Tensor Compiler Engineering (Paperback)

Idioma: inglés

Editorial: Independently Published, 2026

9798182027632

  • Tapa blanda
  • Nuevo
Ver todos los detalles

Librería: Grand Eagle Retail, Bensenville, IL, Estados Unidos de AmericaGrand Eagle Retail

Vendedor de 5 estrellas

Vendedor de IberLibro desde 12 de octubre de 2005

Tapa blanda

Condición: Nuevo

EUR 30,02

 Gastos de envío gratis 
Se envía dentro de Estados Unidos de America

Cantidad disponible: 1 disponible

Añadir al carrito
Devoluciones gratuitas de 30 días

Descripción del artículo del vendedor

Paperback. Tensor Compiler Engineering is your definitive guide to understanding exactly how machine learning models execute on physical hardware. I wrote this book to completely open the black box of AI infrastructure. It breaks down the complete compilation pipeline-from structuring mathematical computation graphs to generating highly optimized binary files-showing you exactly how to squeeze every drop of performance out of your deployment environment.Imagine you just spent weeks training a massive neural network. The mathematics are perfect. But when you push it to a production server, the latency is terrible, and the system crashes from unexpected Out-of-Memory errors.I have been exactly where you are.For a long time, I struggled to understand why perfectly written code ran so slowly. Then I discovered the truth: the bottleneck was not the math. The bottleneck was how the math was being delivered to the physical processor. I realized that mastering the compiler was the only way to break through this invisible barrier. I spent countless hours mapping out exactly how these systems work, step-by-step, tearing down everything from memory caching to hardware threading. I structured this book so you never have to guess what is happening under the hood of your AI deployments ever again.What's insideComputation Graph Optimization: Master operator fusion and Multi-Level Intermediate Representation (MLIR) to streamline your math.Hardware Acceleration: Discover how to manipulate hardware execution using intelligent loop tiling and vector unrolling.Silicon Real Estate: Learn static versus dynamic memory allocation and how to aggressively recycle physical VRAM buffers.Backend Code Generation: See exactly how compilers target and write CUDA, ROCm, and bare-metal CPU assembly code.Production Integration: Seamlessly handle dynamic tensor shapes and debug numerical errors on live servers without crashing.Who it's meant forThis book is written for machine learning engineers, systems developers, and software architects who want to stop relying on bloated software frameworks and start understanding physical execution. If you know how to build a neural network but want to learn how to engineer it to run efficiently on bare silicon, this guide is built for you.The next generation of artificial intelligence belongs to the engineers who control the infrastructure. Do not let hardware limitations choke your software. Grab your copy today and take absolute control of your execution pipeline. This item is printed on demand. Shipping may be from multiple locations in the US or from the UK, depending on stock availability.…

N° de ref. del artículo 9798182027632

Título
Tensor Compiler Engineering (Paperback)
Autor
Timothy A. Sprenger
Editorial
Independently Published
Año de publicación
2026
Estado
new
Encuadernación
Paperback
Idioma
inglés
ISBN 13
9798182027632

Grand Eagle Retail

Bensenville, IL, Estados Unidos de America

Vendedor de 5 estrellas

Vendedor de IberLibro desde 12 de octubre de 2005

Tarifas de envío en Estados Unidos de America

ArtículoDe 6 a 14 días hábilesDe 6 a 16 días hábiles
Primer artículoEUR 0,00EUR 0,00
Los plazos de entrega los establecen los vendedores y varían según el transportista y la ubicación. Los pedidos que pasan por la aduana pueden sufrir retrasos y los compradores son responsables de los aranceles o tarifas asociadas. Los vendedores pueden ponerse en contacto con usted en relación con cargos adicionales para cubrir cualquier aumento en los costes de envío de los artículos.

Métodos de pago

  • Visa
  • Mastercard
  • American Express
  • Carte Bleue
  • Apple Pay
  • Google Pay

Información empresarial del vendedor

APOLLO ONLINE CORP.

605 Geddes Street
Wilmington, DE Estados Unidos de America 19805