Bionic chip VS. Samsung chip

[BTRY B 529th FA BN]

Sorry I couldnt find the proper SoC names for each but how is the IPC comparison between the two SoC chips in these two smartphones?

I've seen the benches but the clock speeds aren't similar so I want a better understanding of the two.

Thanks

 

[runawayprisoner]

Which Samsung chip? The one in the Galaxy S2 or the one in the year-old Fascinate?

 

[BTRY B 529th FA BN]

The Galaxy. Is it referred to as the GSII or GS2 or something like that???

 

[Oyeve]

The Galaxy. Is it referred to as the GSII or GS2 or something like that???

Well, the first line of galaxy's were single core 1ghz chips so they shouldnt even be compared. I assume you mean the GS2 chips which is a 1.5 GHz dual core Qualcomm APQ8060 processor, at least on tmos version. the bionic has an OMAP 4 CPU (dunno the hertz) but was supposedly a tegra previously.

 

[runawayprisoner]

The Galaxy. Is it referred to as the GSII or GS2 or something like that???

GS2 uses Samsung Exynos on AT&T and Sprint variants, but Qualcomm Snapdragon M8 on T-Mobile, and Droid Bionic uses TI OMAP4... or at least that's how I remember it.

In terms of performance, the AT&T/Sprint variant would have the most GPU performance of the bunch. T-Mobile would have the most raw CPU performance based on clock speed, but it's about equal to AT&T or Sprint in reality based on extra features of Exynos.

TI OMAP4 in Droid Bionic is about equal to Exynos in terms of features, but it has a much slower GPU, and from what I have gathered, it runs at a lower clock speed as well (1Ghz vs 1.2GHz on GS2), so it would be a good step below Exynos in terms of performance.

Unfortunately, Verizon doesn't have a GS2 variant. And the Fascinate from last year is definitely inferior to Droid Bionic in terms of performance.

 

[poofyhairguy]

T-Mobile would have the most raw CPU performance based on clock speed

Maybe not. Qualcomm CPU designs all seem behind this round, since they are based off an older ARM design and they have less cache. Look here:

The Sensation with a Qualcomm CPU with the same clock speed as the Exynos gets killed in the pure CPU benchmark. Clock-for-clock that shows that the Qualcomm CPU cores are the worst out there- Tegra and OMAP phones 20% slower in clock speeds almost overtake the Sensation.

Therefore the 1.5Ghz T-Mobile SGS2 will have at best equal CPU performance thanks to the higher clock, worse battery life thanks to the higher clock, and a worse GPU. The T-Mobile version is the worst one.

 

[Oyeve]

Maybe not. Qualcomm CPU designs all seem behind this round, since they are based off an older ARM design and they have less cache. Look here:
The Sensation with a Qualcomm CPU with the same clock speed as the Exynos gets killed in the pure CPU benchmark. Clock-for-clock that shows that the Qualcomm CPU cores are the worst out there- Tegra and OMAP phones 20% slower in clock speeds almost overtake the Sensation.

Therefore the 1.5Ghz T-Mobile SGS2 will have at best equal CPU performance thanks to the higher clock, worse battery life thanks to the higher clock, and a worse GPU. The T-Mobile version is the worst one.

I wonder if we will be able to underclock the tmo if it indeed drains the battery. Thats the only reason I hate my vibrant, I have to charge at least once a day, sometimes twice.

 

[BTRY B 529th FA BN]

Is that chart benchmark comparing 1GHz vs 1GHz chips to get a better understanding of the iterations per cycle performance difference? That's what I'm curios about.

 

[poofyhairguy]

No, all chips are at their sold speeds. It is easy to see what the trend is though- clock for clock Qualcomms are the worst.

 

[runawayprisoner]

Maybe not. Qualcomm CPU designs all seem behind this round, since they are based off an older ARM design and they have less cache. Look here:

The Sensation with a Qualcomm CPU with the same clock speed as the Exynos gets killed in the pure CPU benchmark. Clock-for-clock that shows that the Qualcomm CPU cores are the worst out there- Tegra and OMAP phones 20% slower in clock speeds almost overtake the Sensation.

Therefore the 1.5Ghz T-Mobile SGS2 will have at best equal CPU performance thanks to the higher clock, worse battery life thanks to the higher clock, and a worse GPU. The T-Mobile version is the worst one.

But that benchmark graph was noted as having "something else" at play.

Where we do see big gains from the Exynos' higher clock speed is in our Linpack tests. The single-threaded benchmark actually shows more scaling than just clock speed, indicating that here are other (possibly software?) factors at play here. Either way it's clear that the 20% increase in clock speed can surface as tangible if the conditions are right:

Realistically, it should scale, but the Galaxy S2 boosted itself too far ahead of the pack that it seemed unrealistic. Something else was affecting the results.

Beside that, 1.2GHz to 1.5GHz is still a 25% jump in clock speed, so depending on the condition, that 1.5GHz may still compensate and somehow match or exceed Exynos at 1.2GHz. A 25% boost in performance on paper should bring the HTC Sensation results to the same level as Exynos based on those numbers.

But even then, I did note that it's only on clock speed. Real world results should pit the two on equal footing for the most part.

And I was mistaken. The chip used in the T-Mo S2 seems to be an APQ8 chip instead of the regular M8. Which means Samsung might be using a different modem instead of an in-built modem like on M8 chips. I'd think it has to do with battery life, but we'll see.