Apollo 1240 - Questions about questions

[StompinSteve]

Hello community,

I'm a new member (Intro here: https://eab.abime.net/showthread.php?t=107478)

I managed to get my hands on an Apollo 1240 (to replace my Blizzard 1230 IV). The board has a fresh cap (it has only one) and a fresh oscillator (makes the 040 run at 28mhz). The board did suffer from battery damage in the past but it has been cleaned up and the battery completely removed.
There are plenty of threads discussing alternatives for the battery so no need to repeat that here.

The 1240 board works well. Ok, it has no RTC battery right now but otherwise the system is stable. The fan works fine. I also use a 3D printed trapdoor with massive venting holes.

My first question is: if I'd plug-in a normal clockport-driven RTC clock solution (example: Individual computers RTC-Module), will that conflict with the (battery-less) RTC on the Apollo 1240?
In my thinking, the system will then have 2 RTC's, one of which has a battery and the other (on the 1240) does not. Recipe for trouble?

Second question: My IDE CF HDD does 2.4 MB/s when I use my Blizzard 1230 Mk IV but 1.8 MB/s when I plug in the Apollo 1240? This totally stumped me. How can that be?

3rd Q: I find the Workbench graphics to be slower with the 1240. This is especially visible when moving the Dopus window around for example. The re-draw of the window is just a bit more sluggish. I expected snappier graphics, not slower...
I run a 1024x768 WB resolution (courtesy of the Indivision AGA MK3) with 16 colours.
I did not change anything about my setup when swapping accell.boards (too lazy) so stuff like BlazeWCP and Fblit/Ftext are still loaded.
I also run with Kickstart 3.0 ROM's as the official AF 3.1 ROM's are still in the mail (stuck for weeks now). AmigaOS is already on 3.1 though (and I have 0 issues with this combo and the Blizzard 1230).

Last question: The fan on the Apollo is way too loud for my taste. I don't want to go the 060 route (passive cooling) because that introduces too much compatibility issues with old games from what' i've read and as i've owned a 4000D with a 25mhz 040 in the past, which worked well with everything I wanted to do with it, so a 040 is as high as I want to go.
I know that using a passive heatsink on this board can make it go flaky without at least some fan blowing over it. I did not find too many solutions to the fan-noise issue other than "tower the 1200" or "get used to the noise". Any opinions on this?

Cheers,
Steve

______________________________________
Amiga 1200 (made by Commodore) Rev 1D4, Kickstart 3.0, Workbench 3.1, Blizzard 1230 Mk IV (at least for now) with 32MB Fast, Indivision AGA MK3, CF HDD 4 GB, Modern A1200 compatible power-supply.

[jbenam]

Hello there, fellow Apollo owner!

Plenty of questions, which I am not sure I can answer 100% correctly, but I will try my best:

1) if the RTC circuitry is still on the Apollo (and I am sure it is), then yes, you will have issues. The right thing to do is to fit the Apollo with a battery.

2) If you really have a 28 MHz oscillator on your Apollo 1240, then I have a bad news for you - it's actually running at 14 MHz. The 040 halves the frequency it gets from the oscillator. That's why it's a lot slower - your 030 MK IV is probably running at a full 50MHz. It can be easily fixed with a new oscillator if the old one was socketed (and I truly hope it was socketed by whoever repaired the card originally).

3) Probably same issue as point 2. Also, be sure you have good 040 libs installed. You can find a floppy image with the official Apollo 040 libs here: http://amiga.resource.cx/exp/apollo1260
I _think_ ThoR's MMUlib also comes with an appropriate (and much more updated) 040 library. Not sure how it compares performance-wise to the original Apollo libs. In any case, you can find it on Aminet.

4) A 040 running at 25MHz (thus with a 50MHz oscilloscope) can be easily and safely be passively cooled, unless you have a really early 040 (XC model) - and even so, I ran my Apollo 1240 with an XC 040 without a fan for months without an issue. Good thermal paste plus a smallish heatsink can do wonders.

[StompinSteve]

Ciao Jbenam,

I did write "makes the 040 run at 28mhz" ;-)
It means the (non-socketed) oscillator itself runs at 56MHz which is what is written on the component.
Sysinfo 4.4 reports the speed as being slightly above an original 4000/40 @ 25MHz

I already had the correct 040 lib on my system. Will try the MMUlib you suggested.
I was wondering if any optimisations for the 030 are counter-productive on the 040.
As soon as the Kickstart 3.1 ROM's arrive, I can add the CMQ030 patch.

I'll try to get a low-height heatsink for on the CPU, as the noise is my biggest gripe. Drives me nuts.
I hope the 3MHz overclocking (28MHz) will not cause damage to the 040 when passively cooled. The CPU is a full 040 originally @25MHz.
The RF shield is off by the way and as I wrote, I already use a 3D printed trapdoor with massive venting holes.

[Pollock]

installing MMULibs will work wonders, i think there was a new version released this year. An 040 @28mhz, should be faster than an 030 @50mhz, the 040 i have found, was the greatest generational leap, of all the 68K CPU's

Get an 060 in there, i have an Apollo 040, full fat CPU rated at 40mhz, and it's zippy, but i have had an Apollo 1260 in the past, and have a TF1260 now, and as for issues with game compatibility .... it's an issue that is usually blown out of all proportion. Where there some games that didn't work on an 060? yes, but ... it wasn't many at all, and after using an 060, it's hard to go back to anything else.

[Mathesar]

Back in the days I once upgraded from a Blizzard 1230 IV to a Blizzard PPC 040/25 603/200.
It also felt slower and less zippy than the 030.
I never understood why this is. It might have to with bus translation that makes it less fast because on paper it should be faster. Could you do bustests (chipram!!) On both the 1230 and the 1240?

Ultimately, imho, the Blizzard 1230IV is the best Amiga1200 accelerator ever built though...

[utri007]

Originally Posted by StompinSteve

3rd Q: I find the Workbench graphics to be slower with the 1240. This is especially visible when moving the Dopus window around for example. The re-draw of the window is just a bit more sluggish. I expected snappier graphics, not slower...
I run a 1024x768 WB resolution (courtesy of the Indivision AGA MK3) with 16 colours.
I did not change anything about my setup when swapping accell.boards (too lazy) so stuff like BlazeWCP and Fblit/Ftext are still loaded.
I also run with Kickstart 3.0 ROM's as the official AF 3.1 ROM's are still in the mail (stuck for weeks now). AmigaOS is already on 3.1 though (and I have 0 issues with this combo and the Blizzard 1230).

Last question: The fan on the Apollo is way too loud for my taste. I don't want to go the 060 route (passive cooling) because that introduces too much compatibility issues with old games from what' i've read and as i've owned a 4000D with a 25mhz 040 in the past, which worked well with everything I wanted to do with it, so a 040 is as high as I want to go.
I know that using a passive heatsink on this board can make it go flaky without at least some fan blowing over it. I did not find too many solutions to the fan-noise issue other than "tower the 1200" or "get used to the noise". Any opinions on this?

Cheers,
Steve

______________________________________
Amiga 1200 (made by Commodore) Rev 1D4, Kickstart 3.0, Workbench 3.1, Blizzard 1230 Mk IV (at least for now) with 32MB Fast, Indivision AGA MK3, CF HDD 4 GB, Modern A1200 compatible power-supply.

WB should be a LOT faster, so you have a problems.

You should have a working 68040.library installed on your system. But this is not a cause of speed problem.

You should get new enough setpatch command, it will install 68040 support code. OS3.0 and 3.1 doesn't have that. This might be partially reason.

FBLIT and FTEXT should be much faster with 68040 as those use raw CPU power wich 68040 28mhz has 4x more than 68030 50mhz.


Usually FBlit problem is missing settings file, wich is needed to create manually correct place.

I used to have Apollo 68040 40mhz, I replaced original sink and fan, with bigger sink and fan. You need to modify sink and move fan more towards rear of amiga, but that would do. Fan does not need to be middle of sink.

[jbenam]

Originally Posted by StompinSteve

Ciao Jbenam,

I did write "makes the 040 run at 28mhz" ;-)
It means the (non-socketed) oscillator itself runs at 58MHz which is what is written on the component.
Sysinfo 4.4 reports the speed as being slightly above an original 4000/40 @ 25MHz

I already had the correct 040 lib on my system. Will try the MMUlib you suggested.
I was wondering if any optimisations for the 030 are counter-productive on the 040.
As soon as the Kickstart 3.1 ROM's arrive, I can add the CMQ030 patch.

I'll try to get a low-height heatsink for on the CPU, as the noise is my biggest gripe. Drives me nuts.
I hope the 3MHz overclocking (28MHz) will not cause damage to the 040 when passively cooled. The CPU is a full 040 originally @25MHz.
The RF shield is off by the way and as I wrote, I already use a 3D printed trapdoor with massive venting holes.

Hello,

Are you sure it's really a 58 MHz oscillator? I have never seen one, and I have recently ordered a bunch of different frequencies. I am googling around for infos and I can't seem to find one, except some SMD 3.3V which definitely aren't safe for use in a 5V environment like the Amiga. In my experience it's either 50 MHz or 60 MHz for the usual DIL-14 package. What does WhichAmiga say about your CPU?

The 3 MHz overclock is most definitely fine, I dare say your 040 could probably also safely run at 33 MHz.

I still suspect some frequency weirdness afoot - a 040 at 28 MHz should be faster than a 030 at 050. That is with stock libs. Right now my A4000D/040@25 has 3.1 on it with no mods and it's plenty fast.

It might indeed be a case of some 030-optimized mod (FBlit et al.) going slower than usual on your 040, but I would double-check that weird 58MHz oscillator first.

You're definitely good on the ventilation part of the thing, a good heatsink with a good thermal paste will let you enjoy a silent A1200 I have run my A4000D with just an heatsink on the 040 for years and zero issues. And trust me, that A3640 (the GALs, mostly) gets HOT!

[Daedalus]

Maybe it's 56MHz, not 58? 56 is a common enough value for oscillators, and is double 28MHz.

Transferring information to the Amiga's motherboard is tricky because you're transferring data between two systems with different, non-synchronous clocks. This can mean slow IDE and slow chip RAM access, since the accelerator has to sit and wait for the motherboard clock to coincide with that of the accelerator before it can start a transfer cycle. This can make a machine feel sluggish, even if the raw power (and fast RAM speed) are much higher than an '030 board, and is the reason that some accelerators have slowdowns in some games that run fine on a 68000 for example. Being clocked at 28MHz is possibly the worst case for this, because it's very close to the motherboard frequency, and so it might be stalled for several clock cycles waiting to transfer. It might be counterintuitive, but clocking it at 25MHz instead of 28MHz might improve the chip RAM and IDE speeds at the cost of some raw horsepower.

A bus test of both chip RAM and fast RAM speeds would be useful in diagnosing things, especially comparing to the system with an '030 powering it.

[StompinSteve]

Sorry the osc. is a 56MHz part, not 58. I blame beer...

Will update mmulibs, check fblit and run bustest on both boards.

[tFt]

Originally Posted by StompinSteve

if I'd plug-in a normal clockport-driven RTC clock solution (example: Individual computers RTC-Module), will that conflict with the (battery-less) RTC

Originally Posted by jbenam

1) if the RTC circuitry is still on the Apollo (and I am sure it is), then yes, you will have issues. The right thing to do is to fit the Apollo with a battery.

This really sucks. I am in the same situation with my Apollo1260. Just bought the RTC with the temp.sensor thing from Amigakit. What kind of issues do you think will happen here?
(Haven't gotten around to remove the battery or install the RTC yet)

[utri007]

You can move kickstart to fast ram with mmulib's help, but you can also use RemApollo to it. Kickstart to fast ram would also speed up over all performance.

[jbenam]

Originally Posted by tFt

This really sucks. I am in the same situation with my Apollo1260. Just bought the RTC with the temp.sensor thing from Amigakit. What kind of issues do you think will happen here?
(Haven't gotten around to remove the battery or install the RTC yet)

Honest answer? I am not sure. There will most certainly be a conflict since two devices will try to use the same address range and select line. In the worst case your Amiga will just guru, so it's probably worth a try?

[StompinSteve]

Ok lads, I Installed/updated the MMUlibs and 68030 and 68040 libraries. SetPatch was already up2date. All caches and "CPU burst" is active during both tests.
Here are the BusTest results:

First screenshot "Bustest_B1230_PreMMUlibInstall.jpeg" is from the Blizzard 1230 Mk IV with the MMU stuff that came with the stock WB 3.1 CF card (Cloanto Licensed). Strangely enough, the bustest results are consistently better than the same board with the updated MMU stuff.
"CF HDD" Disk-speed is also consistently the best of all (2.45 MB/s compared to 2.24 MB/s with the updated MMU stuff, drops to 1.83 MB/s with the Apollo 1240)

The second screenshot "Bustest_B1230.jpeg" is with the Blizzard 1230 Mk IV after updating MMULibs to version 46.22 (latest on Aminet).

The third screenshot "Bustest_A1240.jpeg" is of the Apollo 1240 board. As you can see, the "fast writel & writem" tests are noticeably slower and "chip readm, writew, writel & writem" are a lot slower compared to the Blizzard 1230 with the same MMU software.

If I interpreted the last two screenshots correctly, 4 of the 6 "chip" tests are roughly twice as slow on the Apollo 1240 (or twice as fast on the Blizzard 1230 MK IV if you look at it that way). This could explain the snappier WB graphics on the Blizzard?

[Mathesar]

Originally Posted by StompinSteve

If I interpreted the last two screenshots correctly, 4 of the 6 "chip" tests are roughly twice as slow on the Apollo 1240 (or twice as fast on the Blizzard 1230 MK IV if you look at it that way). This could explain the snappier WB graphics on the Blizzard?

I think it explains indeed. If I recall correctly (but I am not too familiar with AGA) there are ~1.8 million chip ram "slots" per second available for the CPU. Each slot transfers 4 bytes. Hence the max chipram speed is ~7Mbyte per second. To utilize each slot the accelerator needs to be properly designed with a small write cache. If not, the accelerator will miss every other slot and speed is halved.

Now, it could be that at 28MHz, more slots are missed, slowing down chipram access. Maybe reverting back to 25MHz improves things. It could also be that the Apollo just has slow chipram access.
Like I said before, the Blizzard 1230IV is a very well made card and I regret selling it back in the day. Personally, I would stick with the Blizzard.....

[StompinSteve]

Originally Posted by Mathesar

Now, it could be that at 28MHz, more slots are missed, slowing down chipram access. Maybe reverting back to 25MHz improves things. It could also be that the Apollo just has slow chipram access.
Like I said before, the Blizzard 1230IV is a very well made card

I found some screenshots of other bustest from various cards (fast, chip & rom tests) and a stock A4000/40 @ 25MHz (fast ram test only unfortunately):

http://apollo-core.com/knowledge.php...e=352&z=yYG4ur


Would love to see full bustest results of a stock A4000/40 @ 25MHz because my old trusty "real" A4000/40 felt much snappier than this Apollo 1240...


SideNote:
I also ran a bustest on my Apollo 1240 with a lower resolution (PAL "High Res Laced" instead of Individual AGA MK3's 1024x768 that I normally run at) and this improves the "chip" test somewhat (fast and rom tests are the same as on 1024x768)

[utri007]

Rom test is pointless, because you should map it to fast ram on boot. It will also make things snappier. When you boot your 1200 up, how much chip ram you have left? With FBLit there should be about 1,8 mb free chip ram.

[daxb]

Originally Posted by Mathesar

It could also be that the Apollo just has slow chipram access.

I don't know about Apollo 1240/25 but 1240/40 is one of the fastest cards. My card chipram values:

readw 3.5 MB
readl 7.0 MB
readm 6.3 MB
writew 3.5 MB
writel 7.0 MB
writem 7.0 MB

[StompinSteve]

Originally Posted by utri007

Rom test is pointless, because you should map it to fast ram on boot. It will also make things snappier. When you boot your 1200 up, how much chip ram you have left? With FBLit there should be about 1,8 mb free chip ram.

I use the "hardware" FastRom option from the Blizzard (jumper).

Free chip mem on standard "PAL:High Res Laced" is around 1.83 MB, upping the resolution to 1024x768 (AGA MK3 driver) it sinks to around 1.65 MB.

[Mathesar]

Originally Posted by daxb

I don't know about Apollo 1240/25 but 1240/40 is one of the fastest cards. My card chipram values:

readw 3.5 MB
readl 7.0 MB
readm 6.3 MB
writew 3.5 MB
writel 7.0 MB
writem 7.0 MB

Then perhaps the problem lies in the overclocking to 28MHz?

[StompinSteve]

Hello all,

I've done a wholelottabenchmarking. SysSpeed 2.6 comes with a results-module of a stock Amiga 4000D with a 68040 at 25MHz and 8MB FastRam. This helped me answer the question "how fast are a Blizzard 1230 Mk4 and Apollo 1240 @ 28MHz compared to my trusty stock A4000/40 back in the day". I also benched my A1200 without any board (totally stock) just for shiits 'n giggles.

Here are the results:

The first column is the stock A4000/40 @ 25MHz (this module came with SysSpeed)
These Modules I created:
The second column is my A1200 with my Apollo 1240 @ 28MHz installed
The third column is my A1200 with my Blizzard 1230 MK4 @ 50MHz installed
The last column is my A1200 totally stock (ok, the CF HDD is not exactly stock...)

FastROM was on (stock A4000 unknown) and all Caches where on incl. Burst modes in all profiles (incl. the stock A4000).

Screen Resolution in all Modules, incl. SysSpeed's A4000/40, was PAL High red Laced.

SysSpeed V2.6  - TestResults
============================

 A4K_040     :   CPU: 68040  FPU: 68040   ChipSet: AGA   Kick 3.0 , WB 3.1
 (Amiga 4000/40 @ 25MHz, 8MB FastRam, Connor 300MB IDE HDD)

 Apol1240    :   CPU: 68040  FPU: 68040   ChipSet: AGA   Kick 3.0 , WB 3.1
 (Amiga1200 + Apollo 1240 @ 28MHz, 32MB FastRam, CF-IDE HDD)

 Bliz1230    :   CPU: 68030  FPU: 68882   ChipSet: AGA   Kick 3.0 , WB 3.1
 (Amiga1200 + Blizzard 1230 Mk4 @ 50MHz, 32MB FastRam, CF-IDE HDD)

 A1200       :   CPU: 68020  FPU: None    ChipSet: AGA   Kick 3.0 , WB 3.1
 (Amiga1200, Stock, No FastRam, CF-IDE HDD)

+-------------+---------+---------+---------+---------+
|   Test      | A4K_40  | Apol1240| Bliz1230| A1200   |
+-------------+---------+---------+---------+---------+
|  Memory       Results in MB/s                                  
+-------------+---------+---------+---------+---------+
| Cache Read  |  ----   |  52.80  |  31.38  |  7.71   |
+-------------+---------+---------+---------+---------+
| ReadROMb    |  5.55   |  6.47   |  3.98   |  0.96   |
+-------------+---------+---------+---------+---------+
| ReadROMw    |  10.48  |  12.72  |  8.10   |  1.74   |
+-------------+---------+---------+---------+---------+
| ReadROMl    |  11.79  |  24.60  |  16.81  |  4.12   |
+-------------+---------+---------+---------+---------+
| ReadFastb   |  5.55   |  6.45   |  3.98   |  ----   |
+-------------+---------+---------+---------+---------+
| ReadFastw   |  10.46  |  12.71  |  8.11   |  ----   |
+-------------+---------+---------+---------+---------+
| ReadFastl   |  11.63  |  24.56  |  16.84  |  ----   |
+-------------+---------+---------+---------+---------+
| WriteFastb  |  4.47   |  6.61   |  4.92   |  ----   |
+-------------+---------+---------+---------+---------+
| WriteFastw  |  6.18   |  13.52  |  10.34  |  ----   |
+-------------+---------+---------+---------+---------+
| WriteFastl  |  6.18   |  22.81  |  22.98  |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Fastb  |  2.77   |  4.47   |  3.32   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Fastw  |  3.99   |  8.17   |  6.58   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Fastl  |  4.02   |  13.29  |  12.89  |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Fastm  |  3.72   |  10.18  |  12.73  |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Fast16 |  5.78   |  17.84  |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| ReadChipb   |  0.84   |  1.04   |  1.04   |  0.64   |
+-------------+---------+---------+---------+---------+
| ReadChipw   |  1.69   |  2.11   |  2.11   |  1.33   |
+-------------+---------+---------+---------+---------+
| ReadChipl   |  3.44   |  4.30   |  4.30   |  2.85   |
+-------------+---------+---------+---------+---------+
| WriteChipb  |  0.94   |  1.04   |  1.68   |  1.22   |
+-------------+---------+---------+---------+---------+
| WriteChipw  |  1.89   |  2.11   |  3.42   |  2.84   |
+-------------+---------+---------+---------+---------+
| WriteChipl  |  3.74   |  4.17   |  6.70   |  5.30   |
+-------------+---------+---------+---------+---------+
| Chip2Chipb  |  0.47   |  0.52   |  0.83   |  0.64   |
+-------------+---------+---------+---------+---------+
| Chip2Chipw  |  0.94   |  1.04   |  1.68   |  1.33   |
+-------------+---------+---------+---------+---------+
| Chip2Chipl  |  1.89   |  2.11   |  3.41   |  2.85   |
+-------------+---------+---------+---------+---------+
| Chip2Chipm  |  1.86   |  2.05   |  3.05   |  2.83   |
+-------------+---------+---------+---------+---------+
| Chip2Chip16 |  1.86   |  2.31   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Chipb  |  0.87   |  1.00   |  1.04   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Chipw  |  1.65   |  1.96   |  2.09   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Chipl  |  2.95   |  3.91   |  4.24   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Chipm  |  2.82   |  3.65   |  5.18   |  ----   |
+-------------+---------+---------+---------+---------+
| Fast2Chip16 |  2.89   |  4.26   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
|   Test      | A4K_40  | Apol1240| Bliz1230| A1200   |
+-------------+---------+---------+---------+---------+
|  Drive        Results in Op/s                                  
+-------------+---------+---------+---------+---------+
| Create      |  38     |  90     |  81     |  ----   |
+-------------+---------+---------+---------+---------+
| Open        |  355    |  207    |  183    |  ----   |
+-------------+---------+---------+---------+---------+
| DirScan     |  4172   |  580    |  541    |  ----   |
+-------------+---------+---------+---------+---------+
| Delete      |  252    |  405    |  375    |  ----   |
+-------------+---------+---------+---------+---------+
| Seek/Read   |  1749   |  1162   |  864    |  ----   |
+-------------+---------+---------+---------+---------+
|  Drive        Results in MB/s                                  
+-------------+---------+---------+---------+---------+
| CreateFile  |  0.92   |  0.73   |  0.71   |  ----   |
+-------------+---------+---------+---------+---------+
| WriteFile   |  0.96   |  1.24   |  1.31   |  ----   |
+-------------+---------+---------+---------+---------+
| ReadFile    |  0.80   |  1.61   |  1.93   |  ----   |
+-------------+---------+---------+---------+---------+
| RawRead     |  1.41   |  1.69   |  2.05   |  ----   |
+-------------+---------+---------+---------+---------+
|   Test      | A4K_40  | Apol1240| Bliz1230| A1200   |
+-------------+---------+---------+---------+---------+
|  Intuition    Results in Op/s                                  
+-------------+---------+---------+---------+---------+
| OpenWin16   |  7      |  8      |  8      |  4      |
+-------------+---------+---------+---------+---------+
| OpenWin256  |  2      |  4      |  5      |  ----   |
+-------------+---------+---------+---------+---------+
| OpenWinHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| OpenWinTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| WinLayer16  |  6      |  6      |  7      |  4      |
+-------------+---------+---------+---------+---------+
| WinLayer256 |  2      |  3      |  3      |  ----   |
+-------------+---------+---------+---------+---------+
| WinLayerHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| WinLayerTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| SizeWin16   |  18     |  40     |  48     |  14     |
+-------------+---------+---------+---------+---------+
| SizeWin256  |  6      |  41     |  44     |  ----   |
+-------------+---------+---------+---------+---------+
| SizeWinHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| SizeWinTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| MoveWin16   |  18     |  41     |  46     |  28     |
+-------------+---------+---------+---------+---------+
| MoveWin256  |  3      |  38     |  42     |  ----   |
+-------------+---------+---------+---------+---------+
| MoveWinHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| MoveWinTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| ConOutP16   |  35     |  48     |  57     |  31     |
+-------------+---------+---------+---------+---------+
| ConOutP256  |  16     |  34     |  39     |  ----   |
+-------------+---------+---------+---------+---------+
| ConOutPHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| ConOutPTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| OpenScr16   |  8      |  8      |  8      |  4      |
+-------------+---------+---------+---------+---------+
| OpenScr256  |  5      |  5      |  5      |  ----   |
+-------------+---------+---------+---------+---------+
| OpenScrHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| OpenScrTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| SwapScr16   |  58     |  50     |  50     |  0      |
+-------------+---------+---------+---------+---------+
| SwapScr256  |  58     |  50     |  50     |  ----   |
+-------------+---------+---------+---------+---------+
| SwapScrHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| SwapScrTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
|   Test      | A4K_40  | Apol1240| Bliz1230| A1200   |
+-------------+---------+---------+---------+---------+
|  Graphic      Results in Op/s                                  
+-------------+---------+---------+---------+---------+
| ScrollX_16  |  ----   |  8      |  10     |  6      |
+-------------+---------+---------+---------+---------+
| ScrollX_256 |  ----   |  3      |  3      |  ----   |
+-------------+---------+---------+---------+---------+
| ScrollX_Hi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| ScrollX_Tru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| ScrollY_16  |  ----   |  10     |  15     |  6      |
+-------------+---------+---------+---------+---------+
| ScrollY_256 |  ----   |  3      |  4      |  ----   |
+-------------+---------+---------+---------+---------+
| ScrollY_Hi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| ScrollY_Tru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| PrintTxt16  |  582    |  1712   |  1363   |  492    |
+-------------+---------+---------+---------+---------+
| PrintTxt256 |  139    |  750    |  726    |  ----   |
+-------------+---------+---------+---------+---------+
| PrintTxtHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| PrintTxtTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| RectFill16  |  349    |  610    |  773    |  437    |
+-------------+---------+---------+---------+---------+
| RectFill256 |  76     |  209    |  255    |  ----   |
+-------------+---------+---------+---------+---------+
| RectFillHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| RectFillTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| DrwElpse16  |  283    |  314    |  220    |  63     |
+-------------+---------+---------+---------+---------+
| DrwElpse256 |  70     |  134    |  110    |  ----   |
+-------------+---------+---------+---------+---------+
| DrwElpseHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| DrwElpseTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| DrwCircle16 |  271    |  300    |  210    |  60     |
+-------------+---------+---------+---------+---------+
| DrCircle256 |  67     |  129    |  105    |  ----   |
+-------------+---------+---------+---------+---------+
| DrwCircleHi |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| DrCircleTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| WrtPixel16  |  36034  |  35762  |  21021  |  7283   |
+-------------+---------+---------+---------+---------+
| WrtPixel256 |  10208  |  19315  |  13972  |  ----   |
+-------------+---------+---------+---------+---------+
| WrtPixelHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| WrtPixelTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| HorLines16  |  3495   |  2759   |  2436   |  225    |
+-------------+---------+---------+---------+---------+
| HorLines256 |  747    |  1130   |  1068   |  ----   |
+-------------+---------+---------+---------+---------+
| HorLinesHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| HorLinesTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| VerLines16  |  341    |  305    |  306    |  300    |
+-------------+---------+---------+---------+---------+
| VerLines256 |  74     |  106    |  106    |  ----   |
+-------------+---------+---------+---------+---------+
| VerLinesHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| VerLinesTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| DiaLines16  |  239    |  269    |  270    |  265    |
+-------------+---------+---------+---------+---------+
| DiaLines256 |  54     |  103    |  102    |  ----   |
+-------------+---------+---------+---------+---------+
| DiaLinesHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| DiaLinesTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| AreaElps16  |  145    |  174    |  128    |  42     |
+-------------+---------+---------+---------+---------+
| AreaElps256 |  46     |  97     |  80     |  ----   |
+-------------+---------+---------+---------+---------+
| AreaElpsHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| AreaElpsTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| AreaCir16   |  131    |  158    |  119    |  40     |
+-------------+---------+---------+---------+---------+
| AreaCir256  |  41     |  86     |  72     |  ----   |
+-------------+---------+---------+---------+---------+
| AreaCirHi   |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| AreaCirTru  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| AreaFill16  |  138    |  175    |  171    |  166    |
+-------------+---------+---------+---------+---------+
| AreaFill256 |  37     |  76     |  75     |  ----   |
+-------------+---------+---------+---------+---------+
| AreaFillHi  |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
| AreaFillTru |  ----   |  ----   |  ----   |  ----   |
+-------------+---------+---------+---------+---------+
|   Test      | A4K_40  | Apol1240| Bliz1230| A1200   |
+-------------+---------+---------+---------+---------+
|  CPU/FPU      Results in MioOp/s                               
+-------------+---------+---------+---------+---------+
| MIPS        |  18.07  |  20.09  |  16.43  |  4.62   |
+-------------+---------+---------+---------+---------+
| MFLOPS      |  6.85   |  7.13   |  1.11   |  ----   |
+-------------+---------+---------+---------+---------+

What is noticeable is that the ROM tests are very fast on the Apollo.
The Disk IOPS performance on the Apollo is better than my Blizzard but the Throughput speeds are lower on the Apollo.

Concerning my original remark that I felt that the Blizzard "feels" faster in terms of certain WB graphics operations: Well I'll leave the explaining to the people that are actually capable of explaining.
After working with the Apollo for the last 2 days, I got used to it and it feels fast most of the time. The whole machine certainly feels faster than with the Blizzard installed (a LOT faster in some cases) and I started using Dopus in Screen-Clone mode which negates any sluggishness in moving a Wb Dopus window around anyway.

Sidenote: remember I upgraded the MMUlibs to the latest version? That package includes it's own 68040.library together with a "dummy" library called 680x0.library which detects and selects the correct "real library" and this made the system crash on boot after swapping the Blizzard with the Apollo. I deleted the 860x0.library (so that SetPatch does not read it) and replaced the 68040.library with that from Apollo and then she booted and behaved normally.

Anyway, hope you appreciate the time i've put into this and happy readings