Recent content by SLee

The US is 3.2 trillion vehicle miles per year or 8.77 billion miles per day. https://fred.stlouisfed.org/series/M12MTVUSM227NFWA Daily US consumption of gasoline and diesel in 2019 was over 550 million gallons per day (13.4 million barrels/day * 42 gallons)...

It's not really a fair comparison, the 5900X being compared only scored 20.2K vs the 12900K's >27K. Without LN2, the 5900X can't overclock high enough to reach that score, and even a 5950X will need considerably more watts than stock to reach it, its stock score is only about 25.5K. The 12900K...

Imports already make up a considerable share of California supply in the evening and morning.

And also British Columbia in Canada.

From what I've read, the 4-core Tiger Lakes are not undervoltable, Intel disabled that functionality. It doesn't help you but Intel seems to have re-allowed undervolting on the newly released 8-core Tiger Lakes.

I think for most Canadians, claiming it as a "hobby" will be better. There you don't claim expenses but only half of the mining profit (edit: at the time when you trade the coins for fiat or other coins) is taxed at your marginal rate.

One of the benefits of staking on pools and exchanges is many of them offer a method to unstake if you have an emergency need for ETH or funds. On staking, they convert it to another token that can be traded back for unstaked Ethereum, though not 1:1 until withdrawals are enabled on the...

For Ethereum these days with Pascal cards, you'll likely get better results with overclocking the core and under-clocking memory or keeping around stock. The increased size of the DAG has changed what's optimal now.

I think that's expected, the high CPU power usage comes mainly from applications that use AVX heavily. Gaming don't utilize cores nearly as much.

There's several factors that come into play in determining the mining rewards. 1) The amount of transaction fees, which are added to the default 2 ETH block reward. In periods of increased transactions (such as euphoria during a rapid price increase), this can account for 2 to 4+ ETH. 2) Your...

I find it work best (with a bit of dragging and clicking) to get a flat curve at the desired clock speed. I ended up at 1125 on a Tuf 3090, though at almost identical hash rates and power usage compared to your 3090.

For a 3090, you could reduce core clocks even more. 1100 MHz is enough (with sufficient memory bandwidth) to hit 120+ Mhash/s.

There's a lot of existing knowledge in Arizona since Intel has several fabs there already.

With Windows and not changing the default P2 State on Cuda setting, this is how your card ends up at 10451: Default: 9500 P2 State on Cuda application, downclocks by 250: 9250 Overclock: 9250+1200 = 10450 (rounded to 10451 probably due to the clock generator increments)

So it looks likes even the Pro series is now TLC.