Ag Geek Speak

7.5 Tiny Bytes: From Bits to Petabytes: Understanding Computer Storage

• A Podcast for Precision Agriculture Geeks • Season 2 • Episode 7

Share episode

We explore the fundamentals of digital storage and how it powers precision agriculture, covering basic bytes to the massive petabyte servers at GK Technology that store agricultural imagery and data.

• A byte equals eight binary bits and typically represents a single character or piece of data in computer language
• Digital storage units progress from kilobytes (thousand) to megabytes (million) to gigabytes (billion) to terabytes (trillion) to petabytes (thousand terabytes)
• GK Technology's Halstead office maintains a petabyte of storage containing agricultural data for most of the United States and three Canadian provinces
• Storage includes shapefile formats, NAIP imagery, Lidar data, and two satellite libraries: Landsat (30m resolution back to 1984) and Sentinel (10m resolution from 2016)
• Precision ag software users access this massive data repository when creating drainage maps or automated management zones
• Higher resolution imagery like drone photos requires significantly more storage space, creating challenges for agricultural data management

https://gktechinc.com/

Jodi:

And now it's time for a tiny bite of knowledge In a world with 32 and 64-bit software.

Sarah:

Enter in the byte, that's byte with a Y Roar Roar. So, Jodi, what is a bite when we're talking about computers? What is a byte?

Jodi:

A byte is a measure of digital storage, and typically what this means is that one single byte is equal to eight bits.

Sarah:

And a byte usually represents a single character, such as a letter, a number in computer language.

Jodi:

And fun fact. So like these 8 bits. What those bits refer to are like the binary 0s and 1s, right, so like if you really break it down or you think about like the matrix, where they've got a bunch of 0s and 1s floating down on the screen, all those little zero and one binaries. Each one of them is a single bit and eight of those equals a byte.

Sarah:

Yeah, you commonly hear about bytes used for computer memory or storage terms. So I mean it's pretty often that those of us geeks and precision agriculture are sitting down and we're going, hey, how big is your storage there? Do you get a 64 gigabyte, one terabyte, two terabyte and over in Halstead we even talk about having a petabyte of storage on our servers for imagery data.

Jodi:

How many gigs of storage is a petabyte?

Sarah:

Let's start by backing up and talking about how much is a kilobyte? So you've got a byte, then you've got a kilobyte that's a thousand bytes. Then you've got a megabyte, which is a million bytes. You've got a gigabyte, which is a billion bytes.

Jodi:

And then a terabyte. A terabyte is-hmm, or one quadrillion bytes, and a petabyte is 1,000 terabytes 1,000 terabytes.

Sarah:

I mean, I'm thinking about like I've got a USB stick here that's sitting with one terabyte of data.

Jodi:

That's a lot of storage. You have a USB stick with one terabyte storage.

Sarah:

Mm-hmm.

Jodi:

Yes, I do.

Sarah:

That's amazing, isn't that amazing? It is amazing, yes, very amazing. Got a lot of stuff on there, lots of pictures, that's great, but it's still a terabyte, and a petabyte, you said, is a thousand terabytes. A thousand terabytes, that's a lot. Now, on that petabyte of storage, we actually have our data on the servers backed up. Yes, we do, but you got to realize, to put things in context, for what's over there for data, I think we're currently missing. What is it, Jodi? Eight states in the United States for data on those servers, something like that. So we have the majority of the United States on there.

Sarah:

We've got at least three provinces in Canada, and we're talking Alberta, Saskatchewan and Manitoba, so provinces that have rather large land masses to them. The data that is stored up there includes any shapefile formats that help show us where township sections, ranges, are located, as well as all of the, for example, nape imagery that we would have in the United States. That's a United States program. Unfortunately, in Canada you're not going to have that, but those are images that are flown once a year or once every two years. It's very detailed imagery and it takes up a lot of room, especially these newer images, as the as the newer imagery is more detailed and contain more bands of light than the old ones used to. We also have over there, you know, Landsat data that's 30 meter resolution, and those libraries go back to like 1984. And we've also got Sentinel data that's 10 meter resolution. Okay, so we've got a lot of storage, a lot of storage. You could even say we took a bite out of storage.

Jodi:

So, even though you and I do Tiny Bytes, the GK office has a big bite of storage. It's like a monster, monster bite.

Sarah:

It's a monster bite over there. Let's back up, though, and talk about that petabyte that's over there. So, as I was saying, we've got all of you know, we've got all of that NAIP imagery. In addition, we've got Lidar data. Some of that data, and we pretty much have got Lidar data for wherever our clients would have asked for it around the United States.

Sarah:

We also have two different libraries of satellite information Landsat data that goes all the way back to 1984. Those images were not recorded as frequently throughout the growing season and it's 30 meter resolution data, but it goes back to 1984. Whereas on the Sentinel data library we have more images in season, that library only goes back to about 2016, maybe 17, right in that range, and that data is 10 meter resolution data, so it's more detailed than that Landsat. We have more images, more detail, but fewer years for the Sentinel data and the Landsat library has a much longer history to it and less detailed and fewer images during the growing season and really, when you start thinking about all those imagery components, that really starts adding up to how much what is using the data on our servers over there and, in case you are an ADMS user, when you go into our servers to download data, to make drainage maps or to automate zones.

Jodi:

You are going into that server and getting those data out of there, so of course, like with the Landsat images versus the Sentinel, because Landsat is less detailed it's got a lower resolution, so 30 meters versus 10 meter resolution. It's going to take less bytes of storage for each of those Landsat images versus Sentinel Correct the more detailed the image is, generally speaking, the more data it's going to take to store them.

Sarah:

That's something we need to really think about in precision agriculture. So we think a lot about satellite imagery and those comparisons that Jodi's just making between Landsat and Sentinel data. One of the things that we need to think about is this push towards going to drones, because those pictures that are coming off of drones take up a pile of space. Yeah, we can run them in our software and and work with them in our software. That's great. But, boy, you put one image on a one terabyte jump drive and it doesn't take a whole lot, very many images to take up a lot of that space. So space and what exactly a byte is, what exactly a terabyte is? A megabyte, gigabyte, petabyte it's interesting to really think about and the challenges that come along with that storage. So this might be a tiny byte. So this might be a tiny bite.

Jodi:

Over in Halstad we have a petabyte which is a monster bite. Tune in next time for a tiny bite of knowledge from GK Technology, where we have a map and an app for that.