Igalia Working Towards Faster 2D Rendering For Older Raspberry Pi Boards

[ssokolow]

Vector instructions weren't introduced until the Pentium MMX, which ran at 233 MHz. RAM was definitely like sub-100 MB/s, but I don't remember exactly.

I could be wrong, but I think, when we bought our HP Pavilion 8160 (233MHz P1 MMX) around 1997, I vaguely remember them also offering a 200MHz MMX model.

[blackshard]

The 3D capability of the VideoCore VI actually does differ from the original VideoCore IV and a Pi v1 sure isn't going to run Quake3 at 1080p @ 80 fps!

Videocore VI (6) is available only on RPi4; RPi zero up to RPi 3 have Videocore IV (4); Rpi3 has been shown to run Quake3 @ 80fps in full-hd, on a fully working Vulkan driver also, yet not totally conformant due to some features the hardware miss.

Most probably RPi1 is not able to run Quake3 at 1080p at 80fps because the ARM is too weak, the L2 cache is shared and GPU frequency is lower than on RPi3, but the silicon IP is the same.

BTW, Quake 3 was already 12 years old, when the original Pi launched. Hardly a cutting-edge workload.

That does not really mean anything: how much the workload is heavy depends on the worker.

Running Quake3 at Full-HD in the first 2000s was not the norm, and surely it needed a nearly sub-ghz x86 Pentium III/Athlon machine with possibly a T&L video card to perform decently at higher resolutions (https://www.anandtech.com/show/742/5)

Mali T604 launched in 2010; the T658 in 2011. Both supported OpenGL ES 3.1 and OpenCL 1.1. PowerVR SGX544 and SGX554 both launched in 2010. Looks like Adreno 220 launched in 2010-2011.

When the original Pi launched, its GPU was acceptable, not cutting-edge. The main problem with the Pi's GPU is that even the VideoCore VI isn't much better than the original VideoCore IV. Meanwhile, the industry standard improved by orders of magnitude.

[/QUOTE]

Never said the performance was cutting-edge, even though broadcom claimed 24 GFLOPS of raw computing performance.

Nonetheless no the RPi1 the fill-rate was capped by 400MHz DDR2 memories, so general performance obviously suffered from that.
I said it was extraordinary, meaning it supported things that were not available from competitors because it was designed as a media processor, not just like a GPU.
Flexible hardware compositor, zero-copy video-to-texture facilities (due to h.264 block being integrated into same silicon), full acceleration for 2D (just to say that also stroked paths in OpenVG are accelerated...) were not so common, and surely competitors were (and still are) not offering all the features into the same package: current GPUs don't provide any video encoding/decoding capabilities, nor 2D features, which are different IP blocks.

[coder]

full acceleration for 2D (just to say that also stroked paths in OpenVG are accelerated...) were not so common,

I don't know about Mali, but Wikipedia states that Adreno supported OpenVG 1.1 for most of its history. And I found a press release from 2011 stating that PowerVR supported OpenVG 1.1 on hardware introduced then.

• https://en.wikipedia.org/wiki/Adreno
• https://www.engadget.com/2011-06-02-...-arm-proc.html

[hamishmb]

I remember the GPU being called the "best camera-phone in 2012" at the time, can't remember who said it, or whether that was accurate or not.

[Raka555]

And my Paradise video card had full 2D acceleration with Xfree86 in 1993 with only 1M of RAM ... (Obviously limited to 1024x768 and 256 colours)
But 2D operations felt faster then than my Vega64 feels today.

So to go from 1024x768x8bit (standard for the day) to 1920x1080x32bit (standard for today) you need to go from 1MB memory to 256MB ?

The paradise needed 768kB, had 1MB and XFree86 was a great experience .

The RPIs needs 8100kB, has up to 256MB and Xorg is piss poor experience ...

And then people roll their eyes when I tell them software are getting worse.

[Raka555]

I don't know who these people were, but every time the Pi touched storage you felt it. Until Pi v4, storage was connected via USB2, even if you used a SD Card. I can't comment on desktop graphics performance, since my experience with Pi's has mostly been headless and using them remotely.

I guess one more thing I can say about general CPU performance is that I remember the first time I tried compiling a fairly simple C++ program on my Pi v1. It took so long that I thought the compiler might've hung! It was like 20 seconds, for something that compiled in well under a second on my PC. IIRC, the CPU cores in Pi v2 were each about 50% faster, and there were 4 of them. Still, I'd imagine its poor single-thread performance would significantly hamper its usability as a general purpose desktop.

And then again. I compiled c++ just fine on my 486DX66 at the time ...

Having said that, c++ was a much cleaner and minimalistic back then.

[Raka555]

I could be wrong, but I think, when we bought our HP Pavilion 8160 (233MHz P1 MMX) around 1997, I vaguely remember them also offering a 200MHz MMX model.

You are correct. I had a Pentium 200MMX

[timofonic]

On the contrary, GPUs on the raspberry pi, especially the first Pi1, are the working horse of the whole platform.
Of course they are now aging, but in 2012 and for some years those GPUs (and the whole compositor/GPU/VPU compound) were quite extraordinary in respect to the competitors, either in terms of computational capacity per watt and features.

I absolutely don't agree with "old crap", everything is "old crap" if you're ignorant enough not to know how to use it.

Well, relatively old crap.

It's a joke platform, but easy to tinker with it. Not so open, but rest of ARM alternatives are even more closed.

[blackshard]

I don't know about Mali, but Wikipedia states that Adreno supported OpenVG 1.1 for most of its history. And I found a press release from 2011 stating that PowerVR supported OpenVG 1.1 on hardware introduced then.

• https://en.wikipedia.org/wiki/Adreno
• https://www.engadget.com/2011-06-02-...-arm-proc.html

Never seen Adreno driver blobs honestly - never had the chance to put hands on a board that sported a qualcomm SoC, but surely ARM was publicizing OpenVG support here and there for some GPUs (example: Mali 400/450) I have never seen a single OpenVG driver (and their Linux OpenGL ES drivers were terrible).