Samsung Orion SoC - dual-core A9 + "5 times the 3D graphics performance"

tangey said:
2) The orion narrative does not mention a poly rate, however they do mention a fill rate of 3200M pix/s. My understanding is that Mail-400MP4 does 1 pix per clock per core, giving 1600M pix/s @ 400Mhz. So it looks like samsung are using an equivalent x2 to allow for their culling of otherwise drawn pixels. If true, this is ironic, as ARM posted an article before Christmas lambasting the notion of "virtual" pixels, which I took to be a very thinly veiled critisim of IMG's x2-x2.5 overdraw allowance.

http://blogs.arm.com/multimedia/353-of-philosophy-and-when-is-a-pixel-not-a-pixel/
Click to expand...

Veiled critisism? That veil is so thin it would clothe the most discerning emperor. I fail to see it at all.

Some GPU manufacturers claim a “typical overdraw factor” and use this to multiply the output pixel rate to make some sort of “effective pixel rate”. The philosophy being that the fate of the pixel is predetermined and therefore its existence only needs to be metaphysical in order to make a contribution to the final scene. Of course the problem with a metaphysical pixel is you can’t see it and you can’t prove it exists, so how did it make a contribution? To all intents and purposes, it could just be the sum total of someone’s imagination.
Click to expand...

:)
 
Entropy said:
Veiled critisism? That veil is so thin it would clothe the most discerning emperor. I fail to see it at all.
:)
Click to expand...

I did say VERY THINLY veiled :), the veil was purely NOT saying the phrase, "imagination technologies".

The point is, given ARM's extremely robust dismissal of overdraw as "imaginary" pixels, it seems unlikely that the first implementation of MALI-400MP4 would make use of it in marketing (albeit its not arm's marketing).
 
Polygon rates Hummingbird - Orion

tangey said:
I did say VERY THINLY veiled :), the veil was purely NOT saying the phrase, "imagination technologies".

The point is, given ARM's extremely robust dismissal of overdraw as "imaginary" pixels, it seems unlikely that the first implementation of MALI-400MP4 would make use of it in marketing (albeit its not arm's marketing).
Click to expand...

The real triangles in the GLBenchmark 2.0 feature tests show there is some ground that needs to be covered by Mali just to equal the current Hummigbird polygon performance.

Perhaps this is why Samsung doesn't mention a polygon rate in the Orion PDF?

Mali-400 MP <=> SGX540
Ramos W10 <=> Galaxy S
Triangle test: textured: 4885 <=> 7113 (1.5x Mali) kTriangles/s
Triangle test: textured, fragment lit: 4045 <=> 4768 (1.2x Mali) kTriangles/s
Triangle test: textured, vertex lit: 4458 <=> 5774 (1.3x Mali) kTriangles/s
Triangle test: white: 5426 <=> 8686 (1.6x Mali) kTriangles/s
 
Ramos W10 appears to use AML8726-M, announced here:

http://www.prnewswire.com/news-rele...d-multimedia-and-3d-experience-107271438.html

Nowhere has Amlogic indicated that this SoC has a multi-core Mali-400. In fact, given the omission I'd lean more towards it being single-core, which also goes with the single-core Cortex-A9.

If it is indeed single-core then I'd say that this paints a very good picture of what a Mali-400 MP4 could be capable of, although we don't know clock speed either.
 
rbaker said:
The real triangles in the GLBenchmark 2.0 feature tests show there is some ground that needs to be covered by Mali just to equal the current Hummigbird polygon performance.

Perhaps this is why Samsung doesn't mention a polygon rate in the Orion PDF?

Mali-400 MP <=> SGX540
Ramos W10 <=> Galaxy S
Triangle test: textured: 4885 <=> 7113 (1.5x Mali) kTriangles/s
Triangle test: textured, fragment lit: 4045 <=> 4768 (1.2x Mali) kTriangles/s
Triangle test: textured, vertex lit: 4458 <=> 5774 (1.3x Mali) kTriangles/s
Triangle test: white: 5426 <=> 8686 (1.6x Mali) kTriangles/s
Click to expand...

Frequency is unknown as amount of cores for the Mali400. Since those aren't pure triangle tests (it says textured doesn't it?) it'll come down how many fragment cores and in extension how many TMUs that Mali400 has. Frequency plays also a role. Can't be sure but sounds like a single Mali400 and not more.
 
Ailuros said:
Frequency is unknown as amount of cores for the Mali400. Since those aren't pure triangle tests (it says textured doesn't it?) it'll come down how many fragment cores and in extension how many TMUs that Mali400 has. Frequency plays also a role. Can't be sure but sounds like a single Mali400 and not more.
Click to expand...

My reading of ARM's literature on MP versus 4MP is the number of vertex shaders is the same in both (4MP has 4X the pixel shaders of MP). Assuming this is the case it would appear to me Samsung would need ~1.5X the frequency of the Ramos just to put Orion on par with Hummingbird vertex performance.
 
rbaker said:
My reading of ARM's literature on MP versus 4MP is the number of vertex shaders is the same in both (4MP has 4X the pixel shaders of MP). Assuming this is the case it would appear to me Samsung would need ~1.5X the frequency of the Ramos just to put Orion on par with Hummingbird vertex performance.
Click to expand...

Your correct; Mali400 MP (as it's not a USC) is scaling only fragment processors, while there's always just one vertex shader unit.

While I don't know what the Mali400 is clocked at in the Ramos, Orion's Mali400 MP4 will be clocked according to rumors at 400MHz. Then again by the time Orion ships Hummingbird will be relatively outdated. TI OMAP4430 with the same SGX540 reports a max clock of 307MHz in it's manual and the OMAP4440 is supposed to have by 25% higher graphics performance (up to 380MHz).

What will be interesting to see is how the Orion Mali400MP will fair against a SGX543 MP2.
 
Ailuros said:
Your correct; Mali400 MP (as it's not a USC) is scaling only fragment processors, while there's always just one vertex shader unit.

While I don't know what the Mali400 is clocked at in the Ramos, Orion's Mali400 MP4 will be clocked according to rumors at 400MHz. Then again by the time Orion ships Hummingbird will be relatively outdated. TI OMAP4430 with the same SGX540 reports a max clock of 307MHz in it's manual and the OMAP4440 is supposed to have by 25% higher graphics performance (up to 380MHz).

What will be interesting to see is how the Orion Mali400MP will fair against a SGX543 MP2.
Click to expand...

Will be interesting indeed. Tegra-3 and Armada-628 should be plying the waters in this time frame with upgrades to the GPUs already leading the pack on the GLBenchmark 2.0. Freescale's iMX6 is another wildcard due to come out soon, perhaps also with an over clocked SGX543-MP2?
 
rbaker said:
My reading of ARM's literature on MP versus 4MP is the number of vertex shaders is the same in both (4MP has 4X the pixel shaders of MP). Assuming this is the case it would appear to me Samsung would need ~1.5X the frequency of the Ramos just to put Orion on par with Hummingbird vertex performance.
Click to expand...

It's true that vertex shading performance doesn't scale, but that doesn't mean that the tests you posted are vertex shading limited. Then again, I don't know how else the load balancing works, like with regards to triangle setup.
 
CPU stuff is no surprise, GPU stuff is perhaps the usual case of drivers being not that great yet. Or so one can hope.
 
Still no news regarding official release of the Galaxy S2 or newer benchmarks of Mali-400 MP4. Samsung might have been better off going with SGX 543 MP2 instead
 
Erinyes said:
Still no news regarding official release of the Galaxy S2 or newer benchmarks of Mali-400 MP4. Samsung might have been better off going with SGX 543 MP2 instead
Click to expand...

Well if Samsung has actually tried to keep costs as much down in order to serve Apple's manufacturing needs (which is my very own assumption), then they're entire strategy for this year makes sense IMHO.
 
You must log in or register to reply here.

Similar threads

I
Samsung Exynos 9810
2
Replies
25
Views
14K
Nebuchadnezzar
N
P
Portable console for all PC games - PGS
2
Replies
31
Views
13K
Cyan
Cyan
P
ARM announces Cortex-A50 series
2
Replies
22
Views
16K
silent_guy
S
O
Inertial Sensors: DualShock 4 vs Sixaxis/DS3 accelerometer/gyroscope chips
Replies
0
Views
10K
onQ
O
D
"Obscure" 800MHz dual-core SoC in Samsung Galaxy Ace 2?
Replies
9
Views
5K
Exophase
E
Back
Top