***Official Galaxy Nexus Pre-Release speculation thread**

[OfTheDamned]

Hey all you folks getting RAZR and hoping for ICS in 2012:

Has it occurred to you that you will have 4 USELESS capacitive buttons at the bottom of the phone?

I'm sorry, but if ICS is making hardware control buttons obsolete, I sure don't want any on my next phone. I call it "instant obsolescence". Get RAZR if you wish and wait for ICS if you will.

My money's on button-less phones from here on out.

ON VERIZON

I've been thinking about that. Do you think they will just render the hardware buttons obsolete on "older" phones with ICS or just reconfigure them to work as the three current buttons on ICS do plus a search button?

 

[OfTheDamned]

My concern began when the FIRST unpack event was announced and VZW's name was NOT part of the event.VZW seems to have been the primary network partner in this from the get go, so I am sure they COULD have been included in that initial announcement. Why would have they opted out of the very initial event? Sure the second wasn't in the USA, but this is all watched through the net anyhow. You don't go from primary partner in a project behind the scenes to nonexistent when things go public for any good reasons.

I don't remember any carriers being apart of the Hong Kong event either. They mentioned the Japanese company, but nobody else was listed in the original announcement other than Google and Samsung.

 

[jasaero]

Why? Wafer splits are hard enough - are you saying demand is so low that makes sense?

More that R&D time for two unique dies has to be justified somehow. Usually it takes big transistor count and die size differences to justify this. Enabling and disabling this or that at packaging ends up being much less costly than laying out two unique dies for production.

This is all my novice take as I don't do this stuff.

 

[dmiller2007]

My concern began when the FIRST unpack event was announced and VZW's name was NOT part of the event.VZW seems to have been the primary network partner in this from the get go, so I am sure they COULD have been included in that initial announcement. Why would have they opted out of the very initial event? Sure the second wasn't in the USA, but this is all watched through the net anyhow. You don't go from primary partner in a project behind the scenes to nonexistent when things go public for any good reasons.

Was any carrier tied to the announcement? Also, what kind of role have carriers played in previous nexus announcements? (serious question, i actually don't know)

 

[EarlyMon]

Hey all you folks getting RAZR and hoping for ICS in 2012:

Has it occurred to you that you will have 4 USELESS capacitive buttons at the bottom of the phone?

I'm sorry, but if ICS is making hardware control buttons obsolete, I sure don't want any on my next phone. I call it "instant obsolescence". Get RAZR if you wish and wait for ICS if you will.

My money's on button-less phones from here on out.

ON VERIZON

Obsolete? Ok...

And where are you getting USELESS from?

You honestly believe that anyone would be so stupid as to not build their revision to support the actual hardware existing on the phone?

I'm going to disagree.

If you find that onscreen buttons obsolesces capacitive button for user-friendliness or something - sure, ok, that's an opinion and those are never wrong by definition.

But - onscreen buttons when not needed with capacitive buttons already there? I don't see that happening.

 

[huskerkate]

Hey all you folks getting RAZR and hoping for ICS in 2012:

Has it occurred to you that you will have 4 USELESS capacitive buttons at the bottom of the phone?

I'm sorry, but if ICS is making hardware control buttons obsolete, I sure don't want any on my next phone. I call it "instant obsolescence". Get RAZR if you wish and wait for ICS if you will.

My money's on button-less phones from here on out.

ON VERIZON

you make a good point, Chief. but, i had heard before that ICS was written to take into account that many devices running ICS would have the physical/capacitive buttons; and, those same "on screen" buttons just wouldn't display. instead, the capacitive buttons would continue to work.

now, that being said, i haven't seen any further discussion (let alone any confirmation) of this - it's just what i remember being theorized and discussed early on. you may be right.

 

[The_Chief]

Okay, EarlyMon, you got me.

Not actually USELESS (poor choice of words), but - shall we say - "old school"

I guess, thinking more about it, ICS will simply not show the onscreen buttons, but use the hardware buttons instead. I'm man enough (and been married long enough) to say the magic words:

I WAS WRONG

 

[jasaero]

I don't remember any carriers being apart of the Hong Kong event either. They mentioned the Japanese company, but nobody else was listed in the original announcement other than Google and Samsung.

I don't know that any worldwide carriers involved behind the scenes in development? Most was probably done here in the USA from a software and network testing perspective anyhow so VZW was probably the ONLY Network operator involved in network tests for the ICS aspects. As far as we know HSPA and other versions of the device coulda been tested with privately acquired SIMs and limited involvement of the carriers of those SIMs.

 

[Steven58]

long day.. gotta catch up again. having a happier day than yesterday, see?

<---me

 

[Stigy]

long day.. gotta catch up again. having a happier day than yesterday, see?

<---me

Happy Steven = happy androidforums!

 

[BlueBiker]

Not sure they come from the same wafers, as the 4460 does have pages more extensions. Also the clocks are independant, but not certain whether they are truely independant, or locked to a reference based on multipiler. I'll have to get some clarification on it.

One easy observation: Raising 4430's nominal speed by 50% (1.0 -> 1.5 in 4460) only raises the GPU's speed by 26%. I don't know SoC hardware, but wouldn't it be an obvious choice to have selectable multipliers to keep all of the various I/O components running within spec even when the CPU's cranked?

I would think the extension differences would be something you just turn off and on, but have the same die. Not sure how many pages you are talking about, but doubting even a huge number of pages of extensions would lead to any drastic differences in die size and transistor count to justify creating whole separate designs on separate waffer runs. Not an expert on this stuff by any mean, but pretty sure the norm is to use the same die and enable and disable features for different models of chips unless there is drastic transistor count difference that can happen with separate designs.

Look at sections 4.3.4 and 4.3.5 in the 4460 manual. The Memory Adapter and Cache Management Unit (providing a redundant optimized interface to external RAM) are documented there and all over the manual, but aren't mentioned at all in the 4430 manual. So either it's brand new functionality in the 4460 or it was excluded from the 4430 doc for some reason.

 

[EarlyMon]

More that R&D time for two unique dies has to be justified somehow. Usually it takes big transistor count and die size differences to justify this. Enabling and disabling this or that at packaging ends up being much less costly than laying out two unique dies for production.

This is all my novice take as I don't do this stuff.

I do this stuff.

You design a chip with a specialized CAD software. (Think Photoshop on steroids with lots of built-in math and layering because wafer manufacturing is all about layering of materials and machine processes during manufacturing.) You then map that to a die size and layout. And then generate a wafer map of dice by a relatively simple block-replication step to optimize the number of squares that fit into a given circle (as supported by your wafer-cutting tools, the step from finished wafers to finished dice).

Long story short, that generates an output file(s) that goes to process machinery and wafers are made.

Mixing the two products adds R&D steps, doesn't save them. And then there's the operators in production who have to handle 2 things at a time instead of 1 thing at a time.

They could do it - nothing stopping them. It's been done before. But if they did, it would be driven by some economics that we'll never likely know - not R&D complexity.

 

[dmiller2007]

I don't know that any worldwide carriers involved behind the scenes in development? Most was probably done here in the USA from a software and network testing perspective anyhow so VZW was probably the ONLY Network operator involved in network tests for the ICS aspects. As far as we know HSPA and other versions of the device coulda been tested with privately acquired SIMs and limited involvement of the carriers of those SIMs.

You don't think they tested the phone on every network they're releasing it on? Weren't one of the leaked videos from Germany? Even if the theory that Verizon is the ONLY network it was thoroughly tested on, doesn't that just give Verizon/Samsung/Google more reason to have it released on Verizon as opposed to waiting months to test it on other carriers?

 

[EarlyMon]

long day.. gotta catch up again. having a happier day than yesterday, see?

<---me

As predicted.

Welcome back!

 

[Sandroidfan]

Or until next week if someone releases a 4460 phone running at its normal speed.

Maybe. But that seems unlikely as Razr now seems to have 4430.
I think real Nexus challenger will come from Sammy with faster Exynos and future Moto phones with 4470 or OMAP5. But that's at least half year away.

 

[Kroaker]

Hey all, first time poster, long time lurker. One thing I can't quite figure out is why some keep saying the Razr has better hardware specs than the Nex, but nothing I've seen backs that up. Higher potential storage seems to be the only area that the Razr beats this. Plus, we will have the plus of ICS, no bloat, and no Motoblur. This thing should easily be better than the Razr.

 

[EarlyMon]

Welcome to the forums, Kroaker!

 

[cereal killer]

My concern began when the FIRST unpack event was announced and VZW's name was NOT part of the event.VZW seems to have been the primary network partner in this from the get go, so I am sure they COULD have been included in that initial announcement. Why would have they opted out of the very initial event? Sure the second wasn't in the USA, but this is all watched through the net anyhow. You don't go from primary partner in a project behind the scenes to nonexistent when things go public for any good reasons.

Precisely. Not to mention the CTIA event was cancelled after all the invites went out. That is nothing to shrug off. I said it back then when i heard nothing about potential problems that something serious must have happened or the Nexus was going to be so "hot" that out of respect to steve jobs they just couldnt do it.

Common sense says that something serious happened and what that was is anyones guess. Maybe the handset was buggy and not ready for prime time and they had to loadd and test 4.0.1. Maybe Verizon did back out who knows. Its not totally out of the realm of possibility that the launch will not happen and its not totally out of the realm of possibility that the launch will in fact happen. Anything is possible.

Nobody can claim either or with 100% certainty. Fact of the matter is Verizon was supposed to be their initial launch partner and they have been non-existant. Wishful thinking is great but maybe we should just step back and see how this plays out and not dismiss anything?

 

[EarlyMon]

Okay, EarlyMon, you got me.

Not actually USELESS (poor choice of words), but - shall we say - "old school"

I guess, thinking more about it, ICS will simply not show the onscreen buttons, but use the hardware buttons instead. I'm man enough (and been married long enough) to say the magic words:

I WAS WRONG

No worries. I was more about not letting that get out as the new myth than bustin' yer chops!

Spoiler

Roger that on the man enuf and married enuf.

 

[Stretlow]

Does anyone know how the 1.5 processor in the vigor (rezound) stacks up against the 4460.

Dont think ive read it anywhere

Stret

 

[dmiller2007]

Hey all, first time poster, long time lurker. One thing I can't quite figure out is why some keep saying the Razr has better hardware specs than the Nex, but nothing I've seen backs that up. Higher potential storage seems to be the only area that the Razr beats this. Plus, we will have the plus of ICS, no bloat, and no Motoblur. This thing should easily be better than the Razr.

Welcome to the madness. I agree with you. Couple your info with the probability of it never get the bootloader unlocked, the RAZR isn't even a possibility for me. Even if the nexus doesn't come to Verizon. Which i think it will.

 

[AntimonyER]

Yeah I am halfway tempted to take a trip over to the RAZR forums to see how they are taking the news of the 4430...

 

[jasaero]

I do this stuff.

You design a chip with a specialized CAD software, then map it to a die. (Think Photoshop on steroids with lots of built-in math and layering because wafer manufacturing is all about layering of materials and machine processes during manufacturing.) You then map that to a die size. And then generate a wafer map of dice by a relatively simple block-replication step to optimize the number of squares that fit into a given circle (as supported by your wafer-cutting tools, the step from finished wafers to finished dice).

Long story short, that generates an output file(s) that goes to process machinery and wafers are made.

Mixing the two products adds R&D steps, doesn't save them. And then there's the operators in production who have to handle 2 things at a time instead of 1 thing at a time.

They could do it - nothing stopping them. It's been done before. But if they did, it would be driven by some economics that we'll never likely know - not R&D complexity.

I was referring more to R&D to get to the initial maps you are referring to I suppose, but you could be correct.

Just seems in the PC world that a lot of times you only have one or two die designs/maps for a CPU or GPU and then they pump them out and bin them and enable or disable this or that to get a plethora of models. Actually remember early AMD processors you could jump circuits on the packaging to get new features only intended for higher models if you got a lower model chip. It seems I have read different overall maps if you will were only developed for quite a bit different die sizes where you might be able to fit a good bit more dies on a waffer? That sound normal? I don't know and it's possible 4460 is enough extra circuits to justify another die based on this thinking for all I know.

 

[dmiller2007]

Precisely. Not to mention the CTIA event was cancelled after all the invites went out. That is nothing to shrug off. I said it back then when i heard nothing about potential problems that something serious must have happened or the Nexus wa

Didn't the event go pretty smoothly in honk Kong without mentioning a carrier? If a carrier was the issue, don't you think they could have just changed the script at CTIA to not include the mention of a carrier?

 

[jasaero]

You don't think they tested the phone on every network they're releasing it on? Weren't one of the leaked videos from Germany? Even if the theory that Verizon is the ONLY network it was thoroughly tested on, doesn't that just give Verizon/Samsung/Google more reason to have it released on Verizon as opposed to waiting months to test it on other carriers?

Probably, but I am talking more of earlier developmental testing. Carrier evaluation sorta testing would have been later on as ICS started to reach completion and the device hardware was more or less done and finalized. My impression is that VZW has been involved since before the hardware device was even real well defined and that early prototype devices were actually VZW LTE stuffs on VZW here in the USA. I say this because rumors of this one being first VZW Nexus go WAY WAY back.