LIBRISTO
LIBROAMANTO
obvezno
Pridružite se zajednici ljubitelja knjige iz cijelog svijeta i ostvarite mnoštvo pogodnosti. Izradite besplatni račun
0
Besplatna dostava Overseas kurirskom službom iznad 69.99 €
DPD kurir 4.49 € DPD točka 3.99 € GLS Kurir 5.49 € GLS paketomat 4.49 € Hrvatska pošta 5.49 € Dostava Overseas 5.49 € Box Now 4.99 €

Besplatna dostava putem Box Now paketomata i Overseas kurirske službe iznad 69,99 €.

AI GPU Infrastructure

From Design Runbook to Reference Architecture

Jezik EngleskiEngleski
Knjiga Meki uvez
Knjiga AI GPU Infrastructure Vitali de Mille
Libristo kod: 54579365
Nakladnici Independently published, studeni 2026
Architect, deploy, commission, and operate mission-critical AI data centres with zero guesswork.Dist... Cijeli opis
? points 90 b Pripremamo Pripremamo Novo Novo
36.68 €
Ne jamčimo isporuku do Božića ▾ Očekivane nove zalihe Dobivanje novih zaliha 24. 11. 2026

Do 30 dana za povrat

Architect, deploy, commission, and operate mission-critical AI data centres with zero guesswork.

Distributed AI workloads have pushed data centres to their physical limits. When a single compute rack draws over 100 kW, a single network packet drop stalls an entire training job, and a component failure collapses goodput, the boundary between software and physical facilities disappears. AI GPU Infrastructure: From Design Runbook to Reference Architecture (2026 Production Edition) by Vitali de Mille is a vendor-neutral, standards-grounded field manual for moving multi-megawatt AI estates out of slides and into reliable production.

This handbook treats the AI cluster as a coupled, repeatable operating system spanning facility physics, silicon topology, and orchestrator runbooks.

Technical Coverage Across All Five Infrastructure Planes:

  • Facility & Rack Engineering: Sizing power distribution under continuous and transient step loads (di/dt). Designing air containment, Rear-Door Heat Exchangers (RDHx), and Direct-to-Chip (DTC) liquid cooling loops with Coolant Distribution Units (CDUs) mapped to ASHRAE liquid classes.
  • Compute Node Architecture: Internal hardware mapping of host topologies, NUMA boundaries, and PCIe Gen5/Gen6 root complexes. Maximizing scale-up fabrics like NVIDIA NVLink/NVSwitch and AMD Infinity Fabric.
  • AI Fabrics & Interconnects: Designing non-blocking, rail-optimized Leaf-Spine topologies. Managing congestion via ECN marking, Priority Flow Control (PFC) watchdogs, and Adaptive Routing across InfiniBand (NDR/XDR) and high-speed Ethernet (RoCEv2).
  • AI Tiered Storage & Pipelines: Sizing pipelines for sequential dataset streaming, parallel filesystems, and checkpoint write bursts. Implementing zero-bounce-buffer paths with GPUDirect Storage (GDS) and RDMA storage networking.
  • Software & Governance Operations: Enforcing hard workload partitioning using hardware Multi-Instance GPU (MIG) and secure container runtimes. Tracking full-stack cluster observability (DCGM telemetry, network FEC/PFC counters, facility sensors) while maintaining a strict, signed, and immutable firmware baseline.

Field Manuals, Sizing Calculations, and Blueprints:

  • 5 Worked Case Studies: End-to-end mathematical sizing for a 48U 8-GPU Air Rack, a 1.08 MW liquid-cooled GB200/GB300 NVL72 Pod, a 32-node fabric, a Megawatt 800 VDC Busbar conversion chain, and a Root-Cause Analysis (RCA) tracking a thermal straggler incident.
  • Day-2 Commissioning Runbooks: Comprehensive step-by-step gate procedures tracking Factory Acceptance Testing (FAT), Site Acceptance Testing (SAT), and 72-hour Burn-In testing to flush out latent infant mortality failures.
  • Normative Standards Alignment: Full architectural compliance mappings to ANSI/TIA-942-C (May 2024), ASHRAE TC 9.9 (Liquid Classes W17-W+), IEEE 802.3df-2024 (800GbE), ISO/IEC 30134-2:2026 (PUE metrics), and NIST SP 800-193.

Built for Infrastructure Professionals:

Designed for Infrastructure Architects, Network Engineers, Data Center Operators, SREs, and Technology Executives responsible for governing large capital expenditure budgets and ensuring the day-to-day resilience of high-density computing platforms.

Stop firefighting uncoordinated infrastructure failures. Build platforms that survive reality.

Glumica & Poliglotkinja
EWA KASP za
Pusti video
Ewa Kasp
Libristo ima najveći izbor literature na stranim jezicima. Zato svoje knjige kupujem ovdje.

Informacije o knjizi

Puni naziv AI GPU Infrastructure
Jezik Engleski
Uvez Knjiga - Meki uvez
Datum izdanja 2026
Broj stranica 638
EAN 9798142990815
Libristo kod 54579365
Težina 1247
Dimenzije 203 x 254 x 33
Poklonite ovu knjigu još danas
To je jednostavno
1 Dodajte knjigu u košaricu i odaberite isporuku kao poklon 2 Zauzvrat ćemo vam poslati kupon 3 Knjiga dolazi na adresu poklonoprimca

Prijava

Prijavite se na svoj račun. Još nemate Libristo račun? Otvorite ga odmah!

 
obvezno
obvezno

Nemate račun? Ostvarite pogodnosti uz Libristo račun!

Sve ćete imati pod kontrolom uz Libristo račun.

Otvoriti Libristo račun
Književni savjetnik Libroamiko
Dobar dan, ja sam Libroamiko, mogu li vam pomoći?