Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Jul 2, 2024 1:56:04 GMT -8
Maybe I should have titled this “New Old Technology”. I came across this and thought that it was interesting. Who knows where this may lead. Why This New CD Could Change Storage Checkout the comments section.
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Jul 9, 2024 2:43:14 GMT -8
|
|
4aapl
Moderator
Posts: 3,845
|
Post by 4aapl on Jul 12, 2024 17:41:22 GMT -8
I expect Nvidia as a company to do well for a while.
Nvidia as a stock, I am less sure about. Given the huge gains recently, I'd be a little worried if only expecting a 16% upside, because a retrace of part of these huge recent gains would be much much more than that.
But I was worried about it back at $400-$500, when my son bought a share in his Roth. That has gone up about 200% in 6-9 months. Hard to complain, but also hard to expect things to continue at that rate.
Nvidia has amazed people and investors, and maybe that will continue. But I believe in Apple's strategy, and feel there is a good chance they will outperform the market and Nvidia in the coming 1-3 years, from this point where Nvidia has done much much better than AAPL in the recent past.
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Jul 13, 2024 2:29:37 GMT -8
Thank you 4aapl for your reply. I remember when everyone was predicting Apples decline because of the power of large numbers and that AAPLs price had to plateau at some point, but it continued to grow, and I’m thankful. But investing in Apple was easy for me because I have been a fan boy of the company and their products for a very long time, buying, using and giving as gifts. I understand their appeal. GPU’s are different for me, I lack that personal connection and understanding to Nvidia’s product line. But with Nvidia I see a company that has jumped so far ahead of their competitors that they should be able to dominate that space for some time. But like AAPL, I’m not afraid of pulling the plug on NVDA if I see it getting too risky. After all, I’m sure they would be willing to sell me some more share’s later if I were to change my mind.
|
|
4aapl
Moderator
Posts: 3,845
|
Post by 4aapl on Jul 13, 2024 5:47:24 GMT -8
Thank you 4aapl for your reply. I remember when everyone was predicting Apples decline because of the power of large numbers and that AAPLs price had to plateau at some point, but it continued to grow, and I’m thankful. But investing in Apple was easy for me because I have been a fan boy of the company and their products for a very long time, buying, using and giving as gifts. I understand their appeal. GPU’s are different for me, I lack that personal connection and understanding to Nvidia’s product line. But with Nvidia I see a company that has jumped so far ahead of their competitors that they should be able to dominate that space for some time. But like AAPL, I’m not afraid of pulling the plug on NVDA if I see it getting too risky. After all, I’m sure they would be willing to sell me some more share’s later if I were to change my mind. I think Nvidia will do great for at least a while. The potential problems are lowering profit margins as meaningful competition comes into the space, and lowering P/E as things change. It was hard to imagine change when AAPL was hitting highs in '99 and '00, seeing a P/E up at 60 and sometimes above. But not that many years later, after passing the market cap of big oil and taking the lead, but hitting some problems, they hit a sub-10 P/E. As for the law of large numbers, it's more of a guidance or reminder, much like a non-absolute speed advisory going into a corner. In that 100-50 market cap spot (of largest companies), and even 50-10 largest companies, you have the advantage of scale up but often don't have regulators and general lawsuits going after you. But in those top 10 largest companies you are very visible, leading the pack. Sometimes the politicians, foreign and domestic, try to find ways that you aren't doing things right. The patent trolls come out, seeing a big potential payoff. And the lawyers get hungry, deciding that a change in software behavior of a 5 year old iPhone with an old battery is bad for the consumer, by slowing down to prevent shutting down. Right now the tech market wants AI chips at any cost. It's not always going to be that way. But it's hard to guess at the timing of change, especially as others are trying to guess at the same thing, trying to sell shares before a drop in P/E. Like the fuel cell, EV, and solar companies, the good profitable ones make it through. While others (second time around for Fiskars, a story about bankruptcy says they are bulk selling ~3200 cars to a New York Uber car supplier, with 3 price tranches, good ones going for ~16.5k, others for 3.5k and 2.5k, for $46M total) get crushed or at least have some major issues along the way. Nvidia is leading the pack, and the competitive advantage and lead should keep them at the front for quite a while. It's on the stock side, taking into account future growth, profit margins, meaningful competition, market size, and P/E, that I wonder about. Of course in retrospect, I wish I bought at least a little 1, 5, or 10 years ago. The stock has outperformed just about anything. But like Lucky was saying, one doesn't have to feel there is just one winner. 8th place in the Marathon at the Olympics might not get the gold or even onto the podium, but it's still an amazingly fast time and the 8th best in the whole world.
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Jul 14, 2024 4:12:31 GMT -8
New Disruptive Microchip Technology ExplainedAdvances in technology continue to accelerate at amazing speeds. For me, this helps to explain a lot. About 18 minutes long with an advertisement in the middle of the video. Not sure about this woman's accent, I'm guessing eastern European.
|
|
4aapl
Moderator
Posts: 3,845
|
Post by 4aapl on Jul 14, 2024 9:19:11 GMT -8
New Disruptive Microchip Technology ExplainedAdvances in technology continue to accelerate at amazing speeds. For me, this helps to explain a lot. About 18 minutes long with an advertisement in the middle of the video. Not sure about this woman's accent, I'm guessing eastern European. She explained about 4 different important things, along with a lot of basics that flew right past me. The biggest is that memory, especially cache, isn't scaling as fast as other things, yet is really important. Chiplets help with this, basically throwing a different chip on top of another chip or board. And this helps keep the chips a little simpler, while letting the add-on chips be made with older technology if appropriate. She then mentioned a new technology from researchers at Stanford. My guess is that puts it very far out in the timeline. I wonder how much has been found in creating a memory that isn't just a 1 or 0, but has more states. 3 or 4 options would be a huge gain. If utilizing the crystalline structure and the underlying resistance, as in that new technology, then maybe there is something that has multiple usable crystal structures. I only remember a little of this about metals from material science, or the use in various cooking (tempering chocolate, making things like caramel or ganache). If it were simple it would already be done, but it would be a huge aim in memory density.
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Jul 28, 2024 2:32:14 GMT -8
|
|
4aapl
Moderator
Posts: 3,845
|
Post by 4aapl on Jul 28, 2024 13:45:49 GMT -8
Fun stuff! A few years back you might wonder about someone getting so excited about a Hybrid. But those cooling ideas for battery cells looked more like a nuclear reactor. And like many technical areas, the comments had even deeper info. Heat kills, along with power surges. At least that's my basic knowledge of high powered electric motors we have had that have failed, like pumps, drills, fans, and a buffer. And it's similar to a lot of electronics that eventually give up the ghost, like our cable modem that went out yesterday.
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Jul 30, 2024 1:22:05 GMT -8
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Aug 21, 2024 4:45:14 GMT -8
|
|
4aapl
Moderator
Posts: 3,845
|
Post by 4aapl on Aug 22, 2024 9:03:59 GMT -8
Good stuff, though long. The basics came down to backside chip powering (so a separate maze of traces for power), and 3d traces/parts, first as rectangular channels on the layer "floor" (so 3 sides instead of mostly just one), and then fully encapsulated rectangular channels (so 4 sides). TMSC has already done the backside powering and is adding the 3d stuff. Intel is behind, so is trying to catch up by adding both things at once. That's their "secret plan". And something about Intel's Fab 14, scheduled for 2027. Plus "Angstrom" level deposition, throwing down a single layer of molecules, but the machinery is very expensive and it is slower so would be the rate determining step. TMSC is skipping it for now, and Intel is looking at another method that basically puts a thin layer of polymer down that when it dries it lines up (crystalline pattern), but they also have a way to make it not all line up. Lots of info there, and I have a hard time following her, both due to the accent but also not really knowing nor being current on the semiconductor fabrication or design fronts. But her videos do a good job of presenting it well enough that I can make some sense of it....even if I know I have the terminology wrong. Thanks for the video! (Back to a previous video on bonding multiple things together, in general but especially if they had varying technologies. One cool thing about that is the yields (basically probabilities) and how they would add up. If both parts had a 95% yield, and there was perfect bonding with no failures, they could still have a 95% yield overall, or 19 out of 20 parts. But if they were further steps in the same process, then you'd have a yield of .95 * .95, or .9025. Basically a 90% yield, or just 18 out of 20 parts. That's a huge difference!)
|
|
Dave
Member
"It's tough to make predictions, especially about the future." Yogi Berra
Posts: 4,313
|
Post by Dave on Aug 22, 2024 13:09:04 GMT -8
Another video from Anastasia in Tech. AI Reality Check. What’s Next for AI and ComputingAbout 11 minutes long. Interesting that one of her sponsors is Babbel. Also, I've found some very old silicon wafers on eBay fairly cheap. I may pickup one just for fun.
|
|