Semi back in a vengeance 
@manch pick of AMD crushes @Jil pick of BYND🤗 6.77% only narrowly edge out MU
1 Like
This interview with Jim Keller is awesome. The host doesn’t know much about computer architecture and asked some pretty basic questions, but that makes it accessible to people not familiar with CPU designs. Keller is the main architect who defined the 64 bit extension to x86 and a legend. He gave some pretty insightful big picture of where computer is heading.
Can you summarize in a few lines what did you learn from this video or at least point to the part you think is the key part of this video so that those who do not have an hour to spare can benefit?
1 Like
The part that struck me the most is the reason he gave why Moore’s law has a lot more headroom than most people think. Under the umbrella of Moore’s law it’s really a collection of different technologies, each with its own S curve. Some subset of these technologies reaching maturity doesn’t preclude another set just barely ramping up. Each silicon device is now about 1000^3 atoms big. We can probably get down to 10^3 that means we still 1Mx scaling left to go.
That’s just one of the things he talked about. Watch the whole thing.
Two chip stocks are among the winners. Micron on strong earnings and a better-than-expected forecast. Xilinx on announcement of expecting higher revenue.
Unfortunately, is an underperformer compare to WFH/ cloud stocks. I have MU 
barely recovered.
Another good interview. This time it’s David Patterson, co-creator of RISC architecture and creator of RAID. He talked a good bit about the CISC vs RISC history and pushed back against Jim Keller’s claim that Moore’s Law is not dead. Patterson is firmly in the camp that Moore’s law is rapidly slowing down.
Meanwhile MU is struggling
It is really hard to judge generic timeline how long. I do not know about SOXL too, but providing some visibility with UPRO.
If someone bought UPRO on Dec 24,2018, the good sale time was Jan-feb 2020.
If someone bought UPRO on Mar 23rd 2020, they can continue to hold until bull run is done.
If someone bought last week, their holding time may be less 2-5 days.
It is basically the swing percentage that we need to see. Additionally, common investors it is pretty hard to see the swings. It depends on various factors, the ETF holdings, volatility and the time or buy.
BTW: This is not stock advice or financial advice, but for discussion purpose only.
Notice SOXL is more volatile than both UPRO and TQQQ? Today +8-10% trumps UPRO and TQQQ.
1 Like
I can not easily jump on different ETFs as I need to buy in all-in-one reliability issue. Not losing single $ is very important. I am just staying with 90% UPRO and 10% TQQQ.
1 Like
What? You are in the slowest ETF?
Intel is having huge problem with its 7nm process. Their CEO, who looked really old and tired on the earnings call, said Intel may need to outsource fabbing to TSMC, which is on track to manufacture 3nm in volume in two years. It’s already going full steam on 5nm.
We may see a day when there are no US firms capable of advanced semiconductor manufacturing. TSMC’s only real rival is Samsung.
There are huge geopolitical consequences to this “silicon is the new oil” era. If China invaded Taiwan, the US would have no choice but to go to war with China. We need to protect our semiconductor supply, just like we went to war in the Middle East to protect oil supply.
All the stocks/etfs are cyclical. Anything goes up will come down one day. We can not assign huge amount blindly as one mistake makes our life hell.
When more money involved, better to seek safety than risk. It is like driving airplane vs car.
The best is to stay broad USA market. I am not giving weight-age for TQQQ too, but staying with UPRO as this SPY broad market.
Above all, this is true => It is hard to beat S&P in the long run. I strongly believe this axiom. We need exceptionally clear edge to win over S&P.
I should be extremely careful in choosing the best ETFs to beat S&P.
[edit] I am still trying with TQQQ (10) and will also monitor SOXL(5%). If I make some good progress, i will update after 3 months.
Hand waving? In politics, only benefits matter. There is NO benefit in invading Taiwan. Case closed.
There is no benefit to the Chinese state, but there is personal benefit to Xi. China is on a war path with the West under Xi.
Very sad Intel lost the Moore’s Law race while Moore is still alive! What a shame!
