Are we going to 64 bit Assembly?

[hutch--]

 :bg

Dave,

One person can lift it but you need both handles. Top box is full of transformers, all of which are masterpieces, amp has 100k range and its largely due to the excellent power transformer and output transformers. Its 10k square wave looks like a bipolar transistor profile but without the multiple harmonics on the leading edge.

I laugh at some of the old valve amps, nice warm soft slow sound, the Radford is really crisp and fast. 15 watts RMS (actually 22 watts RMS) is different to a similar transistor output, it will drive transients of much higher amplitude without clipping it as a transistor does. My 100 watt plus amps of many years ago did not outperform it and they were smartarse differential input split power supply models.

You could nearly dupliicate the result if you applied a very slight treble cut on the input side as you took the ring off the transistor models but the different harmonic structure still made them sound different.

A mate of mine bought it new here in OZ in 1966 for 120 Pounds, roughly $240.00 which was a lot of money back then.

I have a complete set of new Mullard valves for it, the EL34s, GZ34s and the small signal valves for the front end but the current set sound fine.

[dedndave]

yah - tubes just handle large transients better
of course, transistor circuits are usually direct-coupled, which eliminates the need for expensive output transformers
i am guessing they are using push-pull circuits with the tubes
most of the old AM modulators were built the same way
it gives you a freebie reduction in odd-ordered harmonic distortion

[hutch--]

Dave,

Its a shame you are not on this side of the world as I would like to get just a littlle maintainance done on it. My older brother is a valve era electronic man but he hates audio and he is on his way to retirement so its hard to get anything out of him at the moment. It still has 1965 electrolytic power capacitors in it which still seem to be working OK but if its left on for a long time it starts to "motorboat" on the right channel so something is going out of value. With a bit of scouring internationally you can get the high voltage bits for it but its no joy to find them.

Other thing is my oscilloscope has not been turned on for years and it would probably go BANG with the age of the capacitors.

[dedndave]

tube circuits are actually much easier to troubleshoot than transistor circuits
that's because, with transistors, you have to do most everything with no voltage applied
tubes will take a beating, so long as the filament voltage and plate voltage are right
you can power them up and make measurements
if you do that with transistors - 9 times out of 10, you will be replacing them - and still not found the problem  :bg

[Rockoon]

How about SSE XMM0 through XMM7? (I dont know exactly what you mean by address line though)....

What about them? 128-bit words does not mean a 128-bit processor.

Even the 8086 + 8087 could manipulate 64-bit integers (via the FPU) ..and thats a 16-bit architecture

128-bit address lines means 128-bit addressing. Able to address 2^128 ( = 3.4E+0038 ) bytes.

We didnt call the 80386 a 32-bit machine because it could manipulate 32-bit words. We called it a 32-bit machine because its memory bus pointers were 32-bits in size, capable of addressing 2^32 individual bytes.

With the advent of the Pentium MMX, we could manipulate 64-bit words natively. That didnt make it a 64-bit architecture. It still only had 32-bit pointers.

The AMD64 architecture has 64-bit pointers. It can conceivably address 2^64 bytes of memory. It also has 64-bit words, but it could just as easily have 128-bit words, or even 256-bit words. It would still be a 64-bit architecture.

[Rockoon]

Well, somehow this is true, but you have to consider that entrepreneurs
tend to exploit tech stuff in a different way than tech people. They just want
to sell the potential it has, even if it is not really usable for the time being.

That reasoning would have brought about 64-bit computing, but it didn't. It is only recently that 64-bit addressing has become marketable.

I think many people are unfamiliar with how long we stood in the 32-bit world. The 80386 came out in 1985, and before that we were using a 20-bit architecture. That push from 20-bit limits through to the 32-bit limit took more than several decades.

At the time many people ran a 20-bit OS (DOS and then 5 years later Windows 3.0) just like many people run a 32-bit OS today even though they have a 64-bit machine.

It is so 1990 all over again. I personally ran XP/64 for years and it is only now, two whole major OS version later, that 64-bit OS's are just starting to take over.

I personally rely on Valve's hardware surveys to get a good grasp of the market. The surveys are taken via their Steam gaming platform, but it isnt just a top-end sampling by any stretch. Most steam games run on 8 year old hardware just fine.

Win XP 32-bit: 32.99%
Win 7 64-bit: 26.65%
Vista 32-bit: 14.01%
Win 7 32-bit: 12.07%
Vista 64-bit: 6.75%
Win XP 64-bit: 0.56%
Win 2003 64-bit: 0.32%

The remaining %'s are for MacOS.

2GB of ram is still the most common configuration.

[dedndave]

Win XP 32-bit: 32.99%   :U

i would have thought more, actually
there are so many mom and pop computers in the world
they check their e-mail and play solitaire, and that's about it - lol
but, there are millions of them

[Rockoon]

Win XP 32-bit: 32.99%   :U

i would have thought more, actually
there are so many mom and pop computers in the world
they check their e-mail and play solitaire, and that's about it - lol
but, there are millions of them

And most of them have probably purchased a new laptop sometimes since the release (January 2007) of Vista :)

These same people buy new machines because their computer is virus infested, and you know how bad Windows XP is for grandma in that regard!

[oex]

there are so many mom and pop computers in the world
they check their e-mail and play solitaire, and that's about it - lol
but, there are millions of them

These arent counted in this index this is the gamers index.... While you'll find all sorts of wild claims as to what an 'average gamer' is many people will have multiple computers and rarely use many of their capabilities:
http://gigaom.com/2010/02/17/average-social-gamer-is-a-43-year-old-woman/

Further there is the assumption that Steam has equal worlwide appeal.... In reality computer specifications are far from evenly spread with the US and UK having many orders of magnitude higher general unused processing power than other poorer countries

[dedndave]

These arent counted in this index this is the gamers index

that makes a huge difference
you must assume that a large percentage of the world population are neither gamers nor programmers   :bg
they usually don't have "the best stuff" - lol
many are using the second-hand computers that were sold after someone else upgraded
because, that is all they can afford or, in many cases, justify spending

as for the average gamer being a 43-yr old woman - that has got to be a skewed statistic
you are talking about little old ladies that have "farms" on facebook - lol
when i think of "serious gamers", i think of a male in 20's or early 30's that plays interactive "combat" or RPG type games
these are the guys that spend the extra bucks for 64-bit machines and graphics adapters that require their own fan

[frktons]

That reasoning would have brought about 64-bit computing, but it didn't. It is only recently that 64-bit addressing has become marketable.

The reverse is what I meant. We had 64 bit processors well ahead of their
real use [64 bit OS and applications].  So we could have 128 bit processors in the same way.
There is no need and no applicable area, but they can be sold nevertheless.

I think many people are unfamiliar with how long we stood in the 32-bit world. The 80386 came out in 1985, and before that we were using a 20-bit architecture. That push from 20-bit limits through to the 32-bit limit took more than several decades.

Actually we are using a 40 bit addressing architecture, if I correctly recall,
the 64 bit are not fully used, but it could have changed without INTEL informing me  :lol
We should consider that things tend to speed up a little in tech field, so instead
of 25 years the next step could take less  :P

I personally rely on Valve's hardware surveys to get a good grasp of the market. The surveys are taken via their Steam gaming platform, but it isnt just a top-end sampling by any stretch. Most steam games run on 8 year old hardware just fine.

Win XP 32-bit: 32.99%
Win 7 64-bit: 26.65%
Vista 32-bit: 14.01%
Win 7 32-bit: 12.07%
Vista 64-bit: 6.75%
Win XP 64-bit: 0.56%
Win 2003 64-bit: 0.32%

Considering Mac, Linux and others 64 bit OS we are at about 50% user/server market.
Don't you think in the next 10 years we are going towards bigger figures?

So the original question and the title of the thread: Are we going to 64 bit Assembly?
could be easily replied: yes, we can in Obama style, or whatever you like.  :lol

[sinsi]

You could just as easily say "are we going to multi-threaded assembly?" If you want, go for it.
ML64 is going back to the days of masm version 1 - basic functionality (no invoke, which pisses off a lot of people).
I like the bare-bones approach of ml64 for writing win64 programs, but I also use fasm for non-win64 things.

We will see 16-bit DOS/win3 programs dying off, because long mode won't support v86 mode, but 32-bit programs
won't die for a long time.

I think we're up to 53-bit addressing now aren't we? And if you add the win7's together they beat xp  :P

[cork]

ML64 is going back to the days of masm version 1 - basic functionality (no invoke, which pisses off a lot of people).

Maybe somebody could write a simple pre-processor that would take in a MASM source file containing invoke instructions, and do nothing except spit out a file that translated the INVOKE to its equivalent lower-level instructions...

Then a person could call the pre-processor, followed by ML64.

[frktons]

You could just as easily say "are we going to multi-threaded assembly?" If you want, go for it.
ML64 is going back to the days of masm version 1 - basic functionality (no invoke, which pisses off a lot of people).
I like the bare-bones approach of ml64 for writing win64 programs, but I also use fasm for non-win64 things.

We will see 16-bit DOS/win3 programs dying off, because long mode won't support v86 mode, but 32-bit programs
won't die for a long time.

I think we're up to 53-bit addressing now aren't we? And if you add the win7's together they beat xp  :P

Yes, we could  :P
ml64 needs some renovation, and Microsoft is going to use it for his 64 bit stuff.
Probably new versions will be shipped as Microsoft feels the need for them.
If not, we have FASM , GOASM and a few more suitable packages.

Well, win-XP is not going to survive more than 5 years, except on Dave's valve computer  :lol

Maybe somebody could write a simple pre-processor that would take in a MASM source file containing invoke instructions, and do nothing except spit out a file that translated the INVOKE to its equivalent lower-level instructions...

Then a person could call the pre-processor, followed by ML64.

yes, a person could after somebody write it.
I guess Microsoft itself has enough persons and somebody to do that.
And many others are capable to do that as well, or have already done it.

[Rockoon]

I think we're up to 53-bit addressing now aren't we? And if you add the win7's together they beat xp  :P

It depends on the chip and motherboard.. the architecture itself is capable of addressing 2^64 bytes of memory even if the hardware isn't..

...much like the first 80386's couldn't be loaded with 4 gigs of ram (4 megs was probably your max) that in no way means that the architecture didn't support it. It just means that those address lines werent connected to anything.