710 karma · joined August 15, 2015
As for streaming between machines, yep, it's awesome, but that's not the primary use case. I installed my wiring during a remodel a little ahead of the curve in the hopes that pretty but otherwise dreadful hippy tweaker heaven backwater Santa Cruz county would eventually catch up with nearby Silicon Valley. Sadly, ISP bandwidth has lagged Silicon Valley in the same way as most home values here (arguably the latter is a good thing for most but in my case it's kept me from selling and moving over the hill).
Hardwired 10 Gb Ethernet throughout the house.
Best ISP: 2.4 Mb to the house.
Sure, it's more mating display/beauty contest than entrepeneurialism, but are we not entertained? In the meantime, how else is one supposed to get the cash to bootstrap their big idea in the first place?
It's kind of like how John Williams wrote the theme to _Lost in Space_ long before he wrote the soundtrack to _Star Wars_, no? Or, ya know, Paypal before Tesla and SpaceX.
Amazon is at its best when it's customer obsessed and at its worst when it puts politics first.
All IMO of course.
A platform that runs AlexNet well has excellent computation performance for the convolution layers but it also has excellent algorithms/communication for parallelizing the model/data by whatever means.
Networks that attempt to minimize computation and/or communication are cool, but they should be considered in that light IMO.
It's also a great estimate of the low-end for strong scaling. There's a lot of bread and butter machine learning at this level in my experience.
Wonder what would happen to that scaling efficiency if those GPUs were P40s?
See also the absence of equivalent AlexNet numbers to further obscure attempts at comparing this to the other guys(tm).
Can't wait for Intel's response to this.
FPGAs are best when deployed on streaming tasks. And one would think inference would be just that, yet the published performance numbers are on par with 2012 GPUs. That said, if they had as efficient a memory architecture as GPUs do, things could get interesting down the road. But by then I suspect ASICs (including one from NVIDIA) will be the new kings.
But do come back when you've figured this all out and you know how to create a new consciousness in any way except for the obvious way.
IMO the answer to whether you leave or not is "it depends."
I've taken the counter once for a 25% raise as I was in the middle of a project and I cared about finishing it, but I usually leave if I find out I'm significantly (20% or more) underpaid (twice).
Glassdoor and meetups do not give personalized information about one's value. They only give average compensation data. The only way to get what you personally are worth is for you personally to get offers. Feel free to disagree.
In an age where variants of "FizzBuzz" are too hard for most interviewees to answer, it's an absolute joy to meet someone competent and make them an offer. YMMV.
Neither conferences nor meetups have ever worked that way for me. They're great for getting intel on the competition and catching up with friends, but we don't talk compensation much.
IMO unless someone is absolutely miserable, they shouldn't take a chance on a new and unfamiliar team unless the benefits (at least seemingly) outweigh the risks. Remember, you can fire them at any time, with or without cause.
Also most recruiters suck (all IMO of course). To which I'll add one final tip: don't ask me my current compensation and don't ask me what it will take to make me move. Instead, just tell me the salary range of the position during the initial phone screen. Problem solved, no?
From:
http://www.nolo.com/legal-encyclopedia/employment-at-will-de...
"An at-will employee can be fired at any time, for any reason (except for a few illegal reasons, spelled out below). If the employer decides to let you go, that's the end of your job--and you have very limited legal rights to fight your termination."
Now after reading the above sentence, do you still feel weird about this?
As for me, I worked at Amazon for several years as an engineer. The frugality is what eventually made me leave because too often it gets implemented as dirt cheapness. But in their defense, I had opportunities to do things there none of the other big 4 would have granted. And I got the chance to do them by being confrontational and confident. The downside is that if one interviews at other companies with one's Amazon personality, one will frequently come off as a total a$$hole.
Amazon will change you. Some of those changes are great, but you really have to keep your Amazon ego in check outside of the place.