Category Archives: Geeky

My Journey to the Cryptomine – A Toastmaster Speech

Last August I gave a Toastmaster speech on bitcoin, which stood at ~$2,700/bitcoin. It went all the way to almost $20K but today is ~$11,000/bitcoin. With that kind of swing, there’s lots of money to be made or lost. There are many ways to profit OR sustain loss from cryptocurrencies. You can buy the bitcoins or other cryptocurrencies outright and cash out when it fetches a high price – buy low, sell high. But I never believe in the get-rich quick scheme, probably because of my past record of painful experiences.

One of the “sure” ways to make money from cryptocurrencies is to mine the coins. What does it mean to mine crypto currency? The mining process is a way to earn the coin without paying for it directly.

Before we go any further, let me back off a bit and give you a bitcoin primer. The methodology applies to other cryptocurrencies. All transactions are recorded in blocks. Each block is like a page of ledger books consists of ~2000 transactions. They are chained together like the binding of a ledger book, hence the name blockchain. This is so no one can insert a “fake” block/ledger page into the book to double-spend or spend more than they have. To promote decentralization, validation or clearance of a block is performed by miners as contests to win a pot of 12 ½ bitcoins plus fees. Bingo! When first one who create a 256-bit hash/checksum that meets a certain format, usually number of leading 0’s, wins the entire pot. Then on to the next block. The difficulty of the game readjust automatically so that each block takes about 10 minutes. The reward got halved every 4 years or so.

How do the miners earn the coin? By doing the hard work or making the computers doing the work of validating the ledgers of each block of transactions once every 10 mins for bitcoin or shorter time for the altcoins. The miners’ job is to bundle all transactions into a block or a page of ledger and perform the “Proof of Work” by trying and retrying to come up with a hash code that meet a certain format: at least some number of leading zeros. In return, the first person or team/pool who comes up with the correct format shout “Bingo” or some other words get 12.5 bitcoins today plus all the transaction fees within the block. The same process goes on every 10 minutes 24×7 nonstop. This is a case for bitcoin. The other coins usually have a similar process.

I decided to try mining over the Christmas holidays because I was bored. But first, I need to decide which crypto coin I want to mine. I looked around and tried to determine which coin provides the highest return with minimal investment. Bitcoin was out of question because in order to be profitable, the equipment must be able to “hash” at TH/s or billions of hashes within a second or you may be paying more in electricity to PG&E than the reward you’re getting. The special equipment costs around $6K, too much upfront cost for my taste.

I finally decided on mining Ethereum, which would require buying a graphic card or GPU, which cost ~$400 for a popular AMD model. They were very difficult to find online or retail shops. Half of the aisle of the GPU cards were empty as of last weekend I checked. Then I bought some special extension cable to bring the GPU card outside of my home PC to supply separate power (another $50). Next, I decided on which pool to join. For Ethereum, the choice was easy. I picked the biggest. The pool aggregates all the miners’ results and winnings and share the profit the entire pool earns. If you go solo, you may be mining for years before getting the pot of money for the black statistically. I followed the directions and downloaded the software the pool requires and ran it. Tried running another recommended software still got the same bad result until I heard on the net about a firmware tweak on the GPU. Flashing the firmware could potentially risk bricking the GPU card. I hesitated and decided to take the bite. I scouted the internet for tips of what memory and CPU timing to tweak to maximize the “hash” rate. Evidently, someone has tried and probably bricked a few GPUs along the way. I tried it and got lucky; I boosted the GPU performance by 25%. Now I was ready for prime time. Let’s get the dough rolling.

For several days, I watched and observed the payout slowly creeping up. After 4 days, I was earning about $13, then the crypto market had a correction and dropped the earning down to $10 or roughly $2.50/day. Taking the electricity cost into account, I was earning roughly $1.50/day. It would take me 300 days to earn back my gear. I could scale up to more GPUs, right? I couldn’t find the same GPU for $400 anymore. It’s now $520 on eBay if I could find it.

I have tried mining one other altcoin, the payout was even worse. Like the gold rush, the winning party may end up being the one providing the shovel: GPU vendors and ASIC chip vendors.

The good news is that I can now sell the GPU on eBay and even make $100 back. But more importantly, I no longer have to hear the loud fan noise in my den. Ah! A peace of mind of no mine!

Raspberry Pi Garage Server – an IoT Project


This is my newfound interest – IoT or Internet of Things. I have been taking in all the Maker movements and decided to do something about it. This particular project Raspberry Pi Zero W was the first of the many IoT projects I plan to do and bring you along in my journey. Raspberry Pi Zero W is one of my favorite controller because it’s compact, inexpensive ($10) and with built-in Wifi and well supported by the Raspberry Pi community.

I wouldn’t say this was challenging to me as I have done more difficult and complex hardware design in my career as a hardware development engineer, designing the big-iron mainframe computer and multi-CPU servers. But this was probably more fun as I got to sense and control the real-life environment and it’s relative inexpensive to do.

I look forward to sharing more of this type of projects with you, in addition to all the “boring” book reviews, according to my daughter, I will continue to make.

Monitor Display Cropped With HDMI Input From PC – How I Fixed It

Recently, I replaced my 24″ monitor that has a VGA and DVI inputs with another monitor ViewSonic VX2453 that has one VGA input and 2 HDMI inputs. All is well when I connect to the VGA input from my desktop PC. But I need to share this monitor with both my PC and my laptop, which has only one HDMI output. So I just need to connect the HDMI output to the HDMI input of the monitor. Simple right? Not so fast. The display turned out to be heavily cropped by the monitor, in other words only ~70% of display was shown; the rest could not be seen.

I suspected there may be three possible causes: 1) my graphic display driver and 2) HDMI cable 3) the monitor settings.

The first thing I did was to update the Intel HD 3000 graphic driver, which was dated 2011. I downloaded the Intel Driver Update Utility 2.2 . The utility failed to update the driver after 3 attempts. Then I tried to download the driver directly from the vendor’s (Acer) update site, which hosted the same old driver I had. Then I Googled and found the exact driver from Intel’s manual download (go to download center and search on “hd 3000”). The worked fine to update to the latest graphic driver: dated 5/16/15. Great! But it did NOT help my monitor cropping problem.

Next I tried to rule out the HDMI cable by swapping the cable with another. It didn’t help either.

Then I Googled around and saw another idea from this article. The suggestion to change the scale seemed to make sense. The first step is to select the Advanced Graphic Mode by right clicking the background wallpaper and select “Graphic Properties”. See below:
Graphic mode
So I change from “Maintain Display Scaling” to “Custom Display Scaling” and adjust to ~ 70% to get out of the “cropped” mode. See below:
BEFORE:
Scale Original
AFTER
Scale After

So this should fix it right? Yes and no. The display was not cropped but it looked like “crap” – lots of rough edges. This was not going to work for me. So I decided to try my last option 3 – adjust the monitor. I went through each menu carefully. And voila! I found the source of the problem. As it turned out, this monitor has two different modes for the HDMI input: AV mode and PC mode.
2015-09-29 22.12.45
2015-09-29 22.13.01
The default is the AV mode which automatically crop out 30% of the display. By selecting the PC mode. The problem went away after changing the Scale to “Maintain Display Scaling” and the display was fantastic! Problem fixed!

By running the display via the HDMI cable, the sound can be redirected to the speaker of the monitor. But I had to manually changing the sound output to the VX2453 by right clicking the speaker icon on the bottom and select “Playback devices” and then select “VX2453 Series” to be the default output.
Sound Output Selection
That was the bonus! Hope you find this tip helpful and wouldn’t have to waste the 2 hours I spent to fix the problem.

Book Review: “How Google Works” by Eric Schmidt and Jonathan Rosenberg

Eric Schmidt and Jonathan Rosenberg were the early hires of Google. Eric came to provide the adult supervision as the CEO before their IPO and Jonathan came to Google to manage the product offering. This book is the combined perspectives of how Google get things done in terms of their hiring, culture and history. It’s a good read if you admire and want to join Google or you’re starting your own company some day.

My key takeaways:

1. Having the “moonshot” or 10x improvement as the goal. Ask what could be true, not what will be true 5~10 years from now.
2. Focus on the end users, not profit or partner’s profit, money will follow.
3. The distinction between the “knowledge workers” vs. “smart creatives.”
4. Optimize for scale, not revenue.
5. Hiring the learning “smart creative” not just experts. Use a committee.
6. We all need a coach if the world’s best athletes need coaches.
7. Review yourself.
8. Do trip reports in the staff meeting.

1. Introduction:
The new era with three technology trends converging: 1. Everything is online, 2. mobile devices are ubiquitous, 3. cloud computing puts practically infinite computing power and storage at everyone’s disposal. As a result: a. Speed of changes is increasing: b. product excellence is paramount. c, consumers have abundant choices, d. cost of experimentation and failure has dropped sharply, e. faster product cycle and flexible process. The contrast between “knowledge workers” and “smart creative” at Google: not confined to specific tasks, unlimited in computer access, not averse to taking risks, not hemmed in role definition or organization structures, don’t keep quiet when they disagree, get bored easily and shift jobs a lot. He/she is an expert in doing – doesn’t just design concepts, she builds prototypes. A power user, she understands her product from the user perspective…

2. Culture – Believe Your Own Slogans:
Smart Creatives place culture at the top of the list in choosing his work. To be effective, they need to care about the place they work. Define the culture you want at the onset of your company. Don’t listen to the HIPPO (Highly-Paid Person’s Opinion). Rule of Seven: Minimum of 7 direct reports per manager. P&L business units should be driven by separate external customers and partners. Do all reorgs in a day. Organize the company around the people whose impact is highest. Give people responsibility and freedom to manage their own time: work vs. life. Establish a culture of Yes. “Don’t be evil” as Google’s mantra.

3. Strategy – Your Plan Is Wrong:
Business plans are often wrong. Invest in the team, not the plan. Have the “foundation blueprint”: “Bet on technical insights (not market research) that help solve a big problem in a novel way, optimize for scale, not for revenue, and let great products grow the market for everyone.” “Tail is wagging the dog when market research becomes more important than technical innovation.” Scaling up platforms (a set of products or devices that bring together groups of users and providers to form multi-sided market) needs to be the core in the internet age, at the expense of profit. Find ways to specialize in areas that has the potential to expand, as Google does in searches. Default to open, not closed. “Open” harnesses the talents of many. Exceptions including Google’s search algorithms, to avoid people’s gaming the system. Don’t follow competition – leads to mediocrity because you can’t deliver anything truly innovative. On Strategy Meeting, start by asking what will be true in five years and work backward. The responses are different if you’re incumbents vs. challengers.

4. Talent – Hiring Is the Most Important Thing You Do:
“Herd effect” = great employees attracts more great employees. Passionate people don’t use “passionate” word. Hiring learning animals not just an expert in one area. Check the “character” by doing the “LAX test” (OK to be stranded in LAX for 6 hours). Interviews: 30-minute long and by committee, hiring packet (with best/worst answers, grades from each interviewer, school GPA and etc.). On retention: trade the M&M’s and keep the raisins (keep the top performers’ jobs interesting. Do’s: hire people who are smarter and more knowledgeable than you are, will add value to the product and culture, will get things done, enthusiastic, self-motivated, and passionate, inspire and work well with others, will grow with your team and company, well rounded, with unique interests and talents, ethical and communicate openly. Hire only when you’ve found a great candidate. Don’t settle for anything less.

5. Decisions – The True Meaning of Consensus: Using the decision to exit China, decide with data, beware of the bobbledhead yes, know when to ring the bell (biased for action), maker fewer decisions (push down decisions to lower level), meet everyday (dictate the calendar), “You’re both right” (win the hearts, not just arguments – Oprah’s Rule). About good meetings: every meeting needs an owner. The decision-maker should be hands on. Manageable in size <=8~10. Attend the meeting - don't use laptop for other purposes. Decision on spending time: spend 80% of your time on 80% of your revenue. 6. Communications – Be a Damn Good Router: Default to open – board presentation is shared with all employees in weekly TGIF meetings. Every employee shares their OKR (Objective, Key Results). Know the details and truths. It must be safe to tell the truth. Weekly “Dory” Q&A sessions with Larry and Sergey at the TGIF meetings. Start the conversation (open office hours). Repetition doesn’t spoil the prayer: ask 1) does it reinforce core themes that you want everyone to get, 2) effective, 3) interesting, fun or inspirational, 4) authentic, 5) going to the right people, 6) using the right media, 7) tell the truth, be humble, and band goodwill for a rainy day. Break the staff meeting monotony with a humble trip report. Review yourself: make sure you would work for yourself – initiate criticism of yourself gives others the freedom to be more honest. Email wisdom: 1) respond quickly, 2) every word matters, and useless prose doesn’t. 3) clean out your inbox constantly. 4) Handle in LIFO (last-in, first-out), 5) Ask who else to route the email to, 6) Don’t BCC except to a large distribution. 7) Don’t yell. 8) Forward to yourself with keywords for future search. Have a playbook. Nice 1on1 format: 1) performance and job requirements, 2) relationship with peer groups, 3) management/leadership, 4) Innovation (best practices). In board meetings, noses in, fingers out. Discuss strategies and products, not governance and lawsuits. Deal with partners like diplomats and deal with press interviews by having a conversation, not message. Have relationship, take the time to know and care about people. And don’t forget to make people smile. When praise is deserved, don’t hold back.

7. Innovation – Create the Primordial Ooze: 3 criteria to determine if Google would pursue an innovative idea: 1) addresses a big challenge or opportunity, something that affects hundreds of millions or billions of people. 2) an idea for a solution that is radically different form anything currently in the market. 3) feasible and achievable in the not too-distant future. 4) technology, how it will evolve. “Innovative people do not need to be told to do it, they need to be allowed to do it.” – needs to be evolved organically. Focus on the user, not customers, then the money will follow as in the case of Google Earth. Think Big. Think 10x. 70/20/10 resource allocation: 70% related to core business, 20% on emerging, and 10% on new things with high risk of failure. Ship and iterate like Chrome. Fail well like Google Wave. Morph ideas, don’t kill them. Management’s job is not to mitigate risks or prevent failures, but to create an environment resilient enough to take on those risks and tolerate the inevitable missteps. A good failure is a fast one. It’s NOT about money.

Conclusion: Imagine the Unimaginable: We’re in a world of “platforms” (back-and-forth relationship with consumers and suppliers. A lot more give-and-take.) instead of “corporations” (more of one-way street from production to consumers). At the corporate level, most innovative new things look like small opportunities and people aren’t rewarded for taking risks and opt for safety. Ask the hardest questions. Understanding what you do about the future, what do you see for the business that others may not, or may see but chose to ignore? (e.g. Google+) Ask not what will be true, but what could be true. What thing that is unimaginable when abiding by conventional wisdom is in fact imaginable?