Intel Z3770 vs Snapdragon 800 in Kraken 1.1

[SiliconWars]

Sigh. Hate to pee on your Wheaties but I'm not fudging this, I expected the results to be pretty close and obviously this benchmark doesn't consider GPU. Obviously there are likely some optimization advantages on the Intel side given x86 in Windows vs ARM in RT but the benchmark was consistent with the subjective user experience on these two tablets, the intel was clearly faster at nearly everything. Not to say the 800 was slow, far from it, it was the fastest tablet I've used without a Core inside but the 3770 was shockingly fast for an atom tablet.

That'll be why the S800 has all the design wins and Intel has...?

 

[Arachnotronic]

Sigh. Hate to pee on your Wheaties but I'm not fudging this, I expected the results to be pretty close and obviously this benchmark doesn't consider GPU. Obviously there are likely some optimization advantages on the Intel side given x86 in Windows vs ARM in RT but the benchmark was consistent with the subjective user experience on these two tablets, the intel was clearly faster at nearly everything. Not to say the 800 was slow, far from it, it was the fastest tablet I've used without a Core inside but the 3770 was shockingly fast for an atom tablet.

Interesting. I expect the S800 to be faster on GPU when you're looking at full-tilt against full-tilt. I wonder what the comparison looks like on a perf/watt basis, though? BYT looks pretty lean on GPU from what we have seen (4EUs).

 

[Arachnotronic]

That'll be why the S800 has all the design wins and Intel has...?

Why don't you wait for the chip launch and you'll see?

 

[SiliconWars]

Why don't you wait for the chip launch and you'll see?

Design wins are mostly determined before launch. Qualcomm already mocked Intel over their lack of wins.

Just use your brains, if it was THIS good it would be the chip in all the cool new tablets launching next month. Guess what? Yep it's in NONE OF THEM.

 

[Abwx]

You know what? Forget it. I'll just let you continue to make these outrageous claims and pick up the pieces when reality hits.

One look at Kraken on IE paints a clear picture quite quickly -

http://uk.hardware.info/reviews/446...-tested-benchmarks-igpu-ie10---mozilla-kraken

We're dealing with one thread, period. So straight away we have best case for the z3770 (who would have believed that?) and its burst turbo.

If at base frequency then the IPC would be 5%
better than a bobcat and 10% below Jaguar,
if at 2.3GHZ then it would be slower than BC by
30% and 45% slower than Jaguar.

Edit : this one use chrome......

 

[VengenceIsMine]

You know what? Forget it. I'll just let you continue to make these outrageous claims and pick up the pieces when reality hits.

One look at Kraken on IE paints a clear picture quite quickly -

http://uk.hardware.info/reviews/446...-tested-benchmarks-igpu-ie10---mozilla-kraken

We're dealing with one thread, period. So straight away we have best case for the z3770 (who would have believed that?) and its burst turbo.

Right. Because there aren't any applications in the real world that run singled threaded.

Obviously there are significant limitations to Kraken as a benchmark, and they can distort results somewhat but this was intended as a data point given the restrictions I was under at the time, short time with the tablets, inability to install applications and available benchmarks cross Windows & RT. My takeway from my time with them was that if you are expecting the 800 to be faster than z3770 you are going to be badly surprised.

 

[VengenceIsMine]

Design wins are mostly determined before launch. Qualcomm already mocked Intel over their lack of wins.

Just use your brains, if it was THIS good it would be the chip in all the cool new tablets launching next month. Guess what? Yep it's in NONE OF THEM.

There are cost issues, design issues & platform support issues to be considered, all which favor Qualcomm in the short term so this isn't much of an argument.

 

[SiliconWars]

If at base frequency then the IPC would be 5%
better than a bobcat and 10% below Jaguar,
if at 2.3GHZ then it would be slower than BC by
30% and 45% slower than Jaguar.

That's in Chrome though, different browsers throwing up hugely different results.

The important thing is Kraken on IE is simply a single threaded benchmark, as shown in that review. It may also favour Intel's architecture on top of that looking at the results.

Maybe it's not quite an AnTuTu this time, but we're seeing the very best of this Atom, ie pure single-threaded performance on Windows. It only gets worse from here.

 

[Abwx]

That's in Chrome though, different browsers throwing up hugely different results.

The OP said that it s slower with Chrome , so theses
results are still relevant in that they are not optimals.

 

[SiliconWars]

Right. Because there aren't any applications in the real world that run singled threaded.

Obviously there are significant limitations to Kraken as a benchmark, and they can distort results somewhat but this was intended as a data point given the restrictions I was under at the time, short time with the tablets, inability to install applications and available benchmarks cross Windows & RT. My takeway from my time with them was that if you are expecting the 800 to be faster than z3770 you are going to be badly surprised.

Based on Windows single threaded performance in a benchmark that already favoured Intel, it's pretty clear I won't be surprised by performance in what is actually relevant.

Did you ever ask yourself why the tablets were set up that way in the first place btw?

 

[mikk]

Z2770 was turbo enabled up to 2.3 Ghz, base clock was 1.46 ghz.

Do you have a Geekbench result?

 

[ChronoReverse]

He said he wasn't able to run benchmarks other than Kraken.

But hey, turns out you can find scores on Geekbench's site: http://browser.primatelabs.com/geekbench2/compare/2259942/2272551

 

[VengenceIsMine]

Based on Windows single threaded performance in a benchmark that already favoured Intel, it's pretty clear I won't be surprised by performance in what is actually relevant.

Did you ever ask yourself why the tablets were set up that way in the first place btw?

They weren't set up that way, I was given casual access to them but I wasn't the target audience so I opted to minimize my foot print. There wasn't anything nefarious, I just didn't want to attract a lot of attention to the fact I was benchmarking NDA hardware that wasn't mine.

 

[Arachnotronic]

Design wins are mostly determined before launch. Qualcomm already mocked Intel over their lack of wins.

Just use your brains, if it was THIS good it would be the chip in all the cool new tablets launching next month. Guess what? Yep it's in NONE OF THEM.

Well, designs based on the chip won't be announced until the chip is launched. There will be plenty of designs from the likes of ASUS, Acer, Samsung, HP, Fujitsu, and many of the other Taiwanese OEMs/ODMs.

 

[VengenceIsMine]

The OP said that it s slower with Chrome , so theses
results are still relevant in that they are not optimals.

Actually Chrome almost always scores significantly better (lower) than IE in Kraken. IE usually wins in Sunspider on identical hardware.

 

[mikk]

IDF is coming, it makes sense for OEMs to announce their devices there.

 

[Arachnotronic]

IDF is coming, it makes sense for OEMs to announce their devices there.

No, that makes too much sense.

 

[bullzz]

@siliconwars
"Just use your brains, if it was THIS good it would be the chip in all the cool new tablets launching next month. Guess what? Yep it's in NONE OF THEM"

we can do without this amount of rage in the forums
if a sucky clovertrail had 5+ design wins, i dont think baytrail will be lower. either way windows oems dont have a choice. its either temash or bay trail

 

[Nothingness]

No, that makes too much sense.

:biggrin:

OTOH I'm surprised there are not many (consumer) OEM present at IDF according to this page: http://www.intel.com/content/www/us...13-san-francisco-sponsors-and-exhibitors.html

Are they in another place or are they just waiting for Intel to lift the embargo?

 

[mavere]

We're still talking an SoC meant for general purpose, non-professional mobile devices, yes?

If so, I don't get why some of you are trying to downplay high single threaded performance, especially for web browsing (in a terrible JS engine, no less!). This is particularly relevant for the common folk.

 

[SiliconWars]

@siliconwars
"Just use your brains, if it was THIS good it would be the chip in all the cool new tablets launching next month. Guess what? Yep it's in NONE OF THEM"

we can do without this amount of rage in the forums
if a sucky clovertrail had 5+ design wins, i dont think baytrail will be lower. either way windows oems dont have a choice. its either temash or bay trail

I'm not raging. Believe it or not I'm trying to help tone down the nonsense on both sides - I just fail at it because I get more irritated when the Intel guys do it for some reason

We all need to stop jumping on every random post/benchmark as "proof" either way, and start thinking before we post, and I include myself in that as well.

 

[SiliconWars]

We're still talking an SoC meant for general purpose, non-professional mobile devices, yes?

If so, I don't get why some of you are trying to spin high single threaded performance, especially for web browsing (in a terrible JS engine, no less!) is a particularly bad thing...

It's not a bad thing at all, it's a good thing. But we need to know exactly what we're seeing before making rash proclamations on overall performance.

 

[VengenceIsMine]

It's not a bad thing at all, it's a good thing. But we need to know exactly what we're seeing before making rash proclamations on overall performance.

With all due respect (and I say that meaning very little) you were the one primary culprit in making this thread acidic and banal. You repeatedly tried to assert or imply that I was either fibbing or biased without the slightest evidence other than you didn't like the data I was presenting. My approach was a balanced as circumstances allowed. I simply presented a data point and while you strenuously tried to denigrate it, it is a better & more valid data point than any other I am currently aware of comparing these 2 SOCs. Most that have been presented to date in terms of comparing these 2 SOC's are on differing platforms, browsers or are using what is obviously early engineering silicon. It is more valid because the code & platform are nearly identical (barring slight differences in optimization) and the silicon involved is either final or closer to it than has been previously reported. Obviously Kraken does favor single threaded performance but guess what, single threaded performance & single threaded javascript performance is still critical to user experience, ESPECIALLY in tablet usage. So in short, you thrashed and squealed and contributed very little other than mentioning that Kraken favors single threaded performance.

 

[VengenceIsMine]

@siliconwars
"Just use your brains, if it was THIS good it would be the chip in all the cool new tablets launching next month. Guess what? Yep it's in NONE OF THEM"

we can do without this amount of rage in the forums
if a sucky clovertrail had 5+ design wins, i dont think baytrail will be lower. either way windows oems dont have a choice. its either temash or bay trail

I don't know total #'s but it is going to be a lot more than 5. You will see it in 8" as well as the 10"+ form factors where it Clover Trail currently is.

 

[Nothingness]

Most that have been presented to date in terms of comparing these 2 SOC's are on differing platforms, browsers or are using what is obviously early engineering silicon. It is more valid because the code & platform are nearly identical (barring slight differences in optimization) and the silicon involved is either final or closer to it than has been previously reported.

I agree with your post except for what I quoted. IMHO the link ChronoReverse posted above is as valid a comparison as yours: http://browser.primatelabs.com/geekbench2/compare/2259942/2272551 (of course one should put FP scores apart given the known bugs that heavily affect results of Intel chips on sharpen and blur). The OS are the same.

Also you can't claim the code & platform are nearly identical. I guess MS has spent much more time tuning Windows and IE10 JIT for x86 than for ARM. Geekbench2 is not necessarily better due to not enabling optimization when compiling neither on ARM or x86 (Geekbench3 is much better).

Your data is certainly interesting, but I'll wait for other results to get a better picture