[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference hydra::amiga_v1

Title:AMIGA NOTES
Notice:Join us in the *NEW* conference - HYDRA::AMIGA_V2
Moderator:HYDRA::MOORE
Created:Sat Apr 26 1986
Last Modified:Wed Feb 05 1992
Last Successful Update:Fri Jun 06 1997
Number of topics:5378
Total number of notes:38326

4845.0. "'turbo' 68000 7/14mhz for ~�40" by TRUCKS::BUSSINK_E (Honey, I iconified the kids !) Mon Jun 24 1991 09:34

    Well, I got last week an 68000 'Turbo' 7/14mhz chip.
    
    - The first I recived from Bytes&Pieces was dead, so it took them a while
      to get me an other one. (I ordered this chip end February�).
    - It's easy to fit in the computer, if you've got the adequate tools.
      AND beware the thin legs !!! :-) had some trouble there too ...
    - This chip ticks faster but it doesn't increase the speed that much.And
      at 14mhz it doesn't reconize the extra memory (I've got a 8-UP 2mega). 
    
    Anyhow, it's working ! Now I got the test from CPU_BENCHMARK.LZH on
    Tape and run them on my computer in the two speed modes. With the small
    program MIPS,I had .82 Mips at 7mhz and .92-.99 at ~13mhz.
    With the CPU_MEMY at normal speed i've got an 1 factor, and 1.05 factor
    at 14mhz. Each different benchmark program gives me another speed for
    my processor � Now with the AIBBv3.lzh the speed the program gives me
    is 4.82mhz when the processor is supposed to work at 14mhz is there a
    display fault ? and the factor is ~84%. Even slower than at 7mhz which
    already makes only ~86% compared with a normal A2000 � what is
    happening ?
    
    		Any clues � 
    		Anybody else having this strange mutant processor ?
    
    						Erik
    
    System :	B2000 (1meg)
    		1084
    		Trumpcard with 46meg Seagate
    		8-UP (2meg)
    		NEC pinwriter 2200
    		& external drive
T.RTitleUserPersonal
Name
DateLines
4845.1did I miss somethingSALEM::LEIMBERGERMon Jun 24 1991 11:114
    I must have missed something here. I did not see you mention changing
    the Oscillator for the CPU. If you install a faster CPU don't you need
    to increase the clock rate to take advantage mof the faster cpu ?
    								bill
4845.2No, I didTRUCKS::BUSSINK_EHoney, I iconified the kids !Mon Jun 24 1991 11:346
    This CPU processor comes on a small board, and has an additional chip.
    There is a small switch that you get out of the machine in the front
    (Near the mouse, doesn't look great but does the job) and a small
    'probe' that connects on the oscillator without soldering.
    
    							Erik
4845.313Mhz - Sound Familiar !.VIVIAN::D_VISTUERMon Jun 24 1991 18:0846
    Erik,
    	I would like to butt in here. I read the Amiga Shopper write-up on
    this beast. Their example did not work either. However, this is not the
    reason I am writing this reply.
    
    I also have a need for speed, so I tried to build the New14Accel.lzh
    project. It did not work. So, I built the CheapAccel.lzh (?) project
    and it did !!.
    
    I then ran as many benchtest against it while it was working (which
    was, to my surprise about 90% of the time). The benchtests ranged from
    SI2 version 0b to Aibbv3. As a comparison I also ran these same tests
    (I this 12 in all) on the standard 68000 @ 7Mhz and a 68010 @ 7Mhz.
    
    It was a close run thing between the 68010/7 and the 68000/14 but
    the Overall winner with a % improvement over a 68000/7 was the 
    68000/14 with 29%. The 68010/7 scored 11%.
    
    However, the 68010 was the clear winner when I ran CpuSped which does
    lots of different types of memory/register manipulations. In some cases
    scoring 80% faster than a 68000/7 when moving a registers contents to
    fastmem. The best score for the 68000/14 in this suite of tests was a 
    score of 66% when doing an unsigned divide in data register.
     
    I intend to write up the results in a file for others to read in the
    near future.
    
    BTW: I have looked into the AdSpeed a it seems to be quite a wicked
    beast for the cash, especially as it offers 100% compatablity and a
    speed increase that is noticeable !.
    
    Before I go, you mentioned that one of the benchtests you ran gave you
    clock speed as 13.xx Mhz - strange, cos the CheapAccel.lzh hack gave me 
    a speed rating of 13.32 Mhz in some tests and Aibb3 gave 7.67 Mhz. The
    mind boggles !!!. Me thinks its cos the 8520s being clocked at twice the
    speed now causes the thing to screw around hence giving erroneous
    results. In closing, would it be possible for you to tell me the chips
    used in your TURBO board cos I think its exactly the same as the
    CheapAccel.lzh project.
    
    Anyway, better go now, running out of electronic ink !.
    
    Regards.
    
    Dave Vistuer.