|
I've had 2 sapphire 4Ls and they've both been crap. Run hot as hell, noisy (lil quieter in a case but still noisy) and unless you run it by itself, you better have good case airflow to assist the blower. Not to mention both cards thus far began IMMEDIATELY artifacting my desktop while mining as the display card.
Maybe I'm just unlucky but that's been my experience. Double d has bad reviews, no personal experience.
|
|
|
|
|
There is little practical upshot to stuffing 8 physical cards on a single mobo rather than 2x4, except potentially a savings of 100w or so for the extra system. And if you're running 8x200w gpus, that's a fairly paltry savings.
|
|
|
|
|
Chip resale is unlikely. The only sellers would be scammers, people with bad/harmed chips, small fry who didn't realize it's hard to DIY, or those who bought before bitcoin crashed/surged in difficulty (meaning no incentive to buy).
There might be a few who bought specifically to resell, but that would be at a premium, compared to today's prices, which might or might not be worthwhile. I wouldn't expect reselling to be a big thing.
|
|
|
|
I am betting on human greed, which means that once it gets harder to mine bitcoin, the price will go up hard.
You didn't read my article, did you? If that is the case, investing in mining is the worst decision you can make. Investing in equipment is a bet that the price will go down, not up. If the price goes up, you'll have a lot more profit from simply buying coins and holding on to them. .b Disclaimer I did not read the article (working, will skim later) but I will address this response, because I've grown weary of seeing it. It is not wrong per se, but it is overly limited. Investing in hardware is a bet that price will go down. Or stay flat, or go up a little bit, or a look at a timeline where you generate more coins than you could have bought for the same price, in which case it is always better regardless of direction. Mining is a risk averse decision, it has a larger spread of options where it is superior, and less risk of ending up holding a bag of nothing (though ASIC is closer to that). Buying/selling is a liquid, high risk high reward strategy. They are different, and dependent on individuals predilections. One would only be better if one could see the future. |
|
|
|
|
For those who sell a soul lightly, remember the lesson of Bart selling his soul to Milhouse.
|
|
|
|
Alot of 7950's do not ship at 1.075v core but rather 1.25v and some are even voltage locked. He will have a hard time getting those to work without flashing the bios. If he wants to run a dedicated miner bamt linux is one of the easiest to use but offers no software voltage controls. Trust me a 7950's at stock setting can use almost 300w by itself. I've seen it. A 660w Seasonic will not break a sweat powering two 7950's. I've been mining for 6+ months with an XFX 650w unit (rebadged Seasonic) and two 7950's at 1075mv vcore, 1000mhz core. Pull from the wall was a bit less than 500w.
PSU is all about getting a quality unit/brand. Crap like OCZ power supplies might not cut it powering two 7950's but a quality single rail 660w Seasonic unit will be fine.
You haven't seen a 7950 using 300W by itself, because the max TDP for a 7950 is 200W. Now, a 7950 will use potentially up to 200W, and your system can use 100-200W (mobo hd CPU etc), so 200+200+200 < 650, which means you'll be fine. If you can undervolt, then even better. Please research before making claims. |
|
|
|
wtf...
you necro'd this thread to say that...? |
|
|
|
|
It will probably be ok, however it is not a good PSU. HEC OEM, old old old model, not even 80+ certified. It probably won't murder your GPU or anything, but definitely don't try to load much else on it.
|
|
|
|
|
Smart Batch 3 owners would re-sell their pre-order for truckloads of cash.
|
|
|
|
|
My only regret in not pre-ordering an ASIC, is that I can't re-sell the pre-order to dummies for way more than it will ever return.
|
|
|
|
Practical query, other concerns aside: I don't follow how investment and returns are related. You say you require 780BTC to order chips. Fine. THen you will require an additional 3500BTC to make the chips go. Ok. So I invest 100 BTC in this project, my return is...  I'm not even talking about calculating profits, simply calculating what 1 BTC of investment is supposedly purchasing. The cost of manufacture being relatively unknown on your end makes for a problem. What happens if you raise 780BTC, to order chips, but you don't raise BTC to create the remaining portion of the project? TS for the investors? Re-sell chips at some point for whatever they are now worth, and return whatever percentage of the initial investment to the people who funded the buy? Would you take a management fee of that as well? |
|
|
|
ROI is quickly going out the window, but the question becomes will people still run the ASIC's even if they are loosing money on power costs.
Well, those who pay electricity $0.3 will shut their rigs off way before than those paying $0.15, and those who do not pay electricity at all... Well, they will shut their rigs of much later. And I can guarantee you that there are quite a few big-size mining set ups that pay 0 for electricity. Keep in mind that even at $0.3, an Avalon unit will be profitable up until about 950Million difficulty (100x current difficulty, or about 6500TH). It won't be making much, but it will be making something. Though if price drops out it will be much lower of course. |
|
|
|
Don't forget...unless you're a trade-in customer or dev "you're not special"  For clarification the shipping priority is as such:
Sorted from Highest to Lowest.
Developers - e.g. bitcoin core dev and other bitcoin contributor we feel deserve this treatment.
Trade-Ins/Old Customers - your unit will ship soon as we receive the FPGA and you've paid the difference.
Regular Batch #2 Customers - Unfortunately you are not special.
If you're not in the first 2 categories you probably shouldn't fret about shipping until you see trade-in customers posting about receiving theirs. I'm guessing mid May??? anyone feel like buying a FPGA just to trade in and get priority shipping?  well while we are waiting we can play around with the bitsyncom website http://bitsyn.com/is it me or is the page blank? I see a bunch of pretty colored lines. Maybe it's a site that uses your imagination, and you see whatever is inside you? |
|
|
|
Kind of OT but related, did Avalon confirm that they had shipped Batch 2? I believe the date for shipping was set to 4/15/13, but they might be staying mum to avoid bribing officials again. The delays on everything ASIC is what kept me from ordering into the Chip batches, which I may or may not regret ~3months from now. Scares me that someone had 16k BTC to drop on chips, plus whatever they will pay to have them made into whatever they will become (unless burnin finds himself with an order for 1600 modules from one source  ) |
|
|
|
sorry to bump an old thread, FullLife, I was wondering what you finally did to get all 4 working with risers? I didn't know that you could have pci powered extenders until I saw this thread and this: http://cablesaurus.com/PCIe-x16-Extender-Cable-w-Molex-ConnectorI think this was the exact problem I am always having. sometimes all 4 cards work, other times the bios only sees 2 or 3, sometimes it doesn't want to load at all. Now that I think it through, i have 1 slot that only runs at x16 and won't drop to x8. If I use only one card, with just a riser plugged into this slot it still won't boot properly, but sure enough, if I plug the card directly into that slot it works like a charm. Then I tried plugging one card into another slot that can drop to x8 with the riser by itself and that works, but then I was having trouble with all the other combinations, and I think the powered riser might be the solution. edit: hell this might even be the reason my cards aren't able to reach their full potential now that I think of it, but it could be wishful thinking I fixed the problem by switching the 4th card from a 16x - 16x extender to a 1x - 16x extender. can you run a card with full potential with a 1x to 16x riser? Yes, no card is powerful enough to oversaturate 1x. I'm not even totally sure why 16x exists tbqh when all the tests show that 4x handles the most powerful cards of today with no trouble. |
|
|
|
Hello WORKS:Gigabyte GA-970A-D3, sempron 145, 1x 4GB DDR3, 1x HDD, Sentey 750W 80+ PFC PSU 4x Sapphire 7850 1GDDR5 (dual cooler) 2 cards running on 16x powered risers (works with and without power) 2 cards running on 1x to 16x powered risers (only works with power) on 1x slots 2 and 3. image:  Win7 x64 and GUIMiner BTC This thing eats 440W from the AC outlet, and produces 1GH on BTC Did you make that housing yourself? I like it, clean and space efficient. I wish I had tools to make my own. |
|
|
|
|
Same. I like cablesaurus and I've been watching him since he opened his business way back when, glad to see him doing well. But still, I can't make myself payout (cheapo!).
The cables look nice, glad they're workin out for ya!
|
|
|
|
|
Apparently Bogart thinks that it is worthwhile.
|
|
|
|
Wow theres a huge markup for being in the USA it seems. $3 shipped from china, or $40 one day shipping in the US  |
|
|
|
|
Show Posts
