Originally posted by threaded
- Visitors can check out the Forum FAQ by clicking this link. You have to register before you can post: click the REGISTER link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. View our Forum Privacy Policy.
- Want to receive the latest contracting news and advice straight to your inbox? Sign up to the ContractorUK newsletter here. Every sign up will also be entered into a draw to WIN £100 Amazon vouchers!
Reply to: Thought for the day
Collapse
You are not logged in or you do not have permission to access this page. This could be due to one of several reasons:
- You are not logged in. If you are already registered, fill in the form below to log in, or follow the "Sign Up" link to register a new account.
- You may not have sufficient privileges to access this page. Are you trying to edit someone else's post, access administrative features or some other privileged system?
- If you are trying to post, the administrator may have disabled your account, or it may be awaiting activation.
Logging in...
Previously on "Thought for the day"
Collapse
-
Originally posted by AtWWell, I am not planning to target anything apart from x86 in it - so code should run on all, but I am actually interested into SSE instructions (supported by pretty much all modern x86s now) - they might just be ideal to speed data processing nicely - next week is going to be interesting for SKA
It is a bad assumption.
Even the size of the L2 cache can have significance.
threaded in "I've achieved superlinearity, more than once, so ner ner" mode
Leave a comment:
-
Originally posted by VectraManThat's just stupid. They're both template container classes to help you out, neither are meant to be the best possible solution. If you really wanted performance, you should have used your own classes and optimised it to do exactly what you needed. And we don't know that you're comparing like with like here.
Frankly I've always thought STL was needlessly over complicated, and that the MFC container classes are much easier to use.
For my recent heap walk code I wrote my own simple list classes as a) they run much faster than STL and b) they were easy to write. And it allowed me to use a simple pooled allocation scheme to avoid mallocs.
IMO the STL is well worth learning and more powerful than MFC.
I tend to roll my own whenever possible, and then save copies of my libraries, and re-use them at the next client.
A lot of the people at my current client are ex-UNIX bods, and there's an awful lot of open source code from Apache and others. It is another world.
Fungus
Leave a comment:
-
Originally posted by scotspinegreat stuff as long as you know which cpu[s] you're targetting
Leave a comment:
-
great stuff as long as you know which cpu[s] you're targetting
or specifically, instruction set
Leave a comment:
-
Originally posted by scotspinei'd have thought that was an ideal area for some nifty assemby.
Leave a comment:
-
i'd have thought that was an ideal area for some nifty assemby. or better still, [why spoil a friday night?], some good old binary...
Leave a comment:
-
I mean this:
public int iPos
{
get { return oBits[oPos]; }
set { oBits[oPos]=( value<POS_MAX ? value : POS_MAX); }
}
oBits is a BitVector32 and oPos is BitVector32.Section - this code works slowly not because its a function call but because of BitVector itself - I already changed a few places in the past to deal with hex bitmasks directly - this code is likely to be inlined since its small, and performance gains were in region of 1000%.
The alternative code that you have written MAY be faster than bogstandard indexing of BitVector32, but it pays to bother calculate masks manually and use them as constants - sure pain in arse to maintain, but performance improvements are way too high to ignore.
Leave a comment:
-
Originally posted by FungusI once wrote two programs to manipulate maps and lists. The first used STL, the second MFC. Guess which was fastest? Yup. The STL program was twice as fast as the MS one.
Frankly I've always thought STL was needlessly over complicated, and that the MFC container classes are much easier to use.
It's as if MS are creating a demand for geeks with specialist MS knowledge.
Leave a comment:
-
You mean this?
public int this[BitVector32.Section section]
{
get
{
return (int) ((this.data & (section.Mask << (section.Offset & 0x1f))) >> (section.Offset & 0x1f));
}
set
{
value = value << (section.Offset & 0x1f);
int num1 = (0xffff & section.Mask) << (section.Offset & 0x1f);
this.data = (uint) ((this.data & ~num1) | (value & num1));
}
}
Leave a comment:
-
Nothing is wrong, its just not as fast as doing bitshifts yourself - it sure is annoying to create bitmasks manually but its a LOT faster - with luck compiler will also inline those shifts, something it definately won't do for normal BitVector32 operations that use indexer - function calls are not cheap on x86.
Now interface really annoyed me by requiring to use accessors for variables (that were declared in interface) - rather annoying since I can't see any good reason for that.
Leave a comment:
-
Perhaps you can be more specific about what is wrong with the BitVector32 class and I'll see if we can fix it for you?
Leave a comment:
-
Originally posted by AtWPerformance of .NET's BitVector32 is disappointing and having to declare access to (interfaced) variables in interface classes seems to hit performance too.
MS are agressive and well marketed, but often not the best. As an example, COM is widely used, and has helped me earn a living. But IMO it can be fiendishly difficult to maintain, especially for those unskilled in the technology. And the debugging support is awful. It's as if MS are creating a demand for geeks with specialist MS knowledge.
Leave a comment:
-
Originally posted by DundeegeorgeI shall carry them in my heart all day tomorrow.
Probably.
Leave a comment:
- Home
- News & Features
- First Timers
- IR35 / S660 / BN66
- Employee Benefit Trusts
- Agency Workers Regulations
- MSC Legislation
- Limited Companies
- Dividends
- Umbrella Company
- VAT / Flat Rate VAT
- Job News & Guides
- Money News & Guides
- Guide to Contracts
- Successful Contracting
- Contracting Overseas
- Contractor Calculators
- MVL
- Contractor Expenses
Advertisers
Contractor Services
CUK News
- Streamline Your Retirement with iSIPP: A Solution for Contractor Pensions Sep 1 09:13
- Making the most of pension lump sums: overview for contractors Sep 1 08:36
- Umbrella company tribunal cases are opening up; are your wages subject to unlawful deductions, too? Aug 31 08:38
- Contractors, relabelling 'labour' as 'services' to appear 'fully contracted out' won't dupe IR35 inspectors Aug 31 08:30
- How often does HMRC check tax returns? Aug 30 08:27
- Work-life balance as an IT contractor: 5 top tips from a tech recruiter Aug 30 08:20
- Autumn Statement 2023 tipped to prioritise mental health, in a boost for UK workplaces Aug 29 08:33
- Final reminder for contractors to respond to the umbrella consultation (closing today) Aug 29 08:09
- Top 5 most in demand cyber security contract roles Aug 25 08:38
- Changes to the right to request flexible working are incoming, but how will contractors be affected? Aug 24 08:25
Leave a comment: