Your interactive trading laboratory!
 • 
6 users online

hardware recommendations for QS?

Hi, I'm fairly disappointed in QS performance when running it on my iMac with Parallels. I'm considering a dedicated PC for QS and related efforts. Anybody have suggestions on processor types or other considerations to aid my decision toward a cost-effective PC for running lots of QS models? What do you run QS on?

Attachment

Cancel

Responses

Here's my configuration, it's a Dell laptop. What kinds of backtests are you running? Universe size? Oh, and forum post 666 … :)

Here's my configuration, it's a Dell laptop. What kinds of backtests are you running? Universe size? Oh, and forum post 666 … :)

Thanks Glitch I think the problem is with Parallels, and it's not the very latest version of that. My machine is a 3.5Ghz i7 Mac, so it's no real slouch in Mac OS, but doing complicated or long running things (like backtests or Excel macros) causes hiccups and even Parallels failure (I think) as well as maxing out my CPU and causing the fan to run full speed. Maybe it would work fine in Base Camp instead, but I need my Mac for most of the day. Maybe I'll try running it in Base Camp and working on other things from my laptop, or maybe even try Base Camp on my laptop (it's a 2.4Ghz i7). But if it runs well on your laptop, that's great, gives me hope. I am running some very large test Universes (2k+ symbols), so that's a likely contributing factor. I'll try more with my Macs before I give in and make a PC purchase.

Thanks Glitch I think the problem is with Parallels, and it's not the very latest version of that. My machine is a 3.5Ghz i7 Mac, so it's no real slouch in Mac OS, but doing complicated or long running things (like backtests or Excel macros) causes hiccups and even Parallels failure (I think) as well as maxing out my CPU and causing the fan to run full speed. Maybe it would work fine in Base Camp instead, but I need my Mac for most of the day. Maybe I'll try running it in Base Camp and working on other things from my laptop, or maybe even try Base Camp on my laptop (it's a 2.4Ghz i7). But if it runs well on your laptop, that's great, gives me hope. I am running some very large test Universes (2k+ symbols), so that's a likely contributing factor. I'll try more with my Macs before I give in and make a PC purchase.

Cores vs Clock: I'm looking at used laptops... is an i7 with 4 cores/8 threads more important than a faster clock speed but only 2 cores/4 threads?

Cores vs Clock: I'm looking at used laptops... is an i7 with 4 cores/8 threads more important than a faster clock speed but only 2 cores/4 threads?

Bumping up the old thread as I have to make a laptop decision in the near future. Re: memory, how much is needed for large universes that go deep backwards? Wealth-lab published this formula: Required RAM = Number of stocks * number of bars * number of data series (typically 6: DateTime + OHLCV) * 8 byte

Is it about the same for QS? Thanks!

Bumping up the old thread as I have to make a laptop decision in the near future. Re: memory, how much is needed for large universes that go deep backwards? Wealth-lab published this formula: Required RAM = Number of stocks * number of bars * number of data series (typically 6: DateTime + OHLCV) * 8 byte Is it about the same for QS? Thanks!

voron77 said: "Required RAM = Number of stocks * number of bars * number of data series (typically 6: DateTime + OHLCV) * 8 byte Is it about the same for QS? Thanks!"

Yes, OHLCV+DT * 8 bytes each * number of bars. Unlike WL, QS leverages multi-core processors during data loading and optimizations, so that should be a consideration as well.

> voron77 said: "Required RAM = Number of stocks * number of bars * number of data series (typically 6: DateTime + OHLCV) * 8 byte Is it about the same for QS? Thanks!" Yes, OHLCV+DT * 8 bytes each * number of bars. Unlike WL, QS leverages multi-core processors during data loading and optimizations, so that should be a consideration as well.
Forum Tips

Please sign in if you want to participate in our forum.

Our forum uses Markdown syntax to format posts.

To embed code snippets, enclose them in [CODE][/CODE] tags.