About This Dataset

The SymbioStreams AI Hardware Hub is a comprehensive catalog of AI-optimized compute hardware. Each record includes structured specs, cloud pricing, performance tier classification, and primary use-case annotation.

Coverage includes:

GPUs: NVIDIA H100/A100/L40S/RTX series, AMD MI300X/MI250, Intel Arc, etc.
TPUs: Google TPU v4/v5, specialized training accelerators
AI ASICs: Cerebras WSE-3, Graphcore IPU, Groq LPU, SambaNova, Habana Gaudi
Edge & Inference: NVIDIA Jetson, Apple M-series, Qualcomm AI, Hailo-8
Networking & Storage: InfiniBand, NVLink, NVMe for AI workloads

Use cases: Infrastructure planning, hardware benchmarking, cloud cost optimization, AI workload sizing, procurement research.

Filter Hardware

Product	Manufacturer	Category	Tier	Key Specs	Use Case
AWS Inferentia 1	AWS	AI Accelerator	Cloud	—	Budget Inference
AWS Inferentia 2	AWS	AI Accelerator	Cloud	—	Low-cost Inference
AWS Trainium 1	AWS	AI Accelerator	Cloud	—	Cost-effective AWS Training
AWS Trainium 2	AWS	AI Accelerator	Cloud	—	AWS-native Training
Cambricon MLU370-X8	Cambricon	AI Accelerator	Data Center	—	China Cloud AI
Cerebras CS-1	Cerebras	AI Accelerator	Supercomputer	—	Scientific AI
Cerebras WSE-2 (CS-2)	Cerebras	AI Accelerator	Supercomputer	—	Large Model Training
Cerebras WSE-3 (CS-3)	Cerebras	AI Accelerator	Supercomputer	—	Massive Model Training
D-Matrix Corsair	D-Matrix	AI Accelerator	Edge	—	Low-latency Inference
Enflame DTU CloudBlazer T20	Enflame	AI Accelerator	Data Center	—	China Cloud Training
Esperanto ET-SoC-1	Esperanto	AI Accelerator	Edge	—	Energy-efficient Inference
Etched Sohu	Etched	AI Accelerator	Data Center	—	Transformer-specific Inference
FuriosaAI RNGD	FuriosaAI	AI Accelerator	Data Center	—	Korean AI Inference
Graphcore Bow IPU (C600)	Graphcore	AI Accelerator	Cloud	—	Sparse Model Training
Graphcore GC200 IPU (Colossus MK2)	Graphcore	AI Accelerator	Cloud	—	Sparse & Dynamic Models
Hailo-10	Hailo	AI Accelerator	Edge	—	Generative AI on Edge
Hailo-8	Hailo	AI Accelerator	Edge	—	Edge AI (Automotive/IoT)
Huawei Ascend 910B	Huawei	AI Accelerator	Data Center	—	China-domestic AI Training
Huawei Ascend 910C	Huawei	AI Accelerator	Data Center	—	China-domestic Large Model Training
IBM AIU (Artificial Intelligence Unit)	IBM	AI Accelerator	Cloud	—	Enterprise AI (IBM Cloud)
Habana Goya (HL-100)	Intel	AI Accelerator	Cloud	—	Inference Acceleration
Intel Gaudi 2	Intel	AI Accelerator	Data Center	—	Cost-effective AI Training
Intel Gaudi 3	Intel	AI Accelerator	Data Center	—	Cost-effective AI Training & Inference
Kunlunxin R200	Kunlun	AI Accelerator	Data Center	—	Baidu AI Training
Lightmatter Envise	Lightmatter	AI Accelerator	Data Center	—	Photonic AI Computing
MatX Arc	MatX	AI Accelerator	Data Center	—	Efficient AI Training
Meta MTIA v2	Meta	AI Accelerator	Custom	—	Meta Internal Inference
Microsoft Maia 100	Microsoft	AI Accelerator	Cloud	—	Azure AI Services
Mythic M1076 AMP	Mythic	AI Accelerator	Edge	—	Ultra-low Power Edge AI
Positron AI ARC-1	Positron	AI Accelerator	Data Center	—	LLM Inference Optimization
Qualcomm Cloud AI 100 Standard	Qualcomm	AI Accelerator	Cloud	—	Efficient Cloud Inference
Qualcomm Cloud AI 100 Ultra	Qualcomm	AI Accelerator	Cloud	—	Power-efficient Inference
Rebellions ATOM	Rebellions	AI Accelerator	Data Center	—	Korean AI Cloud
SambaNova DataScale SN40L	SambaNova	AI Accelerator	Data Center	—	Enterprise AI Full Stack
SambaNova SN30	SambaNova	AI Accelerator	Data Center	—	Enterprise AI
Tenstorrent Blackhole	Tenstorrent	AI Accelerator	Data Center	—	Scalable AI Training
Tenstorrent Wormhole	Tenstorrent	AI Accelerator	Edge	—	Edge & Data Center Inference
Tesla Dojo D1	Tesla	AI Accelerator	Custom	—	Autonomous Driving Training
Untether AI speedAI240	Untether	AI Accelerator	Edge	—	Ultra-efficient Inference
AWS Inferentia	Amazon	ASIC	High	—	Inference
AWS Inferentia (Inf1)	Amazon	ASIC	High	—	Inference
AWS Inferentia2	Amazon	ASIC	Enterprise	—	Inference
AWS Inferentia2 (Inf2)	Amazon	ASIC	Enterprise	—	Inference
AWS Trainium	Amazon	ASIC	Enterprise	—	Training
AWS Trainium (Trn1)	Amazon	ASIC	Enterprise	—	Training
AWS Trainium2	Amazon	ASIC	Enterprise	—	Training
AWS Trainium2 (Trn2)	Amazon	ASIC	Enterprise	—	Training
Baidu Kunlun 2	Baidu	ASIC	High	—	Both
Baidu Kunlun 3	Baidu	ASIC	Enterprise	—	Both
Blaize Pathfinder P1600	Blaize	ASIC	Entry	—	Inference

Page 1 of 7 (338 total)

Dataset Stats

Total SKUs338

GPU134 SKUs

ASIC62 SKUs

AI Accelerator39 SKUs

CPU29 SKUs

NPU24 SKUs

TPU17 SKUs

FPGA10 SKUs

SoC9 SKUs

Cloud8 SKUs

LPU2 SKUs

SourceSymbioStreams

LicenseOpen Data

API Access

Base Endpoint

https://support-local-businesses.com/api/v1/datasets/hardware-hub

# All hardware (paginated)
curl "https://support-local-businesses.com/api/v1/datasets/hardware-hub"

# Filter by category
curl "https://support-local-businesses.com/api/v1/datasets/hardware-hub?category=GPU"

# Filter by manufacturer + tier
curl "https://support-local-businesses.com/api/v1/datasets/hardware-hub?manufacturer=NVIDIA&performance_tier=High"

# Paginate
curl "https://support-local-businesses.com/api/v1/datasets/hardware-hub?page=2&limit=50"

CORS-enabled · Rate limit: 100 req/hr/IP · Free access