Tuesday, November 8, 2022

Practice coding interviews

https://www.techinterviewhandbook.org/grind75

https://www.educative.io/courses/grokking-the-coding-interview

http://educative.io/

https://www.algoexpert.io/product

hackerrank

CoderByte

Book: Cracking the coding interview

https://software-design-coding.blogspot.com/2020/12/learning-software-design-architecture.html


Frontend:

https://www.greatfrontend.com/questions/coding

https://bigfrontend.dev

https://leetcode.com/discuss/interview-question/742791/front-end-interview-questions


Cyber-Security:

https://tryhackme.com


1. Leetcode

Always ask clarifying questions, they are meant to be vague.


Tech Interview Handbook (solve 50 suggested LCs multiple times): https://lnkd.in/dbK9TasP


Grokking Coding Interview:

https://lnkd.in/d4YzR_3d


Solve each question yourself before looking at answers. After struggling with certain topics, found some youtube resources that were really helpful.


- Graph problems: https://lnkd.in/dT4brkwG

- Kahn’s topsort: https://lnkd.in/dc8VzPx7

- Dijkstra’s shortest path: https://lnkd.in/djWD4ncp

- Kruskal’s MST: https://lnkd.in/dUaF7dDr

- Union find: https://lnkd.in/dmZSyGXq

- Dynamic programming: https://lnkd.in/d85BHsQp


I would say 1/2 of my coding questions was graph related, and I used topsort, MST, union find during my interviews so its worth knowing them well.


Great resource on Binary Search (esp. for harder variants): https://lnkd.in/dPVikn9e


Quite a few LC hards involve binary search as part of its solution, but its non-obvious.


At staff level, you are expected to have great communication, come up with one or more high level solutions in reasonable time with space and time complexity, implement mostly correct, and check for errors independently.


2. SYSTEM DESIGN


Always ask clarifying questions, they are meant to be vague.


No. 1 tip: Pay for mock interviews, take extensive notes, study, pay for more interviews. I used igotanoffer.com. Try booking different interviewers, they have different expectations, communication, friendliness.


Grokking system design: https://lnkd.in/dKBtGDDG


It’s good to complete grokking, but in practice its a little too shallow for staff level.


Watch all videos and take notes from this channel: https://lnkd.in/d32sycXJ


Most other channels like Exponent are actually not great examples of a good interviewee for staff level.


Read papers on Dynamo, BigTable, MapReduce, Cassandra, Raft, Lamport Clocks, etc.


A lot of material written by this Medium writer is great: https://lnkd.in/d_4yxAA7


Memorize Jeff Dean’s latency numbers: https://lnkd.in/dq42yWsa


If you have time, read DDIA: https://dataintensive.net


I memorized mnemonics for structure of how I execute my interview, as well as for each topic. Like OSI model, caching strategies, load balancing strategies, database types, etc. You need to make up your own in order to ingrain them.


link of post from blind website https://lnkd.in/dShYNZDf


Deprecated:

https://www.techinterviewhandbook.org/coding-interview-study-plan/

https://www.techinterviewhandbook.org/best-practice-questions/

https://www.teamblind.com/post/New-Year-Gift---Curated-List-of-Top-75-LeetCode-Questions-to-Save-Your-Time-OaM1orEU

https://leetcode.com/discuss/general-discussion/460599/blind-75-leetcode-questions


Monday, November 7, 2022

AI chip companies

Cerebras: Wafer-scale performant chip
Sambanova: High memory on chip
Graphcore: Stacked wafer design
Groq: Performant Inference platform
Tenstorrent
Neuralink
Untether
Tensorchip
Celestials.ai: Optical light interconnect
Nvidia
Intel with Habana acquisition
AMD's new Ryzen
Blaize: Chip architecture that understands graphs
Axelera AI BV: in-memory computation
Kalray SA: Tensor co-processor PE & direct tensor data exchange
Hailo: Low power inference chip without external memory
LightMatter: Light interconnect instead of copper

Free AI Chat tools

https://grok.com https://x.com/i/grok https://chatgpt.com https://copilot.microsoft.com https://chat.deepseek.com https://www.meta.ai https:...