AMD CDNA 3
E1034272
AMD CDNA 3 is a high-performance GPU compute architecture from AMD designed to power data center accelerators and workloads such as AI, HPC, and large-scale parallel processing.
All labels observed (1)
| Label | Occurrences |
|---|---|
| AMD CDNA 3 canonical | 4 |
How this entity was disambiguated
This entity first appeared as the object of triple T13320094 — resolving that mention is where its identity was fixed. The disambiguator weighed these candidate entities and picked the highlighted one (or “None”, minting a new entity). This is how homonymy is resolved: the same surface form can point to different entities.
NED1
Entity disambiguation (via context triple)
gpt-5-mini-2025-08-07
Target entity: AMD CDNA 3 Context triple: [AMD Instinct, usesArchitecture, AMD CDNA 3]
-
A.
AMD CDNA 2
AMD CDNA 2 is a data center–focused GPU architecture from AMD designed to accelerate high-performance computing and AI workloads in its Instinct accelerator lineup.
-
B.
AMD CDNA
AMD CDNA is a GPU architecture from AMD optimized for high-performance computing and data center workloads, particularly in AI and scientific applications.
-
C.
AMD RDNA
AMD RDNA is a modern GPU architecture from AMD designed to deliver high-performance, power-efficient graphics for gaming and compute workloads across desktops, laptops, and consoles.
-
D.
Zen 3
Zen 3 is AMD’s high-performance CPU microarchitecture that significantly improved instructions-per-clock, gaming performance, and power efficiency over its predecessors in Ryzen and EPYC processors.
-
E.
AMD Radeon RX 500 series
The AMD Radeon RX 500 series is a family of mid-range to high-end graphics cards aimed at gamers and PC enthusiasts, offering improved performance and efficiency over its RX 400-series predecessors.
- F. None of above. chosen
- G. Unsure - the case is ambiguous/there is not enough information to decide.
NED2
Entity disambiguation (via description)
gpt-5-mini-2025-08-07
Target entity: AMD CDNA 3 Target entity description: AMD CDNA 3 is a high-performance GPU compute architecture from AMD designed to power data center accelerators and workloads such as AI, HPC, and large-scale parallel processing.
-
A.
AMD CDNA 2
AMD CDNA 2 is a data center–focused GPU architecture from AMD designed to accelerate high-performance computing and AI workloads in its Instinct accelerator lineup.
-
B.
AMD CDNA
AMD CDNA is a GPU architecture from AMD optimized for high-performance computing and data center workloads, particularly in AI and scientific applications.
-
C.
AMD RDNA
AMD RDNA is a modern GPU architecture from AMD designed to deliver high-performance, power-efficient graphics for gaming and compute workloads across desktops, laptops, and consoles.
-
D.
Zen 3
Zen 3 is AMD’s high-performance CPU microarchitecture that significantly improved instructions-per-clock, gaming performance, and power efficiency over its predecessors in Ryzen and EPYC processors.
-
E.
AMD Radeon RX 500 series
The AMD Radeon RX 500 series is a family of mid-range to high-end graphics cards aimed at gamers and PC enthusiasts, offering improved performance and efficiency over its RX 400-series predecessors.
- F. None of above. chosen
Statements (47)
| Predicate | Object |
|---|---|
| instanceOf |
GPU compute architecture
ⓘ
parallel processing architecture ⓘ |
| announcedBy |
AMD
ⓘ
linked to:
Advanced Micro Devices
|
| announcedForProductLine |
AMD Instinct accelerators
ⓘ
linked to:
AMD data center GPU accelerators
|
| architectureFamily | AMD CDNA ⓘ |
| category |
AI accelerator architecture
ⓘ
HPC accelerator architecture ⓘ data center GPU architecture ⓘ |
| competesWith |
NVIDIA Hopper architecture
ⓘ
NVIDIA data center GPU architectures ⓘ |
| designedFor |
AI inference
ⓘ
AI training ⓘ HPC workloads ⓘ artificial intelligence workloads ⓘ data analytics workloads ⓘ data center accelerators ⓘ high performance computing ⓘ large-scale parallel processing ⓘ |
| developer |
AMD
ⓘ
linked to:
Advanced Micro Devices
Advanced Micro Devices ⓘ |
| feature |
advanced compute units for AI and HPC
ⓘ
enhanced memory bandwidth ⓘ hardware acceleration for AI workloads ⓘ improved interconnect fabric over AMD CDNA 2 ⓘ |
| focus | compute performance over graphics rendering ⓘ |
| follows | AMD CDNA 2 ⓘ |
| introducedAround | 2023 ⓘ |
| notDesignedFor | consumer gaming graphics ⓘ |
| optimizedFor |
energy-efficient performance
ⓘ
exascale computing ⓘ large language model training ⓘ |
| precedes | AMD CDNA 4 ⓘ |
| supports |
HBM3
ⓘ
high bandwidth memory ⓘ high-speed interconnects ⓘ matrix operations acceleration ⓘ mixed-precision compute ⓘ multi-GPU scaling ⓘ |
| supportsEcosystem |
ROCm software stack
ⓘ
linked to:
AMD ROCm software stack
open-source AI and HPC frameworks ⓘ |
| targetMarket |
cloud service providers
ⓘ
data centers ⓘ supercomputing systems ⓘ |
| usedIn |
AMD Instinct MI300 series accelerators
ⓘ
linked to:
AMD Instinct MI300 series
AMD Instinct MI300A ⓘ
linked to:
AMD Instinct MI300 series
AMD Instinct MI300X ⓘ
linked to:
AMD Instinct MI300 series
|
| vendor |
AMD
ⓘ
linked to:
Advanced Micro Devices
|
How these facts were elicited
The pipeline generated the facts above by prompting gpt-5.1 with this entity's name + description and the instruction below.
Instruction
You are a knowledge base construction expert. Given a subject entity and a description of it, return factual statements that you know for the subject as a JSON list of dictionaries(triples), where keys must be "subject", "predicate" and "object". The number of facts may be very high, between 25 to 50 or more, for very popular subjects. For less popular subjects, the number of facts can be very low, like 5 or 10. # Requirements - If you don't know the subject at all, return an empty list. - If the subject is not a named entity, return an empty list. - Include at least one triple where predicate is "instanceOf". - Do not get too wordy. - Separate several objects into multiple triples with one object.
Input
Subject: AMD CDNA 3 Description of subject: AMD CDNA 3 is a high-performance GPU compute architecture from AMD designed to power data center accelerators and workloads such as AI, HPC, and large-scale parallel processing.
Referenced by (4)
Full triples — surface form annotated when it differs from this entity's canonical label.