ACA1221LC 40MHz VS ACA1234 25MHz

turrican9 · Nov 22nd 2021

So I noticed from wiki,icomp that the ACA1221LC at 40MHz has higher benchmark scores VS ACA1234 25MHz. Even has higher scores than ACA1234 33MHz aswell. Would this be felt in real life? I believe I read somewhere that the 68030 is about 10% faster per clock VS 68020? Would the ACA1221LC at 40MHz feel faster VS the ACA1234 at 25MHz or even 33MHz? Especially when used with the ACA 500 Plus. I know the dedicated IDE speed booster port on the ACA1234 series and also the ACA 1221LC has speedboosting for the A1200 internal IDE. However, this is not used with the ACA 500 Plus. It has great IDE CF card performance anyway.

I know that the 1221LC is not guaranteed to overclock to 40MHz. I know it has far less RAM VS 1234 and I also know the ACA1221LC will not support cloaking mode with the ACA 500 Plus.

I have a fully unlocked ACA1221 with 64MB RAM and overclocked to 28MHz and I also have an ACA1233 at 40MHz. When I use these in my A1200 I can feel that the ACA1233 is quite a bit faster. However, the ACA 1233 at 40MHz scores in the 9200 range in sysifo VS the ACA 1221 at 28HMz scores about 4900ish. I know the ACA1221LC is faster clock for clock VS the older ACA1221.

I think the ACA 1221LC looks impressive at 40MHz. I think it has enough RAM for my use and I am debating between these accelerators. Only thing I would like is the cloak mode the 1234s will support with the next ACA 500 Plus firmware. But at the same time I rarely use stock A500 mode.

thebajaguy · Nov 23rd 2021

SysInfo tends to be very CPU-clock/cache oriented in it's 'benchmark' of CPU performance. It therefore must be taken with a few grains of salt. Unless you have equally good memory performance, the bias it gives raw clock speed is at times misleading.

Benchmarks (in general) also tend to bias sequential-oriented memory accesses - something the 68030 can accel at if it is supported by an efficient burst-access memory design. Random memory access and I/O, however, tend to defeat the cache, and overall memory performance becomes more important.

That said, one should take into account a memory bus-speed test (bustest by mlev on Aminet is one), and also the AIBB tests offer some better 'system intensive' tests that show how different hardware configurations can affect overall performance.

Jens · Nov 23rd 2021

Quote from turrican9

Would this be felt in real life?

This reminds me of making an AIBB module for the ACA1221lc (or better: Three of them!). If you can think of a test that I can do in addition to that (make an LHA in RAM: over the C: directory?), I can do that, too. I have a stop watch next to that place anyway...

In any case, the ACA1221lc is probably the fastest 68020 card you can make, as it's shaving off one wait state during RAM access in comparison to your older ACA1221. This explains the faster performance at 26.67MHz compared to 28.38MHz of your ACA1221.

68020 and 68030 are not too far apart in terms of performance - the 68030 gains about 10-15% at the same clock speed compared to the 68020. So the ACA1221lc at 40MHz "barely beaten" by the ACA1234-33 is about what I'd expect.