Compiler Optimization Nitpick: -nocona v. -core2 in Cely DC

Discussions Regarding Software

Moderator: Moderators

Post Reply
Young Hen
Posts: 24
Joined: Sat Dec 26, 2009 22:13
Location: Accokeek, MD

Compiler Optimization Nitpick: -nocona v. -core2 in Cely DC

Post by PGHammer » Tue Jun 15, 2010 3:48

Okay; here's the situation - I have a Celeron DC E1200 (basically a cut-down Conroe-based Core2Duo without VT-x). I am considering replacing it with one of the following- a Celeron DC E3300 (still C2D-based, but Wolfdale instead of Conroe; this DC supports VT-x) or a Core2Quad Q9300 (two C2Ds in the same die). The question is in terms of compiler optimizations in Sabayon 5.3 x64 K Edition (which I am running now). I *do* want to start doing compiles of my own software, and at the very least for multicore operation; however, I am unsure as to whether the -core2 CFAG/CXXFLAG (despite that the Celly DC is still Core2-based) makes sense from a compiler standpoint (or should I stick with the Olde Reliable -nocona CFLAG/CXXFLAG due to the smaller on-die cache). If the -core2 CFLAG is worth it for the Celeron DC, then *of course* I'll use it right now (as it certainly would be worth it with a quad-core); the question is, should I make the flag change now, or wait until after the CPU swap? (I won't need to change anything else.)

Post Reply