iPhone/Zune/iPod & More Prediction Thread

Well, talking about the 3GS's chipset only makes marginal sense when we consider that most of the video (not 3D) silicon in it *seems* to remain dormant at this point. In terms of what is currently exposed, Tegra is obviously massively superior for video. If the rumours are true, the 3GS is quite a bit stronger though (and unlike the WiiHD/GPGPU, this is actually believable and has some real evidence :p). Even if the 3GS was theoretically able to do 1080p HP Decode, it nearly certainly could not do 720p encode - so that's one thing NV wins on at least. Same for the camera, where the 3GS needs to offload much more to the camera module (still no integrated ISP).

As for 3D, Tegra's GPU has the same number of TMUs (2) as the SGX535, nearly certainly more raw ALU power, and a bit lower efficiency (TBDR vs IMR; MIMD vs SIMD shaders). In the end it comes down to drivers and clock speeds, but I have no idea what the 3GS is clocked at...
 
SGX535's frequency in the 3GS is unknown, but the highest frequency you can find on a Tegra 650 would be probably 130MHz but that's a netbook variant.

As for 3D, Tegra's GPU has the same number of TMUs (2) as the SGX535, nearly certainly more raw ALU power
Tegra APX variants also have 2 ALUs/TMUs as the 600/650 variants? If that should be true (which I severely doubt) then you might want to rethink the ALU power part (given that both would have a similar frequency).

, and a bit lower efficiency (TBDR vs IMR; MIMD vs SIMD shaders). In the end it comes down to drivers and clock speeds, but I have no idea what the 3GS is clocked at...
Yes to the TBDR advantage but not necessarily yet to any possible shader differences.

Seriously? Is it that bad? I was looking at the whole picture here, cpu+gpu+videoprocessing and so on, i know that the cortex a8 xyz(3gs type) crushes the Arm11 core when it comes to the cellphone types clocking at 550mhz. The Tegra Types are higher clocked so it seems.

Is there any information, benches, comparisons availible?

Benchmarks on smartphones? What you can do is take a Tegra powered smartphone and compare it in real time against an iPhone3GS and judge how responsive each device is, how good the IQ on video playback on either/or, battery life, browsing etc but it still will come down more to the underlying OS/general sw quality then the hw capabilities.

If the smartphone Tegra APX variants are trimmed down variants of the 600/650 netbook variants, then you can still obviously also compare games in real time, but not really benchmark them.
 
Tegra APX variants also have 2 ALUs/TMUs as the 600/650 variants? If that should be true (which I severely doubt)
It's obviously true: on 65nm it's the exact same chip and redundancy wouldn't be worth the trouble (even less so for what they'd presumably hope to be their highest-volume SKU).
 
Well, talking about the 3GS's chipset only makes marginal sense when we consider that most of the video (not 3D) silicon in it *seems* to remain dormant at this point. In terms of what is currently exposed, Tegra is obviously massively superior for video. If the rumours are true, the 3GS is quite a bit stronger though (and unlike the WiiHD/GPGPU, this is actually believable and has some real evidence :p). Even if the 3GS was theoretically able to do 1080p HP Decode, it nearly certainly could not do 720p encode - so that's one thing NV wins on at least. Same for the camera, where the 3GS needs to offload much more to the camera module (still no integrated ISP).
What video silicon is rumored to remain dormant in the 3GS? Do you mean a VXD because of the 1080p decode "hacks"?
 
Yeah, Apple licensed a VXD as well when they licensed the SGX.

Implementing them at the same time would then make sense.
 
It's obviously true: on 65nm it's the exact same chip and redundancy wouldn't be worth the trouble (even less so for what they'd presumably hope to be their highest-volume SKU).

Let me get this str8:

Tegra 650 = ARM11@750MHz, GPU@130MHz
Tegra 600 = ARM11@650MHz, GPU@120MHz
Tegra APX 2600 = ARM11@550MHz?, GPU@???
Tegra APX 2500 = ARM11@???, GPU@???

Someone would have to fill out the blanks here, since a netbook doesn't have a comparable power portofolio as a smartphone.

Given the maximum resolutions supported, how sure are that there isn't some severe difference in fillrates between variants? (and yes my first usual thought goes to unit amounts before frequencies which shows that my reasoning stinks...)
 
Let me get this str8:

Tegra 650 = ARM11@750MHz, GPU@130MHz
Tegra 600 = ARM11@650MHz, GPU@120MHz
Tegra APX 2600 = ARM11@550MHz?, GPU@???
Tegra APX 2500 = ARM11@???, GPU@???
Not really:

APX 2500: ARM11@600MHz, GPU@120MHz, AP15 chip (aka A2?)
APX 2600: ARM11@600MHz, GPU@120MHz, AP16 chip (aka A3? base/silicon layer changes though)
Tegra 600: ARM11@650MHz, GPU@120MHz(?), exposed LVDS/IDE/... interfaces, AP15 *or* AP16 chip (I'm not kidding)
Tegra 650: ARM11@750MHz, GPU@130MHz, exposed LVDS/IDE/... interfaces, AP15 *or* AP16 chip (once again, I'm not kidding; most will be AP16s though, and given the delays wouldn't be surprised if the AP15 designs got canned)

Someone would have to fill out the blanks here, since a netbook doesn't have a comparable power portofolio as a smartphone.
It's just a better bin/higher voltages resulting in higher clocks. AP15/AP16 chips are used for all current target segments of Tegra, from smartphones to smartbooks passing by IPTV and Auto (the last of which has absurdly long design cycles so Tegra1 ones won't come out until 2012+, heh)

Given the maximum resolutions supported, how sure are that there isn't some severe difference in fillrates between variants?
The resolutions are just a matter of exposing the LVDS interface, iirc (mobile phones use another interface which I can't remember the name of).

It's not entirely impossible (although I doubt it) that there is a "APX 2300" SKU with only one TMU/PS pipeline enabled that they'd only sell to large volume phone OEMs in order to get rid of defective parts and get design wins in cheaper phones but all of the publicly announced SKUs are definitely 2 TMUs.
 
Well I sure hope you are confident about the APX frequencies and it isn't another effective fillrate fiasco LOL (I can come with quite a few creative scenarios while looking at the 2500/2600 numbering scheme ;) ).
 
Well I sure hope you are confident about the APX frequencies and it isn't another effective fillrate fiasco LOL (I can come with quite a few creative scenarios while looking at the 2500/2600 numbering scheme ;) ).
Meh, I got the 600MPixel/s number from NV (I only later learned it was 120*2*2.5) *before* they announced any other SKU than the APX 2500, so stop stressing about it :p (I still don't know whether they can achieve that Z-Rate without MSAA though; wouldn't be surprised if they could, or at least 480MPixel/s)
 
Meh, I got the 600MPixel/s number from NV (I only later learned it was 120*2*2.5) *before* they announced any other SKU than the APX 2500, so stop stressing about it :p (I still don't know whether they can achieve that Z-Rate without MSAA though; wouldn't be surprised if they could, or at least 480MPixel/s)

I've been so many times wrong myself on so many accounts even considering IMG's stuff that I wouldn't know where to start the rather long list; hence my scepticism ;)
 
With the high costs associated with having separate designs manufactured, I'd assume nVidia would be using the same chip for multiple markets, so I'd guess, with the rather conservative specs of the Tegra(s) for netbooks, that the variant(s) for mobile phones would have very similar performance for graphics.

What I don't believe, however, is that an efficiency advantage for PowerVR is only modest. I think the more subtle architectural strengths, like getting better data locality from the tile buffer and tile based rendering and having better memory access patterns through contiguous writes to the frame buffer and more orderly processing, widens the advantage beyond the obvious savings from overdraw.
 
on engadget there was a news stating that tegra2 for next ds and zune will be "double the power" with two arm9 core, that must be dual cortex a9
the article was too generic and confused to extract any other usefull info..
 
on engadget there was a news stating that tegra2 for next ds and zune will be "double the power" with two arm9 core, that must be dual cortex a9
the article was too generic and confused to extract any other usefull info..

Even NV's own roadmap shows Tegra2 to be twice as fast as Tegra1, but that's most likely for the entire SoC and yes the more powerful dual-core CPU has been calculated in that one. As for the T2 GPU it's most likely more units and a higher frequency.
 
Meh, I got the 600MPixel/s number from NV (I only later learned it was 120*2*2.5) *before* they announced any other SKU than the APX 2500, so stop stressing about it :p (I still don't know whether they can achieve that Z-Rate without MSAA though; wouldn't be surprised if they could, or at least 480MPixel/s)

If you modify the application accordingly why not? :LOL:
 
Meh, I got the 600MPixel/s number from NV (I only later learned it was 120*2*2.5) *before* they announced any other SKU than the APX 2500, so stop stressing about it :p (I still don't know whether they can achieve that Z-Rate without MSAA though; wouldn't be surprised if they could, or at least 480MPixel/s)

Where does the *2.5 come from? Overdraw or scaling for the number of AA sample points?
 
"

the 8Gb one is the $199 gaming console

The 32Gb and 64Gb (16 has been dropped) uses the same processor etc as in the iphone 3GS

Are you saying that if I buy a 32gb Ipod touch, I get the same hardware as the Iphone 3GS? and if I buy the 8gb one, I get the older Iphone hardware? Is this true? I was gonna buy a Ipod touch to seriously chk out making stuff for the App store , as the Iphone 3GS is not available in my country as of now.

According to this info, buying a Ipod touch 32gb nets me the same hardware as a 3GS minus the camera?

Also, one query, Does the Ipod touch let you use the bluetooth to implement local MP gaming?

:)
 
Back
Top