[french toast]

Originally Posted by Ailuros

A new LG device with an Adreno320 appeared in the GLBenchmark2.5 database:

http://www.glbenchmark.com/compare.j...&D3=LG%20F180K

What's weird is that the LG E971 & LG F180K results are quite similar, while the LG E970 has only one singled out test result which gives each device its according ranking. The E970 probably has some differences f.e. in terms of SoC bandwidth or whatever else, but it still sounds weird so far.

Does the new s4 pro quad core carry 2mb l2?...linpack single thread is off the charts on Anandtech optimus g performance preview...but quad-core scaling tops out at 2.5x....is that the L2cache at play there? Bandwidth? Or something else?

One thing is for sure though, LG have average software optimisations at best, no jellybean and a poor sunspider score on stock LG browser despite the obvious power on tap.
http://www.anandtech.com/show/6305/l...rmance-preview

[Ailuros]

Well 29fps is a LOT higher than the 17/18 fps the two LG devices get in the database in Egypt 2.5 1080p offscreen. The highest result from the LG E970 is at 24.5fps. I wonder when Anand's results will get published in the Kishonti database.

[french toast]

Well if that is the gpu performance we can expect from a phone with a 720p screen, then that is going to take some beating, I doubt even a6 with its various improvements besides 2x iPhone 4s gpu will even match it.

[Ailuros]

Originally Posted by french toast

Well if that is the gpu performance we can expect from a phone with a 720p screen, then that is going to take some beating, I doubt even a6 with its various improvements besides 2x iPhone 4s gpu will even match it.

I don't expect the iPhone5 to match those 29fps, however it still wil remain a highly competitive product on its own merits.

[Wishmaster]

Here we have results from Xiaomi Mi 2 running apq8064. It gets almost 29fps just like Optimus G on anandtech's benchmark results, and those scores are almost 20% higher than ipad3. That is quite impressive if you ask me.

[Ailuros]

Originally Posted by Wishmaster

Here we have results from Xiaomi Mi 2 running apq8064. It gets almost 29fps just like Optimus G on anandtech's benchmark results, and those scores are almost 20% higher than ipad3. That is quite impressive if you ask me.

Xiaomi Mi 2 is now at 30.5 fps meaning 22% ahead of the 25 fps of the iPad3. Impressive for Qualcomm itself definitely. However if you consider that both GPUs have most likely the same amount of ALU lanes and one is clocked at 400MHz while the other at 250MHz, if you normalize performance to frequencies it'll tell you an entire different story. Especially while comparing a 6 months old 45nm to a not yet shipping in devices 28nm SoC.

[french toast]

Originally Posted by Ailuros

Xiaomi Mi 2 is now at 30.5 fps meaning 22% ahead of the 25 fps of the iPad3. Impressive for Qualcomm itself definitely. However if you consider that both GPUs have most likely the same amount of ALU lanes and one is clocked at 400MHz while the other at 250MHz, if you normalize performance to frequencies it'll tell you an entire different story. Especially while comparing a 6 months old 45nm to a not yet shipping in devices 28nm SoC.

This is a major win for Qualcomm....this can't be underestimated.
If you think back to the iPhone 4s...or even every iPhone launch besides 4...apple has had the best mobile phone gpu by quite some margin....they have carried on being ultra aggressive on that front and DOUBLED the performance of last years power house...that's ridiculous in its self...but to think Qualcomm has beaten even the new a6 before its launched is bbq pretty amazing considering the resources and technical expertise apple has thrown into its SOC division.

It actually beats the new iPad...with its quad core/quad channel memory..in a phone...awesome.

Yes ailuros if you compare just alu lanes perhaps you are right...I'm not as familiar as you with regards to the other execution resources available for both gpus...but I would guess sgx mp4 has more of everything else...and also probably if process node being equal...takes up less die area.

Let's not forget part of gaining g an advantage is through process nodes...and Qualcomm manages it with a relatively high clocked quad core krait, 2mb ram IN A PHONE....apples new phone chip can't touch it outside of some memory benchmarks and a blazing sunspider score...good job Qualcomm!.

[Laurent06]

A company puts a chip that's been designed mainly for tablets in a phone, so I guess you shouldn't be surprised it beats the competition

[Ailuros]

Originally Posted by french toast

This is a major win for Qualcomm....this can't be underestimated.

Who's underestimating anything? It's a natural reality that the newer a product the better it usually ends up or at least should be. Qualcomm & Apple have different design cycles and just for the record's sake Qualcomm has its own customers which it's selling its SoCs too, while Apple has none. They're indirectly competing.

Shrink the A5X to 28nm hypothetically, clock its GPU at the same 400MHz as Adreno320 and do the math yourself how the picture would look like then.

If you think back to the iPhone 4s...or even every iPhone launch besides 4...apple has had the best mobile phone gpu by quite some margin....they have carried on being ultra aggressive on that front and DOUBLED the performance of last years power house...that's ridiculous in its self...but to think Qualcomm has beaten even the new a6 before its launched is bbq pretty amazing considering the resources and technical expertise apple has thrown into its SOC division.

I don't recall "every" iPhone to have the best GPU at the time apart from iPhone4S and even then didn't it take too long before its GPU performance had been matched or exceeded by competing solutions.

Yes Adreno320 doubles GPU performance over Adreno225 but so does A6 vs. iPhone4S A5.

It actually beats the new iPad...with its quad core/quad channel memory..in a phone...awesome.

Qualcomm has undeniably excellent execution, but considering the raw specs of the 320 it's delivering expectable performance compared to a 225 with the same frequency and probably half the unit counts. One would expect that a newer architecture would had come with some homework to increase efficiency by quite a bit or that the driver/compiler would had matured in the meantime.

Yes ailuros if you compare just alu lanes perhaps you are right...I'm not as familiar as you with regards to the other execution resources available for both gpus...but I would guess sgx mp4 has more of everything else...and also probably if process node being equal...takes up less die area.

Let's flip that coin into another direction; assuming I'm right and the 320 has "64SPs" as in desktop marketing parlance at 400MHz. I'm putting an upcoming "64SP"@400MHz Wayne ULP GeForce against it and you tell me on which you'll place your bets for which of the two would win with flying colours. And no the point here isn't neither to compare it to future products in theory nor to undermine the 320 from any perspective.

The real point here is efficiency and it shouldn't be too hard to understand and digest.

Let's not forget part of gaining g an advantage is through process nodes...and Qualcomm manages it with a relatively high clocked quad core krait, 2mb ram IN A PHONE....apples new phone chip can't touch it outside of some memory benchmarks and a blazing sunspider score...good job Qualcomm!.

Why are you so hung up on Apple's i-Gear anyway? I don't have even a single i-device nor any other high end smartphone at the moment, despite that I fool around with several of them from time to time. As I said Apple actually competes with itself and they have an insane profit margin for what they're doing.

But if it has to be a direct comparison from so far GLBenchmark2.5 results there are devices with Adreon320s that range from 2000+ up to 3400+ frames. My own guess would place the iPhone5 into the 2300-2400 frame ballpark which still would make the product highly competitive and still quite a bit above the 2.5 score of Samsung's SIII hotseller.

Will any of the above affect in its majority the public's buying decisions? I severely doubt so especially for iPhones since "i-fanatics" (if I may call them that) will buy the 5 for its own merits or if you prefer the total user experience.

[french toast]

Yes you could flip the coin and clock mp4 @ 400mhz @28nm...yes that would obviously win...but die area would be bigger...power consumption would likely be bigger and so on...the point is its about balance....Qualcomm has cut down redundancy instead of just bolting cores together...thus saving die space....and maybe power?...let's flip it another way....Qualcomm could just odd more units to fill up the die space that the A5x would take and clock lower?? Thus prevailing anyway.

None of that hypothetical mumbo jumbo makes any difference though, it's about actual products shipping...and I use I products socs because tests what we were talking about was it not?..and happen to have had the highest performing gpu parts have they not?...so to put a chip into a smartphone that bests both apple solutions....from tablet only 6 month old...and the new fangled a6...is quite an achievement compared to the gpu gap thst existed between adreno and the competition a couplenof yesrs ago. Thsts all I'm saying.

Also, it's now unlikely that a smartphone gpu will top the adreno 320 till at least MWC (galaxy s3)...which how many months away?...and what was my time scale k predicted for adreno to rule smartphknes? 4-6 months.

Tegra 3 plus has not even been announced yet..so you can forget about tegra 4 showing in a smartphone anytime soon.

[french toast]

Originally Posted by Laurent06

A company puts a chip that's been designed mainly for tablets in a phone, so I guess you shouldn't be surprised it beats the competition

If it doesn't suffer with batterylife issues then what's the problem? If it goes into a smartphone before s shipping tablet I guess it makes it a smartphone soc does it not?

[Laurent06]

Originally Posted by french toast

If it doesn't suffer with batterylife issues then what's the problem? If it goes into a smartphone before s shipping tablet I guess it makes it a smartphone soc does it not?

My understanding is that Qualcomm has presented APQ 8604 as a tablet chip. OTOH I agree, and forgot to say in my previous post that we should wait for battery life results. But performance always has a price (especially when using the same process).

[Ailuros]

Originally Posted by french toast

Yes you could flip the coin and clock mp4 @ 400mhz @28nm...yes that would obviously win...but die area would be bigger...power consumption would likely be bigger and so on...the point is its about balance....

More die area and power consumption compared to what exactly? Compared to A5X@45nm with a purely hypothetical shrink two nodes down and not one to 28nm would result into much less die area and at least comparable power consumption despite the higher frequency.

Samsung clocked in Exynos the Mali400MP4 from 266MHz@45nm up to 440MHz@32nm without affecting in any particular way power consumption from what I recall and that only with one node shrink.

Qualcomm has cut down redundancy instead of just bolting cores together...thus saving die space....and maybe power?...let's flip it another way....Qualcomm could just odd more units to fill up the die space that the A5x would take and clock lower?? Thus prevailing anyway.

I'm not aware of neither die area for Qualcomm's 28nm SoC nor its respective power consumption. If you are go ahead I'm all eyes.

None of that hypothetical mumbo jumbo makes any difference though, it's about actual products shipping...and I use I products socs because tests what we were talking about was it not?..and happen to have had the highest performing gpu parts have they not?...so to put a chip into a smartphone that bests both apple solutions....from tablet only 6 month old...and the new fangled a6...is quite an achievement compared to the gpu gap thst existed between adreno and the competition a couplenof yesrs ago. Thsts all I'm saying.

Comparing an older product with an unavailable yet product isn't hypothetical mumbo jumbo. And since one product is "just" 6 months old let's see how long that reign will last for Adreno320 exactly from actual products shipping with it until other solutions appear in the not so distant future. If after 6 months of 320 device availability something better should come along I won't change perspective in that regard. It's the natural course of technology that by the time A appears, within months B is around the corner for which in the majority of cases B is more attractive from many perspectives whether from the same technology source/OEM or from someone else.

Also, it's now unlikely that a smartphone gpu will top the adreno 320 till at least MWC (galaxy s3)...which how many months away?...and what was my time scale k predicted for adreno to rule smartphknes? 4-6 months.

Tegra 3 plus has not even been announced yet..so you can forget about tegra 4 showing in a smartphone anytime soon.

I asked a legitimate question about a next generation GPU and its respective efficiency and I'm also counting Adreno320 to next generation GPUs whether it sounds convenient or not. If the Wayne example doesn't help try ARM Mali T604@500MHz.

If it doesn't suffer with batterylife issues then what's the problem? If it goes into a smartphone before s shipping tablet I guess it makes it a smartphone soc does it not?

I don't think LG or anyone else would integrate a SoC into a smartphone that would empty it's battery in no time for anything 3D. It sounds completely absurd to me. We'll have to wait and see detailed measurings which shouldn't take too long.

In the GLBenchmark2.5 thread I've noted that there are starting battery life tests to get filled in for 2.5. I didn't search each and every device but I've hit so far on the following:

http://www.glbenchmark.com/compare.j...%20iPhone%204S

The SIII battery time here is outstanding but as I said in the other thread I can't figure out battery life goes down with reduced brightness to 50% when it should be exactly the other way around (unless I'm missing something). iPhone4S battery life is too low for my taste but that's just me.

[ToTTenTranz]

Some new features in Optimus G:
http://www.youtube.com/watch?feature...v=PeP2d6YuGlc#!

There's more innovation shown in this 6min clip than in the entire iphone 5 keynote, IMO.

Except for those new earphones "designed to push the sound directly into our ears". Woah that was awesomely innovative.

[RDGoodla]

Originally Posted by Ailuros

Xiaomi Mi 2 is now at 30.5 fps meaning 22% ahead of the 25 fps of the iPad3. Impressive for Qualcomm itself definitely. However if you consider that both GPUs have most likely the same amount of ALU lanes and one is clocked at 400MHz while the other at 250MHz, if you normalize performance to frequencies it'll tell you an entire different story. Especially while comparing a 6 months old 45nm to a not yet shipping in devices 28nm SoC.

http://www.gsmarena.com/
http://www.gsmarena.com/samsung_i930...view-761p5.php
http://www.gsmarena.com/lg_optimus_g-review-814p5.php

Another public result of LG Optimus G. In GLbenchmark 2.1 Egypt offscreen 720p test, Optimus G (adreno 320) only scores 113 (fps), while iPad 3 is 140 fps, Galaxy S III is 103, and Ones (with Adreno 225) is about 56 fps.

Since most mobile games only render 720p HD (or sub-HD) resolution, which result is more representative of real condition? The "2.5 Egypt HD 1080p offscreen" test or "2.1 Egypt 720p offscreen" test?

In 720p offscreen test Galaxy S3 leads Ones very much, while in 1080p offscreen test, Galaxy S3 has minor advantage of only 15%.

[Ailuros]

Synthetic benchmarks like GLBenchmark attempt in their own way to predict how future games might look like. No mobile ISV would be as insane to create an as demanding game today as GLBenchmark2.5 and have the majority of users/devices look at single digit framerates.

The real point here is that Adrenos in general do extremely well with highly complex shaders and start to wind back as complexity shrinks and that's probably due to their still shaky driver/compiler not allowing the GPUs to reach a higher potential the actual hw should actually have.

If you'd ask me as a user to measure two competing GPUs of any kind I would use the most tortering synthetic stress tests along with as many as possible real 3D games and definitely not the best case scenarios in those and from the entire crop of results I'd attempt to reach a conclusion. Each result will have its own merit; it just comes down to how well you're able to interpret them.

When a mobile device without vsync reaches in any sort of 3D application way above the vsync limit (typically 60Hz), then that performance overhead should be better invested f.e. in IQ improving features like multisampling and/or AF.

The unfortunate thing here is that the small form factor market and especially 3D for it is still too young. Games unfortunately don't have any benchmarking functions, as their results would be far more representative than a handful of synthetic benchmarks. Assume there would be a healthy collection of game benchmarks available would you look rather at those case examples where GPUs get an average framerate of way beyond 100fps or rather something that drives the tested GPUs to their edge with 30-20 or even less average framerates? Or better how would you suggest to measure and compare different GPUs in such cases?

[wco81]

Has the market for high-quality 3D games on Android and iOS grown?

Apple showcases 3D games every keynote and NVidia obviously evangelizes development for devices using their SOCs.

In PC gaming, there used to be one or two games which were used as benchmarks for all the video cards. Is there a counterpart for that in mobile games?

[Ailuros]

One or two games isn't nearly good enough, but for the time being it would be a pleasant change. I'd expect that Epic might do something in that direction for its Unreal mobile engine, unless they've already something relevant and I've missed it.

[french toast]

Originally Posted by Ailuros

One or two games isn't nearly good enough, but for the time being it would be a pleasant change. I'd expect that Epic might do something in that direction for its Unreal mobile engine, unless they've already something relevant and I've missed it.

I've not Beemer bothered to learn how o multi quote posts on my mobile as of yet so....

Right you correctly point out that the exynos 5250 would likely beat the adreno 320...fine, but like I said when is Samsung likely to stick that in a phone? Only a high end one, and galaxy note 2 has just been launched with out one, galaxy s3 is only a few months old...so good luck waiting for that to show up...earliest is MWC...which is just what I said.

I don't know die areas your correct, but when I mentioned redundancy (hope I've got the correct terminology for this remark ) I meant that Qualcomm has with adreno 320..not unlike IMG TECH.. with its rogue....formed what appears to be some sort of clusters with one TMU for each...where as the current gen SGX 543 MP4 is literally 4 gpus bolted together...this is not as efficient in die area compared to performance/power consumption at a guess...IMG TECH move to a new uarch in this area posts to that IMO.

So coming back to what I have said, if you shrink down A5x, to 28nm, and clock it @ 400mhz (do we even know adreno clocks??) To match adreno 320...whilst performance would be better for the sgx...I'm guessing that due to the added execution resources of the full 4x sgx gpu cores and lets not forget the quad channel memory controller.....the APQ snapdragon adreno 320 gpu would still be smaller and I'm guessing slightly more power efficient....tests not taking into consideration the 4 kraits and 2mb L2 cache.

Besides..none of that matters for the her and now...it's speculative..snapdragon s4 pro is launching in a smartphone soon, part of processor design/decision making/planning is obviously manufacturing process and judging benefits/yields supply...and this is where Qualcomm gambled correct..and reaped the performance benfits, good luck to them.

Like I said, the snapdragon s4 pro is showing up in phones around October...retail...all round it is a VERY good chip...the best Imo, and I don't see a SMARTPHONE with the backbone and efficiency to rival it for at least 4-6 months....coming from behind as Qualcomm was a while ago, that's some achievement as there is some big players about with big resources...6 months is long time .

[french toast]

Originally Posted by Laurent06

My understanding is that Qualcomm has presented APQ 8604 as a tablet chip. OTOH I agree, and forgot to say in my previous post that we should wait for battery life results. But performance always has a price (especially when using the same process).

Your correct of course, APQ denotes a tablet chip in Qualcomm speak, but I like to point out that if it launches in a smartphone before any tablets...then that makes it s smartphone SOC.

Of course like you say we await power consumption numbers...bit if it's s success..(and I fully expect it to be) then it's a massive win for Qualcomm.

[Ailuros]

I said at least twice that iPhone5 will be highly competitive; now that the first 2.5 results are out you may bother to check them. It's most likely a MP3@=/>300MHz, so that's not any sort of hypothesis anymore and quite a bit closer to a hypothetical MP4@400MHz.

The difference in Egypt 2.5 offscreen is less than 2 fps, while the iPhone5 has the highest performance per pixel right now.

[Lazy8s]

IPhone 5 is the fastest mobile device available, and its GPU is trading wins with upcoming Adreno 320 phones depending on the specific test in GLBench and other benchmarks. Ultimately, Adreno 320 devices and newer devices shouldn't have too much difficulty overtaking it by a good margin in the near future, but remaining even halfway competitive in the lesser year of their two year cycle is pretty incredible. A6X in the next iPad should lead the market by a margin approached only by the feats of the introductions of the 3GS and the iPad 2.

[french toast]

Originally Posted by Ailuros

I said at least twice that iPhone5 will be highly competitive; now that the first 2.5 results are out you may bother to check them. It's most likely a MP3@=/>300MHz, so that's not any sort of hypothesis anymore and quite a bit closer to a hypothetical MP4@400MHz.

The difference in Egypt 2.5 offscreen is less than 2 fps, while the iPhone5 has the highest performance per pixel right now.

We agree, the apple chip is awesome...but my point about adreno 320 being the fastest, most advanced smartphone gpu for 4-6 months still stands.

Sgx is on very mature drivers..and is still slightly behind, doesn't include halti support and also doesn't have access to 4 krait cores and 2mb of cache .

I don't fully believe that the 1.2ghz dual core apple chip would smack around the snapdragon s4 pro in a REAL world scenario...use same software and I expect apple's superior memory subsystem would mean it pulled out some wins, including likely consumption, but I would rather have s4 pro on my android device.

[Ailuros]

Originally Posted by french toast

We agree, the apple chip is awesome...but my point about adreno 320 being the fastest, most advanced smartphone gpu for 4-6 months still stands.

No one said that it isn't an advanced GPU, rather the contrary.

Sgx is on very mature drivers..and is still slightly behind, doesn't include halti support and also doesn't have access to 4 krait cores and 2mb of cache .

The point is GPU efficiency and not how each SoC manufacturer integrates N other units into a SoC. Adreno320 exactly as its predecessors need additional driver/compiler homework and no today's trend didn't start today. Adreno2xx GPUs were also always excellent with high complexity shaders where efficiency weakened quite a bit with less shader complexity.

iPhone5 gets in Egypt 2.1 offscreen 91.6 fps, while the Xiaomi M2 77.8 fps.

The MP3 in the A6 is clocked at 300MHz or slightly above with less ALUs than the 320 which is clocked at 400MHz. If you still can't comprehend efficiency per clock and per unit it's not my fault really.

I don't fully believe that the 1.2ghz dual core apple chip would smack around the snapdragon s4 pro in a REAL world scenario...use same software and I expect apple's superior memory subsystem would mean it pulled out some wins, including likely consumption, but I would rather have s4 pro on my android device.

Who cares? Qualcomm is stil the king of all Android SoC manufacturers when it comes to SoC smartphone design wins and market share and Apple serves its own market as before. It's rather Intel, Texas Instruments, Samsung etc. that should be seriously worried about Qualcomm's execution. They're the ones that are behind compared to Qualcomm, while Apple is give or take on schedule for its own roadmap.

[french toast]

Originally Posted by Ailuros

No one said that it isn't an advanced GPU, rather the contrary.

The point is GPU efficiency and not how each SoC manufacturer integrates N other units into a SoC. Adreno320 exactly as its predecessors need additional driver/compiler homework and no today's trend didn't start today. Adreno2xx GPUs were also always excellent with high complexity shaders where efficiency weakened quite a bit with less shader complexity.

iPhone5 gets in Egypt 2.1 offscreen 91.6 fps, while the Xiaomi M2 77.8 fps.

The MP3 in the A6 is clocked at 300MHz or slightly above with less ALUs than the 320 which is clocked at 400MHz. If you still can't comprehend efficiency per clock and per unit it's not my fault really.

Who cares? Qualcomm is stil the king of all Android SoC manufacturers when it comes to SoC smartphone design wins and market share and Apple serves its own market as before. It's rather Intel, Texas Instruments, Samsung etc. that should be seriously worried about Qualcomm's execution. They're the ones that are behind compared to Qualcomm, while Apple is give or take on schedule for its own roadmap.

Ha ha who said anything about efficiency per clock comprehension??.. do we even know the clock speed of adreno 320?? Do we know the execution units of that chip??...

The answer is from public information..we don't...so all we can go on is early benchmarks and guestimates...efficiency per clock has nothing to do with me buying s smartphone...performance does..which the adreno has the most of in more future based gaming scenarios.

Like I said 4-6 months