{"company":{"slug":"nvidia","name":"NVIDIA","legal_name":null,"aliases":["NVDA"],"primary_layer":"L4","primary_topic_id":"optical-modules","secondary_topics":["spacex-ipo-ai-infra"],"hq_country":"US","hq_city":null,"founded_year":1993,"is_public":true,"ticker":"NVDA","description":"The dominant AI accelerator vendor — GPUs (H100/B200/GB200), NVLink, and the CUDA software stack that anchors AI training and inference."},"profile":{"snapshot":"GPU supplier behind xAI Colossus; CUDA moat faces custom-silicon and memory-cap pushback","stack_layer":"L2","why_join_top_2":["You'd work on the sole GPU platform underpinning xAI/SpaceX's Colossus Hopper-class cluster (signal 40764050, 2026-05-06) — the largest named frontier buildout in this dataset, not a pilot deployment.","The CUDA software moat is why peers like Marvell and Astera can only compete at interconnect, never compute (crowd signal: 'AMD still don't want to invest into software. Nvidia success is all thankful to their CUDA'); you'd own that platform layer."],"peer_comparison":{"substitute":{"name":"Marvell Technology, Inc.","delta":"Marvell enables the hyperscaler custom-ASIC path (the crowd-cited Microsoft/Meta/Google/Amazon silicon) that substitutes for merchant NVIDIA GPUs in captive inference."},"direct_competitor":{"name":"Astera Labs, Inc.","delta":"Astera competes only at the connectivity/retimer fabric layer around the GPU; NVIDIA owns the compute die and CUDA stack Astera plugs into, not a like-for-like rival."}},"self_quality_score":{"notes":"Only one hard business signal and zero talent signals; hiring specialties left empty rather than fabricated, risk/bet sharpened via CBR crowd signals.","originality":7,"evidence_density":5,"anti_generic_pass":true,"reverse_hype_present":true},"strategic_position":{"top_risk":"Hyperscaler in-house XPUs (crowd-named: Microsoft, Meta, Google, Amazon, plus Broadcom/Marvell-built ASICs) visibly displace merchant-GPU share in inference workloads by H2 2026, eroding Blackwell pull-through outside training.","their_bet":"Blackwell-class GPUs become the default substrate not just for hyperscale training (Colossus, signal 40764050) but for sovereign/inference clouds (crowd signal: UK provider relaxAI running latest open models on NVIDIA Blackwell).","what_they_do_best":"Owning the full GPU + CUDA compute substrate for frontier clusters — evidenced by being the named GPU supplier underpinning the entire Colossus Hopper-class buildout (signal 40764050)."},"top_business_signals_90d":[{"summary":"NVIDIA is the sole GPU supplier underpinning the SpaceX/xAI Colossus Hopper-class compute buildout, making it a direct L2 beneficiary.","signal_id":"40764050-7ec6-4eba-9910-95b071259886","evidence_quote":"NVIDIA is the GPU supplier underpinning Colossus (Hopper-class), making NVIDIA a direct L2 beneficiary of the SpaceX/xAI compute buildout."}],"top_hiring_specialties_60d":[]},"recent_business_signals":[{"signal_type":"product_launch","signal_date":"2026-06-30","signal_data":{"kind":"model","tags":["Model Optimizer","safetensors","mistral3","nvidia","ModelOpt","Mistral","Mistral-Medium-3.5","quantized","NVFP4","nvfp4"],"layer":"L7","likes":10,"title":"HuggingFace model: nvidia/Mistral-Medium-3.5-128B-NVFP4","source":"huggingface-hub","downloads":0,"pipeline_tag":"text-generation"},"source":"huggingface-model"},{"signal_type":"product_launch","signal_date":"2026-06-30","signal_data":{"kind":"dataset","tags":["license:other","size_categories:1K