Click Here
home features news forums classifieds faqs links search
5765 members 
Amiga Q&A /  Free for All /  Emulation /  Gaming / (Latest Posts)
Login

Nickname

Password

Lost Password?

Don't have an account yet?
Register now!

Support Amigaworld.net
Your support is needed and is appreciated as Amigaworld.net is primarily dependent upon the support of its users.
Donate

Menu
Main sections
Home
Features
News
Forums
Classifieds
Links
Downloads
Extras
OS4 Zone
IRC Network
AmigaWorld Radio
Newsfeed
Top Members
Amiga Dealers
Information
About Us
FAQs
Advertise
Polls
Terms of Service
Search

IRC Channel
Server: irc.amigaworld.net
Ports: 1024,5555, 6665-6669
SSL port: 6697
Channel: #Amigaworld
Channel Policy and Guidelines

Who's Online
34 crawler(s) on-line.
 22 guest(s) on-line.
 2 member(s) on-line.


 1Mouse,  Yogi27

You are an anonymous user.
Register Now!
 1Mouse:  4 mins ago
 Yogi27:  4 mins ago
 dirkzwager:  6 mins ago
 NutsAboutAmiga:  13 mins ago
 towo2099:  22 mins ago
 eliyahu:  27 mins ago
 freak:  44 mins ago
 Hexaae:  52 mins ago
 PhantomInterrogative:  1 hr 2 mins ago
 Everblue:  1 hr 19 mins ago

/  Forum Index
   /  Amiga OS4 Hardware
      /  Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Register To Post

Goto page ( 1 | 2 | 3 | 4 Next Page )
PosterThread
wegster 
Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 5:49:36
#1 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

It can be had from os4depot here

If you don't mind, please run and either PM or post the results of:

ramspeed -b 1
ramspeed -b 2
ramspeed -b 3
ramspeed -b 4
ramspeed -b 5
ramspeed -b 6

EDIT- please also run -b3 and -b6 tests, I was mistaken on what those tests were doing, explanation in later posts. I've also made several long-term runs showing complete reproducibility on my XE with variances of less than a few percent over many iterations, so a single run of each should be sufficient.

Yes, I've got some other results, but really would like to see the numbers from a uA1-C for comparison. Would also be good to see the same from Peg I and Peg 2 owners, although I'm not sure if MOS can run this version of ramspeed or not. If not, you can get details and grab v2.3.1 from www.alisir.com, but I don't believe it will build out of the box. If it will, you'll still have to comment out these lines from defines.h:
/* blackmark gcc-2.96 and 3.x.x */
#if ((__GNUC__ == 2) && (__GNUC_MINOR__ == 96)) || (__GNUC__ > 2)
#error Refuse to be compiled by GCC3 without I386_ASM
as there's no asm target for PPC in the Makefile or asm PPC code. gcc 3.X evidently at least on Intel 'optimizes' the floating point operations into integer operations. This does NOT appear to be the case on OS X at least, but I can confirm that behavior on Intel.

If anyone 'follows along at home,' there's a clear delineation when you run this of exactly where the CPU cache runs out (you'll see it). It's normal, as you're now going off-chip and over the bus to your socketed RAM. This is also why the -b3 and -b6 tests are semi-useless, they take an average, which may be nice for marketing stuff, specially if you limit the block size, but it's of little use here.

So, what you'll see is something along the lines of:
INTEGER & WRITING 1 Kb block: HIGH # Mb/s
INTEGER & WRITING 2 Kb block: HIGH # Mb/s
INTEGER & WRITING 4 Kb block: HIGH # Mb/s
INTEGER & WRITING 8 Kb block: HIGH # Mb/s
...
...
INTEGER & WRITING 256 Kb block: HIGH # Mb/s -> XE-G4/7455 out of cache after this
INTEGER & WRITING 512 Kb block: LOW-MEDIUM Mb/s -> 7455 cache still giving a bit of boost
INTEGER & WRITING 1024 Kb block: 'REAL' # Mb/s -> most of the data is going across the bus.

The XE/7455 CPUs will hit the low/low-medium point at the 512kb block size, and the 750FX/uA1-C presumably at the 2048k block size, and those are the numbers I'm mostly interested in.

So, can anyone do the above...please? (on uA1-C, Pegs, and SE or XE-G3?)

EDIT- this _is_ a more applicable/general test that the dnetc numbers I posted previously. Just curious to see what the 'real' info is on our systems, as I've yet to see any valid comparison data (doesn't mean it doesn't exist, just *I* haven't found it! )

Last edited by wegster on 05-Mar-2005 at 11:29 AM.
Last edited by wegster on 05-Mar-2005 at 06:02 AM.
Last edited by wegster on 05-Mar-2005 at 05:55 AM.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
mr_homm 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:15:17
#2 ]
Regular Member
Joined: 21-Mar-2003
Posts: 180
From: Seattle

@wegster

When I follow your link and try to get it, the file ramspeed.lha just
gives me the web page back. If I force it to download using the right mouse button, it still just gives the html for the current page. I tried to UNARC the file ramspeed.lha and UNARC says it's empty. So I looked at it in Notepad, and it's just a bunch of HTML.

In short, I think the link is broken. Either that, or I've suddenly forgotten how to use OS4Depot.

--Stuart Anderson

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:20:28
#3 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@mr_homm
Sorry, maybe I'm having one of those days, lemme check again,
link

I think that works. If not, go to http://www.os4depot and search for 'ramspeed', it's the only one there.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
rinaldo00 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:32:50
#4 ]
Regular Member
Joined: 14-Dec-2004
Posts: 301
From: Unknown

@wegster

link works fine

you like OWE me dude ;)

ramspeed -b 1
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

INTEGER & WRITING 1 Kb block: 2925.71 Mb/s
INTEGER & WRITING 2 Kb block: 2925.71 Mb/s
INTEGER & WRITING 4 Kb block: 2925.71 Mb/s
INTEGER & WRITING 8 Kb block: 2968.12 Mb/s
INTEGER & WRITING 16 Kb block: 2925.71 Mb/s
INTEGER & WRITING 32 Kb block: 2968.12 Mb/s
INTEGER & WRITING 64 Kb block: 1484.06 Mb/s
INTEGER & WRITING 128 Kb block: 1484.06 Mb/s
INTEGER & WRITING 256 Kb block: 1473.38 Mb/s
INTEGER & WRITING 512 Kb block: 1484.06 Mb/s
INTEGER & WRITING 1024 Kb block: 288.05 Mb/s
INTEGER & WRITING 2048 Kb block: 123.90 Mb/s
INTEGER & WRITING 4096 Kb block: 123.90 Mb/s
INTEGER & WRITING 8192 Kb block: 123.90 Mb/s
INTEGER & WRITING 16384 Kb block: 123.97 Mb/s

ramspeed -b 2
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

INTEGER & READING 1 Kb block: 2968.12 Mb/s
INTEGER & READING 2 Kb block: 3011.76 Mb/s
INTEGER & READING 4 Kb block: 2968.12 Mb/s
INTEGER & READING 8 Kb block: 3011.76 Mb/s
INTEGER & READING 16 Kb block: 2968.12 Mb/s
INTEGER & READING 32 Kb block: 2968.12 Mb/s
INTEGER & READING 64 Kb block: 1575.38 Mb/s
INTEGER & READING 128 Kb block: 1575.38 Mb/s
INTEGER & READING 256 Kb block: 1587.60 Mb/s
INTEGER & READING 512 Kb block: 1587.60 Mb/s
INTEGER & READING 1024 Kb block: 429.35 Mb/s
INTEGER & READING 2048 Kb block: 198.45 Mb/s
INTEGER & READING 4096 Kb block: 198.45 Mb/s
INTEGER & READING 8192 Kb block: 198.26 Mb/s
INTEGER & READING 16384 Kb block: 198.45 Mb/s


ramspeed -b 3
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

INTEGER Copy: 132.43 Mb/s
INTEGER Scale: 131.75 Mb/s
INTEGER Add: 132.19 Mb/s
INTEGER Triad: 138.07 Mb/s
---
INTEGER AVERAGE: 133.61 Mb/s

ramspeed -b 4
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

FL-POINT & WRITING 1 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 2 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 4 Kb block: 5688.89 Mb/s
FL-POINT & WRITING 8 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 16 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 32 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 64 Kb block: 1932.08 Mb/s
FL-POINT & WRITING 128 Kb block: 1932.08 Mb/s
FL-POINT & WRITING 256 Kb block: 1932.08 Mb/s
FL-POINT & WRITING 512 Kb block: 1914.02 Mb/s
FL-POINT & WRITING 1024 Kb block: 296.81 Mb/s
FL-POINT & WRITING 2048 Kb block: 124.57 Mb/s
FL-POINT & WRITING 4096 Kb block: 124.57 Mb/s
FL-POINT & WRITING 8192 Kb block: 124.57 Mb/s
FL-POINT & WRITING 16384 Kb block: 124.57 Mb/s

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:36:11
#5 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@rinaldo00
Which model system/CPU etc? It looks like it's got 512k cache, not 1M...is that the uA1 or something else?

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
rinaldo00 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:38:29
#6 ]
Regular Member
Joined: 14-Dec-2004
Posts: 301
From: Unknown

@wegster

ramspeed -b 5
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

FL-POINT & READING 1 Kb block: 5851.43 Mb/s
FL-POINT & READING 2 Kb block: 6023.53 Mb/s
FL-POINT & READING 4 Kb block: 6023.53 Mb/s
FL-POINT & READING 8 Kb block: 6023.53 Mb/s
FL-POINT & READING 16 Kb block: 6023.53 Mb/s
FL-POINT & READING 32 Kb block: 5851.43 Mb/s
FL-POINT & READING 64 Kb block: 2133.33 Mb/s
FL-POINT & READING 128 Kb block: 2133.33 Mb/s
FL-POINT & READING 256 Kb block: 2155.79 Mb/s
FL-POINT & READING 512 Kb block: 2155.79 Mb/s
FL-POINT & READING 1024 Kb block: 447.16 Mb/s
FL-POINT & READING 2048 Kb block: 198.07 Mb/s
FL-POINT & READING 4096 Kb block: 198.07 Mb/s
FL-POINT & READING 8192 Kb block: 198.07 Mb/s
FL-POINT & READING 16384 Kb block: 198.07 Mb/s

 Status: Offline
Profile     Report this post  
rinaldo00 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:43:38
#7 ]
Regular Member
Joined: 14-Dec-2004
Posts: 301
From: Unknown

@wegster
"Which model system/CPU etc? It looks like it's got 512k cache, not 1M...is that the uA1 or something else?"

You sure? It is a Micro1-Ac cpu 750Gx

Last edited by rinaldo00 on 05-Mar-2005 at 06:44 AM.

 Status: Offline
Profile     Report this post  
ktadd 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:45:34
#8 ]
Cult Member
Joined: 30-Jul-2003
Posts: 555
From: California, USA

@wegster
If you don't mind, please run and either PM or post the results of:

========
Here are the results from my uA1C 256M RAM GX CPU:

ramspeed -b 1
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04
4Gb per pass mode

INTEGER & WRITING 1 Kb block: 2925.71 Mb/s
INTEGER & WRITING 2 Kb block: 2925.71 Mb/s
INTEGER & WRITING 4 Kb block: 2925.71 Mb/s
INTEGER & WRITING 8 Kb block: 2925.71 Mb/s
INTEGER & WRITING 16 Kb block: 2925.71 Mb/s
INTEGER & WRITING 32 Kb block: 2925.71 Mb/s
INTEGER & WRITING 64 Kb block: 1473.38 Mb/s
INTEGER & WRITING 128 Kb block: 1473.38 Mb/s
INTEGER & WRITING 256 Kb block: 1473.38 Mb/s
INTEGER & WRITING 512 Kb block: 1473.38 Mb/s
INTEGER & WRITING 1024 Kb block: 314.11 Mb/s
INTEGER & WRITING 2048 Kb block: 121.26 Mb/s
INTEGER & WRITING 4096 Kb block: 121.11 Mb/s
INTEGER & WRITING 8192 Kb block: 120.83 Mb/s
INTEGER & WRITING 16384 Kb block: 120.90 Mb/s



6.RAM Disk:> ramspeed -b 2
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04
4Gb per pass mode

INTEGER & READING 1 Kb block: 2968.12 Mb/s
INTEGER & READING 2 Kb block: 2968.12 Mb/s
INTEGER & READING 4 Kb block: 2968.12 Mb/s
INTEGER & READING 8 Kb block: 2968.12 Mb/s
INTEGER & READING 16 Kb block: 2968.12 Mb/s
INTEGER & READING 32 Kb block: 2968.12 Mb/s
INTEGER & READING 64 Kb block: 1563.36 Mb/s
INTEGER & READING 128 Kb block: 1575.38 Mb/s
INTEGER & READING 256 Kb block: 1575.38 Mb/s
INTEGER & READING 512 Kb block: 1563.36 Mb/s
INTEGER & READING 1024 Kb block: 260.56 Mb/s
INTEGER & READING 2048 Kb block: 184.84 Mb/s
INTEGER & READING 4096 Kb block: 186.01 Mb/s
INTEGER & READING 8192 Kb block: 191.94 Mb/s
INTEGER & READING 16384 Kb block: 191.40 Mb/s



6.RAM Disk:> ramspeed -b 4
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04
4Gb per pass mode

FL-POINT & WRITING 1 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 2 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 4 Kb block: 5688.89 Mb/s
FL-POINT & WRITING 8 Kb block: 5688.89 Mb/s
FL-POINT & WRITING 16 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 32 Kb block: 5851.43 Mb/s
FL-POINT & WRITING 64 Kb block: 1914.02 Mb/s
FL-POINT & WRITING 128 Kb block: 1896.30 Mb/s
FL-POINT & WRITING 256 Kb block: 1896.30 Mb/s
FL-POINT & WRITING 512 Kb block: 1896.30 Mb/s
FL-POINT & WRITING 1024 Kb block: 286.83 Mb/s
FL-POINT & WRITING 2048 Kb block: 121.76 Mb/s
FL-POINT & WRITING 4096 Kb block: 121.98 Mb/s
FL-POINT & WRITING 8192 Kb block: 121.98 Mb/s
FL-POINT & WRITING 16384 Kb block: 121.98 Mb/s



6.RAM Disk:> ramspeed -b 5
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04
4Gb per pass mode

FL-POINT & READING 1 Kb block: 5851.43 Mb/s
FL-POINT & READING 2 Kb block: 5851.43 Mb/s
FL-POINT & READING 4 Kb block: 5851.43 Mb/s
FL-POINT & READING 8 Kb block: 5851.43 Mb/s
FL-POINT & READING 16 Kb block: 6023.53 Mb/s
FL-POINT & READING 32 Kb block: 5851.43 Mb/s
FL-POINT & READING 64 Kb block: 2133.33 Mb/s
FL-POINT & READING 128 Kb block: 2133.33 Mb/s
FL-POINT & READING 256 Kb block: 2133.33 Mb/s
FL-POINT & READING 512 Kb block: 2111.34 Mb/s
FL-POINT & READING 1024 Kb block: 497.09 Mb/s
FL-POINT & READING 2048 Kb block: 193.03 Mb/s
FL-POINT & READING 4096 Kb block: 193.03 Mb/s
FL-POINT & READING 8192 Kb block: 193.03 Mb/s
FL-POINT & READING 16384 Kb block: 192.48 Mb/s



Last edited by ktadd on 05-Mar-2005 at 06:47 AM.

_________________
Kevin - X1000 First Contact / uA1

 Status: Offline
Profile     Report this post  
rinaldo00 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 6:51:13
#9 ]
Regular Member
Joined: 14-Dec-2004
Posts: 301
From: Unknown

@ktadd

Ok, yours are about the same as mine, just a TAD slower.

What CPU is it? uboot will tell you, or %c in the WB titlebar

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:00:47
#10 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@rinaldo00
Hmm, that's interesting. Obviously a 1024Kb block won't fit entirely in your cache, but most of it will, so I'd expect the numbers going from 512k to 1Mb blocks to be closer than this:
FL-POINT & READING 512 Kb block: 2155.79 Mb/s
FL-POINT & READING 1024 Kb block: 447.16 Mb/s

But it looks like the other uA1-C does the same.

Numbers from my 'hardware fixed' A1XE-G4 933MHz, CPU 7455 (256k L2 cache):

5.Workbench:> ramspeed -b 1
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

INTEGER & WRITING 1 Kb block: 2968.12 Mb/s
INTEGER & WRITING 2 Kb block: 2968.12 Mb/s
INTEGER & WRITING 4 Kb block: 3011.76 Mb/s
INTEGER & WRITING 8 Kb block: 3011.76 Mb/s
INTEGER & WRITING 16 Kb block: 2968.12 Mb/s
INTEGER & WRITING 32 Kb block: 2968.12 Mb/s
INTEGER & WRITING 64 Kb block: 1505.88 Mb/s
INTEGER & WRITING 128 Kb block: 1484.06 Mb/s
INTEGER & WRITING 256 Kb block: 1356.29 Mb/s
INTEGER & WRITING 512 Kb block: 271.26 Mb/s
INTEGER & WRITING 1024 Kb block: 221.17 Mb/s
INTEGER & WRITING 2048 Kb block: 201.38 Mb/s
INTEGER & WRITING 4096 Kb block: 192.66 Mb/s
INTEGER & WRITING 8192 Kb block: 188.24 Mb/s
INTEGER & WRITING 16384 Kb block: 200.39 Mb/s
5.Workbench:> ramspeed -b 2
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

INTEGER & READING 1 Kb block: 3471.19 Mb/s
INTEGER & READING 2 Kb block: 3471.19 Mb/s
INTEGER & READING 4 Kb block: 3531.03 Mb/s
INTEGER & READING 8 Kb block: 3531.03 Mb/s
INTEGER & READING 16 Kb block: 3471.19 Mb/s
INTEGER & READING 32 Kb block: 3531.03 Mb/s
INTEGER & READING 64 Kb block: 1651.61 Mb/s
INTEGER & READING 128 Kb block: 1651.61 Mb/s
INTEGER & READING 256 Kb block: 1505.88 Mb/s
INTEGER & READING 512 Kb block: 247.64 Mb/s
INTEGER & READING 1024 Kb block: 207.29 Mb/s
INTEGER & READING 2048 Kb block: 204.60 Mb/s
INTEGER & READING 4096 Kb block: 204.60 Mb/s
INTEGER & READING 8192 Kb block: 204.39 Mb/s
INTEGER & READING 16384 Kb block: 204.39 Mb/s
5.Workbench:> ramspeed -b 4
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

FL-POINT & WRITING 1 Kb block: 2327.27 Mb/s
FL-POINT & WRITING 2 Kb block: 2327.27 Mb/s
FL-POINT & WRITING 4 Kb block: 2354.02 Mb/s
FL-POINT & WRITING 8 Kb block: 2354.02 Mb/s
FL-POINT & WRITING 16 Kb block: 2354.02 Mb/s
FL-POINT & WRITING 32 Kb block: 2354.02 Mb/s
FL-POINT & WRITING 64 Kb block: 2068.69 Mb/s
FL-POINT & WRITING 128 Kb block: 2048.00 Mb/s
FL-POINT & WRITING 256 Kb block: 1845.05 Mb/s
FL-POINT & WRITING 512 Kb block: 534.73 Mb/s
FL-POINT & WRITING 1024 Kb block: 435.74 Mb/s
FL-POINT & WRITING 2048 Kb block: 416.26 Mb/s
FL-POINT & WRITING 4096 Kb block: 411.24 Mb/s
FL-POINT & WRITING 8192 Kb block: 410.42 Mb/s
FL-POINT & WRITING 16384 Kb block: 425.78 Mb/s
5.Workbench:> ramspeed -b 5
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

FL-POINT & READING 1 Kb block: 6826.67 Mb/s
FL-POINT & READING 2 Kb block: 6826.67 Mb/s
FL-POINT & READING 4 Kb block: 7062.07 Mb/s
FL-POINT & READING 8 Kb block: 7062.07 Mb/s
FL-POINT & READING 16 Kb block: 7062.07 Mb/s
FL-POINT & READING 32 Kb block: 7062.07 Mb/s
FL-POINT & READING 64 Kb block: 2155.79 Mb/s
FL-POINT & READING 128 Kb block: 2155.79 Mb/s
FL-POINT & READING 256 Kb block: 1969.23 Mb/s
FL-POINT & READING 512 Kb block: 254.73 Mb/s
FL-POINT & READING 1024 Kb block: 212.89 Mb/s
FL-POINT & READING 2048 Kb block: 209.62 Mb/s
FL-POINT & READING 4096 Kb block: 208.34 Mb/s
FL-POINT & READING 8192 Kb block: 210.05 Mb/s
FL-POINT & READING 16384 Kb block: 210.05 Mb/s

And just for giggles:
5.Workbench:> ramspeed -b 3
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

INTEGER Copy: 159.56 Mb/s
INTEGER Scale: 159.19 Mb/s
INTEGER Add: 154.68 Mb/s
INTEGER Triad: 168.37 Mb/s
---
INTEGER AVERAGE: 160.45 Mb/s

5.Workbench:> ramspeed -b 6
RAMspeed (UNIX) v2.3.0 by Rhett M. Hollander (Alasir Enterprises), 2002-04

4Gb per pass mode

FL-POINT Copy: 189.63 Mb/s
FL-POINT Scale: 164.96 Mb/s
FL-POINT Add: 151.37 Mb/s
FL-POINT Triad: 153.03 Mb/s
---
FL-POINT AVERAGE: 164.75 Mb/s

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:03:23
#11 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@rinaldo00
I'd bet money it's the same exact CPU and speed as yours. A few % are generally within tolerance of most benchmarks, so for a '100$ accurate test' it should be run over many iterations and your results would likely be within 1 point of each other.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:08:48
#12 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@wegster
So, what this basically boils down to:
1. Throw out the in cache results, and the 'mostly in cache'>results.

2. average sustained transfer rate for my A1XE-G4 7455/256K L2 cache:
INT READ: ~205MB/second
INT WRITE: ~190MB/second
FP READ: ~209MB/second
FP WRITE: ~415MB/second

I'd wager that the compile here for the OS4depot test may also be optimizing one way or the other, note the similarity between INT and FP read speeds. Either way, it's a baseline for comparison.

Now, because I'm in IBrowse with only one window allowed at a time (grr!), I need to post this then edit while looking at the uA1-C numbers..sigh.

Last edited by wegster on 05-Mar-2005 at 07:13 AM.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:11:42
#13 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@ktadd
So, if we throw away the in cache results, the uA1-C:
INT READ: ~188 MB/sec?
INT WRITE: ~121 MB/sec
FP READ: ~193 MB/sec
FP WRITE: ~122 MB/sec

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
sundown 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:13:29
#14 ]
Elite Member
Joined: 30-Aug-2003
Posts: 5115
From: Right here...

@wegster

A1 750GX here & I got the same numbers as the other 2 A1's.

Wouldn't programs running in the background take up some space in the L cache?

_________________
Hate tends to make you look stupid...

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:16:14
#15 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@sundown

Yes- I think I wrote originally it's bet to do this on a clean boot with minimal processes running (ie stop dnetc especially if running at > priority 1, or anything that remains _active_)

From watching the same test under Linux, the ramspeed test pretty much takes over running priority, so unless you have some oddball (like dnet running high priority) processes going, I'd expect the results to remain relatively close regardless.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:21:10
#16 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@thread

For clarity (and b/c I can't open two browsers at the same time..):

uA1-C, G3-750GX 800MHz/1MB cache CPU:
INT READ: ~188 MB/sec>
INT WRITE: ~121 MB/sec
FP READ: ~193 MB/sec
FP WRITE: ~122 MB/sec

XE-G4 7557/256K cache CPU:
INT READ: ~205MB/second
INT WRITE: ~190MB/second
FP READ: ~209MB/second
FP WRITE: ~415MB/second

So those are _reasonably_ close (I've run these several times, not seen more than 1-2MB/difference per iteration), but:

1. WTF is up with the INT write speeds on the uA1-C? 121MB/sec versus 190MB/sec??! That's more than different grade of RAM can produce.

2. FP writes- I may need to look at the source or do some homework on this one, that's obviously a huge difference, but may be an optimization on the G4 CPUs.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:43:33
#17 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@thread

For comparison:
Apple PowerMac dual 7557 CPU/256L2 cache/2MB L3(enabled), PC133 CL3, OS X 10.3.8.
Bear in mind this is intentionally the single CPU/non SMP version of ramspeed (the SMP version basically picks up an extra hit from the second CPU's cache size so runs a lot higher 'in cache' and a bit higher when normally above the cache siz), but this should be comparable:
rotten:~/bench/ramspeed-2.3.1 wegster$ uname -a
Darwin rotten.local 7.8.0 Darwin Kernel Version 7.8.0: Wed Dec 22 14:26:17 PST 2
004; root:xnu/xnu-517.11.1.obj~1/RELEASE_PPC Power Macintosh powerpc
rotten:~/bench/ramspeed-2.3.1 wegster$ gcc -v
Reading specs from /usr/libexec/gcc/darwin/ppc/3.3/specs
Thread model: posix
gcc version 3.3 20030304 (Apple Computer, Inc. build 1495)

rotten:~/bench/ramspeed-2.3.1 wegster$ ./ramspeed -b1
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER & WRITING 1 Kb block: 3169.55 Mb/s
INTEGER & WRITING 2 Kb block: 3200.91 Mb/s
INTEGER & WRITING 4 Kb block: 3214.13 Mb/s
INTEGER & WRITING 8 Kb block: 3217.17 Mb/s
INTEGER & WRITING 16 Kb block: 3221.49 Mb/s
./ramINTEGER & WRITING 32 Kb block: 3173.75 Mb/s
speed -b2INTEGER & WRITING 64 Kb block: 1638.60 Mb/s

./ramspeeINTEGER & WRITING 128 Kb block: 1627.42 Mb/s
d -b3
./ramsINTEGER & WRITING 256 Kb block: 1468.12 Mb/s
peed -b4
./ramspeINTEGER & WRITING 512 Kb block: 1105.83 Mb/s
ed -b5
./ramspeed INTEGER & WRITING 1024 Kb block: 983.57 Mb/s
-b6
INTEGER & WRITING 2048 Kb block: 667.91 Mb/s
INTEGER & WRITING 4096 Kb block: 382.34 Mb/s
INTEGER & WRITING 8192 Kb block: 338.77 Mb/s
INTEGER & WRITING 16384 Kb block: 334.41 Mb/s
rotten:~/bench/ramspeed-2.3.1 wegster$ ./ramspeed -b2
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER & READING 1 Kb block: 3721.81 Mb/s
INTEGER & READING 2 Kb block: 3742.92 Mb/s
INTEGER & READING 4 Kb block: 3768.74 Mb/s
INTEGER & READING 8 Kb block: 3767.62 Mb/s
INTEGER & READING 16 Kb block: 3770.51 Mb/s
INTEGER & READING 32 Kb block: 3757.59 Mb/s
INTEGER & READING 64 Kb block: 949.24 Mb/s
INTEGER & READING 128 Kb block: 881.12 Mb/s
INTEGER & READING 256 Kb block: 769.92 Mb/s
INTEGER & READING 512 Kb block: 782.73 Mb/s
INTEGER & READING 1024 Kb block: 632.11 Mb/s
INTEGER & READING 2048 Kb block: 526.47 Mb/s
INTEGER & READING 4096 Kb block: 415.98 Mb/s
INTEGER & READING 8192 Kb block: 386.23 Mb/s
INTEGER & READING 16384 Kb block: 382.56 Mb/s
rotten:~/bench/ramspeed-2.3.1 wegster$ ./ramspeed -b3
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER Copy: 360.57 Mb/s
INTEGER Scale: 360.82 Mb/s
INTEGER Add: 352.76 Mb/s
INTEGER Triad: 330.25 Mb/s
---
INTEGER AVERAGE: 351.10 Mb/s

rotten:~/bench/ramspeed-2.3.1 wegster$ ./ramspeed -b4
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT & WRITING 1 Kb block: 2493.42 Mb/s
FL-POINT & WRITING 2 Kb block: 2494.37 Mb/s
FL-POINT & WRITING 4 Kb block: 2510.86 Mb/s
FL-POINT & WRITING 8 Kb block: 2516.95 Mb/s
FL-POINT & WRITING 16 Kb block: 2519.05 Mb/s
FL-POINT & WRITING 32 Kb block: 2513.77 Mb/s
FL-POINT & WRITING 64 Kb block: 2074.60 Mb/s
FL-POINT & WRITING 128 Kb block: 2064.71 Mb/s
FL-POINT & WRITING 256 Kb block: 1949.36 Mb/s
FL-POINT & WRITING 512 Kb block: 1630.62 Mb/s
FL-POINT & WRITING 1024 Kb block: 1424.60 Mb/s
FL-POINT & WRITING 2048 Kb block: 900.23 Mb/s
FL-POINT & WRITING 4096 Kb block: 510.10 Mb/s
FL-POINT & WRITING 8192 Kb block: 438.91 Mb/s
FL-POINT & WRITING 16384 Kb block: 432.32 Mb/s
rotten:~/bench/ramspeed-2.3.1 wegster$ ./ramspeed -b5
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT & READING 1 Kb block: 7161.11 Mb/s
FL-POINT & READING 2 Kb block: 7450.39 Mb/s
FL-POINT & READING 4 Kb block: 7500.69 Mb/s
FL-POINT & READING 8 Kb block: 7516.66 Mb/s
FL-POINT & READING 16 Kb block: 7537.24 Mb/s
FL-POINT & READING 32 Kb block: 7505.45 Mb/s
FL-POINT & READING 64 Kb block: 2319.06 Mb/s
FL-POINT & READING 128 Kb block: 1449.04 Mb/s
FL-POINT & READING 256 Kb block: 1370.64 Mb/s
FL-POINT & READING 512 Kb block: 853.01 Mb/s
FL-POINT & READING 1024 Kb block: 722.99 Mb/s
FL-POINT & READING 2048 Kb block: 570.13 Mb/s
FL-POINT & READING 4096 Kb block: 424.05 Mb/s
FL-POINT & READING 8192 Kb block: 387.12 Mb/s
FL-POINT & READING 16384 Kb block: 382.42 Mb/s
rotten:~/bench/ramspeed-2.3.1 wegster$ ./ramspeed -b6
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT Copy: 383.73 Mb/s
FL-POINT Scale: 384.73 Mb/s
FL-POINT Add: 401.58 Mb/s
FL-POINT Triad: 402.89 Mb/s
---
FL-POINT AVERAGE: 393.23 Mb/s

Ok, so summary again:

uA1-C, G3-750GX 800MHz/1MB cache CPU:
INT READ: ~188 MB/sec>
INT WRITE: ~121 MB/sec
FP READ: ~193 MB/sec
FP WRITE: ~122 MB/sec

XE-G4 7557/256K cache CPU:
INT READ: ~205MB/second
INT WRITE: ~190MB/second
FP READ: ~209MB/second
FP WRITE: ~415MB/second

Apple PowerMac Quicksilver, Dual G4-1GHz/7557 CPU/256L2/2MB L3, PC133 RAM, but non SMP benchmark:
INT READ: 394 MB/sec
INT WRITE: 352 MB/sec
FP READ: 397 MB/sec
FP WRITE: 460 MB/sec

NOTE- I know this has been discussed and supposedly the L3 cache wasn't useful on our CPUs, but evidently it certainly affects this benchmark, so I included what's obviously the L3 cache out from the PowerMac results, only using the values for the 4096, 8193, and 16384 blocks for the 'average.'

1. WTF is up with the INT write speeds on the uA1-C? 121MB/sec versus 190MB/sec??! That's more than different grade of RAM can produce.

2. FP writes- I may need to look at the source or do some homework on this one, that's obviously a huge difference, but may be an optimization on the G4 CPUs.

3. PowerMac with similar hardware, same CPU (but L3 cache enabled)...seems to _smoke_ the A1s by almost double throughput, with the exception of the XE G4 FP writes. I'd say the preliminary results would so far point torwards we do in fact have a POS memory controller, thanks MAI. I'm open to other interpretations, anyone?

Still would like to see some SEs, other XEs, and Pegs in here.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 7:51:45
#18 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@thread

Ok, so here's my XE-G4 under Debian Linux:
amigaone:~/bench/ramspeed-2.3.1# uname -a
Linux amigaone 2.4.26 #1 Thu Nov 11 16:16:20 EST 2004 ppc GNU/Linux
amigaone:~/bench/ramspeed-2.3.1# cat /proc/cpuinfo
processor : 0
cpu : 7455, altivec supported
revision : 2.1 (pvr 8001 0201)
bogomips : 930.61
machine : AmigaOne G3SE / AmigaOne XE / Teron CX / Teron PX
RAM : 700 MB
clock : 933 MHz
fsb : 133 MHz
l2cr : 0x80000000

amigaone:~/bench/ramspeed-2.3.1# gcc -v
Reading specs from /usr/lib/gcc-lib/powerpc-linux/3.3.4/specs
Configured with: ../src/configure -v --enable-languages=c,c++,java,f77,pascal,ob
jc,ada --prefix= /usr --mandir=/usr/share/man --infodir=/usr/share/i
nfo --with-gxx-include-dir=/usr/include/c++/3 .3 --enable-shared --w
ith-system-zlib --enable-nls --without-included-gettext --enable-__cxa_ate
xit --enable-clocale=gnu --enable-debug --enable-java-gc=boehm --enable-j
ava-awt=xlib --enable-o bjc-gc --disable-multilib powerpc-linux
Thread model: posix
gcc version 3.3.4 (Debian 1:3.3.4-13)
amigaone:~/bench/ramspeed-2.3.1# ./ramspeed -b 1 | tee benchmarks.ram

amigaone:~/bench/ramspeed-2.3.1# ./ramspeed -b 1
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER & WRITING 1 Kb block: 2923.32 Mb/s
INTEGER & WRITING 2 Kb block: 2936.97 Mb/s
INTEGER & WRITING 4 Kb block: 2921.35 Mb/s
INTEGER & WRITING 8 Kb block: 2917.78 Mb/s
INTEGER & WRITING 16 Kb block: 2964.96 Mb/s
INTEGER & WRITING 32 Kb block: 2935.96 Mb/s
INTEGER & WRITING 64 Kb block: 1520.81 Mb/s
INTEGER & WRITING 128 Kb block: 1132.77 Mb/s
INTEGER & WRITING 256 Kb block: 658.07 Mb/s
INTEGER & WRITING 512 Kb block: 273.32 Mb/s
INTEGER & WRITING 1024 Kb block: 210.21 Mb/s
INTEGER & WRITING 2048 Kb block: 180.75 Mb/s
INTEGER & WRITING 4096 Kb block: 172.56 Mb/s
INTEGER & WRITING 8192 Kb block: 198.41 Mb/s
INTEGER & WRITING 16384 Kb block: 204.28 Mb/s
amigaone:~/bench/ramspeed-2.3.1# ./ramspeed -b 2
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER & READING 1 Kb block: 3480.48 Mb/s
INTEGER & READING 2 Kb block: 3499.87 Mb/s
INTEGER & READING 4 Kb block: 3518.01 Mb/s
INTEGER & READING 8 Kb block: 3519.59 Mb/s
INTEGER & READING 16 Kb block: 3513.54 Mb/s
INTEGER & READING 32 Kb block: 3476.41 Mb/s
INTEGER & READING 64 Kb block: 1649.22 Mb/s
INTEGER & READING 128 Kb block: 1647.48 Mb/s
INTEGER & READING 256 Kb block: 616.47 Mb/s
INTEGER & READING 512 Kb block: 252.09 Mb/s
INTEGER & READING 1024 Kb block: 208.63 Mb/s
INTEGER & READING 2048 Kb block: 205.95 Mb/s
INTEGER & READING 4096 Kb block: 206.00 Mb/s
INTEGER & READING 8192 Kb block: 205.74 Mb/s
INTEGER & READING 16384 Kb block: 206.22 Mb/s
amigaone:~/bench/ramspeed-2.3.1# ./ramspeed -b 4
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT & WRITING 1 Kb block: 2322.78 Mb/s
FL-POINT & WRITING 2 Kb block: 2342.00 Mb/s
FL-POINT & WRITING 4 Kb block: 2343.23 Mb/s
FL-POINT & WRITING 8 Kb block: 2350.77 Mb/s
FL-POINT & WRITING 16 Kb block: 2339.72 Mb/s
FL-POINT & WRITING 32 Kb block: 2349.73 Mb/s
FL-POINT & WRITING 64 Kb block: 2144.02 Mb/s
FL-POINT & WRITING 128 Kb block: 2122.87 Mb/s
FL-POINT & WRITING 256 Kb block: 1060.42 Mb/s
FL-POINT & WRITING 512 Kb block: 550.47 Mb/s
FL-POINT & WRITING 1024 Kb block: 411.07 Mb/s
FL-POINT & WRITING 2048 Kb block: 383.63 Mb/s
FL-POINT & WRITING 4096 Kb block: 373.02 Mb/s
FL-POINT & WRITING 8192 Kb block: 378.36 Mb/s
FL-POINT & WRITING 16384 Kb block: 377.05 Mb/s
amigaone:~/bench/ramspeed-2.3.1# ./ramspeed -b 5
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT & READING 1 Kb block: 6901.49 Mb/s
FL-POINT & READING 2 Kb block: 6990.93 Mb/s
FL-POINT & READING 4 Kb block: 6993.56 Mb/s
FL-POINT & READING 8 Kb block: 7028.35 Mb/s
FL-POINT & READING 16 Kb block: 7015.39 Mb/s
FL-POINT & READING 32 Kb block: 6799.21 Mb/s
FL-POINT & READING 64 Kb block: 2153.24 Mb/s
FL-POINT & READING 128 Kb block: 2149.58 Mb/s
FL-POINT & READING 256 Kb block: 697.17 Mb/s
FL-POINT & READING 512 Kb block: 246.51 Mb/s
FL-POINT & READING 1024 Kb block: 203.22 Mb/s
FLFL-POINT & READING 4096 Kb block: 199.85 Mb/s
FL-POINT & READING 8192 Kb block: 210.40 Mb/s
FL-POINT & READING 16384 Kb block: 209.63 Mb/s
FL-POINT & READING 2048 Kb block: 202.45 Mb/s

Let's compare this to the same system under OS4, for 'sanity':

XE-G4 7557/256K cache CPU: (AOS4)
INT READ: ~205MB/second
INT WRITE: ~190MB/second
FP READ: ~209MB/second
FP WRITE: ~415MB/second

XE-G4 7557/256K cache CPU: (Debian 2.4.26 Linux):
INT READ: ~206 MB/sec
INT WRITE: ~200 MB/sec
FP READ: ~205 MB/sec
FP WRITE: ~378 MB/sec

OK, so those numbers check out at this point...some discrepancy with the FP WRITE values (415MB vs 378, AOS run actually did _better_), but the rest are within tolerance of +/- 5%.

So there's either no OS specific issue here, or both Linux AND AOS4 have issues in the same way here, or our memory controller is a relative POS so far.

The _theoretical_ max of PC133 RAM is ~1GB/second, but that number is far off from reality. Going to go check one more box, but if anyone has comments or another run on a different box, please add your data points.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 8:37:16
#19 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@wegster
FreeBSD 6, 2x667Mhz PIII CPU/256K L2 cache, PC133 Registered ECC RAM, xSeries 45
00R
[root@freeb] /tmp
[0] # uname -a
FreeBSD freeb 6.0-CURRENT FreeBSD 6.0-CURRENT #0: Mon Jan 10 08:10:37 EST 2005
root@freeb :/usr/obj/usr/home/freeb/src/sys/SMP53 i386
[root@freeb] /tmp
[0] # cpuid
Vendor ID: "GenuineIntel"; CPUID level 2

Intel-specific functions:
Version 00000681:
Type 0 - Original OEM
Family 6 - Pentium Pro
Model 8 - Pentium III/Pentium III Xeon - internal L2 cache
Stepping 1
Reserved 0

Brand index: 2 [Pentium III processor]

Feature flags 0383fbff:
FPU Floating Point Unit
VME Virtual 8086 Mode Enhancements
DE Debugging Extensions
PSE Page Size Extensions
TSC Time Stamp Counter
MSR Model Specific Registers
PAE Physical Address Extension
MCE Machine Check Exception
CX8 COMPXCHG8B Instruction
APIC On-chip Advanced Programmable Interrupt Controller present and enabled
SEP Fast System Call
MTRR Memory Type Range Registers
PGE PTE Global Flag
MCA Machine Check Architecture
CMOV Conditional Move and Compare Instructions
FGPAT Page Attribute Table
PSE-36 36-bit Page Size Extension
MMX MMX instruction set
FXSR Fast FP/MMX Streaming SIMD Extensions save/restore
SSE Streaming SIMD Extensions instruction set

TLB and cache info:
01: Instruction TLB: 4KB pages, 4-way set assoc, 32 entries
02: Instruction TLB: 4MB pages, 4-way set assoc, 2 entries
03: Data TLB: 4KB pages, 4-way set assoc, 64 entries
82: 2nd-level cache: 256KB, 8-way set assoc, 32 byte line size
08: 1st-level instruction cache: 16KB, 4-way set assoc, 32 byte line size
04: Data TLB: 4MB pages, 4-way set assoc, 8 entries
0c: 1st-level data cache: 16KB, 4-way set assoc, 32 byte line size
[root@freeb] /tmp
[1] # ./ramspeed.fbsd -b1
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER & WRITING 1 Kb block: 2179.03 Mb/s
INTEGER & WRITING 2 Kb block: 2186.06 Mb/s
INTEGER & WRITING 4 Kb block: 2224.35 Mb/s
INTEGER & WRITING 8 Kb block: 2243.53 Mb/s
INTEGER & WRITING 16 Kb block: 2250.41 Mb/s
INTEGER & WRITING 32 Kb block: 1662.67 Mb/s
INTEGER & WRITING 64 Kb block: 1654.85 Mb/s
INTEGER & WRITING 128 Kb block: 1667.80 Mb/s
INTEGER & WRITING 256 Kb block: 1600.82 Mb/s
INTEGER & WRITING 512 Kb block: 288.22 Mb/s
INTEGER & WRITING 1024 Kb block: 239.88 Mb/s
INTEGER & WRITING 2048 Kb block: 235.27 Mb/s
INTEGER & WRITING 4096 Kb block: 235.95 Mb/s
INTEGER & WRITING 8192 Kb block: 235.39 Mb/s
INTEGER & WRITING 16384 Kb block: 235.12 Mb/s
[root@freeb] /tmp
[0] # ./ramspeed.fbsd -b2
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

INTEGER & READING 1 Kb block: 2368.60 Mb/s
INTEGER & READING 2 Kb block: 2432.94 Mb/s
INTEGER & READING 4 Kb block: 2452.75 Mb/s
INTEGER & READING 8 Kb block: 2470.70 Mb/s
INTEGER & READING 16 Kb block: 2474.19 Mb/s
INTEGER & READING 32 Kb block: 1543.99 Mb/s
INTEGER & READING 64 Kb block: 1544.54 Mb/s
INTEGER & READING 128 Kb block: 1544.45 Mb/s
INTEGER & READING 256 Kb block: 1478.18 Mb/s
INTEGER & READING 512 Kb block: 465.11 Mb/s
INTEGER & READING 1024 Kb block: 466.17 Mb/s
INTEGER & READING 2048 Kb block: 464.38 Mb/s
INTEGER & READING 4096 Kb block: 466.04 Mb/s
INTEGER & READING 8192 Kb block: 464.89 Mb/s
INTEGER & READING 16384 Kb block: 465.81 Mb/s

[root@freeb] /tmp
[0] # ./ramspeed.fbsd -b4
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT & WRITING 1 Kb block: 2243.45 Mb/s
FL-POINT & WRITING 2 Kb block: 2272.88 Mb/s
FL-POINT & WRITING 4 Kb block: 2263.01 Mb/s
FL-POINT & WRITING 8 Kb block: 2278.99 Mb/s
FL-POINT & WRITING 16 Kb block: 2279.75 Mb/s
FL-POINT & WRITING 32 Kb block: 1548.11 Mb/s
FL-POINT & WRITING 64 Kb block: 1549.78 Mb/s
FL-POINT & WRITING 128 Kb block: 1550.60 Mb/s
FL-POINT & WRITING 256 Kb block: 1601.82 Mb/s
FL-POINT & WRITING 512 Kb block: 289.23 Mb/s
FL-POINT & WRITING 1024 Kb block: 239.67 Mb/s
FL-POINT & WRITING 2048 Kb block: 235.09 Mb/s
FL-POINT & WRITING 4096 Kb block: 235.33 Mb/s
FL-POINT & WRITING 8192 Kb block: 235.22 Mb/s
FL-POINT & WRITING 16384 Kb block: 234.83 Mb/s
[root@freeb] /tmp
[0] # ./ramspeed.fbsd -b5
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT & READING 1 Kb block: 2468.33 Mb/s
FL-POINT & READING 2 Kb block: 2431.27 Mb/s
FL-POINT & READING 4 Kb block: 2462.97 Mb/s
FL-POINT & READING 8 Kb block: 2478.36 Mb/s
FL-POINT & READING 16 Kb block: 2477.27 Mb/s
FL-POINT & READING 32 Kb block: 1543.97 Mb/s
FL-POINT & READING 64 Kb block: 1544.42 Mb/s
FL-POINT & READING 128 Kb block: 1544.69 Mb/s
FL-POINT & READING 256 Kb block: 1481.04 Mb/s
FL-POINT & READING 512 Kb block: 466.34 Mb/s
FL-POINT & READING 1024 Kb block: 465.10 Mb/s
FL-POINT & READING 2048 Kb block: 466.20 Mb/s
FL-POINT & READING 4096 Kb block: 466.16 Mb/s
FL-POINT & READING 8192 Kb block: 466.07 Mb/s
FL-POINT & READING 16384 Kb block: 465.04 Mb/s
[root@freeb] /tmp
[0] # ./ramspeed.fbsd -b6
RAMspeed (UNIX) v2.3.1 by Rhett M. Hollander (Alasir Enterprises), 2002-05

4Gb per pass mode

FL-POINT Copy: 354.16 Mb/s
FL-POINT Scale: 363.48 Mb/s
FL-POINT Add: 391.21 Mb/s
FL-POINT Triad: 385.53 Mb/s
---
FL-POINT AVERAGE: 373.60 Mb/s

[root@freeb] /tmp
[0] # exit

Script done on Sat Mar 5 03:17:29 2005

Ok, so again, discarding the cache hits:
INT READ: 465 MB/sec
INT WRITE: 236 MB/sec
FP READ: 465 MB/sec
FP WRITE: 236 MB/sec

(again, looks like we're hit with the gcc Intel optimization issue, FP and INT values match).

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
wegster 
Re: Can someone with a uA1-C (and/or Peg) run 'ramspeed' for me?
Posted on 5-Mar-2005 8:44:40
#20 ]
Elite Member
Joined: 29-Nov-2004
Posts: 8554
From: RTP, NC USA

@wegster

One more data point, from a reasonably high end Intel box-
3.2GHZ CPU, 1MB L2 cache, DDR2-400 RAM: (DDR2-400 has roughly a 5GB/sec _theoretical_ limit, so this is partially to show 'theoretical limits simply aren't reached, just like 100Mbit ethernet..)

INT READ: 3340 MB/sec
INT WRITE: 1610 MB/sec
FP READ: 3700 MB/sec
FP WRITE: 1612 MB/sec

If anyone has an SE, XE, or Peg, please post, even if it's just the average and system specs/cpu version.

_________________
Are we not done with the same silly arguments and flames yet??!

 Status: Offline
Profile     Report this post  
Goto page ( 1 | 2 | 3 | 4 Next Page )

[ home ][ about us ][ privacy ] [ forums ][ classifieds ] [ links ][ news archive ] [ link to us ][ user account ]
Copyright (C) 2000 - 2019 Amigaworld.net.
Amigaworld.net was originally founded by David Doyle