Mobile World Congress 2015

[Nothingness]

C'mon guys, surely you both caught to the sarcasm in his post, right? He even ended it with the standard maniacal laugh (you know, because it would be maniacal to want Intel to have a monopoly, etc.). Pretty sure he gets it if he is making satirical fun of it

I have no sense of humor especially when it deals with x86.

 

[Fjodor2001]

Yes, that is precisely how these things happen.

The engineers develop a product (in this case a process node) within a set of constraints (timeline, cost, etc) while being directed to maximize a set of priorities (higher density, lower leakage, etc.) and what comes out of that is something that reflects the priorities and direction set by the decision makers within management.

What I think you are mistakenly doing is assuming that there is such a thing as "intrinsic properties of 14 nm". The intrinsic properties certainly do exist, but they exist by intentional creation by the engineers themselves.

As a baker, when I bake pumpkin pie the final product (the pumpkin pie) is most certainly going to have the intrinsic property of pumpkin. Not by virtue of it being a pumpkin pie, but by virtue of me (the baker) intentionally putting pumpkin into the pie when I created it.

10nm and subsequent nodes will continue to target whatever electrical properties Intel's decision makers feel the market is most willing to pay for.

When Apple reports to its shareholders that the market was willing to enrich Apple with $18B of profits in just one business quarter, that is a compelling reason for Intel to pursue similar markets.

No business right now is standing in front of their shareholders telling them they just banked $18B profit in 90 days by selling high-clockspeed desktop CPUs. So why would they make it a priority when there are clearly much more profitable markets to go after?

Yes, as I assumed then. Thanks for laying out the details.

However I'd like to add that the point here also is that even if Intel wanted to prioritize performance over low power, they cannot increase performance as substantially as before (~1970-2005 era). It was discussed earlier in this thread (although it only discusses the frequency aspects of it, and not IPC).

Another thing I take away from this is that it becomes even harder to compare characteristics and metrics of competing nodes. Since there seems to be an ever ongoing fanboy-battle on the forum, how can we e.g. make a fair comparison between Samsung/TSMC/Intel 14/16 nm, when they might have optimized the process for totally different targets (density, frequency, low power)?

 

[DrMrLordX]

I don't care if x86 or ARM is the winner, what I care about is the veritable explosion in inexpensive and readily accessible apps that has come with the ARM ecosystem.

With ARM, the app store is the killer app.

Yeah I know, ARM is too entrenched for anyone to really push them off the mobile sector altogether. It would take Android completely destroying iOS which is just not going to happen. So long as there is iOS, ARM is safe.

Now if Intel had agreed to make that phone CPU that Apple asked them to make years ago, things would be different.

Intel shares owner/investor?

Because only for profit one would like to see a monopoly in mobile SoC market like he have right now in the desktop cpu business.

I'm too poor for that. I buy AMD hardware after all.

Yes can't wait for that to happen... and see you cry because you've lost the possibility to choose and have to pay what Intel wants and accept no progress for generations of similar chips.

x86 *is* a bad ISA. Only someone with zero knowledge of its assembly language and another assembly language would claim otherwise.

So you've never heard of schadenfreude? RISC fanatics have been griping about x86 for decades. I saw all the "RISC IS BETTAR" charts back in the PowerPC vs Pentium debates, I saw all the handwringing and gnashing of teeth when commodity x86 server hardware marginalized RISC in the server sector. The triumph of the awful ISA has been an ongoing comedy goldmine. It's like playing Axis and Allies and watching Russia kick Germany's butt early in the game.

Besides, no matter what happens to the Qualcomm, Samsung, Apple, et al. you'll always be able to count on AMD to offer a credible alternative to Intel products. *cough* *snort*

Pretty sure he gets it if he is making satirical fun of it

In truth, Intel taking over mobile would probably stink for consumers. But you have to admit, it would be pretty funny in an evil sort of way.

I have no sense of humor especially when it deals with x86.

Apparently not!

 

[Nothingness]

So you've never heard of schadenfreude? RISC fanatics have been griping about x86 for decades. I saw all the "RISC IS BETTAR" charts back in the PowerPC vs Pentium debates, I saw all the handwringing and gnashing of teeth when commodity x86 server hardware marginalized RISC in the server sector. The triumph of the awful ISA has been an ongoing comedy goldmine. It's like playing Axis and Allies and watching Russia kick Germany's butt early in the game.

Real men play WiF or ASL, not Axis and Allies. And real mean have spent time writing assembly for x86 and RISC. I guess you never played WiF or ASL and never wrote assembly, right?

Apparently not!

You definitely can't see sarcasm unless properly tagged.

So here I go: this post was all sarcasm.

 

[witeken]

You definitely can't see sarcasm unless properly tagged.

So here I go: this post was all sarcasm.

No, this is very serious. The x86-hoax is an awful conspiracy myth. People who still think x86 has poor battery life, are stuck in 2012.

I don't care

I see what you did there.

 

[Arachnotronic]

x86 *is* a bad ISA. Only someone with zero knowledge of its assembly language and another assembly language would claim otherwise.

There is a reason that most universities teach fresh young CS minds MIPS and/or ARM when introducing them to assembly language/computer architecture

Although it does seem that the latest version of Computer Organization and Design covers x86 and ARM, with little MIPS coverage. Hmm...

 

[Nothingness]

No, this is very serious. The x86-hoax is an awful conspiracy myth. People who still think x86 has poor battery life, are stuck in 2012.

We are discussing the instruction set, not its supposed power efficiency or lack of it. x86 ISA stinks, like it or not.

There is a reason that most universities teach fresh young CS minds MIPS and/or ARM when introducing them to assembly language/computer architecture

I had the displeasure of being exposed to x86 in the early 90s. Needless to say I didn't attend the course (especially given that I already knew x86 :biggrin

Although it does seem that the latest version of Computer Organization and Design covers x86 and ARM, with little MIPS coverage. Hmm...

I'm afraid replacing MIPS with x86 makes sense. At least they seem to cover both ARMv7 and v8...

 

[Arachnotronic]

I'm afraid replacing MIPS with x86 makes sense. At least they seem to cover both ARMv7 and v8...

I'm almost tempted to pick up this edition to see what's new from the fourth edition. Maybe when the sixth edition is released and the used value of this one plunges 90%+

 

[Nothingness]

I'm almost tempted to pick up this edition to see what's new from the fourth edition. Maybe when the sixth edition is released and the used value of this one plunges 90%+

I prefer Computer Architecture, A Quantitative Approach
http://www.amazon.com/Computer-Archi.../dp/012383872X

 

[Arachnotronic]

I prefer Computer Architecture, A Quantitative Approach
http://www.amazon.com/Computer-Archi.../dp/012383872X

I like to think of that one as a sequel to the other text. Where I went to school, we used the first book for computer org. (i.e. basics), and then we used second book for the sequel course (deeper dive).

 

[ninaholic37]

What's wrong with x86? Do you just find it cumbersome to code asm for, or does it perform worse? I read two books from the 90s Assembly Language: Step-By-Step and Programming in Assembly Language on the IBM PC and coded all their examples and it was a blast! The 20-bit segment/offset method was kind of silly but I guess coding in 32-bit would fix that.

 

[DrMrLordX]

Real men play WiF or ASL, not Axis and Allies. And real mean have spent time writing assembly for x86 and RISC. I guess you never played WiF or ASL and never wrote assembly, right?

Not many real men these days. Used to be you just had to own a truck.

You definitely can't see sarcasm unless properly tagged.

I can't stand looking at monitors wearing my glasses anymore. That must be it.

There is a reason that most universities teach fresh young CS minds MIPS and/or ARM when introducing them to assembly language/computer architecture

Although it does seem that the latest version of Computer Organization and Design covers x86 and ARM, with little MIPS coverage. Hmm...

MIPS is pretty marginalized these days. The most prolific MIPS (or MIPS clone, I should say) CPUs these days are probably Loongson chips.

 

[xthetenth]

I'm pretty sure that in the year 2015, there are more pressing concerns for an ISA than being easy to write assembly by hand for.

I mean really, my copy of K and R has a trademark date literally a decade before my birth.

 

[Nothingness]

What's wrong with x86? Do you just find it cumbersome to code asm for, or does it perform worse?

Performance is great. It's just how irregular x86 instructions are, with dedicated registers. And those shift instructions that touch or leave the flags alone depending on the shift amount, crazy. I guess x86-64 is better, but I have lost the faith, and won't even try except for reading and tweaking JITs (in that case you still pay for the horrible encoding).

I definitely like RISC-like ISA much more, be it MIPS, Alpha, POWER or ARM.

 

[Nothingness]

I'm pretty sure that in the year 2015, there are more pressing concerns for an ISA than being easy to write assembly by hand for.

Assembly will always have its uses for performance. Also it's used by JIT. And compilers.

 

[Arachnotronic]

MIPS is pretty marginalized these days. The most prolific MIPS (or MIPS clone, I should say) CPUs these days are probably Loongson chips.

MIPS architecture CPUs are common in networking, but they're being displaced by ARM/x86 solutions.

 

[VirtualLarry]

MIPS is pretty marginalized these days. The most prolific MIPS (or MIPS clone, I should say) CPUs these days are probably Loongson chips.

MIPS is used in nearly every SOHO wifi router on the market, except for the most recent/powerful ones, that have switched to an ARM core or core(s).

 

[IntelUser2000]

2.6pts is (which is 1.3x the 2.0 of 5y71

No it isn't. 2.0 is also achieved by the 5Y10. Asus claims their upcoming Transformer Book T300 Chi will get 2.3 points for the 5Y71.

So most of the 5Y70/71 systems get ~2.0 points. 2.3 points for Skylake is absolutely horrendous.

 

[dahorns]

No it isn't. 2.0 is also achieved by the 5Y10. Asus claims their upcoming Transformer Book T300 Chi will get 2.3 points for the 5Y71.

So most of the 5Y70/71 systems get ~2.0 points. 2.3 points for Skylake is absolutely horrendous.

Obviously its going to depend on the individual implementation. There are actually benchmarks of the T300 chi hitting 2.8 (clearly in up TDP mode).

I'd betting Skylake Core-M will range from 2.6 - 3.0