Login
Username:

Password:

Remember me



Lost Password?

Register now!

Sections

Who's Online
81 user(s) are online (39 user(s) are browsing Forums)

Members: 2
Guests: 79

jap, trixie, more...

Support us!

Headlines

 
  Register To Post  

AmigaDiskBench Benchmark SSD/NVME/RAM
Just can't stay away
Just can't stay away


See User information
I'm not sure if I just missed it, but the AmigaDiskBench tool that was recently released on Os4Depot is simply fantastic and up to date. I really like it.

So I thought I'd give this tool a try. I tested it with NGFS/01 Sata SSD Intern (AmigaOs4) NVME PcI-Express x4 (Download) and Ramdisk with my X5000/40.

I used the Bulk Benchmark for this, and maybe the results will be interesting for some of you.

Resized Image

Resized Image


Edited by Maijestro on 2026/2/27 20:49:22
Edited by Maijestro on 2026/2/27 20:50:18
MacStudio ARM M1 Max Qemu//Pegasos2 AmigaOs4.1 FE / AmigaOne x5000/40 AmigaOs4.1 FE
Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Not too shy to talk
Not too shy to talk


See User information
@Maijestro

Preview of whats to come

https://ibb.co/ZRWVRBtZ
https://ibb.co/hxJQZRT8

Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Not too shy to talk
Not too shy to talk


See User information
New version of https://os4depot.net/?function=showfil ... chmark/amigadiskbench.lha is currently in the upload queue.

Version History
---------------

v2.3.7 (Current)
S.M.A.R.Tcolumn auto-fitAttribute Name column now correctly expands to
  show full text without truncation
.
Average method displayBenchmark Control row shows a single combined label
  
(e.g., "Average: Median (Middle Value Only)"next to the Passes count.
Bulk tab Settings textQueue summary now includes the averaging method name,
  
e.g., SettingsSprinter 10 Passes (Median) / 4K.
Bug fixes.

v2.3.4 v2.3.6
Flexible Pass AveragingThree averaging options -- All PassesTrimmed Mean,
  and 
Median.
Average Method always visible on the Benchmark tab next to Passes count.
Disk Info tree restructureddevice name at rootVolumeName (Unit Nat
  partition level
multiple units of the same controller merged.
Bug fixes.

v2.2.16
Disk Information Center: New hierarchical view of all physical storage
  devices
categorized as FixedUSB, and Optical.
True Hardware ScanningEngine directly probes the system to map logical
  partitions to physical driver units
.
Enhanced CD/DVD DetectionSafeguards to prevent system freezes when
  querying optical drives
.
Filesystem DisplayStandardized DOS type formatting to ABC/XX hex notation.
Bug fixes.

v2.2.14
Release OptimizationDebug logging disabled for maximum performance.

v2.2.11
Architectural FoundationMulti-threaded benchmark engineCSV history
  persistence
.

v2.2.10
Advanced GraphingVisualization engine overhauledX-axis Block Size;
  
auto-refresh on tab switch.
Variable WorkloadsRandom I/O tests support dynamic user-selected block
  sizes
.
Visual FeedbackTraffic Light status indicator and Fuel Gauge progress bar.
Bug fixes.

Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Home away from home
Home away from home


See User information
@derfs

thanks for the update

_______________________________
c64-dual sids, A1000, A1200-060@93, A4000-CSMKIII
PiStorm32 & Catweasel MK4+= Amazing
! My Master Miggies-Amiga1000 & AmigaONE X1000 !
mancave-ramblings

Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Quite a regular
Quite a regular


See User information
Great utility, thank you!

AmigaOS3: Amiga 1200
AmigaOS4: Micro A1-C, AmigaOne XE, Pegasos II, Sam440ep, Sam440ep-flex, AmigaOne X1000
MorphOS: Efika 5200b, Pegasos I, Pegasos II, Powerbook, Mac Mini, iMac, Powermac Quad
Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Just can't stay away
Just can't stay away


See User information
@derf

Thanks again for the update and for providing this HD/SSD benchmark tool.

What I forgot to mention in my first post above is that my NVME SDD (Download) is called, just for the sake of completeness.

MacStudio ARM M1 Max Qemu//Pegasos2 AmigaOs4.1 FE / AmigaOne x5000/40 AmigaOs4.1 FE
Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Not too shy to talk
Not too shy to talk


See User information
@derfs

Version 2.5.3 of https://os4depot.net/?function=showfil ... chmark/amigadiskbench.lha is currently in the upload queue on os4depot.

Main Updates
* New logging tab to show what is going on.
* Visualization update - 9 template based charts as standard, and you can make your own.
* IOPS calculation fix.

Version History
---------------

v2.5.3 (Current)
VALIDATE mode improvementsWorkbench validation window now displays columns
  at correct proportions
Message column auto-fits to window width with
  horizontal scrolling 
for long messages.
VALIDATE usageRun "AmigaDiskBench VALIDATE" from Shell, or add the tooltype
  VALIDATE to the program
's icon in the Information editor. The benchmark GUI
  does not open -- only a validation report is shown (text report in Shell,
  ReAction window from Workbench).
- Bug fixes:
  - Blank window on Workbench launch: ReadArgs() was guarded by
    Output() != NULL, but Output() can return a valid handle from Workbench
    when a default console is configured. ReadArgs then blocked waiting for
    interactive input, displaying an empty CON: window. Fixed by using
    IDOS->Cli() (the canonical AmigaOS 4 method) to detect Shell vs Workbench.
  - VALIDATE mode Shell/Workbench detection used the same unreliable Output()
    check. Fixed to use IDOS->Cli().
  - Profile validator falsely warned about DefaultDateRange and MinVersion keys
    in the [Filters] section. Both are valid parser-recognized keys that were
    missing from the validator'
s known-key list.
  - 
Visualization hover tooltip and chart legend now replace underscores with
    spaces in volume names 
(e.g., FFS2_DH8_2 displays as FFS2 DH8 2).
Icon cleanupProgram icon now ships with a single (VALIDATEtooltype
  
(commented out by default). Old template tooltypes removedDrawer and
  
folder icons included for the distribution archive.
Makefile "install" targetNow produces a complete distribution layout with
  drawer icon
program iconVisualizations folder icon, and readme.

v2.5.2
Pluggable Visualization ProfilesChart definitions are now loaded from .viz
  files in the Visualizations
folderNine built-in profiles ship with the
  application
ScalingTrendBattleWorkloadHybridPeak Performance,
  
IOPS SmoothedScaling Curve, and Filesystem.
Profile-driven renderingEach profile defines chart type (line/bar/hybrid),
  
X/Y axesseries groupingdata filterstrend linesreference line
  annotations
custom color palettes, and collapse aggregation -- all without
  recompilation
.
Trend linesLinear regressionmoving average, and polynomial curve fitting,
  
rendered per-series or globallyCoordinate clamping prevents drawing outside
  the chart area
.
Collapse aggregationReduce multiple runs at the same X value to a single
  data point using mean
medianmin, or max.
Reference line annotationsHorizontal dashed lines with labels at user-defined
  Y values 
(e.g., SATA III 600 MB/s).
Custom color palettesUp to 16 hex colors per profile.
Reload Profiles buttonRescan the Visualizationsfolder for new or modified
  profiles without restarting
.
VALIDATE modeShell argument or icon tooltype to validate all .viz files and
  
report errors with line numbers and severity.
Quit confirmation dialogAttempting to close the application while a benchmark
  is running now shows a Yes
/No confirmation prompt.
IOPS calculation correctedIOPS is now computed as total I/O operations
  divided by total elapsed time 
(true ops/second). Previouslyworkloads
  reported a fixed operation count of 1
, and the engine averaged ops per pass
  rather than per second
.
Bug fixes:
  - 
Random 4K Write/Read and Mixed R/W benchmarks failed at 1M block size:
    
uint32 total_bytes overflowed (4096 ops x 1M exactly 2^32wrapping to
    0
). Fixed by using uint64 internally with a cap to UINT32_MAX.
  - 
CSV history records silently truncated when field data exceeded the
    1024
-byte line buffer (producing malformed rows). Increased to 2048 bytes
    with overflow detection 
-- oversized records are now skipped with a warning
    in the session log
.
  - 
Benchmark failure messages in the session log were generic ("check target
    volume"
). Now reports the specific failure reasonsetup failure (file
    creation
device open, or buffer allocation), unknown test type, or all
    passes producing zero bytes 
-- with test nametarget path, and block size.
  - 
Crash on exit while benchmarks are running (pending worker messages not
    drained before freeing reply port
).
  - 
Blank window on Workbench launch (ReadArgs called unconditionallyfailing
    when launched from icon
).
  - 
Crash in CollapseSeriesPoints when series had no data points.
  - 
Chart lines drawn outside the graph area (missing coordinate clamping in
    line
/hybrid renderers).
  - 
Duplicate X-axis labels when all data points share the same block size.
  - 
Color By dropdown removed -- now a read-only label driven by the active
    profile
's GroupBy setting.
  - Unused filter dropdowns removed from the visualization tab.

v2.4.1
- Session Log tab: A new scrollable, timestamped log panel records all benchmark
  activity in real time -- start, per-pass progress, results, and failures --
  with [HH:MM:SS] timestamps.
- Live log updates: Cross-process Exec message passing delivers log entries from
  the worker process to the GUI without polling or unsafe gadget access.
- Log context menu: Right-click for Select All and Copy.
- Copy to Clipboard button: Writes the entire log to the system clipboard via
  IFFParse (FORM FTXT/CHRS format).
- Clear Log button: Wipes the transcript and re-inserts the session header.
- Bug fix: Classic menu selections were silently swallowed when a context menu
  was also present. Fixed by checking WINDOW_MenuType before routing the
  WMHI_MENUPICK event.

Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Just popping in
Just popping in


See User information
@derfs

First, thanks for a very nice looking, full-featured and useful program (though it would be helpful to have a bit more detail on what the different benchmark tests do, so one knows what exactly is being benchmarked).

However, I get a recoverable DSI when I run the two most recent versions of AmigaDiskBench (v. 2.5.3 and 2.5.2) on my X1000 running FE Update 3. The DSI occurs soon after I start the program, apparently while it's scanning the drives. I can click on 'Ignore DSI Errors' and the program then appears to run normally (but see below for a different problem).

AmigaOne X1000 release
Machine model
(AmigaOne X1000)
Dump of context at 0xdfac23e0
Trap type
DSI exception
Current kernel stack pointer
0x296ff00
DSISR
40000000  DAR153d5550
No matching page found
Machine State 
(raw): 0x100000000200b030
Machine State 
(verbose): [Hyper] [ExtInt on] [Super] [IAT on] [DAT on
Instruction pointer0x7fa26144
Crashed process
AmigaDiskBench (0x5ea010c0)
DSI verbose error descriptionAccess to address 0x153d5550 not found in hash or BAT (page fault)
Access was a load operation
 0
7fa260c0 5e457220 00000000 5e457304 153d5550 00000005 00000005 02006624
 8
: 02816968 5e457240 00000000 0227fb88 00000794 5d873418 5e7332ac 5e733280
16
61046a10 5def0018 5d870000 5d86b6fc 5d870000 5e733274 5e7332a4 5e73032c
24
5e73229c 80000006 5da5c634 5d870000 5d870000 5e8a5120 5d86b41c 6fcdd350
CR
48222224   XER20000000  CTR00000000  LR7fa260c0

V0 
00000000000000000000000000000000 ffcfcfcfffcfcfcfffcfcfcfffcfcfcf
V2 
01000100010001000100010001000100 ffcfcfcfffcfcfcfffcfcfcfffcfcfcf
V4 
ff01b947ff01b947ff01b947ff01b947 00000000000000000000000000000000
V6 
ff01cf31ff01cf31ff01cf31ff01cf31 ff000000ff000000ff000000ff000000
V8 
00000000000000000000000000000000 ffb9b9b9ffb9b9b9ffb9b9b9ffb9b9b9
V10
: 001002120414061608180a1a0c1c0e1e 01000100010001000100010001000100
V12
01000100010001000100010001000100 ffcfcfcfffcfcfcfffcfcfcfffcfcfcf
V14
cf31cf31cf31cf31cf31cf31cf31cf31 ffcfcfcfffcfcfcfffcfcfcfffcfcfcf
V16
ce31ce31ce31ce31ce31ce31ce31ce31 ffcfcfcfffcfcfcfffcfcfcfffcfcfcf
V18
01000100010001000100010001000100 ffffffffffffffffffffffffffffffff
V20
00000000000000000000000000000000 00000000000000000000000000000000
V22
00000000000000000000000000000000 00000000000000000000000000000000
V24
00000000000000000000000000000000 00000000000000000000000000000000
V26
00000000000000000000000000000000 00000000000000000000000000000000
V28
00000000000000000000000000000000 00000000000000000000000000000000
V30
00000000000000000000000000000000 00000000000000000000000000000000
VSCR
00000000 VRSAVE00000000

Disassembly of crash site
:
 
7fa2613492410138   stw               r18,312(r1)
 
7fa261389261013c   stw               r19,316(r1)
 
7fa2613c92810140   stw               r20,320(r1)
 
7fa2614093410158   stw               r26,344(r1)
>
7fa261448b440000   lbz               r26,0(r4)
 
7fa26148281a001f   cmplwi            r26,31
 7fa2614c
41810404   bgt-              0x7FA26550
 7fa26150
7f45d378   mr                r5,r26
 7fa26154
38840001   addi              r4,r4,1
 7fa26158
386100c4   addi              r3,r1,196

Kernel command line
DEBUGLEVEL=0 SERIAL

Registers pointing to code
:
r0 AmigaDiskBench:ScanSystemDrives()+0x190 (section 1 0x1D008)
r7 native kernel module kernel+0x00006624
r8 
native kernel module dos.library.kmod+0x006e0308
r11
native kernel module newlib.library.kmod+0x00008fc8
r13
AmigaDiskBench:collapsed_buf.1()+0x5ca8 (section 14 0x7F24)
r14module AmigaDiskBench at 0x5E7332AC (section 1 0x3288)
r15module AmigaDiskBench at 0x5E733280 (section 1 0x325C)
r18AmigaDiskBench:collapsed_buf.1()+0x2890 (section 14 0x4B0C)
r19AmigaDiskBench:ui()+0xbc (section 14 0x208)
r20AmigaDiskBench:collapsed_buf.1()+0x2890 (section 14 0x4B0C)
r21module AmigaDiskBench at 0x5E733274 (section 1 0x3250)
r22module AmigaDiskBench at 0x5E7332A4 (section 1 0x3280)
r23module AmigaDiskBench at 0x5E73032C (section 1 0x308)
r24module AmigaDiskBench at 0x5E73229C (section 1 0x2278)
r26AmigaDiskBench:s_part_vol.11()+0x0 (section 14 0x1F1140)
r27AmigaDiskBench:collapsed_buf.1()+0x2890 (section 14 0x4B0C)
r28AmigaDiskBench:collapsed_buf.1()+0x2890 (section 14 0x4B0C)
r30AmigaDiskBench:IDOS()+0x0 (section 13 0x0)
ip AmigaDiskBench:ScanSystemDrives()+0x214 (section 1 0x1D08C)
lr AmigaDiskBench:ScanSystemDrives()+0x190 (section 1 0x1D008)
ctrunknown (0x0)

Stack trace:
(
0x5e457220AmigaDiskBench:ScanSystemDrives()+0x214 (section 1 0x1D08C)
(
0x5e457390AmigaDiskBench:ScanSystemDrives()+0x190 (section 1 0x1D008)
(
0x5e457900AmigaDiskBench:StartGUI()+0xbfc (section 1 0x60F0)
(
0x5e457d40native kernel module newlib.library.kmod+0x00002614
(0x5e457d90native kernel module newlib.library.kmod+0x000032f0
(0x5e457f40native kernel module newlib.library.kmod+0x00003864
(0x5e457f70AmigaDiskBench:_start()+0x1e0 (section 1 0x1E0)
(
0x5e457fc0native kernel module kernel+0x0005a1a4
(0x5e457fd0native kernel module kernel+0x0005a21c

Disassembly of crash site
:
 
7fa2613492410138   stw               r18,312(r1)
 
7fa261389261013c   stw               r19,316(r1)
 
7fa2613c92810140   stw               r20,320(r1)
 
7fa2614093410158   stw               r26,344(r1)
>
7fa261448b440000   lbz               r26,0(r4)
 
7fa26148281a001f   cmplwi            r26,31
 7fa2614c
41810404   bgt-              0x7FA26550
 7fa26150
7f45d378   mr                r5,r26
 7fa26154
38840001   addi              r4,r4,1
 7fa26158
386100c4   addi              r3,r1,196
Stack pointer 
(0x5e457220is inside bounds
Redzone is OK 
(4)

There is also a second, more fatal, crash. Because it's very similar to a crash in the most recent version of Rave, I posted the details of this crash in the Rave thread, where the problem has been discussed at some length.

Go to top
Re: AmigaDiskBench Benchmark SSD/NVME/RAM
Not too shy to talk
Not too shy to talk


See User information
@msteed

Thanks for the feedback - please can you test https://github.com/derfsss/AmigaDiskBench/releases/tag/v2.5.4 to see if it solves your issues? The 2nd crash will probably still happen for now.

On the Benchmarking tab, if you right-click on a test name you can now use the menu to see a full description of that test.

Go to top

  Register To Post

 




Currently Active Users Viewing This Thread: 1 ( 0 members and 1 Anonymous Users )



Polls
Running AmigaOS 4 on?
AmigaOne SE/XE or microA1 12% (26)
Pegasos2 3% (8)
X5000 22% (48)
X1000 14% (30)
A1222 8% (19)
Sam 440/460 18% (40)
Classic PowerPC Amiga 2% (6)
WinUAE emulation 7% (16)
Qemu emulation 9% (21)
Total Votes: 214
The poll closed at 2025/12/1 12:00
8 Comments


Powered by XOOPS 2.0 © 2001-2024 The XOOPS Project