{"resourceId":"gatech-nexus-hardware-dgx-b200-rtx-pro-6000-inventory","versions":[{"version":"external-f5401e4d07b5ecc1581545a22f6e5663985a31b06ca4eac3d3753ac7e5ebfed5","resource":{"id":"gatech-nexus-hardware-dgx-b200-rtx-pro-6000-inventory","title":"Georgia Tech Nexus docs: 480 NVIDIA GPUs installed (16 DGX B200 nodes / 128 B200 GPUs; 38+6 RTX PRO 6000 Blackwell nodes) with NDR InfiniBand and VAST storage","organization":"Georgia Institute of Technology (Nexus)","sector":"Higher education research computing","geography":"United States — Georgia Tech and NCSA persistent-service nodes","publishedAt":"undated (operator documentation; retrieved 2026-10-08)","publicationDate":null,"eventDate":"2026-10-08","sourceName":"Nexus documentation (docs.nexus.gatech.edu)","sourceLabel":"Operator hardware documentation","sourceUrl":"https://docs.nexus.gatech.edu/en/latest/overview/hardware.html","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["infrastructure","data-security","operating-model"],"finding":"Nexus operator docs describe an installed system with NVIDIA DGX B200 nodes for tightly coupled AI, NVIDIA RTX PRO 6000 Blackwell Server Edition for flexible GPU computing, and dedicated RTX nodes for long-running services. Inventory totals 72 nodes, 480 GPUs, 8,576 CPU cores and 118 TB system memory: 16 DGX B200 nodes (128 GPUs, 1.44 TB HBM aggregate per node class), 38 scheduled RTX PRO 6000 nodes (304 GPUs) and 6 persistent-service RTX PRO 6000 nodes (48 GPUs) at Georgia Tech and NCSA. Networking includes NDR InfiniBand (QM9700), BlueField-3 and ConnectX-7; storage is VAST (~10.8 PB raw NVMe). Docs warn published interface rates are not guaranteed application throughput and that Georgia Tech and NCSA are distinct sites for data and identity planning. Why now: pairs with the Oct 8 Georgia Tech/Emory investment announcement of Nexus launching in October.","sledRelevance":"Gives SLED architects concrete NVIDIA platform composition for a NSF-linked university AI factory—useful for peer planning and for Georgia agencies assessing what 'Nexus access' actually means technically.","evidence":"Opened and read the full Hardware Details page from Nexus docs. Operator-stated inventory and architecture; undated page; no utilization, wait-time or independent audit metrics.","architectureImplications":"Two GPU classes (DGX B200 NVSwitch vs RTX PRO 6000) plus separate persistent-service tenancy imply different partitions for training vs inference/services; NCSA nodes sit outside GT InfiniBand fabrics.","governanceImplications":"Site boundary warnings mean multi-site workflows need explicit data-location and federation decisions before projects span GT and NCSA.","securityPrivacyImplications":"Persistent-service nodes for model-serving endpoints and gateways need separate exposure and auth controls from Slurm tenancy; docs describe placement dependence on security and data requirements.","caveats":"Undated operator documentation; inventory may change. Portal marketing figures (464 GPUs / 8,320 cores) differ slightly from this docs total (480 GPUs / 8,576 cores including management nodes)—use docs for engineering. No performance benchmarks.","streamIds":["nvidia"],"roles":{"sales":"Nexus hardware docs turn Georgia Tech's AI investment story into a specific Blackwell footprint—128 B200 GPUs on DGX plus hundreds of RTX PRO 6000 GPUs—that peer universities and Georgia agencies can discuss concretely. Use it to scope whether a partner needs tightly coupled training partitions or service nodes. Offer an architecture-read briefing. Do not promise performance from link rates, and do not blur GT and NCSA tenancies.","engineering":"Fit for AI training and inference planning against a known DGX B200 + RTX PRO 6000 mix with NDR fabrics and VAST. Architecture integration means matching workloads to Slurm vs persistent-service paths and respecting site boundaries. Prerequisites: allocation account and data-location plan. Proof of value: place one training and one serving workload on the appropriate class and record queue and throughput.","delivery":"Delivery for users is partition selection, storage mounts and multi-site data discipline across Georgia Tech and NCSA. Nexus operations owns the platform; the PI owns workload placement choices and dataset location. Checkpoints are data-location confirmation and a partition-fit review before large jobs. Proposed acceptance: jobs land on the intended GPU class without cross-site data surprises. Risk: designing capacity plans from portal marketing totals instead of the operator hardware docs."},"roleDetails":{"sales":{"customerProblem":"Partners hearing 'Nexus' need a clear GPU and tenancy picture before committing staff or matching funds.","stakeholders":["Research computing architects","Agency AI leads","Partner university CIOs"],"discoveryQuestions":["Do you need DGX-class multi-GPU training or RTX inference/services?","Will data stay at Georgia Tech or also NCSA?","What allocation route will you use?"],"valueHypothesis":"A concrete inventory briefing can prevent mismatched expectations about capacity and access (interpretation).","potentialEngagement":"A one-hour architecture-read using the hardware docs for a single partner use case.","unsupportedClaims":["Guaranteed application throughput from fabric rates","That all 480 GPUs are available to every user","Identical networking at NCSA and GT"]},"engineering":{"fit":"Strong reference architecture for Blackwell university AI factories; verify live quotas before design freeze.","architectureIntegration":"Slurm on DGX B200 and scheduled RTX PRO 6000; managed persistent-service RTX nodes; VAST over dedicated storage fabric; NCSA nodes on separate networking.","prerequisites":["Allocation and identity","Workload class (training vs serving)","Data-location decision GT vs NCSA"],"constraints":["Undated docs may lag install changes","Portal vs docs count differences","Published link rates ≠ app throughput"],"security":"Persistent service endpoints need authz and network exposure controls separate from batch tenancy; confirm before public models.","proofOfValue":["Submit a DGX multi-GPU smoke test","Submit an RTX inference job","Verify storage mount and site for the dataset"]},"delivery":{"work":"Onboard users to the correct partition and enforce data-location checks for multi-site work.","dependencies":["Nexus operations runbooks","VAST mount instructions","Site federation status"],"ownership":"Nexus ops owns platform; PI owns placement; partner security owns data-class.","skillsAdoption":"Users need training on GPU-class selection and site boundaries.","governanceCheckpoints":["Data-location confirmation","Partition fit review","Post-job cost/usage review"],"acceptanceCriteria":["Jobs run on intended GPU class (proposed)","No unintended NCSA/GT data movement (proposed)"],"risks":["Designing to marketing GPU counts","Exposing persistent services broadly","Ignoring oversubscription on RTX storage fabric"]}},"retrievedAt":"2026-10-09T03:21:30Z","enrichedAt":"2026-10-09T03:25:09Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Docs are technical operator material; student/educator access claims live on the portal and GT news pages, not here.","procurementImplications":"Peer institutions can use this inventory as a planning reference but must size to their own awards and power/cooling—not copy totals blindly.","operatingModelImplications":"Separates batch AI, persistent services and multi-site operations into distinct operating lanes that SLED partners must staff differently.","sourceVerification":{"openedUrl":"https://docs.nexus.gatech.edu/en/latest/overview/hardware.html","referenceExcerpt":"The installed system includes NVIDIA DGX B200 nodes for tightly coupled AI workloads, NVIDIA RTX PRO 6000 Blackwell Server Edition nodes","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}