WP_Term Object
(
    [term_id] => 13
    [name] => Arm
    [slug] => arm
    [term_group] => 0
    [term_taxonomy_id] => 13
    [taxonomy] => category
    [description] => 
    [parent] => 178
    [count] => 384
    [filter] => raw
    [cat_ID] => 13
    [category_count] => 384
    [category_description] => 
    [cat_name] => Arm
    [category_nicename] => arm
    [category_parent] => 178
)
            
Mobile Unleashed Banner SemiWiki
WP_Term Object
(
    [term_id] => 13
    [name] => Arm
    [slug] => arm
    [term_group] => 0
    [term_taxonomy_id] => 13
    [taxonomy] => category
    [description] => 
    [parent] => 178
    [count] => 384
    [filter] => raw
    [cat_ID] => 13
    [category_count] => 384
    [category_description] => 
    [cat_name] => Arm
    [category_nicename] => arm
    [category_parent] => 178
)

Dilbert Flopped – But We Still Laugh

Dilbert Flopped – But We Still Laugh
by H.B. on 07-04-2016 at 12:00 pm

This tile is about an old timer talking with a smart ass that questions why experience is relevant in todays “fast paced” technology industry. It has shown up so much on LinkedIn that I thought I should make a separate post and copy my responses into it and just link in next time.


I should first admit that this joke is very funny. I am a Dilbert fanboy because they get it right (almost) every time. Dilbert is so hilarious that it is funny even when it gets all wrong. If you still did not get where it is wrong, here is a list:

1. The New Technology Fallacy
Most “new” technologies are just repackaging of old ones: deep learning has been around since the 70s and has been repackaged+renamed by Google’s hype machine. IOT is powered by 20 year old 8-bit RISC processors that just now were cheap enough to create the Arduino boom. So 99% of the time it is not a new technology but rather a new application of old technology.

Technologies are all interconnected. They are not solitary animals like a mountain lion that meets others once in years to procriate. They are dependable on an entire ecosystem. For example, creating a new CPU is very easy. Just head to OpenCores and browse through the dozens of CPUs available.

However, a CPU needs assembly code to run and assembly is very hard to code for so you need a toolchain: preprocessor, compiler and linker tools. More frequently one would take an extensible toolchain like GCC or LLVM and implement a new backend, and that is far from trivial. Now that the CPU can be run, you need to allow it to access hardware and for that you need an operating system. Usually one would attempt to port older and stripped versions of Linux to the new hardware. And only then you go finally through improving your product. Guess how many man-hours (or man-years) are needed?

Then comes the best part – you have to evangelize, sell and train all your customer user base. 90% of websites like Altera and Xilinx are dedicated to documentation and training. Employees nowadays are already drowning in work harassment and sensitivity training – how are you going to fit the needed hours to make them excel in a completely new technology? Most often managers will chose the product that requires the least amount of training and disruption. Just because nobody has ever been fired for buying IBM, correct?

2. Technology changes in 6-months cycles in a “Youth-Oriented Culture”
The language C, that makes up at least 90% of every Linux distro is over 50 years old. C++, 35 yo. Java/Javascript, 20 yo. FPGAs, 30 years. Linux is 25 years old while UNIX is over 40 years old. In fact if you get a sysadmin that was cryogenic-frozen in mid 70s and thaw him today, he will likely to be still very comfortable in a modern datacenter.

The ARM you have today on your iPhone/iPad is a direct derivative of the 6502 processor that gave life to the Apple I. And the powerful Intel processors of today are just new releases of the old 8080 first produced in 1974. On the other hand you can see a resurgence of ancient technologies like the tiny stack-machine ZPU.

The Android OS is nothing more than our Linux O/S, repackaged. While the Linux OS itself, when created, was nothing more than an implementation of the already-20-yo Unix into the Intel 80386/80486 processor. More, the Linux OS should be in fact called GNU/Linux because more than 99% of its code belongs to the GNU project that started back in 1985. And the GNU project itself was an attempt to clone the Unix standard which is much older than that.

Meanwhile the iPhone/iPad operating system iOS is a shrink of the desktop Mac OSX (in 2002) which was a new version of the Classic Mac OS, that dates back to 1984.

This is all very slow, organic growth. Nothing completely changed in 6 months cycles.

3. The Old Modem Reference

I left this to the end because it is the best flop of the whole strip. Well, do you realize how you (and your business) connect to the internet? Yes, a MODEM! A broadband modem, which mind you fits perfectly the definition of a modem by transmitting data through complex waveforms.

Everyone has a (broadband) modem at home and work. It amazes me how people get surprised when pointed this out. Phone modems are different because they are limited by the range of voice frequencies (200-10kHz) that can traverse a telephony circuit.

Broadband modems have dedicated copper so the signal can be modulated to ultra high frequencies UHF, typically around 600-800MHz. But the underlying transport technology, quadrature amplitude modulation or QAM modulation, is exactly the same.

However knowing all this is perhaps too much to ask from a cartoon writer, isnt it?

Conclusion
It takes zero scientific knowledge to draw a cartoon. But it takes a ton of “old technology” and marketing magic to create a new product. Beware that renaming/rebranding is the main tool of Corporate America’s marketing departments.

Do not be fooled by the hype industry. Companies want you to think that their marginal improvement is the biggest breakthrough in twenty years. In reality, technology adoption and transformation is in fact extremely slow.

Henrique F Bucher, PhD, MSc, BSc

Share this post via:

Comments

4 Replies to “Dilbert Flopped – But We Still Laugh”

You must register or log in to view/post comments.