AI Infrastructure Components and Architecture

DCAI · 90 questions

• A city data center must support nonblocking east-west bandwidth for all-reduce traffic during distributed training. What network-deployment evaluation priority should architects emphasize?
• A public-health GPU hall needs microsecond-class latency between GPU servers for RoCE-based training. Which network evaluation criterion is most critical?
• A 911 records lake fabric must survive a leaf failure without stranding half the GPUs. What network design property should evaluation prioritize?
• A transit AI pod expects GPU count to double next fiscal year without redesigning addressing. Which network attribute should architects evaluate first?
• A port authority DC must isolate PHI inference traffic from public kiosk networks. What network evaluation focus applies?
• Airport ops sees incast toward parameter servers during gradient sync on the AI fabric. What should network evaluation address?
• County courts IT compares classic campus oversubscription ratios with AI spine nonblocking goals for the GPU annex. Why do AI fabrics reject heavy oversubscription?
• A housing authority colo cage wants dual spines so maintenance does not halt overnight training. Which evaluation outcome supports that design?
• A utilities control DC requires encrypted interconnect to a warm DR GPU site. What network evaluation concern is primary?
• A library consortium DC plans leaf upgrades in place while inference SLAs continue. Which network attribute enables that operational growth?
• A school district DC measures whether 100G or 400G leaf uplinks meet a training job’s bisection bandwidth needs. What is the correct evaluation approach?
• A wastewater plant’s AI uplink cannot tolerate lossy drops that stall RDMA training iterations. What network property should evaluation require?
• A police evidence AI zone needs predictable latency for synchronous distributed training. Which criterion should network evaluation emphasize?
• A fire CAD analytics design segments a research sandbox from production 911 inference. What network evaluation principle does that illustrate?
• Parks IoT AI edge backhaul adds a second pod and must extend the AI fabric cleanly. What should architects evaluate?
• A municipal open-data warehouse runs overnight model training on the AI fabric while a concurrent backup window replicates datasets to the DR site. What network evaluation should the architects emphasize when sizing the spines?
• An election systems AI sandbox must keep north-south inference clients responsive while east-west distributed training saturates the fabric. How should border leaf placement and design be evaluated?
• A tax assessor imagery farm documents recovery objectives for a multi-day training job if an AI fabric spine pair fails mid-run. What redundancy evaluation is most appropriate?
• A building-permits computer-vision farm must harden GPU nodes so only approved inference frontends can reach them. Which network security evaluation best fits that goal?
• A fleet telematics AI zone plans to enable an RDMA fabric for GPU training. What network feature readiness should be re-evaluated before cutover?
• A city data center is specifying a node that will run RAG embedding pipelines alongside LLM serving. What compute evaluation should guide the CPU-to-GPU ratio?
• A public-health GPU hall is selecting how many GPUs to place per UCS server for a training job that prefers high-bandwidth NVLink islands. What compute connectivity judgment matters most?
• A 911 records lake fabric stages training batches on GPU servers. What host memory evaluation prevents GPUs from going idle?
• A transit AI pod must host multi-tenant inference for several municipal agencies. Which virtualization support evaluation best fits performance-sensitive GPU inference?
• A port authority DC plans to grow an AI training pool from 8 to 64 GPUs without rewriting job scripts. What compute scalability approach should be evaluated?
• An airport ops DC requires clinic-adjacent inference to survive loss of one GPU chassis. What compute redundancy evaluation is appropriate?
• A county courts annex maps overnight training and daytime inference to different UCS server SKUs. What workload-type to compute mapping should guide that choice?
• A housing authority colo cage evaluates offload options on GPU servers for storage and RDMA paths. Which compute connectivity choice should be assessed?
• A utilities control DC observes GPUs idle despite queued training work. Telemetry shows the host PCIe root complex and NIC are saturated. What compute connectivity conclusion is most accurate?
• A library consortium DC will share GPUs across municipal departments. What virtualization/sharing control should be evaluated?
• A school district DC wants to add another GPU tray to an existing UCS AI server. What compute deployment constraint should be validated first?
• A wastewater plant uplink AI zone must run many small inference models on shared GPUs. Which GPU sharing mode evaluation is most appropriate?
• A police evidence AI zone must keep inference available during 911 peak events. What compute redundancy design should be evaluated?
• A fire CAD analytics VLAN team must grow training capacity. How should they evaluate scale-out versus scale-up GPU strategies?
• A parks IoT AI edge backhaul team selected a UCS platform for GPU servers. What compute platform fitness check is required?
• A municipal open-data warehouse separates CPU-only feature engineering from GPU training. What workload-type placement evaluation does that reflect?
• An election systems AI sandbox must prove servers are ready for GPU AI performance. Which compute readiness step should be evaluated?
• A tax assessor imagery farm wants to double training batch size on existing GPUs. What GPU memory evaluation is required?
• A building-permits CV farm is laying out racks so multi-GPU servers keep NVLink domains valid. What interconnect topology evaluation matters?
• A fleet telematics AI zone needs virtualization that still delivers nearly bare-metal GPU performance for inference. Which trade-off should guide the choice?
• A city data center must keep petabyte-class capacity for years of body-worn camera and CAD archives that will feed future municipal models. Which storage design priority should lead the evaluation?
• A public-health GPU hall checkpoints training every few minutes and needs high IOPS NVMe so writes do not stall the GPUs. Which storage evaluation focus is correct?
• A 911 records lake fabric must keep dual controllers so a disk shelf failure does not kill an active training job. Which storage design attribute is the city prioritizing?
• A transit AI pod expects dataset growth of about 40% per year and must avoid forklift storage upgrades. Which design approach best meets that need?
• A port authority data center separates hot training scratch on block/NVMe from warm shared corpora on file storage. What storage design principle does this apply?
• An airport operations data center wants shared high-throughput model checkpoints across many trainers and is evaluating Fibre Channel SAN. When is that choice appropriate?
• A county courts annex wants GPU servers fed from shared fast storage without creating local-only NVMe islands. Which approach should the storage evaluation favor?
• A housing authority colo cage must size parallel file throughput so many trainers can read the same corpus concurrently. Which storage metric should drive that sizing?
• A utilities control data center requires storage snapshots before each major training run so operators can roll back a bad dataset change. Which capability is essential?
• A library consortium data center cannot allow the inference feature store and the training lake to contend on one slow shared array. What should the design emphasize?
• A school district data center must expand storage with new shelves while training jobs continue. Which scalability attribute is required?
• A wastewater plant uplink places replicas of the 911 feature lake across separate failure domains. What storage design goal does this satisfy?
• A police evidence AI zone needs databases of labels on block storage and unstructured PDFs for RAG on file storage. Which design choice is correct?
• A fire CAD analytics VLAN shows GPUs idle while waiting on storage. What should the storage evaluation conclude?
• A parks IoT AI edge backhaul needs WORM or compliance tiers for evidence that will appear in model audits. Which storage outcome should the design deliver?
• A municipal open-data warehouse bursts temporary scratch to faster media during hyperparameter sweeps. Which storage practice is this?
• An election systems AI sandbox must validate multipathing so a single HBA failure is a non-event for trainers. What is the design objective?
• A tax assessor imagery farm must project when the records lake will need a second namespace or scale-out node. Which planning activity is required?
• A city data center hits facility power limits before GPU rack space is full. What should the AI facility evaluation prioritize?
• A public-health GPU hall must improve cooling to support multi-kilowatt GPU nodes. Which facility design focus is correct?
• A 911 records lake fabric team tracks PUE after adding an AI hall to the city data center. What does that metric help them evaluate?
• A transit AI pod evaluates renewable energy credits for the public-health GPU cluster's carbon report. Which sustainability action does this represent?
• A port authority data center schedules heavy training at night when municipal power tariffs and cooling are cheaper. What efficiency practice is this?
• An airport operations data center must redesign CRAC and airflow because rear-door heat from GPUs overwhelms older rows. Which action addresses the root issue?
• A county courts annex compares liquid-assist options because air cooling cannot meet GPU thermal design power. Which evaluation is appropriate?
• A housing authority colo cage sets a sustainability target that caps idle GPU waste. Which design choice best improves power efficiency for the municipal AI pod?
• A utilities control DC isolates hot AI aisles and then reports whether PUE improved. What does a lower PUE after the change correctly indicate?
• A library consortium DC partners with the city energy office so the AI pod can claim renewable supply. Which evaluation best supports that sustainability design?
• A school district DC finds UPS runtime too short for graceful training checkpoint during power events. What should the AI facility evaluation prioritize?
• A wastewater plant uplink calculates stranded power because PDUs are unbalanced across GPU racks. Which design correction best improves power distribution efficiency?
• A police evidence AI zone must justify GPU expansion against the municipal climate action plan. Which evaluation best meets that sustainability governance need?
• A fire CAD analytics VLAN measures facility overhead watts versus IT load after GPU install. Which statement correctly describes a PUE-relevant measurement?
• A parks IoT AI edge backhaul wants to avoid oversizing chillers by right-sizing simultaneous training concurrency. What efficiency evaluation is most appropriate?
• A municipal open-data warehouse documents heat-reuse ideas from the GPU hall into adjacent civic offices. Beyond raw PUE, what sustainability evaluation does that represent?
• A city data center needs encrypted, policy-controlled connectivity from the civic DC to a cloud GPU region for hybrid AI. Which evaluation best addresses that requirement?
• A public-health GPU hall must sync de-identified training features nightly without exposing raw 911 audio. Which hybrid data-synchronization approach is appropriate?
• A 911 records lake fabric bursts overflow training to cloud when the on-prem pod is full, then brings trained weights home. What hybrid capability is primarily being evaluated?
• A transit AI pod keeps sealed-court corpora on-prem while using cloud for public chatbot fine-tunes. Which hybrid placement judgment is correct?
• A port authority DC requires consistent identity and key management across on-prem inference and cloud experiments. Which hybrid evaluation priority is most appropriate?
• An airport ops DC synchronizes model artifacts bidirectionally with the cloud and must detect corruption in transit. Which data-synchronization control should be evaluated?
• A county courts annex moves a completed cloud experiment back to on-prem inference without rewriting serving graphs. What hybrid pattern is being validated?
• A housing authority colo cage prefers not to run hybrid AI pipelines over the open Internet. Which secure connectivity choice best fits that evaluation?
• A utilities control DC must limit which datasets may leave the municipal boundary during hybrid training. Which hybrid evaluation is most important?
• A library consortium DC tests failing over non-sensitive inference to cloud during civic DC maintenance. What hybrid capability is primarily under evaluation?
• A school district DC detects drift between on-prem and cloud feature stores after a failed sync window. Which hybrid data-synchronization response is appropriate?
• A wastewater plant uplink places a cloud burst under the same change-control as on-prem GPU jobs. Why does that operational discipline matter for hybrid AI?
• A police evidence AI zone evaluates hybrid RAG where vectors stay on-prem and LLM weights sit in cloud. What trade-off should the architecture evaluation emphasize?
• A fire CAD analytics VLAN requires strong mutual authentication between the municipal orchestrator and cloud training API endpoints. Which secure-connectivity control fits that hybrid need?
• A parks IoT AI edge backhaul design proposes copying PHI into cloud object storage for hybrid training convenience. What is the correct hybrid security judgment?
• A municipal open-data warehouse documents RPO for model-registry sync between the civic site and a cloud region. What does that hybrid data-synchronization objective primarily define?