Carl Victor Page, Jr. Updated 2025-07-16
Larry Pages's older brother.
It is hard to find information on this little bugger! Not a single photo online!
As suggested by the "Jr.", he is named after Larry's father, Carl Victor Page.
Carl Jr. is mentioned in a few places in the book The Google Story. The full name "Carl Victor Page Jr." is never given in that source, only "Carl Page Jr." is used. These crazy Anglo-Saxons and their semi-optional middle names!
The Google Story does not cite its sources, but it likely got much of its insider information through interviews, e.g. Chapter 2. "When Larry Met Sergey":
Carl Jr. recalls Larry as an inquisitive younger brother with wide-ranging interests
which suggests the authors actually interviewed Carl Jr., since interviews with Carl Jr. cannot be found anywhere else on the Internet. It would be interesting to know more how they got that level of access.
Chapter 2 mentions that Carl Jr. is nine years older than Larry. Therefore, he must have been born in 1963 or 1964. It also states that Carl studied at the University of Michigan, like his father and like Larry would also do later on:
He also enjoyed helping Carl Jr. - who was nine years older - with his college computer homework when Carl came home from the University of Michigan during breaks.
Their father was a professor at the Michigan State University, which is a different university from the University of Michigan, and not in the same city, so by breaks they mean term breaks.
Chapter 2 also mentions that he was working in Silicon Valley by the time their father died in 1996:
Despite his grief [for the death of their father at the early age of 58], Larry remained enrolled at Stanford. It helped that his older brother, Carl Jr., lived and worked in Silicon Valley. They had each other, so Larry wasn't left to bear the loss alone, and the two spent time together, fondly recalling their dad and reflecting on their childhood memories.
In 1997, Carl co-founded the mailing list management website eGroups together with Scott Hassan, programmer of an early version of Google when he was a research assistant at Stanford University. Carl and Scott presumably met through Larry, but we don't have a source. The company was sold to Yahoo! in 2000. The Google Story Chapter 8. "A Trickle" mentions:
Google's deal with Yahoo!] had special significance for Larry Page, since his brother, Carl Jr., also was in serious negotiations with Yahoo! over a major business transaction. The following day, June 27, Yahoo announced plans to buy eGroups, a technology firm that Carl Page had co-founded, for $413 million.
Carl is listed as a co-founder in the SEC filing: www.sec.gov/Archives/edgar/data/1105102/0000950149-00-000584.txt as "Carl Page". He does not appear on the 5% stockholders however, poor Carl.
In 2006, he brought a company he founded called "Handheld Entertainment" public through a reverse merger with a shell company: archive.nytimes.com/dealbook.nytimes.com/2006/03/21/brother-of-google-co-founder-uses-shell-company-for-handheld-start-up/. "Handheld Entertainment" made an iPod competitor apparently. SEC filing: www.sec.gov/Archives/edgar/data/1309710/000095013606009480/file1.htm.
September 27, 2023 marked Google's 25 th aniversary and the page cirosantilli.com/carl-victor-page-jr had a small surge of views according to Google Analytics. On that day, this page was one of the top Google search results for "Carl Victor Page, Jr."[ref]. Wikipedia also had a large bump in searches for "Larry Page" on the same day: pageviews.wmcloud.org/?project=en.wikipedia.org&platform=all-access&agent=user&redirects=0&start=2023-09-11&end=2023-10-01&pages=Cat|Dog|Larry_Page which must be the root cause, Larry actually managed to beat "Cat" and "Dog" on that day.
Just like the adenine nucleotide translocator moves ATP/ADP in and out, this one moves loose phosphate in.
Both of those together recycle the cellular respiration carriers from/to the mitochondria.
Run MLperf v2.1 ResNet on Imagenette Updated 2025-07-16
Let's run on this Imagenet10 subset called Imagenette.
First ensure that you get the dummy test data run working as per MLperf v2.1 ResNet.
Next, in the imagenette2 directory, first let's create a 224x224 scaled version of the inputs as required by the benchmark at mlcommons.org/en/inference-datacenter-21/:
#!/usr/bin/env bash
rm -rf val224x224
mkdir -p val224x224
for syndir in val/*: do
  syn="$(dirname $syndir)"
  for img in "$syndir"/*; do
    convert "$img" -resize 224x224 "val224x224/$syn/$(basename "$img")"
  done
done
and then let's create the val_map.txt file to match the format expected by MLPerf:
#!/usr/bin/env bash
wget https://gist.githubusercontent.com/aaronpolhamus/964a4411c0906315deb9f4a3723aac57/raw/aa66dd9dbf6b56649fa3fab83659b2acbf3cbfd1/map_clsloc.txt
i=0
rm -f val_map.txt
while IFS="" read -r p || [ -n "$p" ]; do
  synset="$(printf '%s\n' "$p" | cut -d ' ' -f1)"
  if [ -d "val224x224/$synset" ]; then
    for f in "val224x224/$synset/"*; do
      echo "$f $i" >> val_map.txt
    done
  fi
  i=$((i + 1))
done < <( sort map_clsloc.txt )
then back on the mlperf directory we download our model:
wget https://zenodo.org/record/4735647/files/resnet50_v1.onnx
and finally run!
DATA_DIR=/mnt/sda3/data/imagenet/imagenette2 time ./run_local.sh onnxruntime resnet50 cpu --accuracy
which gives on P51:
TestScenario.SingleStream qps=164.06, mean=0.0267, time=23.924, acc=87.134%, queries=3925, tiles=50.0:0.0264,80.0:0.0275,90.0:0.0287,95.0:0.0306,99.0:0.0401,99.9:0.0464
where qps presumably means "querries per second". And the time results:
446.78user 33.97system 2:47.51elapsed 286%CPU (0avgtext+0avgdata 964728maxresident)k
The time=23.924 is much smaller than the time executable because of some lengthy pre-loading (TODO not sure what that means) that gets done every time:
INFO:imagenet:loaded 3925 images, cache=0, took=52.6sec
INFO:main:starting TestScenario.SingleStream
Let's try on the GPU now:
DATA_DIR=/mnt/sda3/data/imagenet/imagenette2 time ./run_local.sh onnxruntime resnet50 gpu --accuracy
which gives:
TestScenario.SingleStream qps=130.91, mean=0.0287, time=29.983, acc=90.395%, queries=3925, tiles=50.0:0.0265,80.0:0.0285,90.0:0.0405,95.0:0.0425,99.0:0.0490,99.9:0.0512
455.00user 4.96system 1:59.43elapsed 385%CPU (0avgtext+0avgdata 975080maxresident)k
TODO lower qps on GPU!
Con of superconducting qubits Updated 2025-07-16
Optical table Updated 2025-07-16
For example, that is how most modern microscopes are prototyped, see for example Video "Two Photon Microscopy by Nemonic NeuroNex (2019)".
This is kind of why they are also sometimes called "optical breadboarbds", since breadboards are what we use for early prototyping in electronics. Wikipedia however says "optical breadboard" is a simpler and cheaper type of optical table with less/no stabilization.
Video 1.
A simple refracting telescope built on an optical bench by plenum88 (2013)
Source.
Arthropod Updated 2025-07-16
David Chaum Updated 2025-07-16

There are unlisted articles, also show them or only show them.