Please create an account to participate in the Slashdot moderation system


Forgot your password?
Slashdot Deals: Cyber Monday Sale Extended! Courses ranging from coding to project management - all eLearning deals 20% off with coupon code "CYBERMONDAY20". ×

Comment Re: funny and sad (Score 1) 403

What differentiates them is the number of instructions per cycle each core can execute, which is informed by the amount of on-die cache (L1, L2, and in the higher-end chips, L3) available. The cores are the same, the glue logic is effectively the same, with changes being necessary for the larger or smaller caches, but the caches are larger and that allows more instructions to be queued for execution...

First, I will freely admit that I couldn't design a modern CPU with a gun pressed to my head. BUT I'm not sure that the design of the "Core" series has much to do with ARM design. Yes, I get what you are saying about larger caches being generally faster/better; but the length of the execution pipeline makes even a bigger difference, as do niceties like lookahead, out-of-order execution, etc. And I seem to remember an Intel document where they SHORTENED the instruction pipeline, because the "cost" of refilling a long pipeline based on a bad-predictive branch decision was higher than refilling a shorter pipe. So sometimes, more isn't always better in CPU design.

It wasn't until the A7, which is a 64-bit chip, that Apple's CPUs became competitive with the high end of the Android market; and at that point they were completely destroying anything in the Android world. That is owed in large part to the larger 64-bit instruction set;

Ok, I'll give you that; especially with ARM, where Opcode and Operand are often combined in the same instruction. (Extrapolating from my experience writing 32 bit ARM assembly language).

A dual core chip can only work on two different things at once and, as a result, will switch execution context twice as often as a quad core chip when more than two processes are simultaneously demanding CPU time. Context switching is expensive, and that's why benchmarks that don't account for it don't matter in the real world.

You are right that context-switching is VERY important. And I seem to remember a recent article that complained the Android handled multiple threads much worse than iOS (which I found amazing, actually, because all the Linux people whine about OS X's context-switch overhead). And since Benchmarks actually test SYSTEMS not CPUs, it kinda all gets lost in the sauce.

Bottom line: I would really like to know what someone on the A[x] Chip Development Team has to say about all this. Do you have any of those friends in your back pocket?

Comment Re: funny and sad (Score 1) 403

Yes, they do design the CPUs ...and I never said they didn't... from ARM block designs. They arrange the blocks, they design the connecting lattice between the blocks, and yes, they've designed a few non ARM-core blocks to reside on the same die for specific functions; those blocks are not ARM, though. They're application-specific processors like audio and video hardware CODECs and such; living on the same die as the ARM CPU makes them ARM blocks in the same way that living on the same die as the x86 CPU makes an Intel GPU and x86 block: it doesn't.

I believe that it goes deeper than that. What you are describing is what most everyone else does with ARM designs. I'm pretty sure that Apple does more tweaking inside the CPU "block" itself. How else would you explain the difference in performance between A-Series ARMs and everyone else's? Seriously, they are getting more performance out of their dual-core designs that competitors are out of their quad (and more) core "equivalents".

All of that can't be attributed to differences in glue-logic and memory controllers.

Comment Re:funny and sad (Score 1) 403

The terms Stupid or gimmicky has never been applied to the Apple ][ series.

That's why I qualified the statement with "other than that". With the "that" being the Apple 2 and original iPhone.

No, you placed "year" restrictions around the Apple ][ that were wholly incorrect.

I'm not sure what metrics you use to determine "Stupidity" or "Gimmicky-ness"

Cost more, did less, tried to justify it with shineyness. Most Apple 2 owners upgraded them for 286's when they came out and never looked back. There is a reason Apple almost went bankrupt in the 90's.

And that wasn't it. It was the fact that Apple tried to go in too many directions at once, and buried themselves in SKUs. And the final blow was Apple Licensing MacOS. THAT almost did them in. But it was never about "cost more, did less", unless the "Less" was Windows-Specific software.

But now all that has changed...

Fast-forward to the present: Apple makes the best laptops

Only since the Macbook became a shinier Wintel box. Not really thinking different now is it?

Just because Apple changed to Intel CPUs (the smartest move they ever made!), does NOT mean they became a "Wintel box". Far from it.

Comment Re:Real bad news (Score 1) 403

Nope, built in volume (100K/yr) in a factory in Zhuhai. Using a TI amplifier ($1.20/100K pieces), AKM DAC ($1.00 for a good quality unit, 100K pricing). Regulators ($0.70 - need 3.3V and 1.8V, LDOs are cheap but not that cheap), passives (another $0.40, driven mainly by caps), connectors, mechanicals (squirt a part, shoot a little paint, you're at $0.50 in 250K pricing). It's not cheap - which is why there is a STRONG market for counterfeit IAP2 chips, and many who don't pay the appropriate licensing fees, and many who use raw plastic finish (screw flow and knit lines), improper connectors (fake Lightning connectors with questionable tolerancing), etc.

This isn't just "an adapter" - it also would need to have a full DAC and amplifier inside, as well as a power supply. So it's more than the simple connector/wiring adapters you're thinking of.

Your quantities are low by AT LEAST an order of magnitude. Plus, at those quantities, and with a company like Apple, price-points for components are nothing like what you can find on a distributor's (or manufacturer's) website.

If Apple really were to do this, it would be purchasing at the "million" pricing level (think about the number of iPhones/iPod Touches/iPads that would be affected), not the piddly 100k level, and would be negotiating special pricing on pretty much every component. So, shave at least another 25-30 percent off the cost, especially since Apple doesn't have to pay as much for the IAP2 chip, and no licensing either. Plus, they would likely simply integrate the controller and DAC on the same die, and maybe even the amp. The quantities are such that it would make sense for Apple to spin-up custom silicon for this.

And if you look at my post, I wasn't contemplating a passive adapter, either.

Comment Re:Stop making super thin phones you idiots! (Score 1) 403

As devices get thinner, batteries get smaller, and battery life shrinks. Batteries aren't subject to "Moore's law". They do not gain power as they shrink in size.

Yes, but because of a corollary to Moore's Law, the devices powered by those batteries get more efficient; thus "effectively" increasing the battery-size.

Comment Re:Yea the proof is in the patent (Score 1) 403

Unfortunately this has come out a while back. Shortly after Apple bought Beats. It does appear Apple is moving away from the standard jack because it limits potential for a thinner design. Apple seems obsessed with thin and the only other thing holding Apple back on thin is bendgate. Apple will either have to find a more rigid material in a thin state, or have materials that withstand bending. So get ready to buy a ugly dongle if you want to use a standard headphone with iPhone's in the future. Or buy a wireless one.

They already developed that improved alloy.

Comment Re:Real bad news (Score 1) 403

I carry around a separate Fiio DAC and headphone amplifier with battery in it for listening to music.

You wanna talk about a company that is NOTHING but HYPE, that's Fiio.

I almost peed my pants from laughing while reading the sales brochure for the X5 Gen II. OMFG the utter bullshit!

First off, you simply can't design something that runs off batteries that has a power supply impedance low enough and grounding and shielding good enough to do a 24 bit DAC justice. And since pretty much ALL digital music is 16 bit, there is only SO much dithering can do with those other 8 bits. Sure you can hunt down a FEW 24 bit downloads; but (unfortunately) they simply are few and far between; so your 24 bit DAC, like lossless encoding formats, is simply assinine on a portable music player or phone.

Comment Re:Real bad news (Score 1) 403

Well, the IAP2 chip and license is about a buck, and then you'd need a DAC and amp. It's around $4-$5 minimum to do Lightning to powered analog for the electrical. Add the packaging required (either an on-cable box, or a separate box) and battery and you're up to close to $10 for the Lightning-to-highend headphone adapter. So figure $40-$50 MSRP. Not an insignificant amount!

I assume you're building these by hand in the U.S. out of off-the-shelf parts from TI or someone similar. I submit that, at scale, these could be manufactured in some faraway land for $2 to $3, landed cost. That puts them in the $20-30 retail world, even with insane margins. Just like nearly every other Apple adapter. They'll build the whole thing into the Lightning connector body, and have a short cable with the 3.5mm jack at the other end.

Comment Re:funny and sad (Score 1) 403

Apple has always been a parody. Except for maybe a breif period in 1983 with the Apple 2, and again in 2007-2009 when the iPhone was leading product of its type, but other than that, Apple products have been mostly stupid and gimmicky.

First off, the Apple ][ pretty much supported Apple from 1977 through 1992, and there wasn't a school from K through University in the U.S. that didn't have several dozen to several hundred of them. During the same time, at least up through the mid '80s, it was relatively rare to see a Wintel PC in a school. The terms Stupid or gimmicky has never been applied to the Apple ][ series.

Then we have the Macs, starting in 1984. I'm not sure what metrics you use to determine "Stupidity" or "Gimmicky-ness" (maybe if a computer doesn't come in a beige steel box, eh?); but Macs have been every bit as "Serious" of computers since at least the advent of the Macintosh II in 1987. Actually, the "toaster" Macs were pretty serious, too, compressing 80% of the software and hardware technology of the $10,000 Lisa (the computer so far ahead of its time that almost no one could understand it) into a cube 1/4 the size of the Lisa, and for nearly 1/10 the price. True, the Mac was SUPPOSED to be a Laptop (a "Dynabook", actually); but what they did was still pretty neat, compared with the competition of the time.

Fast-forward to the present: Apple makes the best laptops, as opined by, well, most everyone, and does a pretty damned good job at their other product lines as well.

Curiously enough, though: The one thing that they have NEVER done is make a beige steel box. The PowerMac 9500/9600 are about as close as they got to that; but even those weren't just generic Wintel PCs.

Comment Re: funny and sad (Score 1) 403

That's two components, at least one of which is based on licensed block designs from ARM, who could simply stop licensing them for future use and BAM, no more A series chips can be made. Go on thinking Apple invented everything in the iPhone, really, go ahead, don't let those pesky facts get in the way. Don't get me wrong, I've got a number of apple products and I do love them, but i don't bullshit myself about what Apple actually does as a company. The AC you replied to is absolutely correct.

Actually, Apple has a "Soft Core" ARM license that actually lets them design their own CPUs (and they do), rather than just licensing IP from ARM. Being a founding member in ARM doesn't hurt, either. There are a few other companies that have such a license (Samsung and Qualcomm being two of them, IIRC); but Apple actually does design A-series chips from the CPU-on out.

Comment Re:Bullshit (Score 1) 403

This would mean a DAC, headphone amp, and batteries in every headphone.

So, IOW, a single, fairly simple ARM microcontroller to negotiate the Lightning protocol and do the D/A conversion, likely sub-dollar at Apple's quantities, with possibly a 5 cent amplifier chip and a few passive components. No batteries, since Lightning has power pins. The most expensive thing would be the connector.

Apple could sell it at $20 and still make a handy profit.

If you're talking about bluetooth, I am less sure what would be required; but yes, that would obviously require batteries and a bit more guts to deal with the RF stuff.

Comment Re:converter (Score 1) 403 audio to the speaker is the future, and then it might as well be wireless.

I've wanted that for the past 25 years. No reason that it can't happen.

Most subwoofers are "powered"; why not the "main" speakers? LOTS of nice things can happen when speakers and amplifiers are "matched". For one thing, passive crossovers (which are responsible for a LOT of bad things in a LOT of speaker systems) become a thing of the past. Another is "tuning" the amplifier's response to "flatten" the system. For another, A/V Receivers stop having to be behemoths, and can actually start being modular control systems, where consumers can easily choose their particular needs for number and type of audio in/outs, virtual tape/processing loops, EQ "plugins", etc. In fact, your entire A/V control system can become an App, with faceless boxes that do your bidding, and can distribute the signals to wherever. Yes, they have these things to some extent already; but to really do it right still costs quite a bit, and is really fairly limited.

I disagree to some extent on the wireless part for phase-coherence reasons; but I suppose the same signal that can carry the audio information can also carry a syncronization/timestamp so that samples are presented for conversion at the "same" time. And wireless SURE makes doing surround-sound setups easier in a LOT of homes (including my own).

Comment Re:converter (Score 1) 403

As with virtually any apple device, there will be a $75 piece manufactured for 85 cents that will be a lightning to headphone jack connection.

As with the other lightning connectors, if you plug it into your mac it will crash when it wakes on sleep.

Excuse me. How do you plug a Lightning connector into a Mac? Are you talking about trying to plug one into a Thunderbolt socket?

Never appeal to a man's "better nature." He may not have one. Invoking his self-interest gives you more leverage. -- Lazarus Long