Linux founder tells Intel to stop inventing 'magic instructions' and'start fixing real problems'

nospam wrote:

J. P. Gilliver wrote:

startled when I spoke to a young computing graduate, to
find he'd never done any assembler.

there is no need for assembler anymore, except in very rare
circumstances.

I know that *I* would have a much poorer understanding of how computers
work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.

J. P. Gilliver (John) wrote:

Char Jackson wrote:

J. P. Gilliver (John) wrote:

I know "could care less" is the US version of this expression

Please don't attribute that mangled expression to all of us over here.

Very sorry! Glad it's not universal in US. But I haven't seen it at all
used in UK.

I must admit I did think all Americans used the "could care less"
variation. Similarly I thought all Americans used "I'm ****ed" (which
in the UK means "I'm drunk") rather than "I'm ****ed off" but I've
started to hear the latter more now.

On 18/07/2020 07:09, nospam wrote:
In article , wrote:


We build little computers with parts. Then went to
CPU. All assembly code!

that doesn't make any sense.


Tiny CPUs often have tiny instruction sets that aren't well suited to
any high level language.

The first project I did on a PIC16C55 we did in assembler.
It was not really too difficult. The whole thing fitted on about 3 pages.

We did later get a "C compiler" for those chips but it was pushing it to
call the language "C". It was mostly the syntax of C but nothing but
static variables. Weird syntax to configure the chip the way you wanted.
I can't even remember how you did I/O on it; must have been another
weird extension to the syntax. It made it easier but not as much as you
might think.

--
Brian Gregory (in England).

On 2020-07-18 1:37 a.m., VanguardLH wrote:
One day a prof arrived just before his class, put his hat on his desk,
but got busy elsewhere. When he got back and very late for class, all
the students had left. The next day he declared, "When my hat is on my
desk, that's the same as I'm here." The next day the students came in,
put hats on their chairs, and left.


Good One VanguardLH :-) :-)

Rene

In article , Brian Gregory
wrote:


We build little computers with parts. Then went to
CPU. All assembly code!

that doesn't make any sense.


Tiny CPUs often have tiny instruction sets that aren't well suited to
any high level language.

he didn't specify tiny cpus with tiny instruction sets or writing
software for them.

he said built computers with parts and *then* went to cpus.

computers built before there were cpus were not little. they were minis
and mainframes, which end users did not build.

The first project I did on a PIC16C55 we did in assembler.
It was not really too difficult. The whole thing fitted on about 3 pages.

We did later get a "C compiler" for those chips but it was pushing it to
call the language "C". It was mostly the syntax of C but nothing but
static variables. Weird syntax to configure the chip the way you wanted.
I can't even remember how you did I/O on it; must have been another
weird extension to the syntax. It made it easier but not as much as you
might think.

in other words, assembly code not needed.

In article , Andy Burns
wrote:


I know that *I* would have a much poorer understanding of how computers
work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.

that doesn't give you a good understanding about modern processors,
which are far more complex than those 8-bit micros.

On 7/16/2020 3:19 PM, VanguardLH wrote:
Already pointed out: your "none of which are being used" is wrong. It
is being used. Video games use it, and those are not rare on Windows
platforms. Any game using DirectX 12 are utilizing AVX2. Scientific,
statistical, financial, encryption, and other programs can use it. Any
program using .NET Framework can use AVX. The latest versions of
Prime95 are optimized to use AVX. While it is used to stress test, that
was not its original or current intent which was to discover prime
numbers. Is prime hunting something that home users do? Of course not,
but it illustrates AVX *is* used.

No, there is a difference between "are utilizing" and "can use". You
used both terms in different sentences up above, probably because you're
not actually sure which one it is, and you wanted to CYA.

"Are utilizing" implies that the games have no choice in the matter, and
they are using AVX even if they don't know it. This would presumably
mean that AVX is being used within the DirectX 12 API itself, and
operates in the background regardless of direct utilization by the game
itself.

That is not how the DX12 API operates. It is a bare-metal API, allowing
the games themselves to control most low-level aspects of the visual
production. This is unlike DX11 the previous API, which was more
hands-on, controlling the low-level aspects. I could see if DX11 had
been further developed, they perhaps might have started to use AVX
within the API itself to help some aspects of performance. But they went
in the completely opposite direction with DX12, I really don't see how
AVX benefits the DX12 low-level API, as it's mainly just a series of
calls to the GPU. For DX12, the "can use" AVX is the appropriate term,
not the "are utilizing".

But even within the game itself over the level of the API, AVX
utilization is very rare. Games could use AVX even under DX11 or even
DX9 or 10 beforehand, but it just didn't use it in the API. There was no
prohibition of using AVX within the application itself. These days they
are more likely to pass off most FP calculations off to the GPU than to
try to do it within the CPU anymore, so AVX is dead in the water.

https://www.tomshardware.com/reviews...de,5461-2.html
"By default, Prime95 automatically selects the newest instruction set
extension, such as AVX, AVX2, or even AVX-512."

Your claim AVX is not used is false.

On the contrary, this exactly proves my point. Torvalds was complaining
about how it's only being used in benchmarks but no real apps. Prime95
is exactly an example of a benchmark and stress testing app. Nobody is
actually using Prime95 for anything other than stress testing and
benchmarking. It's not like as if you're going to be finding any new
prime numbers with a PC anymore, those have now firmly entered the realm
of supercomputers/HPC.

Yousuf Khan

nospam wrote:

Andy Burns wrote:

I know that *I* would have a much poorer understanding of how computers
work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.

that doesn't give you a good understanding about modern processors,
which are far more complex than those 8-bit micros.

But it was a good foundation from where to keep up with progress to
newer processors, and higher level languages.

I just think if you chuck e.g. an rPi at a youngster today they won't
get the same understanding of it from logic gates upwards ...

In article , Andy Burns
wrote:

I know that *I* would have a much poorer understanding of how computers
work if I hadn't cut my teeth with assembler (and BASIC) on 8 bit micros.

that doesn't give you a good understanding about modern processors,
which are far more complex than those 8-bit micros.

But it was a good foundation from where to keep up with progress to
newer processors, and higher level languages.

true, although it's not as relevant as you might think.

I just think if you chuck e.g. an rPi at a youngster today they won't
get the same understanding of it from logic gates upwards ...

that depends on what they do with it, although an arduino would be a
better choice.

nospam wrote:

Andy Burns wrote:

I just think if you chuck e.g. an rPi at a youngster today they won't
get the same understanding of it from logic gates upwards ...

an arduino would be a better choice.

No disagreement there ...