backtop


Print 60 comment(s) - last by Cheesew1z69.. on Aug 16 at 5:09 PM

Galaxy Note 4 will also include a 16MP rear camera with OIS

Well, that didn’t take long. Earlier this week, we were given an early glimpse of the Galaxy Note 4 via some leaked photos. The photos revealed a device that shares a similar design language with the newly announced Galaxy Alpha.
 
Today, the full spec sheet of the Galaxy Note 4 has been revealed courtesy of Indonesian e-tailer erafone. The leak confirms what many have already surmised: the Galaxy Note 4 will be a beast. The smartphone comes with a 5.7” QHD (2560x1440) Super AMOLED display as we already knew, and two processor options. The first processor is the 2.5GHz, quad-core Qualcomm Snapdragon 805, while the second option is the octa-core Exynos 5433. The phone will be available with 16GB, 32GB, and 64GB storage options and a microSD slot is included for further expansion.
 
And in what will be a first for an Android smartphone, the Galaxy Note 4 will reportedly feature 4GB of RAM.

 Leaked Galaxy Note 4 images from earlier this week.

If that isn’t enough, the Galaxy Note 4 will also include a 16MP rear camera with optical image stabilization. Other features include Bluetooth 4.0 LE, 802.11ac, and a microUSB 3.0 port for sync/charging.
 
Lastly, the Galaxy Note 4 will ship with Android 4.4.3 (KitKat).
 
The Galaxy Note 4 will be revealed early next month at IFA 2014 in Berlin, Germany. We’re almost certain to see more leaks between now and the official reveal date, but it sure would be nice to get some clearer shots of the device’s exterior…

Sources: erafone, via Phone Arena



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Really?
By amanojaku on 8/14/2014 9:48:26 AM , Rating: 5
quote:
And in what will be a first for an Android smartphone, the Galaxy Note 4 will reportedly feature 4GB of RAM.
I was under the impression that the highest amount of RAM in a phone so far is 3GiB. In an Android phone, also from Samsung. So 4GiB is a first for ALL phones, not just Android.




RE: Really?
By kingmotley on 8/14/14, Rating: -1
RE: Really?
By danbob999 on 8/14/2014 12:02:27 PM , Rating: 2
Flash is not RAM.

DRAM is a type of RAM, by opposition to SRAM. SRAM is faster but much more expensive. There isn't any phone or even PC with more than 4GiB SRAM. It's used mostly for cache (2-12MiB).
Flash memory is neither SRAM or DRAM.


RE: Really?
By jamescox on 8/14/14, Rating: 0
RE: Really?
By Banana Bandit on 8/14/2014 5:53:22 PM , Rating: 5
man the picking is thick today.

Look guys, cool the jets. Everybody gets the OP is talking about memory (DRAM/SRAM/MRAM/whatever) as opposed to mass storage (flash/NAND/HDD/SDD/whatever).

We don't need to be a rocket scientist or pull a dictionary out of our pocket to figure that out.

Just as an aside though flash memory itself is not a randopm access device. They are actually block-access devices (much like sectors fetched on Winchester hard drives) where fixed-size blocks of data are read/written by their controllers (these days blocks are 4K in size but may be bigger depending on the cluster size of file system used on the device). It is the controllers that deal with fetching the specific bytes requested in a random-like manner. While they make a part of the overall storage device, the controller is not the memory itself. And we are talking about the different KINDS of memory, right?

DRAM and SRAM are are more like true RAM devices where a single byte can be addressed by row and column strobes (though even then a full row consisting of a single 64-bit word is fetched for performance).


RE: Really?
By jamescox on 8/15/2014 2:36:37 PM , Rating: 1
All because of a missing 'D'. I agree with you that the original post attempting to correct the person about DRAM vs. RAM is nitpicking, but I don't like it when an essentially correct post is responded to with "epic fail" and a bunch of people spouting terms that they do not know the meaning of.

quote:
Just as an aside though flash memory itself is not a randopm access device.


As an aside and a reiteration, random access refers to possible access patterns which is determined by latency and how tight the bounds are on that latency; it has nothing to do with block size. As you stated, you can't (directly) access a single byte in any modern DRAM either. As far as the processor is concerned, it never accesses smaller than a last level cache line, which is 64 bytes (not bits). For RAM, you can access any block with similar latency and this is true of DRAM and flash. It is not the same latency, but close enough for the next level not to notice. For disk, you can't guarantee this since it is not a random access device. An access may come from the drives cache or it may require a full seek which is orders of magnitude greater. This requires the rest of the system to make allowances for non-random accesses.

The new NVM express interface was created because flash allows RAM-type access patterns rather than the large and variable latency of a disk. Seek time on a hard disk is in milliseconds, flash is (currently) in microseconds, and DRAM is in nanoseconds. PCM and other technologies are bringing non-volatile memory closer to DRAM latencies though.

These distinctions are important since we will soon be stacking DRAM in the same package as the processors. The DRAM will look more like a last level cache. Are we still going to connect external DRAM? Probably not, since this would be a waste. The external memory will not need to be anywhere near as fast as current DRAM since it will be used differently. It will be used more like disk is currently used in the memory hierarchy; as swap space. There will continue to be some markets for massive amounts of external DRAM. For consumer applications, if you have something like 4 or 8 GB of stacked memory in the package and a PCIe connected SSD at several GB/s bandwidth, then external DRAM will be mostly unnecessary. Or maybe not; who knows.

"The less a man makes declarative statements, the less apt he is to look foolish in retrospect."


RE: Really?
By Warwulf on 8/16/2014 3:46:26 PM , Rating: 3
quote:
All because of a missing 'D'


That's what she said.


RE: Really?
By BRB29 on 8/14/2014 12:15:11 PM , Rating: 2

quote:
No, there have been many phones with more than 4GiB of RAM before. What they haven't had is more than 4GiB of DRAM. The 64GiB of flash storage in a "64GB iPhone" is RAM as well


RAM(such as SRAM, VRAM, DRAM, etc..) are extremely fast, much faster than flash or HDD. They also have a defining characteristic of losing all data in it after it is powered off.

Flash on the other hand does not lose the data. That is why one is used for storage and the other used for system memory. When a computer or phone is shut down. One of the things it needs to do is to save some of the data in the RAM to the disk so it doesn't lose it. That is why shutting down a device isn't instant.

So no, RAM isn't part of the flash memory in a phone. It would be too slow if it was.


RE: Really?
By jamescox on 8/14/14, Rating: -1
RE: Really?
By Banana Bandit on 8/14/2014 6:09:54 PM , Rating: 3
I think everybody on this board reading his post understands exactly what the OP is saying except maybe ... you.

Please put the dictionary away, you are staring to look like an idiot.


RE: Really?
By Spuke on 8/14/2014 6:51:02 PM , Rating: 1
quote:
Please put the dictionary away, you are staring to look like an idiot.
Actually, you're the one looking like an idiot right now. There's nothing wrong with some clarification. Not everyone here is an expert like yourself.


RE: Really?
By amanojaku on 8/14/2014 10:39:16 PM , Rating: 2
Actually, EVERYONE is wrong. RAM means random (more appropriately "direct") access to any BYTE (the smallest useable data quantum), so you don't have to pull a block of data just to gain access to a single byte. Typically, solid-state devices are direct access.

The alternative to direct access is sequential access, in which you pull a block of data (e.g. 4096 bytes) in order to find the desired number of bytes (e.g. 2). This includes hard drives, tape, CDs/DVDs/BDs, etc... Storage admins are familiar with how tuning block sizes affects performance.

Flash is most certainly non-volatile memory, but it is NOT RAM. Flash memory can READ bytes randomly just like RAM. Flash makes an awesome DRAM substitute for read-only data (e.g. flash acceleration cards for SAN/NAS). However, flash is ERASED/WRITTEN TO sequentially; you can't erase and write a single byte, you have to erase a whole block before writing! This means flash is not suitable for lots of small writes due to the performance hit AND the wear-leveling. Neither would be so bad if flash memory could be written to randomly. Er, directly.

And since you can't write a single byte without erasing a whole block, that means memory techniques like bitwise operations are totally inefficient on flash.

Bonus: the "random" in RAM comes from the past when the contents of memory wasn't mapped. Your program was responsible for being able to find data in a range of addresses. Today's RAM contains a map of its contents, which you take advantage of if you use programming languages like C that offer pointers and references. So, today's RAM isn't random at all!


RE: Really?
By Banana Bandit on 8/15/2014 10:52:47 AM , Rating: 2
In some ways correct. Even back in the dawn of personal computers you built memory banks to fetch an entire word at a time and buffer it for the memory controller to come along and do its thing. In the days of 8-bit computers with 16-bit address space that would be considered true RAM since words really were only 8 bits wide. Your row and column select strobes were causing single bytes to be pulled to the memory read buffer from that intersection.

These days the RAS and CAS strobes select 64 bit words rather than 8 bit, so if you want an 8, 16 0r 32-bit value, you still have to pull the full 64-bit word and extract what you want from it. Not totally random, but as random as we can get today.

Mass storage devices are either block (HDD, SDD, Flash, UDF) or sequential (Tape, Non-UDF CD & DVD). For block devices you pull a block from the device and buffer it while you use random access access methods to extract the request ed bytes from it. While at the high level they are Random Access devices the term 'RAM' is not commonly used when referring to them.

And THAT is where all these clarifications can cause confusion.

The layman uses the term 'RAM' to refer to the computer's volatile working memory. The portion of memory used to hold a program and it's working data while it is running. There is a large distinction between this memory and random access storage . You cannot run a program directly random access storage.


RE: Really?
By jamescox on 8/15/2014 2:52:17 PM , Rating: 2
quote:
The portion of memory used to hold a program and it's working data while it is running. There is a large distinction between this memory and random access storage . You cannot run a program directly random access storage.


That distinction is quite arbitrary and changing. At one time, processors worked almost directly with the systems DRAM. Now DRAM is farther removed, which is why the speed of DRAM doesn't make much performance difference. We have L1, L2, and now L3 cache. If we start stacking DRAM in the processor package, then this looks like just another level of cache and the SSD looks a lot like current DRAM modules. Part of a currently running program could be sapped out to disk. What is "direct" is dependent on what level of abstraction you are using to view the system. Modern processors (mostly) execute instructions "directly" on registers only. Anything else requires a load instruction to have the memory system fetch the data and get it into a register. The requested data could be out on disk or anywhere in-between.


RE: Really?
By jamescox on 8/15/2014 4:03:59 PM , Rating: 1
quote:
Actually, EVERYONE is wrong. RAM means random (more appropriately "direct") access to any BYTE (the smallest useable data quantum), so you don't have to pull a block of data just to gain access to a single byte. Typically, solid-state devices are direct access.


If we are having a semantic argument, and everyone else appears to be wrong to you, then this means that you are assigning the wrong meaning.

Current DRAM is far from "direct" access. There is no way to grab a single byte. This is largely dependent on the level of abstraction which you are using to view the system. Between processor and DRAM, accesses take place in entire cache lines, which is usually 64 bytes (not bits). The 'D' stand for "dynamic", as opposed to "static" (SRAM). It is called dynamic since it is capacitor based and leaks charge. It has to be read and rewritten periodically (some amount of microseconds in current implementations, I believe) or the value will be lost. Static RAM retains the value as long as the power is on; it does not decay.

quote:
Flash is most certainly non-volatile memory, but it is NOT RAM. Flash memory can READ bytes randomly just like RAM. Flash makes an awesome DRAM substitute for read-only data (e.g. flash acceleration cards for SAN/NAS). However, flash is ERASED/WRITTEN TO sequentially; you can't erase and write a single byte, you have to erase a whole block before writing! This means flash is not suitable for lots of small writes due to the performance hit AND the wear-leveling. Neither would be so bad if flash memory could be written to randomly. Er, directly.


You can't erase or write a single byte "directly" in a modern DRAM either. Usually you would flush an entire dirty cache line. Internal to the DRAM, it is usually a square array, so it will actually have to write the entire line which is a device dependent size and much larger than a byte. Also, if you write outside of the currently "open" line, you incur an extra latency penalty to close the current line, write it back to the array, and open a new line. There is a lot of confusing terminology which I am trying to avoid here, like page is used in DRAM but also in virtual memory "pages". Basic point: the access block size has nothing to do with whether it is considered RAM.

quote:
And since you can't write a single byte without erasing a whole block, that means memory techniques like bitwise operations are totally inefficient on flash. Bonus: the "random" in RAM comes from the past when the contents of memory wasn't mapped. Your program was responsible for being able to find data in a range of addresses. Today's RAM contains a map of its contents, which you take advantage of if you use programming languages like C that offer pointers and references. So, today's RAM isn't random at all!


I am not even sure what you are talking about here. Any operation, like bitwise operations, require bringing the data into a processor register. Anything not in a register will result in a request to the memory system. The requested data could be anywhere in L1 cache, L2 cache, L3 cache, system DRAM, or even out on disk. The instruction will stall until it is brought into a register.

I think you are confused about how memory systems currently work. The "random access" part of RAM means that random accesses patterns are a reasonable thing to do. This is like comparing an array and a linked list. For the array the desired address of some specific element can be computed with a single addition and possibly accessed with low latency. If I say give me element number 10,000 of a linked list, this will take a long time unless you do some complicated stuff. You would have to start at the first element and access each element, counting elements, until you get to the 10,000th element. This will be slow even on a modern system which is why you generally do not index (directly access) elements of a linked list.

This is similar to DRAM and flash vs. a hard disk. If I ask for some block in DRAM or flash I have a narrowly bounded latency which is relatively low compared to the next level of memory hierarchy out. If I ask for some specific element on a hard drive, the hard drive must move the read head to the location where the data is. This is similar to the linked list metaphor. The access could be really fast if the data is in the hard drive cache or at the current head position. If it requires a full seek, then it could be on the order of milliseconds, even higher if the drive needs to spin up first. Flash is currently on the order of microseconds and DRAM in nanoseconds. Hard drives can optimize accesses under heavy loads by reordering them. In this case they just sweep the read/write head back and forth and deliver which ever requested data comes up under the read/write head first. This is similar to magnetic tape drives, except obviously faster than seeking to the end of a tape. We can't put a very narrow bound on the latency of hard drive accesses, so the rest of the system must be designed to cover up this latency. This is not the case with flash. In fact, accessing it like a hard drive was holding it back. The new NVM express (non-volatile memory on pci-e) interface allows flash to be treated as RAM instead of limiting it to a disk interface, which assumes high latency and extremely variable latency.

It is a common problem in software design that random access is needed throughout a block of memory which is larger than the cpu cache. This decreases performance significantly compared to random access within a size that fits in cache. If you have random access required outside of the size of your DRAM memory, then this will get ridiculously slow, and could push the system into what I call thrashing, where the system is doing nothing but swapping memory pages and not accomplishing any useful work. I have an old Core2Duo laptop which is limited to 3 GB of DRAM. It wasn't thrashing, but it was noticeably swapping pages to disk when switching applications. I put an SSD in; since it essentially allows random access, jut a little slower than DRAM, swapping is almost unnoticeable now. In fact it seems faster than some newer systems with hard drives now.

Sorry for the long post. If you are a computer programmer (or CS student), I would recommend taking a computer organization and design class, which may be included in computer engineering departments instead of CS. My class (which was a long time ago) used the textbook "Computer Organization & Design: The Hardware/Software Interface" by Patterson and Hennessy. I think they have kept this updated, so it may still be used. I would recommend this to CS students since it helps understand performance limitations in software design.


RE: Really?
By Banana Bandit on 8/15/2014 10:27:16 AM , Rating: 1
Did you understand what the OP was trying to say? I am pretty well sure that 99.9% of the people on this board understood.

So why do we need somebody confusing the issue by removing the distinction between the device's working storage and mass storage? So what if they both use the same high level access paradigms? Sorry, but I don't call that 'clarification'. I call it being a grammar nazi. There was no need for it.


RE: Really?
By jamescox on 8/15/2014 4:26:06 PM , Rating: 2
quote:
I think everybody on this board reading his post understands exactly what the OP is saying except maybe ... you.


Please note that I am not the original person who nitpicked the difference between DRAM and RAM. kingmotley is. I just took issue with a bunch of people posting "epic fail" and a bunch of other incorrect BS in response. I would not have attempted to correct someone for saying RAM instead of DRAM in this case. The responses to kingmotley were obviously wrong though, regardless of your opinion on the meaning of RAM.

quote:
Please put the dictionary away, you are staring to look like an idiot.


Perhaps both kingmotley and I are idiots but if many post on an internet forum bother you, then you probably shouldn't read them, and you defiantly shouldn't respond to them and go about calling people idiots. Calling people idiots probably isn't going to stop a useless flame war.


RE: Really?
By CZroe on 8/14/2014 7:30:29 PM , Rating: 2
No. NVRAM is typically CMOS (Complementary Metal-Oxide Semiconductor) for storing PC BIOS settings without external power (hence, "non-volatile"), usually powered by a battery just like the SRAM memory in old Nintendo game paks. Though the BIOS itself is often FlashROM or EEPROM, the "flashing" of a BIOS refers to clearing before programming, which was once done on windowed PROMs and EPROMs erased by being "flashed" with a UV light (EEPROMs do not require this). FlashROM is Flash Memory packaged like an EEPROM and typically addressed with similar logic. When later N64 games started using this instead of SRAM, EEPROM, or dual EEPROM, Nintendo called it "FlashRAM," but that is the only place I have ever seen it called that. It was probably part of the console's memory map, like the ROM in the game pak, so they were justified in calling it that.

If being able to be accessed randomly makes Flash storage RAM, then re-writable optical discs, HDDs and floppy disks would be too. They aren't. You simply must distinguish between workspace memory and storage. It's regretable that flash storage is already commonly referred to as "memory cards" and "flash memory" because the word "memory" was once exclusively used to mean "RAM" in the tech sectors.

RAM is workspace memory. Storage is not. There's a reason your page file for virtual memory is paged back into real system memory before processing can continue. Heck, many Nintendo game paks had SRAM for no other reason than to use as additional system memory (ROM was mapped straight to system memory anyway)!


RE: Really?
By Banana Bandit on 8/15/2014 11:01:44 AM , Rating: 2
Just to add to this, a computer's bios is not used directly by a processor. It is actually block-copied into the computer's volatile memory for fast access at runtime during the bootstrap process. The same applies to those game cartridges. The random access happens only when the NVRAM is bulk-copied into DRAM.

This is the reason you can flash your computer using a Windows or DOS-based executable. When you update your bios, you are updating the chip - not the running copy of it. It is also the reason you have to reboot when you are done.


RE: Really?
By CZroe on 8/15/2014 8:14:50 PM , Rating: 2
"The random access happens only when the NVRAM is bulk-copied into DRAM."

I think you meant ROM, but even that wasn't copied into system memory until most cartridge games were compressed (N64). Prior to that, the resource data on most cartridge ROMs were directly mapped to memory and not copied into RAM unless it needed to be modified. CD consoles had to copy everything into RAM. Of course, the ROM instructions used system memory to write all the dynamic values to, hence, the Pro Action Replay letting you edit/freeze your life count, timer, etc. If, by "NVRAM," you meant SRAM, well, the NES and SNES often used that as additional system memory on the cart to expand system memory even further. Some games that didn't even have a battery had SRAM for this reason. This was even used as a copy protection in many games (certain memory addresses ranges beyond base system memory which are normally ROM are checked to see if they are writable).

The system BIOS in a PC is copied into memory on modern systems, but it was once optional (CMOS Setup>BIOS Shadowing>OFF).


RE: Really?
By jamescox on 8/15/2014 1:35:22 PM , Rating: 2
quote:
No. NVRAM is typically CMOS


"The best-known form of NVRAM memory today is flash memory." from wikipedia.

I think most people would disagree with you. Flash is NVRAM by many peoples definition. This is a semantic argument, so you can assign any meaning you want to these terms though.

quote:
If being able to be accessed randomly makes Flash storage RAM, then re-writable optical discs, HDDs and floppy disks would be too. They aren't. You simply must distinguish between workspace memory and storage. It's regretable that flash storage is already commonly referred to as "memory cards" and "flash memory" because the word "memory" was once exclusively used to mean "RAM" in the tech sectors.


This has to do with what level of abstraction you are viewing the memory from, so there is little point in arguing frame of reference. From a certain level, DRAM could be viewed as block access. You can not reach out and get a single byte, bit, or whatever size you are looking for. On any memory access, an entire cache line will be read and brought into cache.

quote:
RAM is workspace memory. Storage is not.


That is a rather arbitrary distinction. A processor only works "directly" with registers or L2 cache in some cases. "Directly" dependent on level of abstraction. Otherwise it is just different size blocks being swapped out by the memory system. Cache lines get swapped out to the next level of cache, until they eventually get pushed out to DRAM. DRAM gets swapped out to disk (SSD or hard drive). DRAM was essentially accessed "directly" at one time, but this has long passed.


RE: Really?
By Banana Bandit on 8/15/2014 2:27:00 PM , Rating: 2
quote:
I think most people would disagree with you. Flash is NVRAM by many peoples definition. This is a semantic argument, so you can assign any meaning you want to these terms though.


By that same token, the acronym RAM is commonly accepted and used to refer to the working memory (memory that is directly used by the memory controller to execute programs and access program runtime data) of a computing device. While mass storage devices like disks and flash devices can be seen as random access devices, the acronym RAM is rarely applied to them in general usage.


RE: Really?
By jamescox on 8/15/2014 4:15:13 PM , Rating: 2
quote:
While mass storage devices like disks and flash devices can be seen as random access devices, the acronym RAM is rarely applied to them in general usage.


That is the exact distinction we are going for here though. A hard disk can not reasonably be used as a random access device while flash can be. This exactly why the new NVM express memory interface was created; to actually treat flash as RAM instead of sequential storage. I would agree that RAM is obviously often used by laymen to refer to DRAM only, and I would not have attempted to correct someone for using RAM instead of DRAM. I am not the original poster who did so. I would take issue with using the term "direct" though in your post. Direct has little to no meaning in modern load/store system unless you are talking about processor registers.


RE: Really?
By CZroe on 8/15/2014 8:34:04 PM , Rating: 2
I am perfectly aware of this. My first PC BIOS with no onboard battery for CMOS was an Intel D850GB in 2001, though laptops were doing this much earlier for security reasons (couldn't just pull the battery to reset the BIOS password on a stolen laptop). That's why I carefully worded it. FlashROM replaced EEPROM + CMOS in many modern motherboards. It doesn't change anything because it is still referring to the SRAM that is being simulated in system memory and shadow-copied to FlashROM. Calling it NVRAM today does not mean NVRAM = Flash, it is simply a legacy term that is now often applied to a type of Flash in a VERY specific implementation. While NVRAM can be Flash, Flash! = NVRAM. Get it?

As for system memory not being RAM because it is external and accessed through cache and buffers or in blocks: LOL! It is DIRECTLY addressed by the CPU by memory address as part of a memory map. No matter how many layers of abstraction, it is still system memory directly addressed by the CPU. It is the most NON arbitrary way to define it and yet you call it "symantics?!" HA! You've got a lot of reading to do, mister. ;)


RE: Really?
By CZroe on 8/15/2014 8:38:02 PM , Rating: 2
*semantics


RE: Really?
By ritualm on 8/14/2014 2:12:01 PM , Rating: 2
quote:
The 64GiB of flash storage in a "64GB iPhone" is RAM as well.

Epic fail.

That 64GiB of flash storage is NAND, not RAM. The former is non-volatile and slow, the latter is volatile (loses data on a power loss) and fast.

Until new NAND-like RAM e.g. racetrack memory are commercially feasible to be produced in mass quantities, NAND and RAM are not the same.


RE: Really?
By jamescox on 8/14/14, Rating: -1
RE: Really?
By Banana Bandit on 8/14/2014 6:06:40 PM , Rating: 2
NAND (a memory device - not a controller) is block-access just like a hard drive.

I don't call the hard drive in my computer RAM though at the highest levels it can be seen as a RAM device. Nor do I call my SSD or microSD cards RAM even though again at the highest levels the are accessible randomly. I call them storage.

You can pick nits on access patterns till you turn purple, but when someone tells me they have 16 gig of ram, I know that they are NOT talking about their hard drive, ssd, sd card, bernoulli box, floppy drive or cassette tape recorder.

I know they are talking about working storage that their processor's memory controller directly access rather than mass storage.

I don't think any of us needs a translator ;)


RE: Really?
By jamescox on 8/15/2014 4:33:58 PM , Rating: 2
I am not the original poster who nitpicked the difference between DRAM and RAM. Take it up with that guy.


RE: Really?
By retrospooty on 8/15/2014 8:33:08 AM , Rating: 2
Epic fail fail failed. (sorry, I feel like a douche for saying that, but it's in direct response to your post).

That isnt how it's set up in a phone. Put extremely simple without all the jargon, it's exactly like geekman said below

"RAM in smartphones is just like RAM in PC, it's where programs are loaded into, when it is to be processed by the CPU... ...Flash Memory in smartphones is like the HDD or SSD in a PC, it's used to store programs and files."

A typical high end smartphone these days has 2-3gb RAM to run the OS and apps, and 32gb storage to store files and apps.


RE: Really?
By jamescox on 8/15/2014 4:31:34 PM , Rating: 2

I do not want to continue arguing semantics. NAND is not directly opposed to DRAM, as the original "epic fail" poster implies. That was all I was trying to get across. I know exactly how flash storage is organized in different systems.


RE: Really?
By geekman1024 on 8/14/2014 11:10:54 PM , Rating: 1
RAM in smartphones is just like RAM in PC, it's where programs are loaded into, when it is to be processed by the CPU.

Flash Memory in smartphones is like the HDD or SSD in a PC, it's used to store programs and files.

Simple as that.


RE: Really?
By jamescox on 8/15/2014 4:37:14 PM , Rating: 2
quote:
RAM in smartphones is just like RAM in PC, it's where programs are loaded into, when it is to be processed by the CPU. Flash Memory in smartphones is like the HDD or SSD in a PC, it's used to store programs and files. Simple as that.


It isn't anywhere near that simple, but if you are not a computer hardware or software engineer, then you don't need to know. If you don't know, then you shouldn't be posting BS in response to people who do know.


RE: Really?
By danbob999 on 8/14/2014 12:04:11 PM , Rating: 2
Finally a phone with 4GiB. It has been long since the Galaxy S3 with 2GiB. We stalled for a few years. They should have put more RAM before increasing the CPU core count to 4 IMO.


RE: Really?
By atechfan on 8/14/2014 8:04:41 PM , Rating: 1
Can the 805 address 4GB of RAM? I am a little bit skeptical of that spec in the leak. Looks like a very nice phone though. I wish that Samsung would make a Windows version. With Onenote, a stylus equipped Windows Phone seems like an obvious choice.


RE: Really?
By Banana Bandit on 8/15/2014 11:26:00 AM , Rating: 2
32 bits can address 4,294,967,295 bytes (4GB).

The reason you rarely see systems with that much memory is that some of it is inaccessible to the process due to device I/O communications buffers and registers (which need to be mapped into processor address space).

It may be that ARM32 architecture allows a full 4GB to be used, but somehow I doubt it. With Intel x86-32 you have to get really tricky using PAE to access that full 4GB. Some programs can use it but most don't.


RE: Really?
By Banana Bandit on 8/15/2014 11:29:23 AM , Rating: 2
(correction before the correctness nazi's have at me - all references to GB in my post should be GiB -- fk all hard drive marketing asshats for creating that confusion!!)


RE: Really?
By jamescox on 8/15/2014 4:42:27 PM , Rating: 2
quote:
It may be that ARM32 architecture allows a full 4GB to be used, but somehow I doubt it. With Intel x86-32 you have to get really tricky using PAE to access that full 4GB. Some programs can use it but most don't.


AFAIK, ARM does have PAE so it should be possible to map the full 4 GB and other memory mapped devices. I am not going to complain if only 3.5 GB is accessible to programs on my phone, so it is a curiosity but not really relevant in actual use.


Mini-HDMI
By FITCamaro on 8/14/2014 11:23:07 AM , Rating: 1
When Android phones first started coming out, many seemed to have mini-HDMI ports. I'd love to see this return. Yes wireless display technology is out there and things like a Chromecast potentially make it unnecessary, but I'd still appreciate one. In a high end phone like this, you'd think it'd be a possibility.




RE: Mini-HDMI
By FITCamaro on 8/14/2014 11:23:25 AM , Rating: 2
And yes I realize it costs money to license HDMI.


RE: Mini-HDMI
By ProZach on 8/14/2014 11:44:25 AM , Rating: 2
I got to agree with you there.

When looking at phablets I check for mini-HDMI and USB-OTG along with the obligatory microSD. The idea is that I could use be a handheld that can be hooked up to keyboard+mouse and/or monitor (i.e. twitch games, A/V presentations).

I saw some phablets with mini-HDMI and USB-OTG but none were made by companies I recognized and a majority of them did not export to the Western world.


RE: Mini-HDMI
By Spuke on 8/14/2014 11:58:56 AM , Rating: 2
Are we sure this doesn't have HDMI through the USB port?


RE: Mini-HDMI
By Spuke on 8/14/2014 12:02:31 PM , Rating: 2
The linked article says it has MHL 2.0. Doesn't MHL 2.0 support HDMI?


RE: Mini-HDMI
By FITCamaro on 8/14/2014 12:21:20 PM , Rating: 2
Apparently you're correct. It does. I didn't know newer micro USB specs could do HDMI out.


RE: Mini-HDMI
By quiksilvr on 8/14/2014 1:28:13 PM , Rating: 2
Standard micro-usb ports are 5-pin connection and is used in an either-or setup. When plugged into a USB port, its used for data transfer and charging. When plugged into an HDMI port, its used for 1080p video and 7.1 surround sound but needs external power in order to do this.

Unfortunately, Samsung thought it would be a good idea to make their micro-usb ports 11-pin (translated: proprietary). This offers no benefit whatsoever because it still needs external power to operate and even more unfortunately is the fact that there is no adapter that works with both, you have to get the right one.

Standard MHL to HDMI:
http://www.amazon.com/dp/B006V7F380/

Samsung's proprietary nonsense:
http://www.samsung.com/us/mobile/cell-phones-acces...

The day HDMI from TVs are properly configured to provide power is the day I switch to MHL. Until then, I'll stick with my mini-HDMI and avoid the nonsense.


RE: Mini-HDMI
By RjBass on 8/14/2014 3:10:01 PM , Rating: 2
Interesting comment. I use the HDMI adapter from Samsung with my Galaxy Note 3 (I linked to it just below) and I don't need an external power source as you claim I should. Previously with my Galaxy S 1 I did need the power source, but I haven't needed it since my Galaxy Note 2.

If the phone has MHL 2.0 the external power source is not required anymore. When it comes to the Galaxy series, that includes just about everything except the Galaxy S 1, Galaxy Note 1, and a few various models of the Galaxy S 2 but not all of them.


RE: Mini-HDMI
By ProZach on 8/14/2014 3:21:10 PM , Rating: 2
Ok, I'm not directly responding anyone's comment. I knew about the MHL and it seemed like complete nonsense when the standard would change between models.

I'm not a fan of kludgy adapters (unless they're bundled for free, lol) but the other issue is there's no chance of connecting input peripherals with USB-OTG at the same time with a MHL adapter.

So I keep searching.


RE: Mini-HDMI
By NAVAIR on 8/14/2014 4:34:13 PM , Rating: 2
<Unfortunately, Samsung thought it would be a good idea to make their micro-usb ports 11-pin (translated: proprietary).>

The 11 pin micro USB port you talk about is the standard USB 3 Micro port, the 5 pin ports are USB 2 only. I have a Seagate external drive that use's it that is 3+ years old.


RE: Mini-HDMI
By ritualm on 8/14/2014 9:07:41 PM , Rating: 2
Not exactly. There are two distinct MHL implementations: first one is what everyone else uses, second one is specifically unique to Samsung. All standard MHL adapters that normally work for the likes of LG, HTC and others simply do not work when plugged into a Samsung device.


RE: Mini-HDMI
By RjBass on 8/14/2014 12:45:31 PM , Rating: 2
It will. Samsung makes one that works with the vast majority of their Galaxy smartphones. I use one of these in my Galaxy Note 3.

http://www.samsung.com/us/mobile/cell-phones-acces...


RE: Mini-HDMI
By Reclaimer77 on 8/14/2014 12:35:26 PM , Rating: 2
Why not just use an adapter?

I think dedicated HDMI ports have gone away because nobody really uses/needs them. You can do it through USB anyway.


RE: Mini-HDMI
By gookpwr on 8/14/2014 3:59:27 PM , Rating: 2
Um dude they still have the MHL that IMO is better because it charges the phone at the same time if you have the adapter. My MHL has worked from my S3 when they changed the spec for better video output to my S5 before I got the Allcast app to work without issues with my Chromecast. So tell me again why a micro HDMI would help at all?


Shame on you Samsung
By anactoraaron on 8/14/2014 12:39:46 PM , Rating: 2
...for shipping your 'latest and greatest' with 4.4.3. IIRK, there was a security flaw in 4.4.3 which is why 4.4.4 came around so quickly. Who the hell is in charge of this? They need to be removed from their position... Samsung's newest flagship will ship with an OS that has a security flaw... brilliant...




RE: Shame on you Samsung
By xti on 8/14/2014 12:49:28 PM , Rating: 2
im more mad that the note3 att flavor hasnt had their bootloader cracked yet. makes me hesitant on upgrading soon.


RE: Shame on you Samsung
By Cheesew1z69 on 8/14/2014 12:58:06 PM , Rating: 2
quote:
IIRK, there was a security flaw in 4.4.3 which is why 4.4.4 came around so quickly.
Proof?


RE: Shame on you Samsung
By anactoraaron on 8/14/2014 11:09:18 PM , Rating: 2
Is your google broken?

http://www.computerworld.com/s/article/9249251/And...

http://www.gottabemobile.com/2014/08/02/android-4-...

"For Nexus users, Android 4.4.4 KitKat delivered but one security patch for an OpenSSL vulnerability. For other users, it delivered Android 4.4.3 KitKat bug fixes on top of that security fix."

That enough? Or would you like me to google some more for you??


RE: Shame on you Samsung
By Cheesew1z69 on 8/16/2014 5:09:13 PM , Rating: 2
Pay attention, everything you link, is for Nexus. So again.....


wow
By zodiacfml on 8/14/2014 11:34:14 AM , Rating: 2
the Note is beasting again. I wonder how soon this device throttles during heavy use.
Samsung should also add whether or not they will upgrade the device to Android L.




Micro USB Port
By Spookster on 8/14/2014 12:50:41 PM , Rating: 2
Hopefully they have redesigned the micro USB port so it's more sturdy. The one on the Note 2 I have is very flimsy and when you take the phone apart you can tell it's not really secured to anything to really hold it in place.




Boring phone
By kwvdc on 8/14/2014 12:53:37 PM , Rating: 1
6 or 6.3 would be interesting to replace my Note 3 .... now it seems to buy a dual sim Note 3 is the next step ..... Sorry Sammy .... Note 5 might make it if you think ahead




"There is a single light of science, and to brighten it anywhere is to brighten it everywhere." -- Isaac Asimov

Did You Partake in "Black Friday/Thursday"?
Did You Partake in "Black Friday/Thursday"? 





0 Comments












botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki