64-bit ARM coming

[Atreidin]

I guess you tried via Python shell but couldn't understand the result, so you are dropping the argument. Nice way out

After reading everything you wrote, I hope you have been trolling. If so, good job. If not, then to everything you wrote, *facepalm*.

It is funny when someone is so incredibly confident about something they are so incredibly wrong about.

 

[Uarch]

... Never mind. Not worth the effort.

 

[Rebel_L]

ncalipari, you have entertained me for a good chunk of time today, thank you. On the off chance that you are not just trolling I suggest that you forget worrying so much about math and worry about your English skills a little more so that you understand that no matter what you thought you were saying, what you said was that 10 is approximately 10 times greater than 2. Until you come to terms with that there is no point in talking about it.



As for ARM, I dont see how it can take over without having backwards compatibility. I always thought most people attribute the x86 long term dominance to backwards compatibility, so until all the computing is done in the cloud and the devices dont run the programs themselves how is ARM's lack of backwards compatibility a design advantage? Unless it can take over inside a single generation of hardware and then not worry about efficiency because it no longer has a competitor I would think that ARM will be way less efficient when they have to start integrating compatibility for the first time without the years of experience to learn the tricks for doing it well.

 

[kevinsbane]

Order of magnitudes confront exponents, not digits, of numebrs stored in scientific notation.

Oh, I finally understand where you went wrong in your understanding. Ok, glad we have that cleared up.

Would a 64 bit ARM processor consume more power than a 32 bit ARM processor with the same compute throughput?

 

[cbn]

Finally, their (Intel) ability to scale their manufacturing processes so quickly means that they will be more power effecient than arm not solely due to their architecture (x86) but do to their ability to be consistently to be (at least) one manufacturing node ahead of their competition .

Yes, but several writers claim Intel requires very large profits in order to consistently fuel those process node advances.

If that is true, then ARM doesn't necessarily need to beat Intel outright. All ARM would need to do is lower Intel's profitability *just enough* over time in order to slowly starve them out.

 

[Soulkeeper]

big question is will we EVER see something on newegg/amazon with this tech in it

I doubt it ...
But I would definitely be interested

 

[cbn]

big question is will we EVER see something on newegg/amazon with this tech in it

I doubt it ...
But I would definitely be interested

Maybe we will see this: http://en.wikipedia.org/wiki/Project_Denver

 

[ncalipari]

Guess what the result is - and no, obviously no arithmetic substitution here.

you are right. The problem is that you used a number too big. Try to add some zeros.

You're just confusing one particular popular fp representation with.. well no idea what this even has to do with the whole topic. There are dozens of different ways to represent floats.

Error propagation doesn't depend on the base you use to store the FP. Error computation does.

So it was fun to guess what you come up next with, but time for me to get some sleep. Although I really enjoy seeing how you can throw completely unrelated and different topics (fp precision, significant digits,..) into one topic!

They are unrelated to someone who doesn't understand how CPUs handles FP

 

[ncalipari]

After reading everything you wrote, I hope you have been trolling. If so, good job. If not, then to everything you wrote, *facepalm*.

It is funny when someone is so incredibly confident about something they are so incredibly wrong about.

the confidence come from the bread I find on my table

 

[ncalipari]

Oh, I finally understand what I did wrong in my understanding. Ok, glad we have that cleared up.

Yes, You don't use logarithms to confront magnitude. You only need to confront Exponentials.

Scientific notation also enables simpler order-of-magnitude comparisons. A proton's mass is 0.0000000000000000000000000016726 kg. If this is written as 1.6726×10−27 kg, it is easier to compare this mass with that of the electron, given above. The order of magnitude of the ratio of the masses can be obtained by comparing the exponents instead of the more error-prone task of counting the leading zeros. In this case, −27 is larger than −31 and therefore the proton is roughly four orders of magnitude (about 10000 times) more massive than the electron.

 

[jones377]

And I thought this thread was about 64-bit ARM?!? Where are the mods when you need them.......

 

[ncalipari]

And I thought this thread was about 64-bit ARM?!? Where are the mods when you need them.......

we dont need mods here. We just need someone to have the humility of taking up a physics book.

I disagree. There's too much going on today while I was gone to ding everyone as necessary, so I'm going to make this simple: you are hereby dismissed from this thread. Do not post in here again.

As for the rest of you, we're done arguing about physics and scientific notation. Take that to Highly Technical. This thread is for discussing 64bit ARM.
-ViRGE

 

[Tsavo]

Yes, but several writers claim Intel requires very large profits in order to consistently fuel those process node advances.

If that is true, then ARM doesn't necessarily need to beat Intel outright. All ARM would need to do is lower Intel's profitability *just enough* over time in order to slowly starve them out.

First of all, Intel spends more in paper towels and tampons for its workers than their closest competitor earns in a year.

Beyond that, Intel recruits within its own ranks for their leadership jobs and looks internally first and only at the very best of anyone else externally.

Going beyond all this, Intel has a Grovian mindset that only the paranoid survive because Intel was caught sleeping with the P4 and it is not going to repeat that corporate mindset that led them down that road.

Worse yet, is that this paranoia has displaced complacency on a corporate level and it has become the tacit mantra of everyone at Intel.

I'd like to say they are like the borg, but their ideal is different and vastly more frightening in that the Borg would preserve your uniqueness and add it to its collective; Intel will just mow your ass down and not bother to look back because you aren't all that smart and distinctive after all.

 

[zsdersw]

Check your facts. IBM had a higher performing processor that they could have used for the IBM PC, but they chose the intel's 8088 instead because it was cheaper, not because it performed better. It was good enough.

IBM's PC took off, not because of performance, but because it was cheap and modular. With the old style of PCs, you bought everything, all the hardware was made by one manufacturer. With IBM's PC was designed from the beginning to be completely modular. It was the modularity that lead to IBM PC's being cheap and the modularity that ultimately caused IBM PCs to be the dominate desktop arrangement.

http://en.wikipedia.org/wiki/Intel_8088
http://en.wikipedia.org/wiki/IBM_801

That has nothing to do with why x86 became the architecture of choice for content creation.

 

[zsdersw]

And now everything is changing again by going with GPGPU

Slowly, yes.

 

[DeathReborn]

So... in an effort to get this thread back on track...

Is ARMv8 64bit the same ISA as being used in Project Denver or are nVidia doing something different again?

Would just be interesting to know if nVidia will get a leg up on the competition come 2014 or not.

 

[soccerballtux]

Slowly, yes.

if only we could gbu acceleration the transition, maype it would habben faster.

 

[ViRGE]

So... in an effort to get this thread back on track...

Is ARMv8 64bit the same ISA as being used in Project Denver or are nVidia doing something different again?

Would just be interesting to know if nVidia will get a leg up on the competition come 2014 or not.

I don't think NV has ever announced what ISA they're using, but since ARM v8 is brand spanking new I doubt Denver is using it.

 

[cbn]

http://www.anandtech.com/show/5027/appliedmicro-announces-xgene-arm-based-socs-for-cloud-computing

I didn't see the Anandtech article posted in this thread

According to that article ARMv8 will be sampling 2nd half of 2012.

Target is up to 3Ghz clockspeed with up to 2 watt TDP per core.