Anyone got a Snickerdoodle yet?

[djsb]

I've been having a quick look at these.

https://www.crowdsupply.com/krtkl/snickerdoodle

and I'd be interested to know if anyone has received theirs yet?

Is it worth the risk of me pre-ordering a Snickerdoodle black?

[agehall]

There are people claiming to have thiers on the Krtkl-forum. I'm still waiting for mine even though.

They do claim they've had production issues that were just resolved a week or two ago, so who knows - maybe they will actually start shipping soon.

[Zermalmer]

according to this latest update
https://www.crowdsupply.com/krtkl/snickerdoodle/updates/signs-of-life
It looks like they run into an long time known but not traceable bootrom problem which they can't determine for sure....but now identified

Maybe they want to get rid of that before really shipping.

[chickenHeadKnob]

I have been periodically checking these guys since they made the $65 initial offering. I thought as did others "too low to be true". Well the price is now $95 (snickerdoodle one) and the shipping date is still Orders placed now ship Oct 31, 2017. There is a resemblance to the iter fusion project which also has a continually receding delivery horizon. 

[Mr. Scram]

I have been periodically checking these guys since they made the $65 initial offering. I thought as did others "too low to be true". Well the price is now $95 (snickerdoodle one) and the shipping date is still Orders placed now ship Oct 31, 2017. There is a resemblance to the iter fusion project which also has a continually receding delivery horizon. 

That's the risk of crowdsourced products. They generally are small start-ups that will encounter issues and roadblocks. Even though the time to market is obviously long, the product still seems very viable and relevant in today's market and at the current price.

[agehall]

Is this just some particular version of the board that is just now maybe getting into production, or everything that they've been developing basically?  It seems to me that it was 1-2 years ago when I was thinking "hmm maybe I should pre-order a snickerdoodle" and they were expecting to ship like in June or July of ... 2016?  2015?
I think that was basically their initial model or 2-3 different variants of the initial model.
Is that what's still not yet out?  Or a different model?

What has been the problem?  Are there originally expected features / capabilities that are not now planned to be supported because of some redesign / design problem or something?

I still think some time I should get a Zynq board and I thought maybe a SnickerDoodle or MicroZed maybe.
I didn't realize they weren't shipping yet / still.

I think they've shipped some boards, but due to the boot ROM problem (or so they say at least), their yield has been terrible. I think most of the boards that have gone out the door has been the regular version and not the black version.

I pretty much see the money I spent on my order as a sunk cost at this point and if anything eventually shows up at my doorstep, I'll be happy, but until then, I'll remain a bit sceptic that I'll ever see a Snickerdoodle.

[mikerj]

The stupid name puts me off, never mind boot problems.

[thm_w]

I'm actually surprised there could be such an issue with boot rom reliability causing hopelessly bricked boards given that I would have assumed there would be no mass programmed OTP / user specified mask ROM involved.  Without having looked into it I would have just assumed that the FPGA SOC would be able to boot / configure from a variety of interfaces, Q/SPI, JTAG, MCU/parallel, whatever, and there'd be checksums and auto retries so it'd recover from errors and that it'd be possible to just select another boot source from the (I would have presumed) native Xilinx bootloader code's options.  If I was designing a devkit I don't think I'd have made the HW so that it could only boot / preload from one possible interface particularly if that wasn't well vetted in a reference design....

They are not even saying the board is bricked though. From how I interpreted it, the FPGA was slower than expected to start up and failed their production test (presumably some time limit in their script is used).
Then they mention its only at low temperature, but I doubt they environmentally testing each board. Maybe a cool breeze in the factory that day, and Y5V caps?
It's odd, but I see how if the production issue is in China, no matter how small, poor communication could delay everything.

As it turns out, manufacturing variations in the PCB fab and 1.8V power supply passives can, under low-temperature conditions, cause the BootROM to consistently delay processor boot on some boards. This condition was being trapped during our rigorous final test and was negatively effecting our yields. The good news is it is easily addressable with a minor passive component change and we expect yields to substantially increase.

[Mr. Scram]

That's awfully sad and to me unexpected.  Although I don't think making a FPGA board is simple for "just anyone" I seem to recall they said they had a lot of relevant experience with such technology and given that and the extensive amount of Xilinx documentation and published reference design information I would have thought that making a Zynq dev kit with minimal additional peripheral HW would be quite simple for them to have had success with.  I'd probably have ordered a couple had I been in the market for such things a while back.

It seemed / seems to me like the perfect opportunity since the FPGA ICs are quite inexpensive in moderate quantities and the PCB / assembly is also, but the costs of those things in low quantities make it impractical to build one economically DIY but a less expensive module / break out like Snickerdoodle seemed like just the perfect option.  Analogous to Raspberry PI in those regards, cheap to mass produce, expensive to prototype,
and very powerful / attractive even as a SBC module.

I'm actually surprised there could be such an issue with boot rom reliability causing hopelessly bricked boards given that I would have assumed there would be no mass programmed OTP / user specified mask ROM involved.  Without having looked into it I would have just assumed that the FPGA SOC would be able to boot / configure from a variety of interfaces, Q/SPI, JTAG, MCU/parallel, whatever, and there'd be checksums and auto retries so it'd recover from errors and that it'd be possible to just select another boot source from the (I would have presumed) native Xilinx bootloader code's options.  If I was designing a devkit I don't think I'd have made the HW so that it could only boot / preload from one possible interface particularly if that wasn't well vetted in a reference design....

Oh well.  Hopefully they get it working and can recover some usability of whatever quantities of boards they've already fabricated by firmware / software-ish means once the problem is understood and work-arounds pursued.  Then again after many months of delay...maybe not.

It might very well be a case of Murphy's Law. No matter how well prepared you are, when everything hangs in the balance, something stupid will turn up that might very well ruin it all.

The stupid name puts me off, never mind boot problems.

Your loss, I guess

[mikerj]

The stupid name puts me off, never mind boot problems.

Your loss, I guess

More like a lucky escape judging from the issues it has,

[Mr. Scram]

More like a lucky escape judging from the issues it has,

That's only if you treat crowdsourcing as normal web shopping. A lot of people do that, but obviously that's not how it works. You make an investment, which might or might not pay off. If you cannot accept that in advance, crowdsourcing projects might not be for you.

Honestly, the original $65 dollar is not much for supporting a cool start up and a fair chance of getting what looks to be a sweet FPGA board.

[donotdespisethesnake]

More like a lucky escape judging from the issues it has,

That's only if you treat crowdsourcing as normal web shopping. A lot of people do that, but obviously that's not how it works. You make an investment, which might or might not pay off. If you cannot accept that in advance, crowdsourcing projects might not be for you.

Honestly, the original $65 dollar is not much for supporting a cool start up and a fair chance of getting what looks to be a sweet FPGA board.

With crowdsourcing, it is not an investment, it is a donation. The reward is a "perk". I think the fact that the process is identical to web shopping is what create a misleading impression.

Of course, you could just until the product is on general sale, and then support the cool start up by buying the product and get a sweet FPGA board, except with buyer protection.

[brucehoult]

I've been having a quick look at these.

https://www.crowdsupply.com/krtkl/snickerdoodle

and I'd be interested to know if anyone has received theirs yet?

Is it worth the risk of me pre-ordering a Snickerdoodle black?

These seem to have been a long time coming.

The Snickerdoodle One is the same chip as the Arty Z7-10 and the Prime LE is the same chip as the Arty Z7-20. For $50 less in both cases. But you can get the Arty off the shelf.

[Mr. Scram]

With crowdsourcing, it is not an investment, it is a donation. The reward is a "perk". I think the fact that the process is identical to web shopping is what create a misleading impression.

Of course, you could just until the product is on general sale, and then support the cool start up by buying the product and get a sweet FPGA board, except with buyer protection.

It's not quite a donation and not quite an investment, but whatever the case, it's certainly closer to both than it is to a regular purchase People inevitably get disappointed when there are delays, the product is not quite what's promised or things are simply not delivered. All while that's intrinsic to the format.

[brucehoult]

With crowdsourcing, it is not an investment, it is a donation. The reward is a "perk". I think the fact that the process is identical to web shopping is what create a misleading impression.

Of course, you could just until the product is on general sale, and then support the cool start up by buying the product and get a sweet FPGA board, except with buyer protection.

It's not quite a donation and not quite an investment, but whatever the case, it's certainly closer to both than it is to a regular purchase People inevitably get disappointed when there are delays, the product is not quite what's promised or things are simply not delivered. All while that's intrinsic to the format.

So far the only thing I've bought from such a site is the HiFive1 RISCV arduino-compatable, also on CrowdSupply. https://www.crowdsupply.com/sifive/hifive1/

"$75163 raised of $1 goal". lol.

I ordered in early December last year and it shipped before Christmas. It was pretty clear that the product was already developed and they were using CrowdSupply only to outsource order fulfilment.

[Mr. Scram]

So far the only thing I've bought from such a site is the HiFive1 RISCV arduino-compatable, also on CrowdSupply. https://www.crowdsupply.com/sifive/hifive1/

"$75163 rained of $1 goal". lol.

I ordered in early December last year and it shipped before Christmas. It was pretty clear that the product was already developed and they were using CrowdSupply only to outsource order fulfilment.

I've looked at the HiFive1 multiple times, but wasn't too sure about the toolchain. I'm also trying to avoid swamping myself with too many platforms. I mean, swamping myself even more

[brucehoult]

So far the only thing I've bought from such a site is the HiFive1 RISCV arduino-compatable, also on CrowdSupply. https://www.crowdsupply.com/sifive/hifive1/

"$75163 rained of $1 goal". lol.

I ordered in early December last year and it shipped before Christmas. It was pretty clear that the product was already developed and they were using CrowdSupply only to outsource order fulfilment.

I've looked at the HiFive1 multiple times, but wasn't too sure about the toolchain. I'm also trying to avoid swamping myself with too many platforms. I mean, swamping myself even more

Toolchain is fine. RISCV is a standard target in the main gcc & binutils repos now. llvm getting there. OpenOCD works fine. gdb.

The easiest thing is to just add the URL from the SiFive web page into a standard Arduino IDE preferences dialog and everything gets automagically downloaded and installed. Select the board and port and you're away. Or you can use the toolchain downloaded by Arduino from the command line if you prefer.

[osholt]

I got one of the first batch months ago and while the hardware is there and it seems to work, the toolchain, documentation and support are non-existent. USB JTAG doesn't have a software implementation / instructions yet and the board files provided for use in Vivado etc. are pretty basic so it's pretty difficult to use right now. They been promising a documentation update for months now and none of it has materialised. So much potential...

The only projects anyone seems to have done with them so far are linux based and you could easily have done all that with a Raspberry Pi with much less hassle.

I hope they get their act together and start shipping and documenting.

[SMB784]

I received one of the fabled Black versions back in mid may of this year.  The problems they had with boot were a yield issue, not bricked devices that they actually shipped.  As a result, there were a few of us who were able to get our hands on working devices despite this yield problem.  They shipped around a thousand units, maybe 2000 if you count all the snickerdoodle black versions that they shipped.

My main complaint has been the inconsistent/near total lack of documentation for the device.  However, I understand that they have a team of 3 guys trying to put together a raspberry pi version of an FPGA.  Things are going to be slow, especially given their yield problems that they only recently were able to address.

I am hoping that now that they have all the bugs worked out of the actual device, they can focus on support and documentation.  However with their pending alpha release of the pismasher (which I am very patiently waiting for), this may cause a further delay.  It's a device with great potential, and I love the fact that it has built in wifi/bluetooth.  That was the selling point for me.

[Mr. Scram]

Pismasher, Rasperry Pis... The name isn't really great in the plural form, is it?

[KL27x]

The stupid name puts me off, never mind boot problems.

I finally clicked on this link after weeks of thinking a "Snickerdoodle" was a "Drawdio."