* Posts by bbulkow

6 posts • joined 22 Nov 2012

Mohawk Valley, NY to become the new Silicon Valley in fab fab deal


Upstate NY: do you even know what a fab is? Or a research fab?

I recently got to visit one of the world's premiere research fabs - Boise, ID. Engineers in silicon valley ( and all over the world ) control robots that execute the processes at the research fab. They fly in all kinds of silicon production equipment from all over the world (including Santa Clara). The research fab has a very large number (100's++) of "process gasses" of various toxicities but admittedly small amounts, because they're doing research. After the process is proven, some engineers visit for a few weeks, then they move the process to one of their production fab lines somewhere else in the world ( such as Singapore or Japan or Italy ).

We don't put fabs in silicon valley anymore because silicon valley is about people, brainpower, capital.... not toxic processes. The cynical might say anywhere there is a fab is CERTAINLY NOT silicon valley because no wise engineer would move their family to the shadow of a toxic plant.

Fab lines, in order to be clean, have very, very few people .... and a lot of toxics .... sure there will be a few thousand jobs as add-ons for shipping equipment and hotel stays for traveling workers....

Let's talk about the (real) price of flash and spinning disks


Prices for flash are wrong

By lining up consumer rotational drives against enterprise "SAS" flash, this artical does the reader no favors.

At time of printing, the first 3D Flash has just come available, with 1T SATA drives costing between $600 and $350, where the rather excellent Samsung 850's are on the higher end. I have seen drives like this - SS 840's, 843's, and Intel 320's bought years ago and still running. Lower power means lower strain.

In 3 years, we will have hybrid systems - but the two layers will be RAM and Flash. The next drop in Flash prices will shock anyone with a major investment in "hybrid".

Job for IT generalist ...


my old title: "Utility Infielder"

Answer first, questions later.

Pitch yourself as having specific skills. Choose a few of your skills, pitch them. Like, you do your homework about a job, you find out it's a c# networking job, and you pitch yourself as a c# and networking guy. You get the job, you spend 6 months doing the specific job (which will further hone your skills), then you pitch some other job within the group.

This is an essential problem in marketing. I have it in my company (a very fast NoSQL database called Aerospike). Our marketing deparment wants to talk about the speed of the database, because it's fast. And they want to talk about the reliability, because it's NoSQL that does ACID. And they want to talk about our Flash and SSD optimizations, because those are great for lowering in memory database cost. And they want to talk about our happy customers. And the real-time analytics that beat Hana. Or they talk about replacing redis and memcache and hbase. Or our incredibly robust clustering algorithm.

What happens? We end up with rambling statements without a point, confusing everyone, because they throw in the kitchen sink of buzzwords.

Get specific in each resume, cast yourself as something they understand, and once you're in, they'll use you for more and more.

( One word though: you really do have to be an excellent generalist. I run into a lot of people who say they're generalists but know so little about any given thing that they're useless for everything. They hide behind the "generalist" label because they can't be bothered to learn any single thing. If that's you, nut up, pick one thing, and spend two months without a job learning the living daylights out of that thing, then present yourself as an expert, and win jobs at that. You think you're a problem solver? Solve the problem of getting a job as a...I dunno... high scale Python expert, android hacker, whatever you fancy. That's actually another key to the tech industry that people seem to forget, and gets lost in the comments of ageism. Most older techies start to feel entitled and don't learn new things. I was taught you need to spend 20% of your time always learning new languages & skills, otherwise you get stale. Thus I am an over employed late-forties techie.)

Almost everyone read the Verizon v FCC net neutrality verdict WRONG


American consumers aren't dumb

Your several arguments are quite weak, and the court's ruling does have some strong points. We can only hope that, by putting the regulatory body on firm ground, we can get regulations that everyone abides by.

First, a duopoly isn't a wide open market, and the real feet-on-the-ground experience of American internet users is clear : there's rarely a reasonable choice. The court reasons - correctly - that the duopoly power we've created in internet service might be better than a pure monopoly, but it isn't a free market. Your statments that a customer "would just switch" are simply laughable to most of America.

In my area, I can get service from Comcast/Xfinity, and from ATT/Uverse. Right now, Xfinity is providing much better quality of service - my Uverse friends say they have multiple outages per day, periods when ATT's core routers seem congested and broken, and switching to Xfinity cleared it up. In my area, Xfinity offers speeds to 150Mb/sec, Uverse offers 20Mb/sec. Wireless in my area is possible (10Mb/sec), but cells seems to get jammed up now and then, hard to keep a VPN connection live - and wireless data fees are high (roughly $10/GB).

This is, at least, the duopoly at work. Many people I talk to are living with Uverse's poor service, but some are switching. Some stick because of the visceral loathing of Comcast, some can't be bothered.

If we add another axis, where each provider blocks some set of websites, good luck to me for figuring out who I'm going to use as a provider. One partners with Amazon, one with Netflix. What, I'm supposed to get both turned on? The court argues this is ridiculous, that limited set of provicers is reasonable for physical infrastructure, and thus is rightfully regulated.

Finally, we - internet users of america - aren't worried so much about blocking of a single media website like Netflix. Your straw man paints an unreasonable light. The real fear is google being held up to ensure quick response, and every other media company. The Reg certainly has a dog in this fight - they might be required to wrap their content in a larger company's livery (imagine: Yahoo), pay that company for the privilege, that company is paying off the ISPs for reasonable carriage, and The Reg has to squeeze all its contributors.

A poor way to run a railroad.

Leave the duopoly as it stands, but require "common carriage" style rules, allowing these companies to compete in a marketplace without a torture of 'if I want this movie series I have to choose that ISP'. That'll be fine.

IBM blesses NoSQL upstart MongoDB


welcome to the party, IBM

IBM has finally realized developers are let down by 10gen's NoSQL implementation, and IBM can provide an enterprise solution to address 10gen's failings. This API needs standardization - but make no mistake, this is IBM's move to cannibalize 10gen's enterprise sales effort by offering DB/2.

At Aerospike, we've delivered a fast and scalable ACID NoSQL solution, and seen huge market uptake - we understand IBM's interest in playing catch up in this market.

The fact that Aerospike has several ex-DB/2 engineers and advice from Don Haderle has helped.

NoSQL's CAP theorem busters: We don't drop ACID


this isn't about SQL

SQL is about the language, FoundationDB is about scalable transactions. Foundation seems to have some good technology, it will be interesting to see where they get traction as they have not yet been deployed at scale - unlike, say, http://aerospike.com/

The statement about the CAP theorem makes good press, but is misleading. It is absolutely possible to build at-scale transactions, and the CAP theorem doesn't prevent that. What happened was the CAP theorem allowed developers and architects to talk openly about relaxed consistency, which they had already given up by deploying and relying on memcache and sharded MySQL (memcache gives an inconsistent view of the data, shards remove the ability to transactionalize many queries, and MySQL is often run in a non-durable mode).

Once you've given up transactions, you have more choices in which database to use. But Foundation is right, the next step is to build safe transactions into scale-up databases. There is - absoluetly - a new world of databases, and the savvy architect needs to keep up with the times. When millions of transactions per second is an easy step, new businesses and problems can be attacked.


Biting the hand that feeds IT © 1998–2020