Comments Locked

41 Comments

Back to Article

  • heartinpiece - Tuesday, September 16, 2014 - link

    Why on earth did they decide to lower the memory controller frequency (leading to bandwidth drop?)
    This'll sacrifice GPU performance as well as CPU performance...
  • Andrei Frumusanu - Tuesday, September 16, 2014 - link

    The Mali's T760's AFBC feature should provide a lot of room to bandwidth utilization. We still don't know the real-world effect until we test it out, but don't dismiss the chip's bandwidth quite yet.

    http://www.anandtech.com/show/8234/arms-mali-midga...
  • GC2:CS - Tuesday, September 16, 2014 - link

    It is powering QHD !

    Unless there is some "Crystalwell" cache inside like in Apple A7 (which I heavily doubt) it just won't be enough.
  • MrCommunistGen - Tuesday, September 16, 2014 - link

    Getting A53/A57 to production as quickly as possible, Samsung may have had some issues in their first production revision design or with the manufacturing itself. This could also be a reason for not running AArch64(?). Perhaps the chips aren't fully functional in that regard. Rather than letting the chips go to waste, might as well sell them in a few select markets, right?

    Also, they've been making and tuning the A15/A7 memory controllers for quite some time and across multiple manufacturing processes. I imagine they've got that optimized pretty well. As their first A57/A53 memory controllers on the also new 20nm node I imagine they still have quite a bit of performance tweaking they can do to improve memory bandwidth down the line.
  • MrCommunistGen - Tuesday, September 16, 2014 - link

    Posted before finishing my thought...

    5430 was likely a "clearing the pipes" chip for the 20nm process using the very familiar A15/A7 designs. It fits in the product stack and is certainly a nice "last hurrah" for A15. Less aggressive clocks on a smaller process also help ensure its not constantly thermally throttling.

    Similarly, 5433 could also be their "clearing the pipes" chip for 20nm + A57/53 ahead of a more wholehearted launch of Exynos 7xxx down the road.
  • coburn_c - Wednesday, September 17, 2014 - link

    Well hopefully memory bandwidth isn't a problem, because LPDDR4 is still using 32bit channels. Between that and the power overhead I doubt throughput will increase much throughout this generation.
  • AppleCrappleHater2 - Tuesday, September 23, 2014 - link

    Exynos chips have always tended to outperform Snapdragons released at about the same time.

    Problem is that Cortex-A15 has been a dud in terms of power consumption, which is why it hasn't been seen in anything but a small percentage of mainline phones. Even in tablets where its poor power efficiency is less of an issue, Qualcomm Kraits have been dominating.

    The truth is - at peak performance, Krait is far more power efficient than the A15, which means that it takes longer for a Krait to hit thermal throttling limits and it can run for longer on battery. What's the point of having that insane peak performance if you can't sustain it without overheating or killing the battery?

    Snapdragons didn't have such a clear advantage when it came to Scorpion vs. Cortex-A9, or even dual-Krait vs. quad-A9 which is why Exynos 4210 and 4412 dominated Samsung's handset lineup in their heyday - but the A15 has been a failure outside of the tablet and Chromebook market.

    Samsung will claim the reason for Krait dominating their handset lineup is LTE in public documentation, but the Note 2 and international LTE GS3 (I9305) proved that Samsung can hang a Qualcomm LTE modem off of an Exynos CPU.
  • ddriver - Tuesday, September 16, 2014 - link

    I see samsung is still very dedicated to strangling its chip with narrow 64 bit memory bus. 13.2 GB/sec - that's quite the step back from the 5430, seems like samsung intends to starve those powerful cores so that in a few months it can release yet another "faster" chip without breaking too much sweat at R&D instead of doing the smart thing and launch A57 products with 128bit memory bus from the start.
  • GC2:CS - Tuesday, September 16, 2014 - link

    Anybody who expected samsung to deliver a fully thought 64-bit solution is rather foolish.

    The are rushing to 64-bit as fast as they can, look a this chip, a great example ! They used alpha as an experiment to kick up the 20nm process a bit and then they swapped the cores for those 64-bit ones, they don't mind to improve anything else over the 5430.
  • Death666Angel - Tuesday, September 16, 2014 - link

    So they are taking a page out of Intels book? Or AMDs (4770 for 40nm)? How is that a bad thing? Do minor architectural changes for process shrinks and to major architectural changes on the same process.
  • MrCommunistGen - Tuesday, September 16, 2014 - link

    Interesting... my go to example of this concept is also the HD 4770 even though I'm sure that there are more modern examples of the concept.
  • frostyfiredude - Tuesday, September 16, 2014 - link

    If I'm remembering correctly Apple make some die shrunk 32nm A5s for the iPod Touch and Apple TVs mid-cycle a handful of months prior to the A6's release. Same idea.
  • melgross - Wednesday, September 17, 2014 - link

    But they didn't put them in new products. Just swapped them into current products that were a year old at the time.
  • BoneAT - Tuesday, September 16, 2014 - link

    I'm also surprised why did they go quiet about, though with the Mobile Phone and other divisions being so far apart they are almost like different companies, I see how they didn't make a fuss about it, when they should've. Here they had the chance to equip ALL Note 4s and Edges with a genuine 64-bit nex-gen ARM-based SoC with the Android L update already in testing, but for some reason they sticked with the dual Snapdragon/Exynos release, and probably capped the 5433 bacause of it. Bit silly cause both the Octa-core CPU and the T760 GPU comfortably beats out the 805 and Andreni 420, both in performance AND efficiency and probably in heat management.

    I had the pleasure to play around with the 5430-based Galaxy Alpha with the T628 for a few days, and the hardware stayed cool, low-consuming and incredibly consistently top-performing under all circumstances. The 20nm process indeed brought major benefits, and without having to manufacture the A8 chips for Apple, I find it odd why didn't they go all out producing 5433s for all Notes and Edges and probably Tab S models too, especially with the integrated Intel Cat.6 LTE modem cause LTE was the only advantage of Snapdragons before.
  • Krysto - Tuesday, September 16, 2014 - link

    They went quiet because they don't want people to know what a dumb decision they made to continue using a 32-bit Qualcomm chip in US markets.
  • bigstrudel - Tuesday, September 16, 2014 - link

    I've been warning people not to buy Android right now for just this reason. Brand new flagships that aren't even able to use ARMv8 in Android L.
  • djvita - Wednesday, September 17, 2014 - link

    tell them to buy a phone after CES/MWC 2015 when we will have Samsung S6, HTC M9, Xperia Z4, LG G4 announced with L pre-installed. even the nexus 6 will use 32bit according to rumors.

    same goes for windoge 9
  • jjj - Tuesday, September 16, 2014 - link

    Qualcomm has pretty much a monopoly in the US phone market.
    If the US had any regulators maybe they would be curious to check IF carriers have any role in that and how legal that is.
  • melgross - Wednesday, September 17, 2014 - link

    Nonsense. Qualcomm simply makes better performing chips than Samsung. Performance means more here than it does in most other places. That why Samsung uses their chips here, and their own in less critical markets.
  • lilmoe - Wednesday, September 17, 2014 - link

    Very uninformed.
    Samsung Mobile is sort of a "separate" division from Samsung Electronics/Semiconductor and they're not too loyal about using Exynos themselves even when they've always been generally better than Snapdragons (contrary to what you believe). Exynos has always had better CPUs, better ES 2.0 performing GPUs, better DACs, and more features (like fast charging since the GS4).
    Qualcomm has always offered Samsung Mobile great pricing and good offers (to the point that LG and other OEMs publicly complained), and they also have the somewhat more advantage in the US and European in carrier approval for their modems and radios.

    Benchmarks don't tell the whole story, and it is only recently that Samsung's software is being better optimized for Exynos to make better use of its features in general, and better utilize big.LITTLE in particular.
  • tuxRoller - Wednesday, September 17, 2014 - link

    Since benchmarks don't tell the whole story how can we confirm your views? Especially the fairly significant idea of Exynos ALWAYS being better.
  • lilmoe - Thursday, September 18, 2014 - link

    By actual real world usage... Especially comparing both SoCs on the exact same hardware/software (Galaxy+Touchwiz). The GS3, GS4, GS5, and Note3 all came in 2 variants. I've used both variants of the GS4 and Note3 first hand and can tell you that the Exynos versions run faster in general usage and gaming and they last longer on battery, but they don't perform better in benchmarks.

    Also, the Snapdragon GS4 took 2.5 hours to charge from 0-100, but my Exynos GS4 takes about 1.5/1.7 hours...
  • Rama TT - Friday, September 19, 2014 - link

    Its not that Samsung chips are less performing just that by integrating the processor as well as LTE modem and selling them as a single package Qualcomm is able to compete on price compared to Exynos
  • michael2k - Tuesday, September 16, 2014 - link

    There is nothing to brag about, as per ARM:
    http://anandtech.com/show/7995/arm-shares-updated-...

    On the same process in 32 bit mode there is only a 15% to 30% improvement in performance. The article you're replying too even says this chip may never boot into 64 bit mode, even when Android L becomes available. It's not that they are capping the 5433, it's that they haven't gone through the trouble of testing in 64 bit because Android L isn't out yet. I bet they will release a new SoC after Android L is released. They did it before with their first Exynos 5410, which suffered from a cache coherency bug that negatively impacted performance and battery life, but was fixed with the 5420. I expect an updated 64 bit SoC early next year with higher clock and switched on 64 bit mode.
  • lilmoe - Wednesday, September 17, 2014 - link

    "The article you're replying too even says this chip may never boot into 64 bit mode, even when Android L becomes available"

    That's a pretty confident assumption. With all due respect to Andrei, he was also confident about his assumption that Exynos 5433 was 32 bit. I seriously wouldn't be in a hurry to make another similar assumption if I was him.

    Everyone is assuming that Samsung didn't make any adjustments in their new chips except for the die shirnk. I have no idea on what ground they're basing their opinions on. But even if it was true, isn't that exactly what Intel has been doing for years in their "tic" updates? Why is Samsung considered a culprit now?
  • Andrei Frumusanu - Tuesday, October 28, 2014 - link

    I didn't make assumptions about the 5433, I had hard evidence that were pointing out to A15/A7 cores. That Samsung would *fake* their software stack to make the chip appear as a A15/A7 is something completely different which nobody could have predicted.
  • jjj - Tuesday, September 16, 2014 - link

    Can't wait for some numbers.
    The clocks do seem rather low, makes you wonder if it's a choice or the process is not allowing more than that.
  • banvetor - Tuesday, September 16, 2014 - link

    Are you sure about the Samsung 20nm process? I thought that Samsung was skipping 20nm and going directly to 14nm... At least comparing with TSMC processes, where 16nm is "just" the addition of finfets with 20nm metalization.

    In my opinion this chip is either Samsung 14nm or TSMC 20nm, but of course I could be wrong... ;)
  • Andrei Frumusanu - Tuesday, September 16, 2014 - link

    Yes. The voltages point out to a 20nm chip.
  • Achtung_BG - Tuesday, September 16, 2014 - link

    Samsung USA fab S2 Austin -> 20nm ->14nm, fab S1 SK Gi Heung ->14nmLPE, fab S3 SK Hwa Seong -> 14nmLPE. Exynos 7 use custom core in 14nmLPP.
  • Exophase - Tuesday, September 16, 2014 - link

    Samsung themselves say that Exynos 5430 is 20nm:

    http://www.samsung.com/global/business/semiconduct...

    No reason why 5433 can't be as well. No way are they using TSMC for their own SoCs. Note they didn't strike 20nm off the list here: http://www.samsung.com/global/business/semiconduct...
  • JoshHo - Tuesday, September 16, 2014 - link

    Remember, 14nm Samsung is also a 14nm FinFET process with 20nm BEOL. This means area gains don't scale as much compared to Intel's 14nm process.
  • Krysto - Tuesday, September 16, 2014 - link

    > The Exynos 5433 is definitely a surprise that many didn't expect

    Um, no. Note 4 using Exynos 5433, which is a 64-bit chip has been rumored for months. Also, the fact that it's Cortex A57/A53 on 20nm and with Mali T760 was known weeks ago, too.
  • Andrei Frumusanu - Tuesday, September 16, 2014 - link

    At this time last year the 5430 was rumored to be a 64bit 14nm chip shipping in the Galaxy S5. Rumours don't always pan out and there's a lot of misinformation being reported as fact by many.
  • lilo777 - Tuesday, September 16, 2014 - link

    Most likely it just did not make any sense for Samsung to waste their efforts on 64-bit support in the absence of 64-bit Android. Also, given the circumstances aren't they actually getting the best technical solution where the performance of new CPU benefits from having new architecture (more registers?) without OS/apps having to waste memory on 64-bit pointers (with RAM being smaller than 4GB anyways)
  • redtfsi - Wednesday, September 17, 2014 - link

    You are writing: The chip's firmware and drivers are running on a "CAL".
    So I suppose it is not hardware limitation.
    Could this change when they (or someone else) update their firmware, so that it becomes 64bit capable?
  • darkich - Wednesday, September 17, 2014 - link

    Soo, some of us were right and AT was wrong..

    Man this chip might actually make me buy the Note 4 since it already basically is what I expected and wanted in the Note 5 - a truly future proof 20nm A5x Mali T760 SoC.

    The Snapdragon 805 was barely an improvement over the Note 3's Snapdragon 800AB on 1080p, but this is an entirely different beast.

    If tests show that this chip isn't crippled by its initial release status, I'll be almost half a year ahead of competition.
  • tuxRoller - Wednesday, September 17, 2014 - link

    Not sure how great that Mali (judging from the AT article and improvements made to adreno) will be but those those big.LITTLE cores look incredible.
  • patbean - Wednesday, December 10, 2014 - link

    Dose anyone think the gap will be filled with an exynos 6xx with A7/A17 cores and Mali 760 at say 14nm to round off the 32bit range. For devices that will never havev more than 4gb of ram. Or have we seen the end of 32bit from samsung?
  • TaiHT - Friday, December 26, 2014 - link

    Any chance Samsung will update chip's firmware to support 64-bit when Android L is out for Note 4?
  • LingerBringer - Sunday, January 11, 2015 - link

    I've been thinking of purchasing a smartphone, and I know 64 bit is the way to go. There aren't a lot of high-end 64 bit smartphones on the market as of now. Anyway I've been brought to the Exynos 5433 version of the Note 4 and am curious if it will ever actually get its 64 bit mode activated.

    Can anyone confirm for me if the Exynos 5433 is exactly the same as the recently-announced 64 bit Exynos 7 Octa (with the exception of software and drivers), and that the Exynos 5433 CAN have its 64 bit mode enabled when Android L comes out? Please?

    Would really appreciate any help. Thanks.

Log in

Don't have an account? Sign up now