What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Worth Getting a KNC Titan?

Pansyfaust

VIP
VIP Supporter
Rating - 100%
133   0   0
Joined
May 8, 2010
Messages
5,602
Reaction score
984
Points
7,165
Location
Durban/Queensburgh
Hi guys, posted this on Bitcointalk forums in regards to the 100MH/s KNC titan scrypt miner, thought I might as well post it here as well

Hi guys, just want to discuss some hard numbers with you. This is not a Asic bash, even If I dont like the centralising of it, I understand the benefits of getting in early, as I already showed interest in coming late to the party for an Asic buy, but I just want to develop some conversation around it and realistic targets.

I realise their may be errors in this post in regards to calculations and projected figures, but its basically a hypothetical situation extrapolated from current data.

If there are 2500 of these units KNC are going to add over 250000MH/s to the current network alone. To put it in perspective, the current network is less than the amount this ONE set of mining equipment can provide, not to mention the addition of the rest of the proposed mining manufacturers.

If the difficulty keeps increasing at a moderate 3% up until the pre-order date of delivery then the difficulty when the miner arrives (no delay) means you are mining at a difficulty of 16k in around July/August at the earliest.

Now the network/difficulty rate adjustment seems to be that there is 1000 points of difficulty per ~30mh/s. Just did some quick calculations about diff adjustments and total network hashrate for a few spot points over time and it seems to stick to around 30Mhs (eg December 30 2013: diff was 3168 and network hash rate was 90697 = roughly 30Mhs per 1000, March 09 2014: diff 4738 net hash is 135660 ~28Mhs per 1000.

Now the point I am making is that when the Titans eventually start coming online around the world the difficulty (based on current projections) is going to be somewhere around 16k in July/August sometime, extrapolate that to Hash rate and we get ~480000 Mhs.

Next add the combined potential hash rates from Titan's and we have an extra 250000Mhs pouring into the market for a total of ~730000Mhs. You may get the Titan earlier, but the difficulty will be in the region of 24333 by the time all Titans are online and adding to the network. Based on that one would only make roughly 4 LTC a day once the network has adjusted for the massive influx of hash rate.Now I have no idea how much the hash rate will increase after this massive influx and what the result will be for the difficulty, but at the best scenario presented within this context I have typed out the Titan will make $2000 a month or less, and as we all know, this will steadily drop off as hash and difficulty inevitably increase. The one saving grace may be if the price of LTC doubles or even triples, but even that means only $3700 or less after one month. Tripled the price? You would make back half or less in a month.

Now I know these numbers are possibly worst case scenarios and that the hash rate might be wildly different, although I actually suspect I have been quite conservative in my calculations as this scenario only deals with the Titan introducing its hash rate, never mind the gridseed miners, Fibonacci, Zeus and Alpha.

The point is if the price of LTC does not increase rapidly or even before the impact then its a pretty big risk at current trajectory.

Discuss?
 
Elaborate :p? I'm not saying I want (or can afford) to get one, just want to discuss why or if people should

You're thread title asks if it's worth getting one.

The answer is No. ROI won't happen and you'll lose money.

Also, wall of text crits me for over 9000.
 
You're thread title asks if it's worth getting one.

The answer is No. ROI won't happen and you'll lose money.

Also, wall of text crits me for over 9000.

I think, with my "wall-o-text" calculations, have come to the same conclusion as you so eloquently put it = No.

Also, now I know your weakness. Paragraphs and paragraphs...
 
Litecoin isn't the one to be mining for ROI on hardware.
 
Well now you get 250 Mh/s per Titan. If it does come out in July, which I'm almost certain it won't, it should break even in 1.5 months or so.
 
Arnt afew coins (especially doge) considering changing algorithms in the case that scrypt miners come out?

Sent from my GT-I9500 using Tapatalk 2
 
Care to elaborate on your 30mh/s statement? I'm not following your math. Shouldn't it be 30 000 mh/s ? Or such? or is your mh not equal to 1000 kh?
 
Also, the KNC miner is as Roligov has stated, now 250mh/s

It seems quite strange, everyone seems to happily be inflating the numbers.
Aka
Company A (lets call them fibonacci :p) came out saying they have 1mh/9mh/18 mh miners
Everyone oohed and aahed

Then Company B (lets call them ... alpha) came out saying they had 5mh/15mh/30mh

Then Company C (lets call them ... KNC) came out saying they had 100mh...

Company A announced --- Suprise, you all get free upgrade to 2/18/36
Followed by B going --- sorry typo, ours actually does 10/30/60
And now company C suddenly does 250....

Fishy fishy
 
Care to elaborate on your 30mh/s statement? I'm not following your math. Shouldn't it be 30 000 mh/s ? Or such? or is your mh not equal to 1000 kh?

You are right somewhat. I meant to say 30Mhs per 1 point of difficulty, not a 1000. It is 30 000Mhs per 1000 :)

I think my calculations are still correct though?

For every point of difficulty increase there it is due to ~ 30Mhs bump in hash rate. So with 2500 Titans * 250Mhs now estimated = 625000mhs instead of ~250000mhs.

But yeah. How can they just "have" an extra 150% or extra performance just lying about with the ASIC? By the time we get to release date they are going to be announcing 500Mh/s Asics probably.

I'm wondering if they *bumped* up the hash on paper because they just did the same calculations I did and saw people are going to do the same calculations I did and arrive at the same conclusion: Spending 10k on an ASIC to maybe get a profit after a few months.
 
Last edited:
Oh, and as a sidenote [MENTION=470]Pansyfaust[/MENTION],

I think its probably worth it just to hedge my bets .

Hypothetical situation
100mhash gpu farm
100mhash usb asic farm

Value: Probably 1.5mil?

Now, say the KNC is a flop -00> hoax whatever, thats R100k down the drain, boohoo the other 200mhash stays profitable and the loss is recouped in 2-3 months

Now say the KNC is NOT a haox --> R100k spent, that brings the hash rate from 200mhash to 450mhash.... you stay in the game. If you didn't you would now be mining for peanuts with expensive hardware that takes a lot of power.

My 2c
 
Oh, and as a sidenote [MENTION=470]Pansyfaust[/MENTION],

I think its probably worth it just to hedge my bets .

Hypothetical situation
100mhash gpu farm
100mhash usb asic farm

Value: Probably 1.5mil?

Now, say the KNC is a flop -00> hoax whatever, thats R100k down the drain, boohoo the other 200mhash stays profitable and the loss is recouped in 2-3 months

Now say the KNC is NOT a haox --> R100k spent, that brings the hash rate from 200mhash to 450mhash.... you stay in the game. If you didn't you would now be mining for peanuts with expensive hardware that takes a lot of power.

My 2c

True, if its a flop the network stays relatively mild and increases in hashrate are stable.

But like you say you can afford to hedge your bets as you have a farm up and running to absorb the risk and pay for the scammed hardware. As for people who dont have the capital to absorb it, its more risky.

I take it you are wanting to hedge your bets :p
 
Also, if you are investing 1.5 Mil into something like mining, 100k is something you can afford to risk.

An Asic makes sense to this demographic, but effectively shuts out the rest of the "average" miners and creates the perpetual centralization of mining power in a few hands worldwide. I dont think on this too much, just hope those hands take the coin somewhere profitable for the rest of us.
 
GPU Miners > Bitcoin
ASICS > Bitcoin

GPU Miners > Litecoin
ASICS > Litecoin

GPU Miners > Scrypt-N / Keccak
ASICS > ????
 
I reckon they will stop at at scrypt-n There is no reason to develop another algo at this stage.

Guys look at what happened before. Look at BTC for example

BTC launched > GPU mining > ASICS launched > No more gpu mining, everyone on ASICS
Now replace the top with LTC instead of BTC

LTC will go through exactly the same stages and its already happening

My advice for GPU miners, start looking at Scrypt-n as Scrypt on GPU is nearing its end and if you don't jump on the waggon now for ASICS on scrypt then you have officially missed wave two of the Crypto market. Bye bye

I also suggest holding coins, I am sure most of us expect a nice little bump in price when the network hash goes ups
 
GPU Miners > Bitcoin
ASICS > Bitcoin

GPU Miners > Litecoin
ASICS > Litecoin

GPU Miners > Scrypt-N / Keccak
ASICS > ????

Scrypt-n is assic resistant, so not sure if they will get around that easily :)
 
Litecoin was meant to be asic resistant too and was for a long while, but where there is profit to be made, the Chinese will find a way around it. I think most GPU's miners that are mining for profits have already moved over to Scrpyt-N, are in the process of moving over and just the speculators are left mining LTC/DOGE in the hopes it skyrockets. I think Scrpyt-N will be to GPU mining what Litecoin was to GPU mining when the BTC asics came out.
 
Last edited:
Litecoin was meant to be asic resistant too and was for a long while, but where there is profit to be made, the Chinese will find a way. I think most GPU's miners have already moved over to Scrpyt-N, are in the process of moving over or are just mining LTC in the hopes it skyrockets. I think Scrpyt-N will be to GPU mining what Litecoin was to GPU mining when the BTC asics came out.

I think the devs for vertcoin have stated that they will resort to hard forking the coin in they ever need to counter ASIC development for this coin in particular, not sure about other scrypt-n coins.

And since we are here, is anybody looking to get a scrypt ASIC?
 
All algorithms will ultimately go the same route.

It's GPU proof!
Wait, maybe not.
At least it's still ASIC proof!
Or not.
 
Seems someone is adding mountains of Mhs to the system. Mining with the Asics a month/s ahead of delivery? This is why I hate ASIC company's, not for what they do to the network, but how they mine with their hardware before they sell it.

I was contemplating selling my hardware and getting a Gridseed 5.2Mhs blade, but with the difficulty increasing at such a rate it would be hard to make a ROI at current LTC price. I know the gridseed blades are nearly coming to market, but dam, having nearly 80GHs in the space of 9 days smells so hard of ASIC company mining shill.

Definitely time to jump to scrypt-n
 
The only true winners are the hardware manufacturers, it's better to buy the coins cash and holding then invest in mining, unless mining turns you on. :)
 

Users who are viewing this thread

Back
Top Bottom