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

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.

I didnt get you there, what costs are you referring to and how do you think they've kept costs down?

The fab business is profitable no doubt, but surely selling more of your own phones is much more profitable? I would have thought they would try to sell as much of their own hardware as possible, while continuing to make money from fabbing chips for apple.

or they even would have been better off with keeping the good old sgx540 and bumped frequency!
look at the OMAP 4 benches
I supposed they fell for ARM's marketting datas... unless there the mali400 benches only reflect poor drivers..
http://www.anandtech.com/show/4179/ti-omap4-and-lg-optimus-3d-tested

True, the SGX 540's performance is still impressive and it trades blows with Tegra 2. Note that the GPU in the OMAP 4 is clocked higher than Hummingbird. Also according to Rys, those performance numbers are with the old driver. Performance improves 20-30% with the new driver :oops:
 
I didnt get you there, what costs are you referring to and how do you think they've kept costs down?

The fab business is profitable no doubt, but surely selling more of your own phones is much more profitable? I would have thought they would try to sell as much of their own hardware as possible, while continuing to make money from fabbing chips for apple.

Rumors have it that Apple has quadrupled wafer orders at Samsung; if true it would mean nearly half of Samsung's manufacturing capacity. Orion is on damn short supply and that's probably the reason why Samsung ordered a healthy amount of Tegra2 SoCs to equip both the Galaxy 10.1 tab and the GalaxyS2 Superphone in parallel with the Orion based GalaxyS2 smart-phone.

A hypothetical SGX543MP2 for Samsung would have meant quite a bit more performance compared to Mali400MP4, but on the other hand that performance wouldn't had come for free either. More die area for the GPU block for sure.

True, the SGX 540's performance is still impressive and it trades blows with Tegra 2. Note that the GPU in the OMAP 4 is clocked higher than Hummingbird. Also according to Rys, those performance numbers are with the old driver. Performance improves 20-30% with the new driver :oops:

S5PC11x = SGX540@200MHz, OMAP4430 = SGX540@307MHz, OMAP4440 (+25% graphics performance) = SGX540@384MHz
 
Rumors have it that Apple has quadrupled wafer orders at Samsung; if true it would mean nearly half of Samsung's manufacturing capacity. Orion is on damn short supply and that's probably the reason why Samsung ordered a healthy amount of Tegra2 SoCs to equip both the Galaxy 10.1 tab and the GalaxyS2 Superphone in parallel with the Orion based GalaxyS2 smart-phone.

A hypothetical SGX543MP2 for Samsung would have meant quite a bit more performance compared to Mali400MP4, but on the other hand that performance wouldn't had come for free either. More die area for the GPU block for sure.

Yea those rumours are possibly true. With the increase in die size of A5 and expected higher sales those numbers are quite plausible. Now this is just my opinion but im sure Samsung would have prioritized Orion production over Apple as that would help them sell more of their own products. But the fact that they have ordered a lot of Tegra 2's also goes against my theory. Apart from the supply issue, maybe Orion was just not ready early enough for them to use it in the Galaxy Tab 10.1. After all the rumoured date for availability for the Galaxy S2 is end March/early April.

I dont have much knowledge about mobile GPU's so i have no idea about the die size differences. How much would that difference be? Only a few mm2 right?
 
Yea those rumours are possibly true. With the increase in die size of A5 and expected higher sales those numbers are quite plausible. Now this is just my opinion but im sure Samsung would have prioritized Orion production over Apple as that would help them sell more of their own products. But the fact that they have ordered a lot of Tegra 2's also goes against my theory. Apart from the supply issue, maybe Orion was just not ready early enough for them to use it in the Galaxy Tab 10.1. After all the rumoured date for availability for the Galaxy S2 is end March/early April.

I recall reading a newsblurb that Samsung has supposedly ordered 1.5Mio T2s, but I have no idea if that's true or not. If it should be true however T2 sounds like a gap filler until Orion can be produced in reasonable quantities.

I dont have much knowledge about mobile GPU's so i have no idea about the die size differences. How much would that difference be? Only a few mm2 right?

ARM claims for a single core Mali400 4.7mm2 under 65nm. http://www.arm.com/products/multimedia/mali-graphics-hardware/mali-400-mp.php A Mali400MP4 should be less than that amount times 4x since its scaling only fragment cores. With 4 fragment cores (MP4) it still has only one VS unit.

IMG has claimed that a single core SGX543 is at 8mm2 under 65nm. MP2 would be at 16mm2.
 
Isn't it weird that ARM made such a grave misjudgement by limiting so much the vertex shading performance in their top-end solution for 2010-2011?

I've seen dumber design decisions than that. Does Larabee work as an example?
 
Keep in mind that the VS unit is slightly overspecced; not only in relative performance terms, but also as a single PS unit cannot use/read as much data as the VS unit can output per cycle. Also remember that PS demands are increasing much more rapidly than VS in handhelds because screen resolutions are increasing (and that's before we even consider tablets!)

So I'm really not convinced that's a big deal. The Mali-400 MP1's performance is very competitive for its die size, so overspeccing on the VS there doesn't seem like it hurt them too badly. The Mali-400 MP4 benchmarks on Orion/Exynos are disappointing and imply subpar multicore scaling but I doubt that has much to do with being VS-limited. We'll see if it can be fixed or at least improved in drivers - if not, that might be yet another reason why Samsung isn't pushing Exynos as hard as some people were expecting.
 
A lot of the characteristics of the Mali design and even some of the benchmark sub-tests show potential individually, but I'm wondering if they didn't give perf/power enough of a priority over perf/area.

Plus, IMG really thought through the multi-core approach with their ability to also scale vertex work, keeping MP designs efficient despite unavoidable overhead/redundancy.
 
Keep in mind that the VS unit is slightly overspecced; not only in relative performance terms, but also as a single PS unit cannot use/read as much data as the VS unit can output per cycle. Also remember that PS demands are increasing much more rapidly than VS in handhelds because screen resolutions are increasing (and that's before we even consider tablets!)

The triangle throughput of it's VS unit may be slightly over specced for a single pipe, but it's shading throughput is marginal, particularly when they go to 4 rasteriser pipes.
 
Think it's 3D so it's actually 1080p @ 60Hz x 2, however kinda looks OGLES1.1 content.

http://blogs.arm.com/multimedia/425-mobile-game-graphics-going-high-end/

Excitingly enough, a few pioneering game developers have lately created some mobile content that tries to do something beyond the basic fixed function shading which we all grew sick of in the 90’s. Rage by Id pushes the bar of OpenGL ES 1.1 using Carmack’s black magic texture streaming tricks. Infinity Blade, based on the Unreal Engine, features some basic per pixel bump shading the low resolution versions for iPhone. In the higher resolutions however, the game seemingly is more or less surviving on the great artist work alone.

Dunno but I'd take the Rage tech-demo or Infinity Blade any day over that model that seems to suffer from constipation in that link above.
 
http://www.samsung.com/global/business/semiconductor/productInfo.do?fmly_id=844&partnum=Exynos 4210

Exynos 4210

General Description

Exynos 4210 is a system-on-a-chip (SoC) based on the 32-bit RISC processor for smartphones, tablet PCs, and Netbook markets. Exynos 4210 provides the best performance features such as dual core CPU, highest memory bandwidth, world's first native triple display, 1080p video decode and encode hardware, 3D graphics hardware, and high-speed interfaces such as SATA and USB.

Exynos 4210 uses the CortexA9 dual core, which is 25% DMIPS faster than the CortexA8 core. It provides 6.4GB/s memory bandwidth for heavy traffic operations such as 1080p video en/decoding, 3D graphics display, and native triple display. The application processor supports dynamic virtual address mapping. This feature will help the software engineers to fully utilize the memory resources with ease.

Exynos 4210 provides the best 3D graphics performance and native triple display. The native triple display, in particular, supports WSVGA resolution of two main LCD displays and 1080p HDTV display throughout HDMI, simultaneously. This is possible due to the capability of Exynos 4210 to support separate post processing pipelines.

Exynos 4210 lowers the Bill of Materials (BOM) by integrating the following IPs: world's first DDR3 interfaces that will prepare bit cross with DDR2; 8 channels of I2C for a variety of sensors; SATA2; the GPS baseband; and a variety of USB derivatives (USB Host 2.0, Device 2.0, and HSIC interfaces with PHY transceivers to be connected with 802.11n, Ethernet, HSPA+, and 4G LTE modem). The application processor also supports industry's first DDR based eMMC 4.4 interfaces to increase the file system's performance.

Exynos 4210 is available as FCMSP Package on Package (PoP), which has a 0.45mm ball pitch with LPDDR2 configuration. The MCP will depend upon the customer's requirement.

512535Exynos4210.jpg


Features

  • ARM CortexA9 dual core subsystem with 64-/128-bit SIMD NEON
    blank.gif
    - 32KB (Instruction)/32KB (Data) L1 Cache and 1MB L2 Cache
    blank.gif
    - 1.0GHz Core Frequency: Voltage 1.2V
  • 64-bit Multi-layered bus architecture
  • Internal ROM and RAM for secure booting, security, and general purposes
  • Memory Subsystem:
    blank.gif
    - SRAM/ROM/NOR/NAND Interface with x8 or x16 data bus
    blank.gif
    - OneNAND Interface with x16 data bus
    blank.gif
    - 2-ports 32-bit 800Mbps LPDDR2/DDR2/DDR3 Interfaces
  • 8-bit ITU 601/656 Camera Interface
  • Multi-format Video Hardware Codec: 1080p 30fps (capable of decoding and encoding MPEG-4/H.263/H.264) and 1080p 30fps (capable of decoding MPEG-2/VC1)
  • JPEG Hardware Codec
  • 3D and 2D graphics hardware, supporting OpenGL ES 1.1/2.0, and OpenVG 1.1
  • LCD single or dual display, supporting 24bpp RGB, MIPI
  • Native triple display, supporting WSVGA LCD dual display and 1080p HDMI, simultaneously
  • Composite TV-out and HDMI 1.3a interfaces
  • GPS baseband integration with GPS RF interface
  • 2-ports (4-lanes and 2-lanes) MIPI DSI and MIPI CSI interfaces
  • 1-channel AC-97, 2-channel PCM, and 3-channel 24-bit I2S audio interface, supporting 5.1 channel audio
  • 1-channel S/PDIF interface support for digital audio
  • 8-channel I2C interface support for PMIC, HDMI, and general-purpose multi-master
  • 3-channel high-speed SPI
  • 4-channel high-speed UART (up to 3Mbps data rate for Bluetooth 2.1 EDR and IrDA 1.0 SIR)
  • USB 2.0 Device 1-channel, supporting FS/HS (12Mbps/480Mbps) with on-chip PHY
  • USB 2.0 Host 1-channel, supporting LS/FS/HS (1.5Mbps/12Mbps/480Mbps) with on-chip PHY
  • USB HSIC 2-channel, supporting (480Mbps) with on-chip PHY
  • Asynchronous Direct Modem Interface with 16KB DPSRAM
  • 4-channel SD/MMC interface, supporting SD 2.0, HS-MMC 4.3, and 1ch HS-MMC 4.4 DDR 4-bit interface muxed with HS-MMC 4.3
  • SATA AHCI 1-channel, supporting SATA1 (1.5Gbps) and SATA2 (3.0Gbps) with on-chip PHY
  • 32-channel DMA Controller
  • 14x8 keypad support
  • 10-channel 12-bit multiplexed ADCs
  • Configurable GPIOs
  • Real time clock, PLLs, timer with PWM, and watchdog timer
 
I recall reading a newsblurb that Samsung has supposedly ordered 1.5Mio T2s, but I have no idea if that's true or not. If it should be true however T2 sounds like a gap filler until Orion can be produced in reasonable quantities.

So their internal design isn't ready and they're sourcing from an external vendor? The new Galaxy tabs aren't due until June right?

Would they have some other tablet product using their own SOC later this year?

Isn't T3 tablets suppose to be out later this year with quad-core CPU? If that's the case, Samsung may have to use that as well, unless their SOC can at least match the T3.


Seems like every major tablet except iPad is using T2, at least for the first half to 8 months of 2011.
 
Tegra 2 is in all of the Honeycomb designs; since it's the only chip Honeycomb is optimized for currently.

Others will come but they'll come later than T2 tablets.
 
Back
Top