> Juggling which banks are open and closed has been traditionally a CPU's memory controller job. But communicating this information over a many-cm long PCB trace incurs latency. So it makes more sense to put this logic as close to the RAM as possible (speed of light, capacitance, inductance, etc. etc. slow down the signal).
The problem with this is that putting 'complex control logic' on the DIMM inevitably brings up the cost of a component which has a shorter lifespan than the CPU. In many applications, the RAM DIMMs are almost a 'consumable resource', in the sense that they are expected to be regularly replaced as heavy read/write cycles lead to their deterioration and eventual demise.
I specifically kept it vague for reasons of anonymity, as these types of statements have cost people their jobs in the past, and also due to the fact that what I described was work habits of people I talk to across different engineering roles (software, hardware, infrastructure) in the company I work for.
You can work less than 60 hours, but it will show in your output, as deadlines are aggressive and set tops down.
Examples of the kind of thing someone is answering emails in the early morning for is bugs discovered in your code, a regression that's failing, or something found that's blocking another coworkers' progress (since they were working late into the night as well).
It's definitely not an ideal life, but seems to be increasingly the norm among individual contributor positions at the big guys (Google, Apple, Amazon, etc). What's more is that on paper, these same companies espouse 'work-life' balance, while those who try to leave there laptops at work after hours end up on probation like the individual in this article.
I understand where you're coming from as I've worked both in and outside of SV (Amazon isn't in SV but shares the culture). Though it may look like they're asking for a lot, it has to do with the demanding nature of the jobs they're working.
To answer your question about being woke up in the middle of the night, there is no such thing as a "shift". Most of us are expected to work at the very least 60 hour work weeks (no additional pay, everyone is salaried), with many jobs beings 'always on', extending our work into nights and weekends with take home laptops. A typical (though not universal) day at Amazon is like follows:
- Wake up and open laptop to check emails and get some work in.
- Shower, eat, etc. before going into work.
- Work until around 7-8pm.
- Come home and eat, then open up laptop to work (maybe with the tv on in the background) up until going to bed.
Waking hours are pretty much consumed by work, and on weekends a lot of people will do a couple hours here and there and between other things, to stay current with issues.
If people in SV look like they're asking for a lot or getting upset over petty things, it's because of the sheer soul sucking amount of work that's being imposed on them. Small conveniences like a catered cafe makes the difference in helping them work through it.
Can you elaborate on what impact they had? The vikings reaching North America early is pretty neat, but doesn't seem to have changed the course of history much. Their colonies West of Greenland were pretty much failures, as they were systematically unable to compete with natives for resources. They were eventually pushed back east by skirmishes with the Thule and harsh, underprovisioned winters.
Please no. That's exactly the type of suggestion I would expect to see a startup looking to do things "differently" to get attention, without bringing any really value to the overall system.
The article is pretty shallow in the sense that it just casually glosses over a few key characteristics that make the server/datacenter space an interesting battlefield for the current round of architecture wars.
It will be an interesting next couple of years. Intel didn't sit on their hands with this one, like they did with the mobile market. They saw the potential for ARM SoCs, using a large number of smaller/lighter cores to take market share in things like simple web servers or hosting static data. Their response was their Atom based "Centerton" which is cannibalizing their own Xeon line.
As for more CPU performance intensive tasks, ARM has yet to prove itself in the performance/watt arena, especially in the 64-bit realm of server features, like error correcting code or RAS. Most sys admins building infrastructure like to play things safe, since they have to live with their decisions for years.
It seems like ARMs biggest advantage will be price, and the SoC business model of custom tailoring silicon for a customer's needs.
The problem with this is that putting 'complex control logic' on the DIMM inevitably brings up the cost of a component which has a shorter lifespan than the CPU. In many applications, the RAM DIMMs are almost a 'consumable resource', in the sense that they are expected to be regularly replaced as heavy read/write cycles lead to their deterioration and eventual demise.