Hm. So, for my university course, since we're talking about algorithms and GPUs and shit, my group has been given the topic of bitcoin and crypto mining, to make a presentation for. And since I intend to be as impartial as possible, and my cursory impression is that this whole thing is a massive clusterfuck, I expect that my presentation is going to end up involving at least a solid five minutes of continuous dunking on the blockchain, miners, and the like. What are good places to start looking for how this works and some stats on things like power wasted by bitcoin mining and such?
Hm. So, for my university course, since we're talking about algorithms and GPUs and shit, my group has been given the topic of bitcoin and crypto mining, to make a presentation for. And since I intend to be as impartial as possible, and my cursory impression is that this whole thing is a massive clusterfuck, I expect that my presentation is going to end up involving at least a solid five minutes of continuous dunking on the blockchain, miners, and the like. What are good places to start looking for how this works and some stats on things like power wasted by bitcoin mining and such?
From earlier in the thread, I liked these two sources:
Hm. So, for my university course, since we're talking about algorithms and GPUs and shit, my group has been given the topic of bitcoin and crypto mining, to make a presentation for. And since I intend to be as impartial as possible, and my cursory impression is that this whole thing is a massive clusterfuck, I expect that my presentation is going to end up involving at least a solid five minutes of continuous dunking on the blockchain, miners, and the like. What are good places to start looking for how this works and some stats on things like power wasted by bitcoin mining and such?
Probably the most important question: How does your prof feel about bitcoin?
Hm. So, for my university course, since we're talking about algorithms and GPUs and shit, my group has been given the topic of bitcoin and crypto mining, to make a presentation for. And since I intend to be as impartial as possible, and my cursory impression is that this whole thing is a massive clusterfuck, I expect that my presentation is going to end up involving at least a solid five minutes of continuous dunking on the blockchain, miners, and the like. What are good places to start looking for how this works and some stats on things like power wasted by bitcoin mining and such?
Probably the most important question: How does your prof feel about bitcoin?
Honestly, I don't know for sure, but he seems pretty neutral on the whole thing. But he's a reasonable enough dude that I do think that he wouldn't fail me for disagreeing with him.
Besides, it's one topic of many. Topic options were stuff like branch predictors, different CPU architectures, graphene and its uses in computer design, and so on.
Branch predictors: how ten plus years of cpus have fundamental security flaws built in
ehh sorta. it's more that no one bothered to notice that those branches were executable memory space that needs to be protected/managed like executable memory space in main memory.
like all this speculative execution stuff is old exploits, that don't tend to work easily these days, but applied to a new part if the computer.
but... yeah.
They moistly come out at night, moistly.
0
Options
Descendant XSkyrim is my god now.Outpost 31Registered Userregular
I picked up Attack of the 50 Foot Blockchain yesterday and am already halfway done reading it. It's quite interesting and is a pretty good crash course in cryptocurrency. I recommend it for anyone who's interested in the subject and wants an entertaining history of all this bollocks.
Garry: I know you gentlemen have been through a lot, but when you find the time I'd rather not spend the rest of the winter TIED TO THIS FUCKING COUCH!
Branch predictors: how ten plus years of cpus have fundamental security flaws built in
ehh sorta. it's more that no one bothered to notice that those branches were executable memory space that needs to be protected/managed like executable memory space in main memory.
like all this speculative execution stuff is old exploits, that don't tend to work easily these days, but applied to a new part if the computer.
but... yeah.
A lot of these exploits weren't all that relevant until the cloud became a thing too. Nobody's using a branch prediction attack on your PC, when they've already had to compromise your OS's ring-0 security to do so in the first place.
A coworker of mine has an amusing blog-post he wrote about when he was at MS debugging a crash failure in... either original xbox or 360. There was a bug deep down in their software, I think due to a specific version of malloc doing something unsupported by the cache? Anyway, they couldn't get off that version so they rewrote their code to never, ever call the offending command.
Only the branch predictor would venture into code that would never get executed checking to see if it might get executed and *boof* crash goes the system.
I'm glad I didn't have to debug that.
PSN,Steam,Live | CptHamiltonian
+12
Options
ShimshaiFlush with Success!Isle of EmeraldRegistered Userregular
A coworker of mine has an amusing blog-post he wrote about when he was at MS debugging a crash failure in... either original xbox or 360. There was a bug deep down in their software, I think due to a specific version of malloc doing something unsupported by the cache? Anyway, they couldn't get off that version so they rewrote their code to never, ever call the offending command.
Only the branch predictor would venture into code that would never get executed checking to see if it might get executed and *boof* crash goes the system.
I'm glad I didn't have to debug that.
This reminds me of one of the Trenches tales, where basically a few weeks from launch of a game, (OXbox or 360, not sure) a tester discovered a bug that would brick the system. It was obscure. In the end I think they pushed the release as-was.
Branch predictors: how ten plus years of cpus have fundamental security flaws built in
Nah, the branch predictors are fine. It's the caches that make everything unsafe. Disable those and you're fine
I thought it was a combination of predictors running ahead and calculating results before resolving security instructions, and placing the results in insecure caches because the sec check hadn't resolved yet and then another process divines the data before the cache was flushed.
HMRC's decision on the taxation of Bitcoin and other cryptocurrencies was published this week. Linky-poo
MASSIVE DISCLAIMER: I am not, nor have I ever been, a tax consultant or tax adviser, nor do I or have I played one on TV.
The short of it for traders is that if you buy and sell bitcoin, it's not gambling, but investment activity. As a result, the profits are liable to Capital Gains Tax. This also means that if you sell a cryptocurrency at a loss, you can offset that loss against other capital gains for the year. The electricity cost of mining is also not allowable as a cost of acquiring the assets for Capital Gains Tax purposes (but may be if the mining itself constitutes a trade).
Branch predictors: how ten plus years of cpus have fundamental security flaws built in
Nah, the branch predictors are fine. It's the caches that make everything unsafe. Disable those and you're fine
I thought it was a combination of predictors running ahead and calculating results before resolving security instructions, and placing the results in insecure caches because the sec check hadn't resolved yet and then another process divines the data before the cache was flushed.
One variant yes, the overarching thing for those vulns is speculative execution can cause attacker-controlled memory access (in several variants) but the way every vulnerability in the last little while is observed is through the cache. Every one of them involves doing something that causes a cache line load to be taken and then you observe that by constantly testing access times. No cache, no information
Yeah, not sure how you do that though without sacrificing significant amounts of speed.
You want the information cached just in case it was actually necessary/allowable in the first place, otherwise there was no real point to the speculative execution.
Ok so this is sort of crypto currency related, but exactly so.
A company that I'm invested in outlined in their end of year video that they're looking to do some sort of equity token - current shares are going to be linked on a 1 to 1 basis to a token, which can then apparently be bought and sold like any other share.
My questions are:
1) While this seems legal, from the brief amount of reading that I've done on it, it's not common. Therefore - are there any actual places that things like this can be bought and sold?
2) Is there any real advantage to doing this, as opposed to a traditional IPO? I'm not sure.
Wisdom of the forums, lay it on me!
0
Options
OrcaAlso known as EspressosaurusWrexRegistered Userregular
Ok so this is sort of crypto currency related, but exactly so.
A company that I'm invested in outlined in their end of year video that they're looking to do some sort of equity token - current shares are going to be linked on a 1 to 1 basis to a token, which can then apparently be bought and sold like any other share.
My questions are:
1) While this seems legal, from the brief amount of reading that I've done on it, it's not common. Therefore - are there any actual places that things like this can be bought and sold?
2) Is there any real advantage to doing this, as opposed to a traditional IPO? I'm not sure.
Wisdom of the forums, lay it on me!
Personally I would view that as evidence the company's a scam, but YMMV, the devil is in the details, etc.
This is also why I'm all about index mutual funds instead of individual stocks...
+2
Options
Ninja Snarl PMy helmet is my burden.Ninja Snarl: Gone, but not forgotten.Registered Userregular
Sounds to me like they're trying to avoid dealing with shares by using tokens as not-shares. Maybe it lets them skirt tax laws or something?
But it seems real sketchy to me. If you own a share, you own a share. The company for that share can't up and decide to do anything funny with that share, and it sounds to me like this "token" idea great for abuse because stocks have all sorts of protections and "tokens" wouldn't. And I would think you would have to redeem the token with the company to get the shares if you ever want to actually sell the shares, because I've never heard of "token" exchanges.
It seems really weird to me. If a token is worth a share, why not just issue/sell shares? This just sounds like a scam with extra steps.
Posts
Also they’re hella cheap compared to other office chairs.
My razor mouse is one of the longest lasting mice I have ever had.
Based on my experience trying to build a gaming pc without any LEDs, there's not much of a market there.
I tried to part one out for a high-end work PC that didn't cost $5000 from a major vendor.
The non-LED version was something like $300-400 more than the LED one. So now I have an RGB PC at work for absolutely 0 reason.
$5 wire cutters will mostly solve this problem.
Rock Band DLC | GW:OttW - arrcd | WLD - Thortar
In fairness my boss did think it was neat because that was the first time he was exposed to RGB.
He asked what my thoughts were on doing something for everyone and making them our company's colors.
I do wonder why I can't find a laptop with a decent graphics card without it being colored like a Sith Lord.
I want there to have been a meeting where someone proposed mining after hours as an additional revenue stream.
"The systems are sitting there unused anyway"
Da red wunz go fasta.
https://www.evga.com/products/product.aspx?pn=516-34-1833-T1
My baby. Got her for 1k earlier this year when the new NVIDIA cards were announced or for sale or something.
Your first PC case: A blinged out case with all the LED's.
Your second PC case: A straight up black box.
This must be why I'm at the black box stage.
I'll always like Razer for making a lefty Deathadder, but this crypto thing is dumb as balls
Inquisitor77: Rius, you are Sisyphus and melee Wizard is your boulder
Tube: This must be what it felt like to be an Iraqi when Saddam was killed
Bookish Stickers - Mrs. Rius' Etsy shop with bumper stickers and vinyl decals.
From earlier in the thread, I liked these two sources:
https://davidgerard.co.uk/blockchain/table-of-contents/
Probably the most important question: How does your prof feel about bitcoin?
Enjoy retaking the class next semester!
PSN / Xbox / NNID: Fodder185
Because it is your job to read undergrad papers?
How to get away with bitcoin
Honestly, I don't know for sure, but he seems pretty neutral on the whole thing. But he's a reasonable enough dude that I do think that he wouldn't fail me for disagreeing with him.
Besides, it's one topic of many. Topic options were stuff like branch predictors, different CPU architectures, graphene and its uses in computer design, and so on.
ehh sorta. it's more that no one bothered to notice that those branches were executable memory space that needs to be protected/managed like executable memory space in main memory.
like all this speculative execution stuff is old exploits, that don't tend to work easily these days, but applied to a new part if the computer.
but... yeah.
A lot of these exploits weren't all that relevant until the cloud became a thing too. Nobody's using a branch prediction attack on your PC, when they've already had to compromise your OS's ring-0 security to do so in the first place.
Only the branch predictor would venture into code that would never get executed checking to see if it might get executed and *boof* crash goes the system.
I'm glad I didn't have to debug that.
This reminds me of one of the Trenches tales, where basically a few weeks from launch of a game, (OXbox or 360, not sure) a tester discovered a bug that would brick the system. It was obscure. In the end I think they pushed the release as-was.
I wonder if the two could be related somehow?
Nah, the branch predictors are fine. It's the caches that make everything unsafe. Disable those and you're fine
I thought it was a combination of predictors running ahead and calculating results before resolving security instructions, and placing the results in insecure caches because the sec check hadn't resolved yet and then another process divines the data before the cache was flushed.
MASSIVE DISCLAIMER: I am not, nor have I ever been, a tax consultant or tax adviser, nor do I or have I played one on TV.
The short of it for traders is that if you buy and sell bitcoin, it's not gambling, but investment activity. As a result, the profits are liable to Capital Gains Tax. This also means that if you sell a cryptocurrency at a loss, you can offset that loss against other capital gains for the year. The electricity cost of mining is also not allowable as a cost of acquiring the assets for Capital Gains Tax purposes (but may be if the mining itself constitutes a trade).
One variant yes, the overarching thing for those vulns is speculative execution can cause attacker-controlled memory access (in several variants) but the way every vulnerability in the last little while is observed is through the cache. Every one of them involves doing something that causes a cache line load to be taken and then you observe that by constantly testing access times. No cache, no information
You want the information cached just in case it was actually necessary/allowable in the first place, otherwise there was no real point to the speculative execution.
A company that I'm invested in outlined in their end of year video that they're looking to do some sort of equity token - current shares are going to be linked on a 1 to 1 basis to a token, which can then apparently be bought and sold like any other share.
My questions are:
1) While this seems legal, from the brief amount of reading that I've done on it, it's not common. Therefore - are there any actual places that things like this can be bought and sold?
2) Is there any real advantage to doing this, as opposed to a traditional IPO? I'm not sure.
Wisdom of the forums, lay it on me!
Personally I would view that as evidence the company's a scam, but YMMV, the devil is in the details, etc.
This is also why I'm all about index mutual funds instead of individual stocks...
But it seems real sketchy to me. If you own a share, you own a share. The company for that share can't up and decide to do anything funny with that share, and it sounds to me like this "token" idea great for abuse because stocks have all sorts of protections and "tokens" wouldn't. And I would think you would have to redeem the token with the company to get the shares if you ever want to actually sell the shares, because I've never heard of "token" exchanges.
It seems really weird to me. If a token is worth a share, why not just issue/sell shares? This just sounds like a scam with extra steps.