Data ports

הרשומה הזו היא באנגלית בעיקר כי הכתיבה באנגלית היא שריר שכבר מזמן לא תירגלתי. כשהתחלתי את הבלוג הזה, לפני כמעט אלף רשומות, החלטתי שהוא יהיה בעברית למרות שזה מצמצם מאד את קהל היעד, כי אז, אחרי עשרים ושתיים שנים של עבודה בחברה אמריקאית (בישראל), כתיבה בעברית הייתה שריר שהתנוון אצלי והייתי צריך לתרגל. הכתיבה באנגלית הפעם היא גם כי הנושא הספציפי הזה הרבה יותר טבעי לכתבו בשפה האנגלית. זו השפה של התחום ואין מה לעשות. אז סליחה ממי שקשה לו בלע"ז.

Lately, I've been introduced to the term "dark silicon"' – a term denoting the area in an integrated device ("chip" from now on) that has to stay "dark", that is – inactive, while other parts of the chip area are active, due mainly to power dissipation constraints in the modest thermal envelope of modern computers.

That “dark silicon” is one way of describing the lack of scaling from one process generation to the next because the transistor count is growing faster than the ability to run all of them concurrently at full speed (or at all, sometimes, due to leakage) without them overheating.

Back in the day, when I was doing computer architecture research, I had an idea for something I termed “data ports”, which now, in retrospect, I think might have been able to provide extra scaling for processors. 

I adopted the idea from the domain of special-purpose accelerators and goes something like this:

Build, on the side of the processor, a configurable data pipeline that delivers data to the processor’s execution units and takes the data computed by these execution units and writes them out.

The idea is that for code in compute loops, a few instructions in a preamble to the loop will set up a few control registers, for example, the starting memory address for the data that is input to the loop and the address “stride” in-between consecutive data elements. Following that setup stage, the data pipeline starts to run for a preset number of iterations, without any instructions being repeatedly fetched, decoded and executed by the processor. The processor is mostly idle while the loop is running.

This kind of computation paradigm required new code to be generated by a compiler, or by hand. New code, especially one for a new compute paradigm, is something that Intel, my employer, was not fond of. Especially after the phenomenal flop of the Itanium series of processors that were based on a new architecture handed down from the HP labs, in which the compiler, rather than the processor, was supposed to expose all the parallelism in the code. This was perhaps the main reason the idea did not take hold, back then. Also, 20 years ago, Dennard's law still was believed valid – the fact that operating voltage does not go down at the same rate as the device size was not yet internalized to the degree it is today.

Those data ports were supposed to be very efficient in power for code in loops and could have addressed a wide range of applications, especially in which the data is organized in vectors or matrices. However, with the current understanding that all the silicon cannot be used concurrently anyway, a setup in which these data pipelines "waste transistors" and have their own separate set of execution units, and do not use the main processor units, makes more sense.

So – while executing code in loops, the rest of the processor can cool down. Thus – the surplus transistors in that “dark silicon”  are being put to good use that way, with power being dissipated at different areas of the chip for different types of code.

As I'm writing this post, I realize that I still have hope that someone will re-invent this idea and implement it. Perhaps someone already did – I wouldn't know – for the last 11 years I've been mostly reading philosophy and have lost almost all contact with the state of the art in the field of computer architecture. Oh well…

att: The Google maps development team

I'm writing this blog post in English, on the off-chance that someone at Google might "google" this post, and communicate its contents to the team at Google developing google maps. (or google photos, or both)

This post is about (what I view as) a missing link between my geo-tagged photos and google maps.

Every now and then, when I take a picture in my phone, google maps pops up and asks me if I'd be interested in adding this photo to google maps. I assume this happens mostly in places that lack photographic coverage in the image base that google maps is building over time, for as many locations as possible, perhaps hoping to get coverage of the whole globe eventually. A worthy goal, no doubt.

Google could do it simpler and better. Because all the while I, as well as billions of others, have a huge stock of geotagged photos just sitting there, that could be harvested for that same purpose. On the other hand, google maps could add a lot of functionality to the management of people's photos, and travels. Here is how I view this:

For photos stored locally: Provide a way to display the locations of   geotagged photos on google maps. Picasa, now defunct, used to be able to do that on a map insert, but its link to maps seems to be broken. And it needs to be implemented on a full-scale map, not an inset.

For photos in Google photos: Provide a way to add locations or edit the estimated locations google photos found, using a map interface
Also, google photos currently clusters photos based on geotagged location, but does so in a very clumsy way, so that sometimes "concentric" clusters are formed with the same photos. Doing it with a map interface would solve these issues easily.

Search photos by map:  Google maps allows for searching places such a hotels around a specific location. Why not allow for searching by location my own photos, stored locally or in the cloud? Useful when I want to find those photos I took in that village in Ethiopia, but I can't remember when I took it or where I put it.
Feature should allow for specifying the search radius.

Auto geo-tag: Find photos that are not geo-tagged but photographed at the same time, and thus, presumably the same place, as photographs that are geo-tagged, and allow for automatically geo-tagging them. Typically the later would be photos taken by a phone while the former would be higher quality photos takes by a camera. One gets a much richer and better stock of photos that way. Perhaps "photos" already does that to some degree, but it's unclear how these "estimated locations" are generated.

Editable location history: I maintain my own "loaction history" in a document, but others may have other ways of knowing where and when they went. If I could edit the location history google maintains for me, add, remove and correct locations, it would also help geo-tag photos, especially from earlier years, where there is not machine generated location history.

Adding the above rather simple features, "maps" gets a huge stock of photos that it could ask users to add. Users get a significant usability boost from these features/ Perhaps all it takes is just a  better synergy between "maps" and "photos".

And while I'm at it: wish I had an easy way to plot my travels on google maps. To create private maps  I could share, with dates, photos and some text associated with geographic locations. How about it?

Perhaps also auto-include travel data harvested from Gmail, but added optionally and modifiable, not like those calendar entries that show up in my calendar automatically, entries I have no way of modifying, even when they miss.

busy

Lately, I happen to run into quite a few people I used to worked with, at Intel corporation, that have been separated from Intel in a “voluntary” manner over the last year. This, as part of an Intel effort to reduce its HC, especially the costlier (and experienced) part of it. Most are in their fifties, a few older, all of them worked hard and contributed a lot over the years, but became too expensive to the company. Some of them started appearing in Facebook, some in the "real world" outside the Intel's Haifa industrial park campus.

Almost in all these meetings, one thing stood out: the need, sometimes desperate need, for all of them to emphasize how busy they are, how hard they are working, either for no income or for way less than they used to earn. This is more or less the first thing they tell about themselves. Maybe this is not so odd.

At work place like Intel or in the high tech jobs, it is forbidden to be viewed as non-busy person. Just sitting in the office pondering things, thinking about stuff, work related or otherwise, is a signal for idleness. You always need to look as if you are in a middle of a thing. Preferably an important thing. And you shouldn’t  be distracted. In a meeting, it is ok for you to sit and ponder things if you manage to do that while you pretend to be interested in the meeting subject or answering to very important email. The key word is: I am busy.

And then you stop working, and it implies that you don’t need to look busy anymore. But this vacum is scary. All of a sudden the entire day is handed to you free of any restrictions. There is not a single meeting in your calendar. Your inbox has gone from 100s of "very important" emails that you need to read and even answer some, to near empty. Facebook is only a partial short-term answer, as it takes time for a person to convince (fool?) himself that Facebook interaction or blogging are worthy substitutes to invest time in.

As an interim measure, for a few weeks/months, you can be really busy fixing and organizing things around the house. Maybe digitizing all those analog photos, maybe fixing that lamp that broke down half a year ago. Everyone has enough tasks like these to fill the initial spare time. But this is temporary. Soon enough you realize that you cannot fill your life with boring maintenance chore, not even by maintaining your body (gym etc).

School/study is a good answer for a while until you exhaust this option or get tires of it. Besides stimulating your mind, lectures are given at fixed times and locations, and they do provide a good way to fill up an empty calendar. If you also follow the off classes assignments then it further fills your calendar.

Parkinson's law: "work expands so as to fill the time available for its completion" may help. But this could be dangerous and you could easily deteriorate as a person.

So what is the correct answer? Being not-busy.

It is ok not to be busy (or not busy is ok). Memorize this.

This why when people ask me (retired from Intel almost 6 years): “so what do you do?”, I usually answer: “nothing”.

It is a very similar answer I give as a vegan person to the question: “so, what do you eat?”, I usually answer: “Humus”.

I answer that way, as I don’t see a point to detail the rich variety of food I can eat and do eat as a vegan. The same goes to the answer to the “what do you do?” question. I don’t need to detail the incredibly rich tapestry of activities that I, as a non-busy person, do.

I am not going to list them here either. Hebrew readers may be able read my other blog posts (almost all are in Hebrew) to get a better idea. But I would proudly mention one activity: I take a 40 minute afternoon nap in most weekdays. I am not busy, and I like it.

———————-

This blog post was translated to English by a friend, a former co-worker and a proud non-busy person these days. Apparently this post resonated with quite a few people in the same phase in life as me, trying to cope with the anxiety of not being busy.
The Hebrew version is here.

Barrel distortion

A regular stop in the tours that the settlers take you through the occupied territories (Judea and Samaria to them) is "Sharon's balcony". From this vantage point on the western side of the Samaria hills range, one can see "all of Israel", the whole coastal plain where most of the population lives, all (both) of Israel's power generating stations and Israel's single international airport. The message is self-explanatory; it illustrates the strategic "control" that the Palestinians will have if Israel stops controlling these territories. This "control" is meaningful  however, only if a line-of-sight is actually required in order to launch a missile or fire an artillery shell, and that is something that may have last been  true in the times of the bow and arrow, less so today.

Sharon, Israel's former (now comatose) prime minister, had quite a few of these "balconies" to which he took foreign visitors and journalists. The one I was taken to, on two tours, was in Peduel, a settlement of nice private houses, whose inhabitants commute to work across the "green line" to the Tel-Aviv area, 30 minutes away.

There,  I noticed for the first time the collection of rusting  metal barrels strewn in the open area around the settlement. As I passed near some of these on the foot trail to the vantage point, I peeked in to see what's inside. In most of them, there wasn't anything to see. In some, one could discern the remnants of a dead seedling. None had any irrigation, drip or otherwise. No seedling was tied to a pole, and as far as I could see, most were either standing on rocks or on a pile of stones with precious little soil underneath. Not a setting that would allow for high survival  rates for the seedlings planted in those barrels.

At that time I thought that this was a remnant or a Tu-B'shvat tree planting ceremony. In these ceremonies, trees are planted to celebrate the Jewish day of trees but occasionally some of  the trees die later for lack of care. I also assumed that the barrels are there to prevent goats and sheep from eating the fresh seedlings, and the barrel as a whole may serve as a rudimentary drip irrigation scheme where the barrel gets filled with water, and the water drain slowly through holes in the bottom, but then I noticed that the bottom was completely missing in most barrels, so that option was unlikely.

Later in the tour I noticed similar clusters of barrels in many locations in the occupied territories, but I got around to ask about those only on a later tour, of the southern Hebron mountain, with "breaking the silence" – an organization of reserve soldiers committed to tell the rest of us what they were forced to do as part of the occupying army.  There I finally understood what "barrel farming" is all about, and it works like this:

Israel never applied the Israeli law in the occupied territories. The Israeli law is applied only to the Israeli citizens living in the occupied territories, the "settlers". A post (in Hebrew) about the anomalies this creates, is here.

The Arab population in the occupied territories is governed by a combination of military laws, enacted on a need-to basis, and the laws that were in effect in those territories prior to the Israeli occupation: the Jordanian law, the British Mandate law and before that, the law set by the Ottoman sultan. All these law systems are concurrently in force in the occupied territories, and Israel has the option to choose, in each case and for each situation separately, which law in that set of laws suits it best, or more accurately, suits the Palestinians less. And it does exactly that. A must-see documentary on exactly this topic is "the law in these parts".

The law for land ownership was pulled from the Ottoman law. An old Ottoman bylaw, designed to maximize taxes paid to the sultan, returns ownership of agricultural land back to the Sultan if it was not cultivated for more than 3 years, and on the other hand, gives land ownership to any farmer that has cultivated a plot of land more than 10 consecutive years. This applied to "Mawal" land, defined as  plots of land far enough from the village to that the call of a cock at the edge of the village cannot be heard in them. Gives you inkling as to when these laws were written…

This Ottoman law is interpreted very loosely when applied to Jews in the territories, and a lot more narrowly when applied to Palestinians living there.

If a Palestinian does not cultivate his land for three years, due for example to the IDF preventing him from doing so, on a pretext that the plot is too close to a settlement and cultivating it poses a security risk, he loses ownership.

On the other hand, if a Jewish settler places some barrels in a plot of land (as in the picture below) and perhaps even plants a seedling in each, if he keeps the barrels on the same plot for ten consecutive years, the land becomes his. Even if no plant survives. Even if he has not spent a single minute watering, providing fertilizer or taking care of the plants in any way or form. Even if he "planted" the trees not on a 3 by 3 meter grid as typically done in orchards, or even on a 10 by 10 meter grid. As few as 4 barrels per Dunam (1000 square meters) will typically do in order to fake cultivation of the land.  

The interpretation of this ottoman law is extremely liberal for Jews, but consider that Jews in the occupied territories are supposed to be solely under Israeli law, which does not support any form of Ottoman style land grab anyway.

So, why can't the Palestinians do the same, toss some barrels across a hilltop in order to gain ownership? Well, they can't. Definitely not in the 60% of the area under direct Israeli administrative and military control (area C). Their barrels will not survive a day. The settlers will remove them at gun point or paint them over and claim them to be theirs, again by force. The settlers are armed and have the backing of the army, the Palestinians can try to throw stones, but that's about it. Even if the barrels survive, and they manage to cultivate the trees, the settlers will likely cut the trees down on the first occasion they get, as they have done in olive orchards all over the territories.

The Palestinians will also not be able to water the trees. Water from wells is supplied only to Jewish farmers and if the Palestinians dig cisterns to collect rain water, these cisterns are considered "building without permit" and the IDF will send a bulldozer to plug them. Building permits are issued as a matter of course only to Jewish construction as the Israeli governor cannot spare the time, for almost 50 years now, to draw a plan for the Palestinians that would enable issuing building permits.

That's how it works.

The name of this post, "barrel distortion", is a term in optics. Camera lenses have it, and in modern digital cameras, post processing of the images can effectively correct such distortion. But no amount of post processing will be able to eliminate the "barrel distortion" in the images one sees daily in the occupied territories.

This post is an English version of an earlier Hebrew post. I thought it worthwhile to translate this one so that people who cannot read Hebrew can read this – I was that bothered by this cynical  and abominable form of "farming".