Jump in the discussion.

No email address required.

I have 256GB of ram in my desktop. This is just laptopcels posting their L's.

Jump in the discussion.

No email address required.

wtf do you need that much ram for

Jump in the discussion.

No email address required.

data analysis and trading simulations.

Jump in the discussion.

No email address required.

Imagine having enough RAM to run an entire small business' IT infrastructure and using it to not beat the S&P500

Jump in the discussion.

No email address required.

First sign you dont know anything is you think 'beating the s&p 500' is what people with billions of dollar to invest care about.

Jump in the discussion.

No email address required.

Post returns

Jump in the discussion.

No email address required.

:#marseydeadcatbounce:

Jump in the discussion.

No email address required.

:mjlol:

Jump in the discussion.

No email address required.

I dont trade for myself and returns are only part of the story. Let me just give everyone a pro tip: if you read something in the press about trading and it talks only takes about returns but not risk, correlation to othet assets or drawdown profile, you are reading the words of an r-slur.

Jump in the discussion.

No email address required.

Just get an AWS subscription and use spark at that point, shit

Jump in the discussion.

No email address required.

No offense bro, but this is bad advise. This entire system, with 18 cores, 256 GB or ram, 4TB of NVME and a high end (at the time) GPU was under 5K. If would be even cheaper if you build it yourself. You will never replicate this performance for this price on AWS and spark is bad for any kind of exploratory analysis. Too slow to distribute and collect. Throw on a decent NAS with at least 2.5GB ports and some 10+TB disks and you will to sleep anyone who reaches for AWS for this kind of setup. Cloud can get you to the next level of scale, but most people don't actually need it and reach for it way too early.

Jump in the discussion.

No email address required.

Jokes on you, I have a foolproof way to get better trading returns at a lower price point than either a homelab or public cloud (I'm running for Congress)

Jump in the discussion.

No email address required.

I have a server setup with the RAM and processor you're describing, but it's commercial server hardware and cost me $3500. I am absolutely not using it as a desktop computer. My gaming PC cost well under half of that. Are you sure you're not 100% full of shit?

Jump in the discussion.

No email address required.

The high end GPU probably made the difference

Jump in the discussion.

No email address required.

You may be right about cloud compute for price, my compute is all paid for so I've never actually budgeted it out. spark is only slow (even for eda) if you don't know how to use it or youre data is pretty small. Seems like a pain to get that shit together compared to just using a tool meant for dealing with it

Jump in the discussion.

No email address required.

Right tool for the right job. Most trading applications arent big data. Its small to medium and very heterogeneous. There are cases where i would still want to run on the cloud because i need access to many more cores or gpus, which is generally at model fitting time.

Jump in the discussion.

No email address required.

Not op but I have the same because I do a lot of high end computer simulations and they need a boat load of ram.

Jump in the discussion.

No email address required.

I load my entire 2TB drive into RAM every time my computer boots. When you're a speedy hackster like me, SSD's can't keep up. The 2 millisecond delay reading files from disk costs me the equivalent productivity of 3.2 man hours of a normal person's time. When I am waiting for vi to load from a hard disk, I can physically feel my hair growing longer

Jump in the discussion.

No email address required.

I can see it is very hard for people to imagine that applications exist beyond basic word processing and bideo games. The spectrum of users here seem to range from neets all the way up to admin assistants. It is much easier and much faster to do any kind of data analytics if can fit the data in memory. In the last couple of years, polars and duckdb have made this a lot better, but its stll way better to be in memory. With pandas or R, its especially important.

Jump in the discussion.

No email address required.

When you grow up and get tired of playing with pythons and pandas, you can join the adults by investing in a real analysis machine. In my world, we inspect packets at the individual electron level to get the jump on low-tech chumps like you.

I don't have time for low-energy nerds. Stay in the slow lane and out of my way

Jump in the discussion.

No email address required.

Thanks for the tip Brent.

https://media.giphy.com/media/GI5lkXChv0VB0z1glJ/giphy.webp

Jump in the discussion.

No email address required.

I'm always happy to give tips to promising youngsters. Hope my tough love style didn't get under your skin too much

Jump in the discussion.

No email address required.

What for? Ramdisk? What's on it?

Jump in the discussion.

No email address required.

Reported by:
  • Zizo : Too dumb to live

:wow: how'd you manage that? I don't know computer hardware well but I imagine that must take up tons of space.

Jump in the discussion.

No email address required.

I imagine the ram sticks are basically all the same size. He just spent a lot of money and slotted them in? Are you fricking r-slurred? Do you think that powerful processors take more space in the case than weaker ones? Are you too stupid to remember how to breathe? Keep yourself safe.

Jump in the discussion.

No email address required.

>Are you fricking r-slurred?

Yes.

>Do you think that powerful processors take more space in the case than weaker ones?

Yes.

>Are you too stupid to remember how to breathe?

Yes.

>Keep yourself safe.

:marseyletsfuckinggo2: I AM NOT AFRAID TO KEEP ON LIVING

I AM NOT AFRAID TO WALK THIS WORLD ALONE :marseywindmill:

Jump in the discussion.

No email address required.

:#gigachad2talking:

Jump in the discussion.

No email address required.

its simple really. more memory takes up more space :tayscrunch:

how do you not get it?

Jump in the discussion.

No email address required.

I don't remember.

Jump in the discussion.

No email address required.

RETVRN TO VACCUM TUBES


Give me your money and I'll annoy people with it :space: https://i.rdrama.net/images/16965516366194396.webp

Jump in the discussion.

No email address required.

Best comment :marseysoypointdubz: I've seen on this site in years

Jump in the discussion.

No email address required.

Its just 4 slots of RAM at 64 GB each.

Jump in the discussion.

No email address required.

It's "just" that if you're deciding on your setup using only a calculator, but here in reality, there is nothing you can buy to throw in your PC that will support 256 GB.

Jump in the discussion.

No email address required.

https://media.giphy.com/media/hPPx8yk3Bmqys/giphy.webp

Jump in the discussion.

No email address required.

Big dimms. I didnt build it, got it from a hedt builder.

Jump in the discussion.

No email address required.

No you don't you liar. There isn't a single consumer motherboard on the market that supports that.

Jump in the discussion.

No email address required.

This neighbor ain't never heard of a threadripper workstation. Shit supports 2TB ram for about 1000 dollars.

Jump in the discussion.

No email address required.

Jump in the discussion.

No email address required.

https://media.giphy.com/media/3oz8xLd9DJq2l2VFtu/giphy.webp

X299-pro. I didnt realize this post would blast so many bussys with jealousy.

Jump in the discussion.

No email address required.

Link copied to clipboard
Action successful!
Error, please refresh the page and try again.