Quote from: reactor on June 06, 2014, 03:46:10 PM

Also, since people are borderline begging for the update on the community forum I'll add it here so Bitcointalkers can see - news on the firmware update?

I see Eric saying people need to email him at ZenMiner for an update for Fury issues, why isn't it posted on the site? Stickied on the community section? As-is I don't feel comfortable running the Fury with the included power supply without sitting there watching it, but I'd like to see this update pushed out to the community rather than folks who happen upon it in a GAW Community thread on your website.

Please, do the community some service and fix this stuff. It's about to hit shipping for week 3 and people still can't get this stuff working out of the box. In all honesty this hardware as well as the Zen service were not ready for public consumption and needed a much more thorough QA round on all sides.

There was an e-mail sent out to GAW customers, and I could swear it was also posted on the gawmunity site, but I can't find it there. Anyway, see below, it mentions firmware issues with GAW's typical "blame someone else" attitude.

Quote

Thank you all for hanging in there during this bumpy launch for week 2. A mixture of events have occurred in the past few days concerning power, firmware, and general bugs which I want to cover in this notification.

First of all, we would like to thank you for your business. We couldn't have made it this far without you and we are humbled by your support and belief in us as a market leader. We now have a couple of thousand hosted customers which is quite a bit more than are original limit of 100 at launch just a couple of months ago. A large percentage of these customers have multiple rigs (some close to 40 different rigs). With this type of scale comes bugs and UI issues. This also forces us to implement better alerts and notifications for rogue nodes. In the past couple of weeks we have hired new Full Stack Developers along with other technical CSR's to assist with growth. We are not stopping there though. In the beginning we had a rough idea of how fast things would grow and boy were we on the low end of that forecast. We have since then planned for an even larger scale and deployment by adding resources (as mentioned above) and giving more effort to making our internal systems run more efficiently.

The issues week 2 hosted customers have experiencing in the past few days are related to possibly 3 items with the first being the most probable.

1. Power issues

2. Firmware/Mining software issues

3. UI bugs and cloud service issues

1. The power issues for 75% of the week 2 hosted units have been resolved. We have learned that whenever the units are tested in our lab, they are usually connected to pools that we use which are only higher difficulty and have more consistency as far as uptime and actual jobs. Some of the pools that are being used by customers have very low difficulty (like 16 - 128) which causes the units to draw more power. The draw is not entirely related to the difficulty but it plays a big factor. So an extra power draw of 1-2 more amps is significant when working with thousands of these units. Obviously we have planned for variance but we are right on the limit when factoring these numbers. We had a similar issue with the gridseeds although on a much smaller scale at the time.

New hardware = new issues but at least we can now see the light at the end of the tunnel. To resolve this issue on the remaining 25% of units, we are having the power company upgrade transformers to larger a much larger kw as well as bringing in new power legs to spread the load in multiple data centers.

2. Recently, we received new firmware/cgminer from Zeusminers for the Generation A Chips. After applying the firmware we realized that there we no improvements and in fact it made things worse as far as power consumption and fluctuating hash rate. We are in the process of reverting to the older firmware and have been for the past day. When we tested the firmware on our lab units, it seemed stable and more efficient. We are, again hiring more resources to work on Xminer and the controllers themselves for better performance and reliability.

3. Some issues in the UI we are aware of can be anywhere from your units not displaying properly, to the actual hash rate not being displayed properly. 50% of these issues are related to just reporting issues from the mining software to the UI and the other 50% are related to the actual miners either not responding or frozen. This is a serious issue because the only thing worse than not being able to manage your miner, is not being able to see that it even exists. We have immediate plans to work out this specific issue and confident that we can have this resolved in a timely manner.

Since we started this business as BTC and Scrypt miners just like yourself, we know what is most important to you and are working hard to achieve customer satisfaction at it's fullest. The speedy delivery of products is actually one of the main key components here and we realize that. This is why we try to do things like launch 1 day early or provide compensation for downtime. We know the score and understand the importance of uptime (especially in the beginning).

All of this said, this is why we want to offer a 150% compensation for possible downtime during the last few days from the beginning of week 2 launch. When we have mitigated the issues at hand, we will send another message with a link to a compensation form for the Generation A Week 2 Hosted customers asking for specific information.

I believe in providing an unmatched customer experience along with continued innovation in every area of our business. This is the foundation our company was built on and this is how it will continue.

Thank you very much for your support.

-- Joe Mordica CTO