I’m in for the crowdfund at 1 BTC.
1 Like
I’ll add 1BTC if deliverables are compatible with both NVIDIA/AMD and support both Windows/Linux.
1 Like
Akiron is one of those with a GPU miner.
$320 used GPU:
8 GB on board RAM at 20x to 30x higher bandwidth allows for 2.5x more dedicated cores than the 4 cores on a CPU. This would naively mean 20 x 2.5 = 50x to 75x more than a CPU, but it’s not that high because its cores are at 1/3 the GHz and generally not nearly as good individually as CPU cores. So it’s more like 10x to 15x better than CPUs if they just copy over the CPU algorithm and do not optimize it or make it parallelize the sorting to use more of the GPU cores. It will be a little more efficient in terms of electricity per block due to not needing to parallelizing the code, and running at lower clock speeds is more efficient, better than the ratio of the speeds. A new 300 W GPU should be capable of getting 10x than a new 40 W CPU at the same 22 nm or 14 nm process.
This has nothing to do with the 4x in the Equihash paper because that was about parallel sorting on the same block of memory, not working on several blocks of memory at the same time. Zcash has been forced to use 10x less RAM than what modern GPUs have on board, giving GPUs the clear advantage in a way that was not considered in the Equihash paper.
BTW the paper was written this year, but they only had 2011 GPU and CPU sort comparisons.
1 Like
@zawy From my conversation with akiron, his miner isn’t ready for the public and he currently only plans to support NVIDIA. However, ampy is ready. He’ll release benchmarks if he agrees to our deal, and you’ll have a chance to look over these details and ask questions before you make any financial contributions.
1 Like
With all this talk of a GPU miner, how will this impact the “slow start” mining method of zcash? Will it really give those that invest that big of an advantage?
Why would there be an impact on the slow start?
What about AMD, all of my cards are AMD cards.
@blackhatzw
Yes, AMD would be supported.
@MrJaekin
If Ampy agrees to deal, there will be an incentive to invest, however, we will need to work out some numbers to provide investors with specifics. But first, we need Ampy to agree to our proposition.
I guess impact was probably the wrong word. Sorry, just trying to understand. 
As I understand it, there is a number of blocks after the genesis block where the block reward isn’t that much, 12 or so I think. If we crowd source a GPU miner and it’s available at launch, only to the investors, would we be wasting our time on not that many rewards since the miner would be available to others when the rewards were raised?
@MrJaekin I have thought of another way to incentivize investors. Most likely the miner will leak to some degree. But the bonus for supporting the crowdfund may be worth it in the end. More details to come.
@coin_artist Gotcha, thanks! Hope Ampy is willing, excited to see where this goes!
1 Like
The slow start will occur during the first 20000 blocks / roughly 35 days. The block reward will increase from almost zero to 10 zec in that time.
The purpose of this is to allow an anticipated influx of new and inexperienced miners to get up to speed and figure out what works best for them without the added pressure of mistakes costing them so much per block.
4 Likes
Makes sense. Thanks for clarifying! I will need all the time I can get to brush up on my Linux skills! Now to get this GPU miner going! 
I’d be happy to contribute 1BTC. Great idea. Linux support is important for me.
1 Like
I agree with Zampys comment. I dedicated probably more then 1BTC to the cause and a fixed pre-order.
btw how can I join the slakk channel?
Thanks
1 Like
Here’s an invitation link to the zcash slack channel https://inviteme.z.cash
I previously stated one of the GPU code writers was on the Zcash team. That is not the case. Sorry.
thanks coin_artist! very very nice of you.
I will commit 1btc. Send to which address ?