I think I spent like $15 so if it doesn't pan out no big loss. But it does seem like it would be a decent print server. And then I can set up software defined radio because why not.
$15 for the specs (when comparing to the Pi) is a price point I can see them hitting though (if they sign the right deals and get enough orders in for bulk buying, which is looks like they have). I'm not in much doubt that they'll deliver the product. What I do think is that the computer's price is being subsidized by shipping and accessories.
Though they'd be nice for 'smaller' tasks. I have had Pis running multiple tasks which I guess the CHIP would be good for spreading out.
The WiFi chip is nice I guess, though I would want a ethernet port ideally.
Another thing, they say you can browse the web on their promotional video and their 'has tons of apps' section is em, aspirational to say the least.
Chrome on those specs? Dream on.
I guess they got a stock image for each and you can run programs in those categories, not necessary the programs that the images depict, though that's one of the three things I have doubts about with that Kickstarter
I had my mouse pointer over the Kickstarter funding button for a while, decided against it for these reasons. So waiting for general sale.
A Few Forum Topics
https://forum.teksyndicate.com/t/getting-started-with-your-raspberry-pi-2/76316?u=kai
@Kat you might like this one as it walks you through inital setup.
I originally posted this one on the Tek Syndicate blog where it got a fair bit of attention (I think it was @wolfleben's idea to post it there originally?) however I reposted it on the forums after I had my account wiped.
https://forum.teksyndicate.com/t/raspberry-pi-firmware-upgrade-command/81248?u=kai
After owning Pis for multiple years I actually never realised there was a dedicated firmware upgrade command. This thread details the commands to use to update your Pis package lists, packages and firmware.
Also GCHQ did this pretty damn nice Raspberry Pi cluster 'mini supercomputer' at the NEC in Birmingham.
This may be an absurd question but Gnome based distros are out of the question with the Raspberry Pi 2?
You can get Ubuntu MATE which is basically GNOME 2.
Pidora is a good choice if you'd like a dock:
Ubuntu MATE for Pi 2 is a bit sluggish though I only tried the first build. Haven't tried Pidora yet.
I'd start with Rasbian though.
1 Like
When I get around to it I will do a guide to doing SDR on a Raspi cause who doesn't want to listen to police/fire and airports?
So this has been along time coming. I got a Pi 2 pretty much immediately after launch and I was very curious to see how the new Quad Core Pi stacked against its older slower brother.
Notable changes between the Model B and 2 are:
- Switch to an Cortex A7 Quad core from the ARM11 (Along with stock clock bump)
- Switch to 1GB of RAM from the 512MB.
Not many, but enough to warrant testing the throughput. As these were the two major changing points, this is what the tests focused on.
First up's was testing the Pi Processor. The major thing here was both the updated underlying architecture along with the addition of MOAR CORES!!!! Other than that the SOC package is unchanged.
Prime Number Test
(Please Note: These test were done synthetically as it was the only reliable method of testing both platforms. Both Pi's were tested with the same version of Raspbian, running headless with as little memory dedicated to the GPU as possible. All tests were repeated three times with what being shown here being the average)
Sysbench was used (0.4.12) with the CPU being tested with the following parameters.
sysbench --test=cpu --cpu-max-prime=2000 --num-threads=1 runA single threaded test to show the differences at an architectural level. Calculating the prime numbers between 0 and 2000 (as I did not want to be here all day)
=== Model B ===
- Total Time: 54.51s
- Min: 5.35ms
- Avg: 5.45ms
- Max: 11.93ms
=== Model 2 ===
- Total Time: 31.86s
- Min: 3.17ms
- Avg: 3.18ms
- Max: 10.61ms
This works out as the newer Pi being 71% faster core per core than its predecessor! Talk about bang for the buck. Not only is it faster to finish, but it is also faster to respond and process with the minimum time taken on each event in the test and the average time taken falling by 2ms!
So what happens if you really let it run (with all four cores).
sysbench --test=cpu --cpu-max-prime=2000 --num-threads=4 runSame params, just enabling all four cores.
=== Model 2 All Cores ===
- Total Time: 7.97s
- Min: 3.15ms
- Avg: 3.18ms
- Max: 5.15ms
Impressive! The additional cores certainly help the bench run with the max time taken also falling drastically. (584% faster than the Model B if you want a percentage)
Thread Scheduler Test
The thread test is more of a scheduler test than anything and tests its handling of threads. Read the full Sysbench documentation if you want a better insight to it. But to put it simply, its how many things the CPU can handle and how well it can switch between them all.
sysbench --test=threads --thread-locks=2 --max-time=20s runThe params above test to see how many threads it can successfully spawn, run, lock and finish within 20 seconds.
=== Model B ===
- Total Number: 6825
- Min: 2.75ms
- Avg: 2.93ms
- Max: 4.84ms
=== Model 2 ===
- Total Number: 9632
- Min: 2.01ms
- Avg: 2.07ms
- Max: 3.49ms
As can be seen, there is a pretty vast improvement between the old and new.
Memory Write Test
So the memory test is simply to see how fast the memory will write, I stupidly did not test the read however I will update this when I actually have my Model B to hand.
sysbench --test=memory --memory-total-size=10G runOriginally I planned to run this with 10GB worth of writes.... however the Model B crashed out during the test and could not actually complete the run. So the test value was lowered to 2GB
sysbench --test=memory --memory-total-size=2G run=== Model B ===
- Operations per Second: 89790.84
- Transfer Rate: 87.69 MB/s
- Total Time: 23.36s
- Total Events: 2097152
=== Model 2 ===
- Operations per Second: 186396.8
- Transfer Rate: 182.03 MB/s
- Total Time: 11.25s
- Total Events: 2097152
Each event in the operation equals 1KB of information. Moving to 1GB of RAM has increased the write speed by over a factor of 2x.
Storage Test
I did test the storage, however as the rest of the SoC in both the Pi's is pretty much the same. I was doubt full that anything would change. Turns out I was correct. Using two write tests (one writing all zero's, the other doing random). The zeroing averaged 9.5MB/s on both Pi's with the Random writes dropping to an average of 1.7MB/s. The Reads were much zippier at and average of 18.5MB/s on both Pi's.
This is probably the exact reason that other devices like the CubieBoard have SATA connectors so they can use much faster drives. The only place where this may become a major issue with this device is media center applications.
Model B vs 2
There you have it, a pretty quick and rough benchmark comparison between the old and new Pi's. So if you were looking for a reason to pick up the newer one, here you go. If you have any other micro computer (BananaPi, CubieBoard, BeagleBone....etc) feel free to compare.
2 Likes
They are a wonder box for Hackers :)... I love them... in fact hackers love them because they can setup a device thats small and easy to buy as a burner.. where they can just have it mine packets off of a wifi network or target network.. for weeks or months and crack its encryption in two seconds at home... and boom.. they got it and they can do man in the middles.. I know thats a dark hat use of it.. but hey it is a powerful little device
disclaimer: I have not done such attacks.. I am merely presenting this for educational purposes
1 Like
I actually used my RPi for two projects. The first project I ended up turning it into a DHCP Server for a LAN Party that I was running. For the other project, I put retro pie onto a seperate SD card to play some old school games.
Erhmmmmmm...I wouldn't use a Pi as a router...period
The Pi 2 maybe but not the original Pi
He asked if it were possible. I said it was, i never said he should though lol
They're perfectly fine for router on a stick applications, but they tend to be quite slugish when dealing with firewall duties.
I just mean since the USB/Ethernet share the same controller, the bandwidth will be limited (don't expect 100mbps constantly). :P
They ARE nice if you just need a "router on a stick" I've just seen too many people assume they'll rival a cheap Celeron box in performance as a "real" firewall/router
yeah this isn't really going to beat a PFsense router. at all. but it's good enough for basic use or if you made a tiny cluster it could be theoretically possible.
This is exaly very interessting stuff.
i'm playing with skyjack and taking down drones that fly near my house. It deauth owner and auth my raspberry pi as new owner ; from there i use node.js to create http java based window with controls - from there i can pretty much control whole thing, and kill it or steal it ;D
I plan to go on some RC plane airport or something or drone show ... then shut them all down.
Wow CyklonDX
So you like breaking others toys little brat.Hope your plan fails .Yours parents are on vacation ?
You seem to be very creative and destructive. Grow Up
1 Like
If you didn't put RC in that sentence you'd be on a watchlist.
Either way - don't be a dick...
Hey you didn't show the raspberry pi 2?