buying hardware to speed up your software

  • (3 Pages)
  • +
  • 1
  • 2
  • 3

30 Replies - 3454 Views - Last Post: 24 July 2018 - 12:28 PM

#16 modi123_1   User is online

  • Suitor #2
  • member icon



Reputation: 14685
  • View blog
  • Posts: 58,683
  • Joined: 12-June 08

Re: buying hardware to speed up your software

Posted 26 May 2018 - 07:47 AM

Yaaaawn.. I know it's the internet.. and I know it's hard, but let's not all get too far down the happy trail in the Windows/Apple fight. ;)
Was This Post Helpful? 0
  • +
  • -

#17 bobsmith76   User is offline

  • D.I.C Regular

Reputation: 11
  • View blog
  • Posts: 343
  • Joined: 14-February 17

Re: buying hardware to speed up your software

Posted 26 May 2018 - 02:23 PM

View PostSkydiver, on 22 May 2018 - 06:02 PM, said:

My gut feel is that if you used a database, you don't have to check all the combinations every time you pick up new predicates. You only need to check combinations of the old predicates with the new predicates added. You shouldn't have to re-check the combinations of the old predicates in combination with the old predicates.


Yea, I've realized that. I really don't need to test the predicates in advance, I can theoretically test them while I'm trying to calculate the sentence's consistency, it's just much easier to do it in advance.
Was This Post Helpful? 0
  • +
  • -

#18 FusionDude   User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 5
  • Joined: 26-July 12

Re: buying hardware to speed up your software

Posted 13 June 2018 - 04:16 PM

Why the focus on 10 seconds vs 30?

I don't see what problem you are trying to solve. The task is essentially monumental and indefinite as described. Word meanings derive from context and usage. This seems to be computation without purpose. If you have a subset of vocabulary you are examining for a particular use, such as cooking or political debate; it would begin to make sense. However, it is also an indefinite task except within a time frame, shaped by examining some particular aspect of the vocabulary in use and your analytical purpose.
If you are trying to solve a problem, then 10 seconds of computation or 10 minutes doesn't make any difference, if you achieve your analysis. In a work sense, the better your algorithm and the better your gear then, of course, the more work you could get done.

Offhand, it looks like you are kicking out lists of possibly meaningful sentences for some casual display. In that case, there is no driving need for you to buy or change anything.
Was This Post Helpful? 0
  • +
  • -

#19 Skydiver   User is online

  • Code herder
  • member icon

Reputation: 6657
  • View blog
  • Posts: 22,740
  • Joined: 05-May 12

Re: buying hardware to speed up your software

Posted 13 June 2018 - 05:23 PM

View PostFusionDude, on 13 June 2018 - 07:16 PM, said:

Why the focus on 10 seconds vs 30?

...

If you are trying to solve a problem, then 10 seconds of computation or 10 minutes doesn't make any difference, if you achieve your analysis. In a work sense, the better your algorithm and the better your gear then, of course, the more work you could get done.

So we should just continue to use bubble sort even when we have thousands of items as long as the result is a sorted list?

Or it okay for this web page to load in 5 minutes instead of 5 seconds as long as the page eventually loads?

Or for backup of your machine to take 3 days as long as a good backup is created? Or better yet, for a restore of your machine to take one week as long as a good image is restored -- nevermind that your term paper that was backed up is due tomorrow?
Was This Post Helpful? 1
  • +
  • -

#20 FusionDude   User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 5
  • Joined: 26-July 12

Re: buying hardware to speed up your software

Posted 13 June 2018 - 05:28 PM

That;s a nonsense response. Skydiver.
Was This Post Helpful? 0
  • +
  • -

#21 Skydiver   User is online

  • Code herder
  • member icon

Reputation: 6657
  • View blog
  • Posts: 22,740
  • Joined: 05-May 12

Re: buying hardware to speed up your software

Posted 13 June 2018 - 05:56 PM

I was just trying to see where your pain tolerance for waiting as long as task is accomplished vs. need to find a better algorithm and/or hardware.
Was This Post Helpful? 0
  • +
  • -

#22 FusionDude   User is offline

  • New D.I.C Head

Reputation: 0
  • View blog
  • Posts: 5
  • Joined: 26-July 12

Re: buying hardware to speed up your software

Posted 13 June 2018 - 06:00 PM

Well, sorry, I was just trying to get a better idea of what he was doing.

Some problems are worth no more than the initial try, others keep popping up and deserve effort.

Of course, a lot depends on what you are getting paid and when, or, of course, personal interest in tweaking something forever.
Was This Post Helpful? 0
  • +
  • -

#23 jon.kiparsky   User is offline

  • Beginner
  • member icon


Reputation: 11374
  • View blog
  • Posts: 19,406
  • Joined: 19-March 11

Re: buying hardware to speed up your software

Posted 13 June 2018 - 07:55 PM

So I think there's a couple of reasonable points here. First, if the question was just about thirty seconds here or there, you're right, we wouldn't care. That's one thing. On the other hand, if this is thirty seconds that's part of your regular workflow and you're held up by it several times an hour, then it starts to feel frustrating and like a roadblock and maybe it takes on an importance on that basis - either an emotional importance, just from the sense of being impeded, or perhaps a more significant cost of context recovery. So that's a case where that thirty seconds could start to matter.
By what's interesting is that things don't typically scale up by ten seconds, then twenty, then thirty, and six months from now you're up to a minute and a half. If a process hits thirty seconds to complete, it's hard not to suspect that it's going to hit the upswing of that curve shortly and become totally unusable - to borrow from Skydiver's example, if you're happily sorting the words in various text files and you find that sorting Robert Frost's poems and E. B. White's essays is going along great, and then you maybe run into a hitch when you try The Old Man and the Sea, that should make you a little nervous that things are going to go badly when you get to Tolstoy, and maybe that's a good time to reconsider your use of bubblesort as your algorithm.
Was This Post Helpful? 1
  • +
  • -

#24 bobsmith76   User is offline

  • D.I.C Regular

Reputation: 11
  • View blog
  • Posts: 343
  • Joined: 14-February 17

Re: buying hardware to speed up your software

Posted 13 July 2018 - 10:17 AM

I had to get a new computer. Here are the performance stats for the two different computers. I ran the same program in both computers.



macbookpro 8,2
intel core i7
2.3 ghz
memory 16 gb
4 cores, late 2011

program speed: 15 seconds


macbookpro 11,4
intel core i7
2.2 ghz
memory 16 gb
4 cores, late 2015

program speed: 9 seconds
Was This Post Helpful? 1
  • +
  • -

#25 Salem_c   User is offline

  • void main'ers are DOOMED
  • member icon

Reputation: 2278
  • View blog
  • Posts: 4,362
  • Joined: 30-May 10

Re: buying hardware to speed up your software

Posted 14 July 2018 - 08:17 AM

To a first approximation, it looks like you've doubled the speed.

I wonder what memory modules are in the old and new machine?
https://en.wikipedia...tandard_modules
Better memory would help if you're processing large volumes of data.

Is it the same OS on both machines?

Is it still written in Python?
What version(s) of the interpreter are running on both machines?

How large are the data files on disk? If they're very large files (10's to 100's of MB or more), then better hard disks (or an SSD) will improve things.
Was This Post Helpful? 0
  • +
  • -

#26 bobsmith76   User is offline

  • D.I.C Regular

Reputation: 11
  • View blog
  • Posts: 343
  • Joined: 14-February 17

Re: buying hardware to speed up your software

Posted 14 July 2018 - 03:30 PM

Hi Salem, thanks for your interest.

View PostSalem_c, on 14 July 2018 - 08:17 AM, said:

To a first approximation, it looks like you've doubled the speed.


I wonder what memory modules are in the old and new machine?
https://en.wikipedia...tandard_modules
Better memory would help if you're processing large volumes of data.

Are you talking about RAM? If so, the RAM is the same, 16 GB


Quote

Is it the same OS on both machines? Is it still written in Python?

yes and yes


Quote

What version(s) of the interpreter are running on both machines?

Python 3.6

Quote

How large are the data files on disk?

The files are quite small. Around 1 MB.

Quote

If they're very large files (10's to 100's of MB or more), then better hard disks (or an SSD) will improve things.

The second MAC does have an SSD drive.
Was This Post Helpful? 0
  • +
  • -

#27 Salem_c   User is offline

  • void main'ers are DOOMED
  • member icon

Reputation: 2278
  • View blog
  • Posts: 4,362
  • Joined: 30-May 10

Re: buying hardware to speed up your software

Posted 15 July 2018 - 01:17 AM

I was referring to the memory technology used to implement the RAM - like DDR2, DDR3 or DDR4

The later modules are higher performance, and thus more likely to keep up with the demands from the CPU.
Was This Post Helpful? 0
  • +
  • -

#28 bobsmith76   User is offline

  • D.I.C Regular

Reputation: 11
  • View blog
  • Posts: 343
  • Joined: 14-February 17

Re: buying hardware to speed up your software

Posted 15 July 2018 - 02:33 PM

View PostSalem_c, on 15 July 2018 - 01:17 AM, said:

I was referring to the memory technology used to implement the RAM - like DDR2, DDR3 or DDR4

The later modules are higher performance, and thus more likely to keep up with the demands from the CPU.



In that case the second Mac is DDR3.

I turned in the first Mac but based on this website

https://support.appl...n-us/ht201165#2

it seems that the memory specs for a Macbook pro 15 inch late 2011 are

- Double Data Rate Small Outline Dual Inline Memory Module (DDR3) format
- 30mm (1.18 inch)
- 2 GB or 4 GB
- 204-pin
- PC3-10600 DDR3 1333 MHz type RAM

So the memory card is DDR3 on both.

I don't know all that much about computer hardware. I was told at the store that GHz and also from this article

https://www.techrepu...-fast-cpu-fast/

that GHz is no longer a good predictor of CPU speed. What I would like to know then is what CPU chip I have in both computers and what differentiates them. They're both intel core i7 but I have a feeling that they are not exactly the same.
Was This Post Helpful? 0
  • +
  • -

#29 Skydiver   User is online

  • Code herder
  • member icon

Reputation: 6657
  • View blog
  • Posts: 22,740
  • Joined: 05-May 12

Re: buying hardware to speed up your software

Posted 15 July 2018 - 05:33 PM

Could be as simple of dual core vs. quad core, or different L1 and L2 cache sizes...
Was This Post Helpful? 0
  • +
  • -

#30 bobsmith76   User is offline

  • D.I.C Regular

Reputation: 11
  • View blog
  • Posts: 343
  • Joined: 14-February 17

Re: buying hardware to speed up your software

Posted 17 July 2018 - 12:51 PM

or it could be the SSD drive that's being used. Also, simple wear and tear does slow the computer down or so I've been told.
Was This Post Helpful? 0
  • +
  • -

  • (3 Pages)
  • +
  • 1
  • 2
  • 3