D5 - Moore's Law and More

0.0(0)
studied byStudied by 0 people
0.0(0)
full-widthCall with Kai
learnLearn
examPractice Test
spaced repetitionSpaced Repetition
heart puzzleMatch
flashcardsFlashcards
GameKnowt Play
Card Sorting

1/22

encourage image

There's no tags or description

Looks like no tags are added yet.

Study Analytics
Name
Mastery
Learn
Test
Matching
Spaced
Call with Kai

No study sessions yet.

23 Terms

1
New cards

Moore's Law

Prediction that transistors per chip double roughly every 2 years, making computing cheaper and faster.

2
New cards

Outcome of Moore’s Law

Exponential growth in chip performance for decades.

3
New cards

Moore’s Law slowing

Physical limits mean clock speeds aren’t increasing; single-thread performance has plateaued.

4
New cards

Parallel performance

Gains now come mostly from multi-core and parallel computing.

5
New cards

AI/ML growth rate

AI performance doubles roughly every 6 months, far faster than Moore’s Law.

6
New cards

AI training cost drop

AI model training went from $1,000 to $7.43 (2017–2020).

7
New cards

Waves of computing

Mainframes → Minicomputers → PCs → Internet → Smartphones → Pervasive computing (IoT).

8
New cards

Pervasive computing

Future devices powered by cloud “remote brains” like smart speakers and autonomous cars.

9
New cards

Cloud computing role

Offloads heavy compute to the cloud, giving devices more power.

10
New cards

Cloud downside

Latency delays from sending data to cloud and back.

11
New cards

Quantum computing

Highly experimental; hyped but not replacing classical computing soon.

12
New cards

Total Cost of Ownership (TCO)

Full hardware cost including maintenance, energy, staffing, upgrades, training, and disposal.

13
New cards

Hardware trend: go small

Mobile devices, wearables, and BYOD; increases flexibility but adds security risk.

14
New cards

Hardware trend: go big

Data centers and mainframes for massive compute workloads.

15
New cards

Solar-powered data center

Example of “go green” hardware for sustainability.

16
New cards

High-performance computing (HPC)

Uses supercomputers or many processors working together for large problems.

17
New cards

Grid computing

Links many distributed machines to act like one big supercomputer.

18
New cards

Cluster computing

Group of linked servers functioning as a unified system.

19
New cards

SaaS

Software accessed through a browser; provider handles everything.

20
New cards

PaaS

Cloud tools and platforms developers use to build apps.

21
New cards

IaaS

Renting cloud servers, storage, and networking instead of owning hardware.

22
New cards

Cloud tradeoff

Less control and customization but lower cost and fewer management responsibilities.

23
New cards