The high performance core in your I7-1265u itself is about 10x faster than the e5500 core. The memory performance is even better (compared to the poor P5020 single core DDR memory performance). Yet the result is 8 times slower under winuae and almost 10 times for QEMU. This means an emulation efficiency of 1-1.25%. I would expect an emulation efficiency of about 10% so something is definitly wrong here. Maybe no JIT like Hans is suggesting?
QEMU does have JIT, but I remember it being rather slow. IIRC, their JIT compiler uses techniques that make it very portable, but obviously doesn't generate the most efficient code. I vaguely remember reading about some technique that uses the C compiler to generate the assembly stubs for individual instructions. If that's what they (still) do, then it's not surprising that it would be slower then a well-crafted JIT whose code contains assembly written specifically for the host CPU.
@Hans QEMU do have KVM : same naming as KVM switches, but different meaning : KVM in QEMU allowing you to run emulation running on the same as emulated CPU platform through real CPU without emulating it: so, theoretically, you should be able someday to run QEMU on PPC Macs with the “same” speed.
But again, emulation issues/bugs and no 3d, still will be there. And PPC Macs is also today not of big interest. But if theoretically somewhere will be PPC on 4GHZ single core, then running QEMU-KVM on such a beast will be cool.
QEMU do have KVM : same naming as KVM switches, but different meaning : KVM in QEMU allowing you to run emulation running on the same as emulated CPU platform through real CPU without emulating it: so, theoretically, you should be able someday to run QEMU on PPC Macs with the “same” speed.
KVM isn't really emulation, but a virtual machine. The code is executed natively by the CPU, but run in a virtual environment.
One thing I'd like to know, could a big-endian PowerPC virtual machine run on a little-endian PowerPC host OS?
I have a question for you, but I am not sure if you are "Hans de Ruiter" who pushed the graphics card support under AmigaOs4.1. If you are, I have already read a few things about you.
To my question:
The Sm502/1 which is also used in the Sam460 as graphics unit does this chip support 32bit output on real hardware or is it limited to 16bit there too. ?
Under AmigaOs4.1 Ranger the chip is listed as supporting the R8G8B8 texture format, but for some reason not yet known the driver does not switch to 24/32 bit mode or AmigaOs4.1 does not offer other modes than 16 bit.
The SM502 data manual says that the chip supports 32bpp for the main and video layers.
The sam460ex manual says the following:
5 – the onboard graphic card has a maximum resolution of 1280x1024 pixels (4:3) or
1440x900 pixels (16:9).
Supports DDC, 2D hardware acceleration, hardware sprites (3 colors) and PIP.
There is no support for 3D or compositing.
Some operating systems may limit the color resolution to 8 and 16 bits.
It doesn't elaborate on "Some operating systems" or why there's this limitation.
Yes, I'm the Hans de Ruiter who wrote the RadeonHD & RadeonRX drivers. I wasn't involved in the SM502 drivers, though. I see that your question has been answered already.
@all
Could someone please cross-post the hardware vs emulation survey (link)to non-English forums? So far only amiga-news.de and amigaportal.cz seem to have picked it up.
The more responses, the more accurate a picture we get.
EDIT: Also, please don't be a jerk and try to submit multiple times. You're just corrupting the results that way, and everybody loses (including you).
You can be sure that our English-language version of the news is read by all, whether or not English is the mother tongue. But I try to do some additional PR...