https://www.kitguru.net/channel/generaltech/matthew-wilson/bitcoin-surpasses-15000-value/Earlier this week, we saw Bitcoin managing to hit the $10,000 value landmark. In the days following, it has shot up well past that number, with the cryptocurrency now sitting at $15,000 in value per coin.
That just leaves one question to answer—how well does it mine cryptocurrency? Pretty darn good, as it turns out. As in, almost double the mining performance (of Ethereum) of an AMD Radeon RX Vega 64 graphics card.
Here's a look at how the TITAN V performed with some other types of cryptocurrencies:
Zcash: 750 SOL / 221W (877 SOL / 244W while overclocked)
Monero: 1,224 H/s / 157W (1,417 Hs, 165W while overclocked)
LBRY Credit: 685 MH/s / 241W
Vertcoin: 88.7 MH/s / 246W (100.3 MH/s / 259W while overclocked)
I guess someone who is dedicated to mining (ie, 10K MH/s range) investing a large amount of capital (millions USD) for the sake of large profits.No, why should be Titan V desirable and what is a comercial miner?
Its merits are just that the hashrate is achieved with signifcantly lower energy consumption compared to the 3 GTX 1070
I guess someone who is dedicated to mining (ie, 10K MH/s range) investing a large amount of capital (millions USD) for the sake of large profits.
This is a significant cost reduction in electricity when you're mining with that much gear. Theoretically mine with 66% less real estate, and significantly less power.
good addIt really doesn't compare that favorably to a 1080ti when price is factored in. Electricity can be expensive, but not *that* expensive. And then consider that there will almost certainly be a consumer product incorporating these efficiency improvements (Volta or not) in a few months and these seem like a pretty bad option.
Here's a look at how the TITAN V performed with some other types of cryptocurrencies:
Monero: 1,224 H/s / 157W (1,417 Hs, 165W while overclocked)
Eventually you run out of capital and you are forced to sell the very coins you mine. But selling them means transaction fees and selling them for below the targeted value.
That being said anything that would delay that, or reduce it would be ideal. In particular, reducing the real-estate, energy costs, and administration costs will go a long way in keeping your miners on for longer before needing to sell.
.
hmm I might have it totally wrong. I would assume they wanted to ride it lol. I'm watching the #1 wallet right now on nanopool => https://eth.nanopool.org/account/0x39d27d66c14f7372553b1ba59833c6ba8981a76a^^
Maybe vega's outstanding performance in CryptoNight algo is due to the HBM (if not, what else would be the reason? i'm struggling past 800 h/s with Rx580, so it isn't anything AMD specific). In time, in that case, we could see the Titan V catching up
AFAIK, and quite contrarly to what you wrote they would sell a significant on a daily basis, covering operating costs and collecting a piece of profit. Not saying they sell everything, i wouln't know
It's risky to wait for price increases that may never come, especially since you put hundreds of Ks $ into hardware. I can attest to that a bit having put > 5k $ into mining hardware; you do feel some pressure swimming in just digital currency only
I saw somewhere an interview with a guy working for such a mining company (sadly couldn't find it briefly) and he emphasized on how time before ROI is important and they plan their hardware purchases according to coin prices, and considering that time. A discussion that includes, oximoronically, both time to ROI and Titan V seems alien to me, hence my posts.
Fiji does perform good in Cryptonight, but not stellar. Note though, that it's basically Cast XMR in conjunction with the blockchain beta driver that enables this kind of output.^^
Maybe vega's outstanding performance in CryptoNight algo is due to the HBM (if not, what else would be the reason? i'm struggling past 800 h/s with Rx580, so it isn't anything AMD specific).
Fiji does perform good in Cryptonight, but not stellar. Note though, that it's basically Cast XMR in conjunction with the blockchain beta driver that enables this kind of output.
https://hothardware.com/news/nvidia-titan-v-volta-ethereumUsing Claymore Miner (v10.2), we were able to hit 69.06 MH/s at default speeds, far outpacing what's possible with the current standard-bearer from AMD and the previous generation Pascal-based TITAN Xp. The Radeon RX Vega 64 was well behind with a 37.72 MH/s rate, while the TITAN Xp was virtually tied at 37.45.
However, when we cranked up the GPU and memory clocks, we saw a significant boost in performance, with a hash rate of 82.067 MH/s.
I thought that Ethereum was almost completely memory bound? If Vega 64 and Titan V are using the same kind of HBM2 memories, running at roughly the same clock speeds, you'd then expect the Titan V to be roughly 50% faster than Vega 64, but non-overclocked it's 83%. That's really surprising, isn't it?NVIDIA Titan V Ethereum Mining Blows Past 82MH/s While Overclocked On Our Test Bench
https://hothardware.com/news/nvidia-titan-v-volta-ethereum
Fiji does perform good in Cryptonight, but not stellar. Note though, that it's basically Cast XMR in conjunction with the blockchain beta driver that enables this kind of output.
I thought that Ethereum was almost completely memory bound? If Vega 64 and Titan V are using the same kind of HBM2 memories, running at roughly the same clock speeds, you'd then expect the Titan V to be roughly 50% faster than Vega 64, but non-overclocked it's 83%. That's really surprising, isn't it?
But I'm also really surprised by the Titan Xp Ethereum hash rate being equal to Vega 64 as well: Vega 64 has a theoretical higher BW and HBM2 should be more efficient than GDDR5X (which is notoriously bad for Ethereum for some reason.)
Maybe there is something wrong with the Vega 64 memory system after all?
Are those hothardware numbers in line with expectations?