Update:
See also conserver family disposable income
MLP hosted a latency presentation by Martin Thompson. One MLP audience asked a question. Here’s what I vaguely recall:
Q: we spent lots of time to tune a system and achieved a few microsec of latency improvement, but someone points at another system (perhaps at a competitor) and say our system can be faster. What’s your view?
Case in point — I told a Goldman Hong Kong interviewer that my orderbook at Rebus can handle 700k mps. He said array-based orderbook can easily handle 1000k mps. Well, 700k mps on Rebus was more than enough for the business needs at RTS 🙂
Answer from Martin : To improve latency, the most important thing is time including developer’s time. Look at the requirement of the system. If your latency meets the requirement, then spending more time tuning is not effective (and not necessary, IMO). A few microsec further saving could be meaningless.
Q2: what if the requirement is to beat the competitor?
A: I think the project could become never-ending. The MLP latency-critical networking team shared personal experiences on the proper approach to network latency engineering. Network engineers would work closely with latency-sensitive traders to implement incremental latency improvements and within 72 hours assess the PnL impact. Invest more time and resources only if justified by PnL. If your trades are already beating the competition in terms of PnL, then having a faster or slower network may be irrelevant. Focus on the end goal (PnL) and stop worrying about secondary yardsticks like latency. Latency is not always the dominant factor in the game.
So that wraps up my three separate examples on latency. Now let’s turn to the other (wide-ranging ) topic of this blogpost — FOMO vs livelihood.
Q4: how much disposable income is enough for an (slightly above) average life, not a life in comfort and style? In terms of brbr (burn rate buffer ratio), how plentiful is enough? ffree^FOMO #9K/M asked a similar but more specific question.
I have achieved bare-bones ffree at a “conserver” level, So one of my two major pains now looks like the rich feeling outpaced by a tycoon… endless greed, mindless peer comparison like Rajat Gupta
— Rajat Gupta was the CEO of McKinsey, after earning IIT and Harvard degrees. (https://en.wikipedia.org/wiki/Rajat_Gupta) He was a multi-millionaire but reportedly wanted to join the billionaire club. Wikipedia said … Gupta reportedly began to express a certain resentment about money, as his peers in Silicon Valley and Wall Street (including McKinsey’s private equity clients) at the time “raking in staggering amounts of money while Gupta soldiered on with a mere senior partner’s millions”
His “mere millions” is more than enough for livelihood, but he was presumably driven by endless peer comparison. This endless FOMO may or may not be related to his disgrace.
— LuckyPlaza .. I envy the LuckyPlaza early investor who made perhaps a million dollar paper profit, but hey, he also envies me for
* my high annual salary. Total salary would exceed his paper profit in a few years. Over a few more years, my total would far exceed his paper profit.
* my wellness + family harmony
He would also envy the bigger investors who made $3M .. low-latency arms race! What’s enough? Perhaps half a million.
— Some cultures (and ethic groups) may be lukewarm about the latency arms race. As stated in some Americans must want to be successful, some Singaporeans must want to be successful.
— In ffree]US=unrealistic4many 华裔 middle-class Henry.Yin said RMB 500k/Y is insufficient for an unmarried tech worker.
What’s his “latency requirement”? I think a standard requirement for this class is owning a home in Shanghai or Beijing, which typically costs RMB 6,000k to USD 1M. Kevin.Chen said RMB 3,000k would buy a very small house in Shanghai. So I guess 99% of Chinese population is too poor by this standard. This standard is probably an affluence or luxury standard, for an exclusive club of rich people.
By this requirement, most low-latency systems are too slow…
— Dating competition .. is similar to the latency arms race. The attractive young ladies would demand “top 0.1% of bachelors” meaning more properties, more fancy cars, more branded degrees. My father told me that nowadays, it’s not enough for a bachelor to have a home in a Tier 1 city as the girl demands a 100 sqm home .. arms race.
— healthcare inflation.. my parents once said that they need a RMB 400k medical fund. If some treatment exceeds that then they would give up and say No to the 无底洞.
China consumers (like my mother) often favor imported products and newer products, often for valid reasons:
- imported products are sometimes superior in some ways, but often not much or not relevant
- older products sometimes hit drug resistance
For some common conditions (like cancer), I think there is “always” more fancy, more expensive, newer solutions, just like the new techniques in latency arms race. Some of these solutions cost a bomb but offer marginal benefits (some observers may say diminishing return). My mom gave examples about extending life by months. If those extra few months cost $1M, then it’s rational to ask “Where can this million dollars have more impact?” These /advanced/fancy/elective/ solutions contribute to medical inflation, but this is CPIx inflation and often driven by exclub, not livelihood.
eg: Nursing home is no “medical” facility but /healthcare/ facility. The minimum cost is very low, as proven in many Singapore community nursing homes. However, my parents in Beijing seem to want something much better. Indeed there are very fancy nursing homes asking RMB 10k/M or more…. smells like arms race.
eg: medishield coverage for private ward.. costs a lot more than the basic insurance that covers B2 ward. Many in my cohort prefer the more luxury insurance, sometimes without due diligence. Some of them probably think they can afford the more expensive insurance, and perceive private ward as important. I think this mentality contributes to medical inflation, driven by exclub, not livelihood.
In many cases, if a consumer doesn’t exercise critical thinking, then her worry about healthcare inflation boils down to blind_FOMO, not livelihood.