Intel Z3770 vs Snapdragon 800 in Kraken 1.1

[SiliconWars]

So the take home message here is that kraken is worthless as a benchmark. Sorry, but the pre-requisite for any benchmark being taken is consistency - kraken gives dramatically different results just from using a different browser. This opens up an entire can of worms when comparing results of different platforms with different operating systems. There is absolutely no consistency here, even when testing on the same system.

I don't think there are any good benchmarks for the ARM SOC platform, and krakn seems to be no exception.

In a nutshell, yes. All I'm trying to say here is that this benchmark shows Intel at it's best (that should be clear given the margin). Running S800 on Windows RT, with a single-threaded benchmark under IE is as bad as it gets for the S800, while it's as good as it gets for the Atom.

Both will be great tablet chips, the Atom will feel smooth and snappy unlike any previous one, but we knew all this already.

 

[SiliconWars]

fair enough!

do you think we'll see s800's in sub 200 dollar tablets? Not a leading question. It seems to me that the s800 is purely a premium end device and the s600 and s400 should be the low end to mid end comp? So you have bay trail which should be according to intel playing in price points from sub 200 to premium end. I just wonder how much gpu performance matters here vs cpu performance and performance per watt?

Hard to say. I believe that Intel will fail to gain any traction vs S800 if they attempt that one, however vs the S600 they could have a very competitive offering. Remember they still lack the modem on the SoC itself, which will increase power and board integration costs.

 

[VengenceIsMine]

it was on chrome, and it's a stock tegra 4 you could have just asked "what browser were you using?"

Relax, I didn't get hostile, just pointed out the # in a vacuum isn't very informative. So was it in a Shield or ....???

 

[Khato]

Hard to say. I believe that Intel will fail to gain any traction vs S800 if they attempt that one, however vs the S600 they could have a very competitive offering. Remember they still lack the modem on the SoC itself, which will increase power and board integration costs.

That has indeed been Intel's primary failing in the smartphone space. Luckily it's a comparatively minimal effect when it comes to tablets as it only affects BoM, the small amount of extra board real estate isn't much of a concern. As well it's usually just WiFi connectivity that's needed, not cellular.

Regardless, it'll be quite interesting if Windows RT continues to come up short even when running on S800. Not that being king of the windows tablet market would be saying much for Intel, but it's better than the alternative of ARM actually getting a foot in the door there. Just have to wait and see how the picture looks in Android... as well as the all-important power consumption figures, of which we currently don't really have a clue for either platform.

 

[VengenceIsMine]

That has indeed been Intel's primary failing in the smartphone space. Luckily it's a comparatively minimal effect when it comes to tablets as it only affects BoM, the small amount of extra board real estate isn't much of a concern. As well it's usually just WiFi connectivity that's needed, not cellular.

Regardless, it'll be quite interesting if Windows RT continues to come up short even when running on S800. Not that being king of the windows tablet market would be saying much for Intel, but it's better than the alternative of ARM actually getting a foot in the door there. Just have to wait and see how the picture looks in Android... as well as the all-important power consumption figures, of which we currently don't really have a clue for either platform.

The RT experience was night and day vs the current platform. The snapdragon 800 was very responsive and quick even on the heavier desktop Office applications, Word 2013 launched in under 3.0 seconds while it is more like 2-3x that on a surface or vivotab rt. What was noticeable between the two was that the 3770 was faster at almost everything, not crushingly so like the difference between the Surface RT and this 800 tablet but still it was present and I wasn't the only one who handled the tablets that noticed it. 800 basically completely solves RT's hardware issues, it was a completely satisfactory tablet experience from a speed standpoint and was subjectively faster than an ipad 4 or Nexus 10, though the screen was slightly lower res than those 2. RT still have big ecosystem issues but RT on an 800 or a Tegra 4 will be a very different experience than the current hardware.

 

[Khato]

RT still have big ecosystem issues but RT on an 800 or a Tegra 4 will be a very different experience than the current hardware.

Thanks for the summary of impressions. It's just yet another instance of Microsoft shooting itself in the foot eh? Now that RT's name has been drug through the mud for almost a year killing whatever momentum they'd built up in the ecosystem for windows on ARM they finally are going to make it decent.

I'll freely admit that I'm hoping it'll be too little too late. Maybe if Microsoft wasn't wasting resourcing on their ARM experiment they could make the real Windows better.

 

[ChronoReverse]

You'd think that application launch speed is more tied to how slow the eMMC would be. I suspect the main difference is less crappy storage speed.

 

[blackened23]

The RT experience was night and day vs the current platform. The snapdragon 800 was very responsive and quick even on the heavier desktop Office applications, Word 2013 launched in under 3.0 seconds while it is more like 2-3x that on a surface or vivotab rt. What was noticeable between the two was that the 3770 was faster at almost everything, not crushingly so like the difference between the Surface RT and this 800 tablet but still it was present and I wasn't the only one who handled the tablets that noticed it. 800 basically completely solves RT's hardware issues, it was a completely satisfactory tablet experience from a speed standpoint and was subjectively faster than an ipad 4 or Nexus 10, though the screen was slightly lower res than those 2. RT still have big ecosystem issues but RT on an 800 or a Tegra 4 will be a very different experience than the current hardware.

Word 2013 on a snapdragon 800. Say what? Only Office 365 which is cloud based is available on Android. There is a huge difference between that gimped Android version of Office as compared to the full x86 version.

 

[sm625]

I dont think anyone really doubts that atom will be faster than any of the ARM SoCs. But it is irrelevant because the cost is going to keep it from being used. The average SoC ASP of the tablet market is what? $15? Maybe < $10? I'm factoring in all the mediatek stuff and all the noname brands, not just the apples and samsungs. There is no market for intel to come in and charge $30, $50, even $60. Crackheads...

 

[Exophase]

Word 2013 on a snapdragon 800. Say what? Only Office 365 which is cloud based is available on Android. There is a huge difference between that gimped Android version of Office as compared to the full x86 version.

He's testing on WinRT, not Android.

 

[liahos1]

I dont think anyone really doubts that atom will be faster than any of the ARM SoCs. But it is irrelevant because the cost is going to keep it from being used. The average SoC ASP of the tablet market is what? $15? Maybe < $10? I'm factoring in all the mediatek stuff and all the noname brands, not just the apples and samsungs. There is no market for intel to come in and charge $30, $50, even $60. Crackheads...

i think clovertrail pricing is well below that. Kzarnich is saying bay trail in tablets sub 200...as low as or lower than 150. I dont see where 30-60pp's come in?

 

[tential]

Really wish this would come out already so I could see a review on anandtech so I could be happy lol.

Really want to see if intel can land straight dominance in the mobile market in one fell swoop this year. Haswell is a stud in laptops, if Baytrail matches that, they really have shown that when they put their mind to something, it truly is an amazing product they delivered.

 

[Homeles]

Hard to say. I believe that Intel will fail to gain any traction vs S800 if they attempt that one, however vs the S600 they could have a very competitive offering. Remember they still lack the modem on the SoC itself, which will increase power and board integration costs.

Z3770 should handily defeat S800 in CPU perf and perf/watt, and will nip at the heels of 15W Kabini's top performance (the 15w A4-5200 Kabini, that is), while demolishing it in perf/watt. Their modem is no slouch, either; in addition, its successor is launching 1H 2014, and is expected to be very competitive, if not better than Qualcomm's unannounced MDM9x25 successor.

You are forgetting about Intel's enormous process lead -- they're truly a full node ahead and then some, and the situation will only be further in Intel's favor once 14nm lands and their modems are moved over to their own fabs.

Graphics definitely aren't going to be Intel's game, and I expect they'll be pretty far behind here. I don't know what their phone platform will look like, but it'd be pretty damn funny if Intel was the first to have DX11-capable graphics in a phone.

Cost is going to be the biggest question -- we all know Intel loves their fat margins. Second biggest question is availability -- if these devices launch too late, they'll have 20nm competition, which would largely nix Intel's lead.

 

[bullzz]

@sm625
"I dont think anyone really doubts that atom will be faster than any of the ARM SoCs. But it is irrelevant because the cost is going to keep it from being used. The average SoC ASP of the tablet market is what? $15? Maybe < $10? I'm factoring in all the mediatek stuff and all the noname brands, not just the apples and samsungs. There is no market for intel to come in and charge $30, $50, even $60. Crackheads..."

I dont think s800 costs $15 or even s600. if they did, google wud have put that in nexus7 instead of the 1 yr old s4 pro. qualcomm is becoming like intel in PC space. intel only has to match qualcomm s800 coz they both will be going head to head in windows space or highend android tablets

 

[Khato]

I dont think s800 costs $15 or even s600. if they did, google wud have put that in nexus7 instead of the 1 yr old s4 pro. qualcomm is becoming like intel in PC space. intel only has to match qualcomm s800 coz they both will be going head to head in windows space or highend android tablets

Just to provide a reference point - isuppli's Galaxy S4 estimated BoM. They're estimating $20 for the Snapdragon 600 (and $25 for Qualcomm's LTE wireless section, amusing eh?) And they actually touch that $30 SoC cost for the Exynos version. So there are definitely indications of the premium SoC offerings rising in price.

 

[liahos1]

Just to provide a reference point - isuppli's Galaxy S4 estimated BoM. They're estimating $20 for the Snapdragon 600 (and $25 for Qualcomm's LTE wireless section, amusing eh?) And they actually touch that $30 SoC cost for the Exynos version. So there are definitely indications of the premium SoC offerings rising in price.

The intel pricing argument is illogical. They keep their fab margins (qcom generates a healthy gross margin AFTER paying TSMC the production (TSMC also keeps a margin), are going to be running bay trail on fabs that have been depreciated for 2 years (22nm) and will possibly be outputting higher dies per wafer than an s600.

Furthermore, Intel's main motivation is to keep its fab utilization high. They already have the best cost structure in the space, potentially smaller die (higher dies/wafer) and a core market that is continuing to erode (thus opening up more capacity). Do people really think at the end of the day they are going to price themselves out of such a big market?

 

[Nothingness]

Do people really think at the end of the day they are going to price themselves out of such a big market?

Yeah, can't wait to see investor reaction if a price war starts :biggrin:

 

[mavere]

Cost is going to be the biggest question -- we all know Intel loves their fat margins.

Taking the $500 iPad as a baseline and throwing on an extravagant $100 for the Z3770 and maybe another $150 for a decent memory/storage system would still give (IMO) a reasonable price for a high quality, efficient tablet with a full-fledged OS.

I really like the concept of a hybrid laptop, but most options so far have made too many compromises to be a good buy. The missing key is that the hybrid must be strictly excellent in one of its two forms. This new Atom, with its (presumably) good performance at low wattage, has a chance to do that for at least the tablet form.

 

[Khato]

Yeah, can't wait to see investor reaction if a price war starts :biggrin:

Indeed, the majority of the idiotic investor crowd who doesn't care about the long-term implications would go crazy. But I know I'd sure be buying Intel stock in that situation. Intel will win any price war that it enters - their only concern is navigating it correctly to not come out of it as a monopoly for all the regulators to hound. Keep in mind that it's not only the ARM IP companies (Qualcomm, mediatek, NVIDIA, so on so such) that would feel the effect of a price war, but also their foundries. Really I'd be more concerned about the effect it can potentially have on the foundries - even if they don't reduce their pricing per wafer to help out they'll still have to deal with selling however many less wafers that Intel has gained. And with how much the foundries have been expanding lately in order to meet current/future demand it'd be rather bad for them if that demand suddenly evaporates.

Basically, it's not so much about making money for Intel as it would be depriving the competition of sales.

 

[blackened23]

Taking the $500 iPad as a baseline and throwing on an extravagant $100 for the Z3770 and maybe another $150 for a decent memory/storage system would still give (IMO) a reasonable price for a high quality, efficient tablet with a full-fledged OS.

I really like the concept of a hybrid laptop, but most options so far have made too many compromises to be a good buy. The missing key is that the hybrid must be strictly excellent in one of its two forms. This new Atom, with its (presumably) good performance at low wattage, has a chance to do that for at least the tablet form.

Your post is full of speculation - Intel's new CEO stated that they will be price competitive. As well, Bay Trail is usable with both Android and Windows - I feel that most Bay Trail SKUs will be using Android since there is no os license fee associated with it. With windows, there is obviously a large cost due to the OS license.

Your statement with regard to 100$ more for the Z3770 has no basis in reality - this is aside from the fact that there will be a TON of Bay Trail SKUs for everything from 200$ tablets up to 600-700$ Windows 8 tablets. If you're able to find tray pricing on these SKUs, feel free to and provide your sources (intel has not released pricing yet) but Intel are not idiots. Their new CEO knows what it will take to compete and he knows that they can't price themselves out of that market. That's really all there is too it.

 

[Homeles]

You're doing nothing but speculating and exaggerating here. Intel's new CEO stated that they will be price competitive. As well, Bay Trail is usable with both Android and Windows - I feel that most Bay Trail SKUs will be using Android since there is no os license fee associated with it. With windows, there is obviously a large cost due to the OS license.

Your stating with regard to 100$ more for the Z3770 is nothing but hyperbole and has no basis in reality. If you have actual facts to share, feel free to and provide your sources (intel has not released pricing yet) but Intel are not idiots. Their new CEO knows what it will take to compete and he knows that they can't price themselves out of that market. That's really all there is too it, there is no need fo ryou to continue your exaggeration.

I think the takeaway from his speculation is that even if it were priced like that, it would still be a very interesting product.

I feel like you're looking at it the wrong way.

 

[blackened23]

I think the takeaway from his speculation is that even if it were priced like that, it would still be a very interesting product.

I feel like you're looking at it the wrong way.

Yeah, you're correct, my mistake.

I should add that Bay Trail will have multiple SKUs to address this issue: there will be (relatively) powerful Bay Trail SKUs that can run full fledged windows 8.1 -- and there will be lower end Bay Trail SKUs designed for the low cost crowd.

 

[SiliconWars]

Indeed, the majority of the idiotic investor crowd who doesn't care about the long-term implications would go crazy. But I know I'd sure be buying Intel stock in that situation. Intel will win any price war that it enters - their only concern is navigating it correctly to not come out of it as a monopoly for all the regulators to hound. Keep in mind that it's not only the ARM IP companies (Qualcomm, mediatek, NVIDIA, so on so such) that would feel the effect of a price war, but also their foundries. Really I'd be more concerned about the effect it can potentially have on the foundries - even if they don't reduce their pricing per wafer to help out they'll still have to deal with selling however many less wafers that Intel has gained. And with how much the foundries have been expanding lately in order to meet current/future demand it'd be rather bad for them if that demand suddenly evaporates.

Basically, it's not so much about making money for Intel as it would be depriving the competition of sales.

Mediatek might suffer but I have a feeling they aren't that bothered and can go lower than Intel is willing to go on price.

Nvidia is already having a disaster with Tegra but can sort of rely on their graphics business to keep them from having to leave the mobile game altogether. They probably have most to lose to Intel however, depending on how good or bad T4i is. Could be said they already lost everything to Qualcomm anyway.

Qualcomm makes the vast majority of their money from licensing. They could lose all of their chip business to Intel and they'd still be making a huge profit. Obviously before that happened they'd be having a price war instead - one which I believe Intel cannot win - and which is infact the entire basis of my fears for Intels long term future.

 

[Khato]

Qualcomm makes the vast majority of their money from licensing. They could lose all of their chip business to Intel and they'd still be making a huge profit. Obviously before that happened they'd be having a price war instead - one which I believe Intel cannot win - and which is infact the entire basis of my fears for Intels long term future.

I'm curious as to why you believe that Intel cannot win a price war?

 

[SiliconWars]

I'm curious as to why you believe that Intel cannot win a price war?

Qualcomm has more money (zero debt, double the bank balance currently, and rising) and less dependence on chip manufacturing than Intel does.

I covered it in another thread a few weeks ago, but basically put Intel is now spending most of their profit on capex, ie building new fabs. The cost of fab building and process R&D rises while their profits drop.

That's why Qualcomm is making so much money compared to Intel. If Intel took 50% of Qualcomm's chip business, Qualcomm would barely flinch. If the opposite happened Intel would basically be finished.