Work in progress: 82,651 of 181,233 sites checked so far, more every week · updated 2 Nov 2026Missing a blog? Tell us

Top all-time · #363

J

James Hamilton

Personal blog · perspectives.mvdirona.comsystemsprogramming

James Hamilton writes Perspectives, a personal blog at perspectives.mvdirona.com about systems and programming. Since 2008, 148 different Hacker News readers have linked it 278 times, most often “AWS Nitro System”. It is still writing: the last post was on 2026-07-06, and you can follow it by feed.

Visit the blog
278mentions?
8recommend?
148HN people?
2008–2026first – last linked?
Links on Hacker News per year · peak 39 in 2011

Sometimes recommended · 8 of 40 sampled comments recommend it?

Why HN links it

HN readers link this blog for datacenter, cloud infrastructure and distributed-systems writing by an AWS VP. Comments cite specific posts on outages, ECC memory, latency and formal verification.

Recommended for · 8 comments
Called fascinating and insightful; comments called gems; a paper called the de facto way to build distributed services.

Summarised by Claude from a sample of 40 HN comments that link here. Each comment is marked recommending, critical or neutral (citing without judging); how this works.

Most-linked posts?

  1. AWS Nitro System
    7 people · 7 mentions · first linked 2019-02-17
  2. Xen-on-Nitro: AWS Nitro for Legacy Instances
    5 people · 6 mentions · first linked 2021-11-30
  3. /
    5 people · 5 mentions · first linked 2010-11-15
  4. Choose Technology Suppliers Carefully
    4 people · 4 mentions · first linked 2018-11-14
  5. Networking: The last bastion of mainframe computing
    4 people · 4 mentions · first linked 2009-12-19
  6. Zynga returns to the cloud
    4 people · 4 mentions · first linked 2015-05-10
  7. Overall data center costs
    3 people · 4 mentions · first linked 2010-09-19
  8. The Cost of Latency
    3 people · 4 mentions · first linked 2009-11-03
  9. A Decade of Innovation
    3 people · 4 mentions · first linked 2016-03-10
  10. Amazon Aurora Wins 2019 Sigmod Systems Award
    3 people · 3 mentions · first linked 2019-07-07

What Hacker News says?

  1. jedberg · 2020-06-07Recommends
    > The people actually doing the work don’t have time to talk about it and the people spending all their time talking about it aren’t doing the work. I think that's a really sad and cynical view. I assume you don't spend 16 hours a day writing code. I assume you have hobbies and other interests. For some people, educating others about their favorite topics is their hobby. Writing about their projects is their hobby.…

    Comment linking /

    on “Most tech content is bullshit”
  2. Submitted At Scale, Rare Events Aren’t Rare as a story · 218 points

    HN discussion
  3. AWS has stated that there is a "Nitro Card for Instance Storage"[0][1] which is a NVMe PCIe controller that implements transparent encryption[2]. I don't have access to an EC2 instance to check, but you should be able to see the PCIe topology to determine how many physical cards are likely in i4i and im4gn and their PCIe connections. i4i claims to have 8 x 3,750 AWS Nitro SSD, but it isn't clear how many PCIe lanes…

    Comment linking /2019/02/aws-nitro-system/

    on “SSDs have become fast, except in the cloud”
  4. Submitted AWS Graviton2 as a story · 188 points

    HN discussion
  5. You're not wrong, but if you look at the problem that Glacier is trying to solve then it doesn't look so bad: 1. Glacier is competing with tape. So for potential customers, if it's easier to acquire and maintain than your own tape library then that's a win. 2. In the words of James Hamilton the use case for Glacier is referred "jokingly to as Write Only Storage". If you're using Glacier for disaster recovery then pa…

    Comment linking /2012/08/glacier-engineering-for-cold-data-storage-in-the-cloud/

    on “SSD Prices in a Free Fall”
  6. Submitted AWS Nitro System as a story · 161 points

    HN discussion
  7. Link to an article in James Hamilton's blog with actual pdfs of various designs: http://perspectives.mvdirona.com/2011/04/07/OpenComputeProje...

    Comment linking /2011/04/07/OpenComputeProject.aspx

    on “Facebook open sources its servers and data centers”
  8. Submitted Four DB2 Code Bases? as a story · 157 points

    HN discussion
  9. SEJeff · 2016-12-06Recommends
    Not sure how many papers they publish, but I do really use this paper[1] from James Hamilton (literally the Architect of AWS) as the defacto way of building large distributed services: [1] http://perspectives.mvdirona.com/2014/07/challenges-in-desig...

    Comment linking /2014/07/challenges-in-designing-at-scale-formal-methods-in-building-robust-distributed-systems/

    on “Apple to Start Publishing AI Research”
  10. Submitted Tesla Project Dojo Overview as a story · 125 points

    HN discussion
  11. Some more info here: https://perspectives.mvdirona.com/2018/11/1227/

    Comment linking /2018/11/1227/

    on “Amazon’s Chips Threaten Intel”
  12. Submitted ARM server market as a story · 121 points

    HN discussion
  13. Reminds me of a "self-demotion" that worked out well: That of a certain Allan Vermeulen [0] who, by some accounts, was the chief architect of Amazon S3 [1], and by all accounts, an engineer extraordinaire. Vermeulen managed the Platform Services group when he joined [2], and later moved to a CTO role with focus on the then very nascent Web Services initiative [3], but eventually "demoted" himself to an "AWS Engineer…

    Comment linking /2016/03/a-decade-of-innovation/

    on “I followed my dreams to get demoted to software developer”
  14. Submitted Formal Methods in Building Robust Distributed Systems as a story · 109 points

    HN discussion
  15. > we wanted the generators to sacrifice themselves under most every circumstance The rare failure mode that can cost $100m looks like this: when the utility power fails, the switch gear detects a voltage anomaly sufficiently large to indicate a high probability of a ground fault within the data center. A generator brought online into a direct short could be damaged. With expensive equipment possibly at risk, the swi…

    Comment linking /2017/04/at-scale-rare-events-arent-rare/

    on “OVH outage explained”
  16. Submitted Design Lessons & Advice from Building Large Scale Distributed Systems at Google as a story · 83 points

    HN discussion
  17. Annapurna is customer facing now with Graviton. Also, we don't know how much they invest after acquisition. I think Annapurna was more strategic than you think. * James Hamilton in 2010: https://perspectives.mvdirona.com/2010/10/datacenter-network... * Annapurna acquired in 2015. * In 2017 they reveal that every EC2 server has a Nitro (Annapurna) chip for networking: http://www.brendangregg.com/blog/2017-11-29/aws-e…

    Comment linking /2010/10/datacenter-networks-are-in-my-way/

    on “Amazon to pay $1B+ for Zoox”
  18. Submitted AWS Compute Cluster #231 on Top 500 as a story · 44 points

    HN discussion
  19. Facebook runs primarily out of ram via memcached. The last numbers I'm aware of were that they had about 200TB in memcache capacity [1]. They use a variety of data stores, but primarily sharded mysql. I don't have recent numbers there, but they were above 1000 master-master pairs as of 2008. While buying that much ram sound costly, that's only looking at capacity. Assuming typical 1u servers and common pricing at th…

    Comment linking /2010/07/01/Velocity2010.aspx

    on “MongoDB's lead developer: Foursquare outage post mortem”
  20. Submitted ARM Cortex-A9 SMP Announced: High-Performance ARM as a story · 43 points

    HN discussion
  21. NGI SpA, Milan, Italy or REMOTE I'm the CTO of a Italian ISP, we have 2000 radio towers all over the country and 150k active customers. We operate a very peculiar network based on a proprietary fixed wireless access platform and with unusually high backhauling mesh factor [1] which provides both extremely high redundancy and extremely painful headaches :) Frustrated by the limitations of the big vendors [2] we decid…

    Comment linking /2009/12/19/NetworkingTheLastBastionOfMainframeComputing.aspx

    on “Ask HN: Who is hiring? (June 2014)”
  22. Submitted Nvidia Project Denver: ARM Powered Servers as a story · 38 points

    HN discussion
  23. Same topic, same conclusion, even more hard facts. http://perspectives.mvdirona.com/2009/10/you-really-do-need-...

    Comment linking /2009/10/you-really-do-need-ecc-memory/

    on “Should I buy ECC memory? (2015)”
  24. Submitted 2010: The year of the microslice server as a story · 37 points

    HN discussion
  25. Unknown is relative, but here are some of the rss feeds I'm subscribed too: http://nosql.mypopescu.com/ http://lambda-the-ultimate.org/ http://eli.thegreenplace.net/ http://www.stefankrause.net/wp/ http://mechanical-sympathy.blogspot.com/ http://www.cowtowncoder.com/blog/blog.html http://highlyscalable.wordpress.com/ http://hunch.net/ http://perspectives.mvdirona.com/ http://www.scottaaronson.com/blog/ http://www.so…

    Comment linking /

    on “Ask HN: What unknown technical blogs or sites do you read?”
  26. Submitted Google Mechanical Design as a story · 34 points

    HN discussion
  27. wapoamspomw · 2019-02-07Recommends
    Some great reading on the topic from James Hamilton's blog. https://perspectives.mvdirona.com/2013/02/the-power-failure-... https://perspectives.mvdirona.com/2017/04/at-scale-rare-even... https://perspectives.mvdirona.com/2017/12/when-you-cant-affo...

    Comment linking /2017/12/when-you-cant-afford-not-to-have-power-redundancy/

    on “Tell HN: Wells Fargo completely offline”
  28. Submitted Web Search Using Small Cores: A paper on Microsoft Bing Architecture as a story · 28 points

    HN discussion
  29. This reminded me of James Hamilton's (the AWS datacentre guru) writeup of the power outage at the 2013 SuperBowl. Specifically how the default configuration of backup generators is often to protect themselves, even if that means shutting down the things they are meant to keep running. In my own experience running a reasonably sized datacentre I was often surprised at how UPSs and generators reacted to adverse condit…

    Comment linking /2013/02/the-power-failure-seen-around-the-world/

    on “That UPS you bought for your home server may not be as useful as you think”
  30. Submitted Observations on Errors, Corrections, & Trust of Dependent Systems as a story · 27 points

    HN discussion

Search all of HN for perspectives.mvdirona.com

Linked by the same people?