Bitcoinorama:
I've been following your posts on Knc from the beginning, starting with the initial interview. I check the direct link to your posts before I check the general boards. That's how important you are here.
You seem like a really good guy, genuinely interested in Bitcoin and the btc economy and community and not just the money. And I know you trust the guys at Knc, they trust you, and you have a good relationship, and so on...but seriously, how far are you willing to go for these guys? It actually kinda hurts to see you coming under fire, and your integrity questioned.
The truth at the end of the day is that it's all about the result, and while KnC have put in a tremendous effort that's recognized by all, they're not delivering on a lot of counts. Shipping delays, boards catching on fire, voltage regulator cutbacks (8 to 4), firmware isssues ... the list goes on and on. It's like in football/soccer.. the team and manager can put in a great effort, pull out all the stops, but if at the end of the day they lose or get relegated, the fans are still unhappy, possibly even up in arms. And heads still roll, regardless of how hard they tried.
In all sincerity, can I suggest that you take a breather, have a pint, and consider just how much you're willing to give up for these guys? I think you should have a think about it, and be careful not to entangle/embroil yourself in something you wish you could get out of later but can't. You don't work for them, yet are going to great lengths to defend them and answer questions on their behalf. The truth is that THEY should be the ones trading exchanges with people on here and taking the heat, not you. I just don't want to see you turn into an Ineba-like figure and have your credibility and reputation comprimised or ruined altogether. That'd be a shame after the stellar shift you've put in - all on KnC's behalf.
Soapmodem, I take on board and appreciate what you are saying, and likewise please might I suggest you take a break from reading the armchair engineers and FUD spread around maliciously within this thread. Whilst I agree Murphy is yet omnipresent, I have always told it precisely how it is. If you are to receive a product that is being shipped from the day after chips were first received there will be a lack of refinement that will be ironed out. If you ask any real engineer about what is being undertaken here they will tell you unreservedly the magnitude of the work and achievement being undertaken, let alone at 28nm tech.
In essence you cannot have extreme performance, without extreme refinement, let alone at this speed to market. Within one week there is a working product, that has had some hiccups and the firmware revision have reached a point of stability for the most part. If you want a perfect hands-off consumer ready product than you're welcome to wait the standard average times of 12-18 months for 28nm tech. If you are willing to be proactive and responsive, than you'll be hashing sooner than anyone else with 28nm from another company.
The margins in the design, such as the 8vrms were in place to compensate for any design issues so the product could fly out of the door and into your hands. There will be ongoing revisions to the product over time, both physical, and firmware/software, that is a given as that aspect of the standard timeframe has been removed so you can hash with the product immediately. This is why the true ability of the product was consistently under promised. Few that have the product in hand are complaining.
It's also why there is a sudden desperation by competitors who have made unrealistic performance claims and delievry dates to badmouth KnC, and attempt to appease their audience, whilst beginning to backtrack on dates ahead of their dishonest claims being realised. My integrity can be questioned for all anyone likes, go for it. I've never lied. My integrity remains intact. You cannot argue honesty, and you cannot compete with lies.
In conclusion, as and when issues arise, and are being brought to our attention here, the logical responseis to do what any sane party would do, determine cause, and rectify the issue as soon as possible, without hinderence. It's unfortunate the Arctic fans appear to still loosen in transit. The boxes are being replaced with thicker reinforced cardboard, and new courier companies are being considered, however as I stated above they all now contract to third party delivery firms to handle ground collection and delivery, choosing to focus on the air transit and customs side. The guys that UPS send to the manufacturing fab, turn up in a variety of vehicles, and never with UPS branding. The PSU issue is unfortunate, but unless you want KnC to take the time to test every potential consumer PSU for concern, which would delay delivery further, the best that can be offered is an ongoing list of which PSUs create such fault, and which don't. They immediately flew in General Electrics Critical Power division to work on any complications. I have the accurate description the GM guy gave me, but that's at KnC's office on my MacBook. The battery died whilst he was there yesterday and although I had the charger, I had left the UK->EU adapter plugged in the wall at the hotel I'm now at. That said the corsair HX850 is now known to elcit such an undesirable consequence due to the load initiated upon reset of the PSU and should be avoided. Could not have been predicted. Could once the chips were in hand every brand of PSU likely to have been used been tested on working devices running for significant periods of burn in time time whilst firmware is revised to a point of absolute perfection to satisfy a consumer audience? Sure. Does anyone here want to wait the length of time that would involve? No.
The Coolermaster V850 is without a doubt the most tested compliant PSU with so far no complications present as all the dev boxes, and the entire hosting facility is equipt throughout and there have been no issues since.
Hopefully within this next week everyone can be happily hashing as comfortably as the majority of those in possession of their units. Sam, and Andreas are at the hosting facility today and boards, and working with the pool configuration. Of the guys currently hosted, only six devices are under performing and that is being addressed, that includes DHenson, and one, or two of Bargraphics lot. Again this is not hosting all devices and offering a proportional piece of a shared hashrate like certain cloud based entities. This is each individual customer having access to their individual device configured by a rather complex API. Marcus and Michael drove back to the manufacturing facility today to oversee the physical production and deliveries personally, after handing the reigns over previously to concentrate on working alongside General Electric the last few days.