i7 2600 @ 3.4GHz

---------- testing started ----------testing x87Starting matrix multiplication loop...Elapsed time: 2059 mstesting sse1Starting matrix multiplication loop...Elapsed time: 2168 mstesting sse2Starting matrix multiplication loop...Elapsed time: 2028 ms---------- testing finished ----------

AMD Phenom II X4 920 @ 2.8ghz

---------- testing started ----------

Geeks3D.com - x87 / SSE2 Test
-----------------------------
Instruction set: x87
Starting matrix multiplication loop...
Elapsed time: 750 ms
Bye!

Geeks3D.com - x87 / SSE2 Test
-----------------------------
Instruction set: SSE
Starting matrix multiplication loop...
Elapsed time: 710 ms
Bye!

Geeks3D.com - x87 / SSE2 Test
-----------------------------
Instruction set: SSE2
Starting matrix multiplication loop...
Elapsed time: 1070 ms
Bye!
---------- testing finished ----------

Mine is Phenom 9950 BE stock speed


Geeks3D.com - x87 / SSE2 Test
-----------------------------
Instruction set: x87
Starting matrix multiplication loop...
Elapsed time: 827 ms
Bye!

Geeks3D.com - x87 / SSE2 Test
-----------------------------
Instruction set: SSE
Starting matrix multiplication loop...
Elapsed time: 781 ms
Bye!

Geeks3D.com - x87 / SSE2 Test
-----------------------------
Instruction set: SSE2
Starting matrix multiplication loop...
Elapsed time: 1166 ms
Bye!
---------- testing finished ----------

So you are correct. SSE2 is slower. Now the question is: why?

I absolutely love what you have done. Thank you!!!!!!!

Using the newer 94.6 KB download:
AMD Athlon X4 630 @ 3.2GHz with a GTX 560 Ti 1 GB ~ 953 MHz core.

C:\Users\Diego\Desktop>echo off---------- testing started ----------testing x87Starting matrix multiplication loop...Elapsed time: 4368 mstesting sse1Starting matrix multiplication loop...Elapsed time: 5741 mstesting sse2Starting matrix multiplication loop...Elapsed time: 5663 ms---------- testing finished ----------Press any key to continue . . .

Here is my results from the new test

---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 4290 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 6053 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 7847 ms
---------- testing finished ----------

I tried again and got this


---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 4259 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 5912 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 5850 ms
---------- testing finished ----------

New test AMD X2 250 X2 3.0 ghz

QuadCore AMD Phenom II X4 Black Edition 965, 3600 MHz (18 x 200)

I wanna help!

AMD Phenom II x4 965 @3.8 ghz

More intell users plz !

Athlon 64 x2 3800+ 2.0Ghz

---------- testing started ----------testing x87Starting matrix multiplication loop...Elapsed time: 6490 mstesting sse1Starting matrix multiplication loop...Elapsed time: 7925 mstesting sse2Starting matrix multiplication loop...Elapsed time: 7535 ms---------- testing finished ----------

i7 2600k @ 4.5GHz

---------- testing started ----------testing x87Starting matrix multiplication loop...Elapsed time: 2062 mstesting sse1Starting matrix multiplication loop...Elapsed time: 2125 mstesting sse2Starting matrix multiplication loop...Elapsed time: 2015 ms---------- testing finished ----------

i5 2500k @ stock clock (3.3 ghz)

---------- testing started ----------testing x87Starting matrix multiplication loop...Elapsed time: 1950 mstesting sse1Starting matrix multiplication loop...Elapsed time: 2012 mstesting sse2Starting matrix multiplication loop...Elapsed time: 1888 ms---------- testing finished ----------

AMD Phenom II x4 @ 3.2GHz (originally x3 720 @2.8 GHz, if that makes any difference)

---------- testing started ----------testing x87Starting matrix multiplication loop...Elapsed time: 3775 mstesting sse1Starting matrix multiplication loop...Elapsed time: 5102 mstesting sse2Starting matrix multiplication loop...Elapsed time: 4680 ms---------- testing finished ----------

Intel Q6600 @ 2.4 GHz


---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 6193 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 4711 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 4321 ms
---------- testing finished ----------

AMD PhenonII x4 810 (OC 3.1Ghz)



---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 3891 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 5250 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 5016 ms
---------- testing finished ----------

This looks fun!

FX-6100

---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 2574 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 2777 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 2870 ms
---------- testing finished ----------
Press any key to continue . . .


Edit: Is it just me or doesn't it look like most people are getting better times on x87? Besides the q6600.
It looks like sse1 is faster than sse2.. for amd at least, actually it is mixed, it think?

i7 860 Stock Clocks



C:\Users\Wade\Documents\fputest>echo off
---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 3479 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 3463 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 3432 ms
---------- testing finished ----------
Press any key to continue . . .

edit. thanks trira!

Right click, click on Select All and then press Enter

Intel i5-2430M 2.4 ghz (3.0 ghz turbo) in high performance mode:



C:\Users\NathanT\Downloads\fputest>echo off
---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 2808 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 2948 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 2792 ms
---------- testing finished ----------
Press any key to continue . . .

Processor 1 ID = 0
Number of cores 2 (max 2)
Number of threads 2 (max 2)
Name Intel Core 2 Duo E8400
Codename Wolfdale
Specification Intel® Core™2 Duo CPU E8400 @ 3.00GHz
Package (platform ID) Socket 775 LGA (0x0)
CPUID 6.7.6
Extended CPUID 6.17
Core Stepping C0
Technology 45 nm
TDP Limit 65 Watts
Core Speed 3600.0 MHz
Multiplier x FSB 9.0 x 400.0 MHz
Rated Bus speed 1600.0 MHz
Stock frequency 3000 MHz
Instructions sets MMX, SSE, SSE2, SSE3, SSSE3, SSE4.1, EM64T, VT-x
L1 Data cache 2 x 32 KBytes, 8-way set associative, 64-byte line size
L1 Instruction cache 2 x 32 KBytes, 8-way set associative, 64-byte line size
L2 cache 6144 KBytes, 24-way set associative, 64-byte line size
FID/VID Control yes
FID range 6.0x - 9.0x
Max VID 1.225 V


---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 3042 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 2559 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 2215 ms
---------- testing finished ----------
Press any key to continue . . .

I dont know much about all this, but it seems most AMD users are getting lower numbers in SSE1 does that means SSE1 is faster? and SSE3 seems to be faster than SSE2 slightly.

Here's my results from a fresh restart:

CPU: Intel Core 2 Quad Q8200 @ 3.33Ghz (1904Mhz FSB)

---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 3152 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 2698 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 2387 ms
---------- testing finished ----------

(Looking at the above posts SSE2 does seem alot better on Intel then AMD, odd.)

I've been watching the beth forums for ages and this mod is just awesome, benchmark details below

Fast Travled to Markarth.
Vanilla: 33 FPS (http://i43.tinypic.com/z0s5z.jpg w/ FPS Counter)
SkyBoost R3: 44 FPS (http://i41.tinypic.com/mkbper.jpg w/ FPS Counter)


Very NICE work Alex, can't wait for future releases. Maybe there can be two versions? x87 for AMD and SSE2 for Intel, just a thought.

e6550 oc'ed @ 3.01

AMD Athlon 64 X2 3600+

>all.bat

>echo off
---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 6428 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 7894 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 7426 ms
---------- testing finished ----------

Intel E6600 OC 3.42GHZ XP32 Home 2GB RAM

D:\temp>echo off
---------- testing started ----------
testing x87

Starting matrix multiplication loop...
Elapsed time: 3531 ms
testing sse1

Starting matrix multiplication loop...
Elapsed time: 2921 ms
testing sse2

Starting matrix multiplication loop...
Elapsed time: 2515 ms
---------- testing finished ----------
Press any key to continue . . .