Performance and Cost: How Hardware Influences Our Programs
summary
Explains how modern CPU and accelerator hardware (x86-64, ARM including Apple M1, AWS Graviton, RISC‑V, SoCs vs bare CPUs, and DPUs/GPUs such as AWS Nitro, NVIDIA BlueField, Kalray) affect application performance and cost. Covers CPU internals and architectural differences, the role and use of DPUs/GPUs as offload units, compilation examples, and practical guidance for developers to choose and exploit cloud and server hardware for better price/performance.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 3:15Modern hardware evolution
- 13:49System on chip (SoC)
- 20:56Instruction sets and optimization
- 30:23Software and hardware optimization
- 40:49European open hardware innovations
- 44:20Code and ecological impact
- 45:04Q&A
related talks
01h 11mPerformance engineering - being friendly to your hardware
45mEfficient Use of Modern x86-Compatible CPUs
01h 05mMegahertz, Gigahertz, Registers and Instructions: How does a CPU actually work?
01h 02mMegahertz, Gigahertz, Registers and Instructions: How does a CPU actually work?
51mLinux Systems Performance - Architecture
38mPeculiarities of modern hardware
59mBeyond pprof, Hardware Performance Counters
44mOptimizing Software for Modern CPUs and Linux