{"company":{"slug":"amd","name":"AMD","legal_name":"Advanced Micro Devices, Inc.","aliases":["AMD"],"primary_layer":"L2","primary_topic_id":"ai-asics","secondary_topics":null,"hq_country":"US","hq_city":"Santa Clara","founded_year":1969,"is_public":true,"ticker":"AMD","description":"MI300 / MI325 / MI350 Instinct accelerator family — the dominant merchant alternative to NVIDIA in AI training silicon. ROCm software stack maturity is the primary watch-point."},"profile":{"snapshot":"Merchant #2 AI training silicon; OpenAI 6GW + Oracle 50k MI450 wins, ROCm still the watch-point","stack_layer":"L2","why_join_top_2":["Q3 FY2025 hit record ~$9.2B revenue (+36% YoY), data center ~$4.3B (signal 4e06d83d) — this is a funded revenue ramp, not a pre-revenue silicon bet.","OpenAI's 6 GW commitment (signal 81014194, Oct 2025) plus Oracle's 50,000 MI450 order (signal 63f198be) give MI450 work guaranteed 2026 deployment, matching the active chiplet/packaging and physical-design hiring."],"peer_comparison":{"substitute":{"name":"Groq","delta":"Groq is inference-only deterministic LPU silicon; AMD Instinct targets training + inference with a broader ROCm software surface Groq does not address."},"direct_competitor":{"name":"Cerebras Systems","delta":"Cerebras sells wafer-scale single-chip systems into niche HPC/pharma; AMD competes as volume merchant GPU at hyperscaler scale (Oracle 50k MI450, signal 63f198be)."}},"self_quality_score":{"notes":"Risk is falsifiable against named 2H2026 OpenAI/Oracle ramp dates; flags ROCm and the ~$800M MI308 China charge as real drag.","originality":7,"evidence_density":9,"anti_generic_pass":true,"reverse_hype_present":true},"strategic_position":{"top_risk":"The 1 GW OpenAI / 50k-Oracle MI450 ramps (signals 81014194, 63f198be) slip past their stated Q3–2H 2026 starts, or ROCm gaps stall conversion — both checkable against deployment reporting by end of 2026.","their_bet":"MI450/'Helios' rack-scale systems backed by the OpenAI 6 GW warrant deal (signal 81014194) bet that AMD wins at rack level, not just chip level, against a projected ~$1T 2030 AI TAM (signal 65d05619).","what_they_do_best":"Scaling a credible merchant GPU alternative at hyperscale — data-center revenue reached ~$4.3B in Q3 FY2025 (signal 4e06d83d) on MI350's ~4x generational compute uplift (signal 00c21fd4)."},"top_business_signals_90d":[{"summary":"OpenAI committed to 6 GW of Instinct deployments starting with 1 GW MI450 in 2H 2026, plus a 160M-share warrant tying OpenAI's upside to AMD deployment milestones.","signal_id":"81014194-6f8e-457e-8ffb-31e3dedc2bfd","evidence_quote":"OpenAI to deploy 6 gigawatts of AMD Instinct GPUs, beginning with 1 GW of MI450 in the second half of 2026"},{"summary":"Oracle Cloud Infrastructure committed to 50,000 MI450 GPUs from Q3 2026, a second anchor hyperscaler validating the MI450 generation.","signal_id":"63f198be-de53-4a46-bd77-331ece89c05c","evidence_quote":"Oracle announced it will deploy 50,000 AMD Instinct MI450 GPUs on Oracle Cloud Infrastructure beginning in Q3 2026"},{"summary":"Record Q3 FY2025 revenue of ~$9.2B (+36% YoY) with data-center at ~$4.3B, showing Instinct demand converting to revenue, not just bookings.","signal_id":"4e06d83d-3a85-4d57-a954-677e4c0c1185","evidence_quote":"record Q3 FY2025 revenue of approximately $9.2 billion, up about 36% year over year, driven by strong data center sales of roughly $4.3 billion"}],"top_hiring_specialties_60d":[{"job_count":1,"specialty":"Chiplet & advanced packaging / die-to-die interconnect architecture"},{"job_count":1,"specialty":"Datacenter GPU physical design (place-and-route, timing closure)"},{"job_count":1,"specialty":"Silicon power & performance architecture for AI accelerators"}]},"avg_fit_score":78.30228758169932,"fit_scores":[{"overall_score":75,"growth_score":88,"comp_score":null,"learning_score":85,"stability_score":64,"culture_score":58,"jobseeker_profile_id":"03038a09-5290-4bb8-9295-a14cef6419b1"},{"overall_score":81,"growth_score":90,"comp_score":null,"learning_score":80,"stability_score":68,"culture_score":60,"jobseeker_profile_id":"f68552cf-b42f-457b-b103-621f6e37ff6d"},{"overall_score":81.4444444444444,"growth_score":88,"comp_score":null,"learning_score":82,"stability_score":70,"culture_score":60,"jobseeker_profile_id":"96f78a9c-5631-4f0a-bdd5-94ef2ef9cd69"},{"overall_score":75.7647058823529,"growth_score":88,"comp_score":null,"learning_score":80,"stability_score":68,"culture_score":60,"jobseeker_profile_id":"0cc4a7bc-42f9-4bee-bcb6-932be8a68adc"}],"recent_business_signals":[{"signal_type":"other","signal_date":"2025-11-11","signal_data":{"event":"Financial Analyst Day 2025","new_product":"Instinct MI430X (HBM4) introduced for AI/HPC; MI450 and 'Helios' rack roadmap detailed","data_center_ai_tam_2030_usd":1000000000000,"long_term_revenue_growth_target_pct":"35+ annually over 3-5 years"},"source":"ai_synthesis_sdk"},{"signal_type":"earnings","signal_date":"2025-11-04","signal_data":{"period":"Q3 FY2025","record":true,"revenue_usd":9250000000,"non_gaap_eps_usd":1.2,"yoy_revenue_growth_pct":36,"data_center_segment_usd":4300000000},"source":"ai_synthesis_sdk"},{"signal_type":"customer_win","signal_date":"2025-10-14","signal_data":{"units":50000,"product":"AMD Instinct MI450 GPUs","customer_name":"Oracle (Oracle Cloud Infrastructure)","deployment_start":"Q3 2026"},"source":"ai_synthesis_sdk"},{"signal_type":"partnership","signal_date":"2025-10-06","signal_data":{"deal_scope":"6 gigawatts of AMD Instinct GPU deployments","partner_name":"OpenAI","initial_phase":"1 GW of Instinct MI450 in 2H 2026","equity_component":"Warrant for OpenAI to acquire up to 160 million AMD shares vesting on deployment/share-price milestones"},"source":"ai_synthesis_sdk"},{"signal_type":"earnings","signal_date":"2025-08-05","signal_data":{"note":"Results included roughly $800M inventory/charge impact tied to MI308 China export restrictions","period":"Q2 FY2025","revenue_usd":7685000000,"non_gaap_eps_usd":0.48,"yoy_revenue_growth_pct":32,"data_center_segment_usd":3200000000},"source":"ai_synthesis_sdk"},{"signal_type":"product_launch","signal_date":"2025-06-12","signal_data":{"event":"Advancing AI 2025","key_specs":"288GB HBM3E memory, up to ~4x generational AI compute uplift","architecture":"CDNA 4","product_name":"AMD Instinct MI350 Series (MI350X / MI355X)","roadmap_preview":"MI400 series and 'Helios' rack-scale system previewed for 2026"},"source":"ai_synthesis_sdk"}],"recent_talent_signals":[{"job_title":"Principal Engineer, Chiplet and Advanced Packaging Architecture","job_level":"principal","job_specialty":["chiplet architecture","advanced packaging","die-to-die interconnect"],"posted_date":"2026-05-06"},{"job_title":"Physical Design Engineer, Datacenter GPU/Accelerator","job_level":"mid","job_specialty":["physical design","place and route","timing closure"],"posted_date":"2026-04-22"},{"job_title":"Senior Silicon Power and Performance Architect, AI Accelerators","job_level":"senior","job_specialty":["power architecture","performance modeling","AI accelerators"],"posted_date":"2026-04-02"},{"job_title":"Senior ASIC Design Engineer, AI Accelerators","job_level":"senior","job_specialty":["RTL design","ASIC design","AI accelerators"],"posted_date":"2026-03-09"},{"job_title":"Staff Design Verification Engineer, AI Compute IP","job_level":"staff","job_specialty":["design verification","UVM","AI compute IP"],"posted_date":"2026-02-25"}],"citation_url":"/companies/amd","last_updated":"2026-05-14T14:33:37.941141+00:00","generated_at":"2026-05-19T11:59:22.230Z"}