or model for execution on multiple TSP modules configured for accelerating deep learning workloads.
What this signalsA new patent often shows where a company is putting its R&D budget.
Tags
Technologies named
patents-companypatentGrantGet every patent signal for Groq and the companies you sell to, in the tools you already use.
Connect Autobound to Claude, Claude Code or Cursor with MCP. Then ask: “What changed at Groq this week?”
The Signal API returns patent signals for any list of companies as JSON, for your CRM, warehouse or app.
Sign up and spend your free credits on the companies you sell to.
This page shows a preview. The full patents-company record in the Signal API and MCP can also have these 8 fields. Some fields are empty for some signals.
Company
linkedin_urlValue in the APIindustriesValue in the APIemployee_count_lowValue in the APIemployee_count_highValue in the APIrevenueValue in the APIdescriptionValue in the APISignal
signal_nameValue in the APIassociationValue in the APIGET /v1/signals/723875a2-89bf-5130-9258-ac7fe7104e9b returns this record as JSON. POST /v1/companies/enrich returns every signal for groq.com.
{
"signal_id": "723875a2-89bf-5130-9258-ac7fe7104e9b",
"signal_type": "patents-company",
"signal_subtype": "patentGrant",
"detected_at": "2026-09-17T21:43:59+00:00",
"company": {
"name": "Groq",
"domain": "groq.com"
},
"data": {
"cpc": [
{
"code": "G06F 9/5077",
"label": "Electric digital data processing"
},
{
"code": "G06F 9/5072",
"label": "Electric digital data processing"
}
],
"tags": [
"Intellectual Property",
"Research & Development",
"Artificial Intelligence",
"Machine Learning",
"Software"
],
"detail": "In July 2026 Groq patented compiler technology for running large deep-learning models across multiple tensor streaming processors in AI computing systems.",
"summary": "Groq was granted a patent for splitting AI models across processors.",
"event_at": "2026-07-21",
"inventors": [
{
"name": "Kyeong Mo Kang",
"status": "unresolved",
"is_primary": true
},
{
"name": "Yuxi Cai",
"status": "unresolved",
"is_primary": false
},
{
"name": "Naif Tarafdar",
"status": "resolved",
"is_primary": false,
"linkedin_url": "https://www.linkedin.com/in/naif-tarafdar"
},
{
"name": "Andrew Chaang Ling",
"status": "unresolved",
"is_primary": false
},
{
"name": "Pao-Sheng Chou",
"status": "unresolved",
"is_primary": false
}
],
"tech_area": "software / computing",
"event_kind": "grant_date",
"grant_date": "2026-07-21",
"source_url": "https://patents.google.com/patent/US12688075",
"cpc_primary": {
"code": "G06F 9/5077",
"label": "Electric digital data processing"
},
"patent_title": "Graph partitioning and implementation of large models on tensor streaming processors",
"patent_number": "US12688075",
"uspto_assignee": "Groq, Inc.",
"patent_abstract": "A graph partitioning compiler partitions an AI program or model for execution on multiple TSP modules configured for accelerating deep learning workloads.",
"precision_class": "structural",
"invention_explanation": "A compiler divides an AI model into parts and assigns those parts to several specialized processors so they can execute the workload together.",
"technologies_mentioned": [
"AI model compilation",
"deep learning",
"tensor processors",
"graph partitioning"
]
}
}
curl https://signals.autobound.ai/v1/signals/723875a2-89bf-5130-9258-ac7fe7104e9b \
-H "X-API-KEY: $AUTOBOUND_API_KEY"curl -X POST https://signals.autobound.ai/v1/companies/enrich \
-H "X-API-KEY: $AUTOBOUND_API_KEY" \
-H "Content-Type: application/json" \
-d '{"domain":"groq.com","limit":20}'Long text fields are shortened on this page.
Looking up one signal by its id is free. Enrich costs 2 credits per signal returned; a call with no results is free.