[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-31 Thread jason
On Monday 19 January 2009 19:32:39 ja...@njkfrudils.plus.com wrote: > Presently for a 20x20 mul_basecase we have cycle counts of > 2082 gmp-4.2.4 > 1461 mpir trunk > 1153 mpir k8-branch > 1103 pipeline > 1033 perfect > > "pipeline" is my fully pipelined version of mpir k8-branch and "perfect" is >

[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-25 Thread David Harvey
On Jan 25, 9:50 am, ja...@njkfrudils.plus.com wrote: > Another way is to break the mul_basecase rectangle into columns rather than > rows.Here we get a much better basic block of > > mov (src1),AX > mul (src2) > add AX,t1 > adc DX,t2 > adc 0,t3 > > which is 6 macro-ops  , hopefully leading to 2c

[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-25 Thread jason
On Monday 19 January 2009 19:32:39 ja...@njkfrudils.plus.com wrote: > Presently for a 20x20 mul_basecase we have cycle counts of > 2082 gmp-4.2.4 > 1461 mpir trunk > 1153 mpir k8-branch > 1103 pipeline > 1033 perfect > > "pipeline" is my fully pipelined version of mpir k8-branch and "perfect" is >

[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-19 Thread jason
On Monday 19 January 2009 21:54:51 Bill Hart wrote: > It's not clear from the above whether Jason has checked in the fastest > version he has. Beware of that when checking out. The k8 branch he Not checked in yet > refers to is not in trunk yet. It's in a branch called > k8-experimental. > > Bi

[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-19 Thread Bill Hart
It's not clear from the above whether Jason has checked in the fastest version he has. Beware of that when checking out. The k8 branch he refers to is not in trunk yet. It's in a branch called k8-experimental. Bill. On 19/01/2009, ja...@njkfrudils.plus.com wrote: > > On Monday 19 January 2009 2

[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-19 Thread jason
On Monday 19 January 2009 20:39:15 David Harvey wrote: > On Jan 19, 2:32 pm, ja...@njkfrudils.plus.com wrote: > > Presently for a 20x20 mul_basecase we have cycle counts of > > 2082 gmp-4.2.4 > > 1461 mpir trunk > > 1153 mpir k8-branch > > 1103 pipeline > > 1033 perfect > > I'm curious to try this

[mpir-devel] Re: Some thoughts on mul basecase for AMD

2009-01-19 Thread David Harvey
On Jan 19, 2:32 pm, ja...@njkfrudils.plus.com wrote: > Presently for a 20x20 mul_basecase we have cycle counts of > 2082 gmp-4.2.4 > 1461 mpir trunk > 1153 mpir k8-branch > 1103 pipeline > 1033 perfect I'm curious to try this code, but I can't seem to check out the repo right now. Where is it lo