FinallyFoolin August '24

The beginning of August was NOT good but came back somewhat over the last part of the month. I haven’t been making many moves lately but August ended up being one with some changes.

YTD: 35.45%

Transactions:
The biggest change I made was increasing my allocation in NVDA. In the process, I sold out of BILL & Sentinel One. I also upped my Crowdstrike position and iniated a tiny Confluent position.

First week of month
Trimmed tiny bit off all other positions to add to NVDA position

8/16
Sold $BILL after earnings (2% loss), put the money into $NVDA & opened small position in $CFLT

8/21
$S popped after buyout potential rumors, sold half the position, put it into $CRWD & NVDA

8/23
After NVDA earnings, sold the rest of my Sentinel One and put it all into NVDA.

Nvidia:
I’m convinced that Nvidia is set up for much more growth than they’ve already displayed. I think there will be a new normal as far as GPUs to support AI but its going to be much much higher than today’s. I also believe that GPUs will replace a lot of the hardware currently in datacenters.

Their lead in this space is enormous. If they dominate this initial AI entry phase then their software becomes a big differentiator when/if competitors do get to a closer to even level.

54 Likes

Blockquote
“Their lead in this space is enormous. If they dominate this initial AI entry phase then their software becomes a big differentiator when/if competitors do get to a closer to even level.”

NVDA started life as a video card chip company. Graphics math requires lots of matrix math. Matrix math implements combinations of multiply/add operations in fast loops. The graphics accelerators optimize that kind of math. Artificial Neural Networks (computer models of brain cells) use the same kinds of matrix math operations. The AI research community discovered they could use those same cards and chips to accelerate their computations. NVDA helped with a programming model CUDA that has become a standard in that computing space.

23 Likes