WeAreDevelopers Logo WeAreDevelopers
Sign in
For You LIVE
Aug 22, 2024

Leveraging Moore’s Law: Optimising Database Performance

Behrad Babaee

Traditional database caching introduces unpredictable latencies and wastes modern hardware. Adopt a hybrid memory architecture to achieve in-memory speeds while drastically reducing your infrastructure footprint.

Pause
Mute Enter Fullscreen
#1 about 2 min

Applying modern hardware trends to optimize database performance

Applying historical hardware scaling trends optimizes structural data storage and backend operations.

#2 about 2 min

Understanding Moore's law and early hardware processing capabilities

Historical increases in transistor density reliably doubled hardware throughput every two years.

#3 about 2 min

Shifting software engineering focus to scalability and distributed computing

Backend software architecture transitioned from single-thread optimization to scalable distributed engineering systems.

#4 about 2 min

Examining legacy server hardware constraints and storage capabilities

Historical server configurations demonstrated severe bottlenecks in available system memory and rotating disk speed.

#5 about 2 min

Historical database architecture based on limited memory hardware resources

Limited server resources previously forced legacy database systems to overflow memory indexes onto slower physical disks.

#6 about 5 min

Analyzing modern server improvements in memory and storage

The adoption of solid state drives and dense memory banks radically shifted essential storage execution metrics.

#7 about 2 min

Utilizing excess modern server memory as a database cache

Deploying expanded hardware capabilities upon legacy database systems primarily utilizes excess memory as a caching layer.

#8 about 3 min

Unpredictable production failures caused by simplified database caching

Development testing strategies frequently evaluate volatile cache behavior instead of measuring underlying backend resiliency.

#9 about 4 min

Why database caching fails at improving parallel application latency

A single processing cache miss negatively impacts overall application speed during parallel retrieval tasks.

#10 about 4 min

Designing direct disk storage systems with memory index management

Maintaining the entire dataset query index within large system memory allows precise disk access operations.

#11 about 2 min

Leveraging constant memory indexing and direct hardware disk access

Dedicating hardware memory architecture entirely to index queries guarantees consistent read latency across changing network loads.

#12 about 3 min

Achieving faster database query performance by reducing resource consumption

Refining hardware operational sequences safely shrinks infrastructure workload footprints while multiplying database query response speed.

Matching moments

2:31 min

Accelerating data processing via resilient in-memory computing strategies

Markus Kett Markus Kett · World Congress 2022

4:44 min

Adopting purpose-built storage and compute instances

Ricardo Sueiras Sueiras · LIVE

1:29 min

High-performance data processing and IO capabilities

Igor Todorovski · World Congress 2023

3:26 min

Analyzing speed differences between memory and database queries

Markus Kett Markus Kett · World Congress 2023

5:37 min

Critical factors driving modern application performance requirements

Markus Kett Markus Kett · World Congress 2023

2:34 min

Simplifying enterprise architecture with database-less data processing

Markus Kett Markus Kett · World Congress 2023

Related videos

In-Memory Computing - The Big Picture thumbnail
30:36
World Congress 2023 • about 3 years ago

In-Memory Computing - The Big Picture

Markus Kett

Scaling Databases thumbnail
21:09
World Congress 2024 • about 2 years ago

Scaling Databases

Tobias Petry

Leveraging Real time data in FSIs thumbnail
41:51
LIVE • almost 3 years ago

Leveraging Real time data in FSIs

Tim Faulkes

Making Data Warehouses fast. A developer's story. thumbnail
42:16
JS Congress • almost 5 years ago

Making Data Warehouses fast. A developer's story.

Adnan Rahic

How building an industry DBMS differs from building a research one thumbnail
30:34
World Congress 2023 • almost 3 years ago

How building an industry DBMS differs from building a research one

Markus Dreseler

Build ultra-fast In-Memory Database Apps and Microservices with Java thumbnail
44:05
World Congress 2022 • over 4 years ago

Build ultra-fast In-Memory Database Apps and Microservices with Java

Markus Kett

Related articles

From learning to earning

Jobs that call for the skills explored in this talk.