| Monday, June 15 2009 |
NVIDIA today announced its latest line of mobile graphics processors (mGPUs) based on the new 40 nm manufacturing technology. The GPUs also come as a surprise, as they embrace two technologies very few expected them to, this early/late. The GPUs support the DirectX 10.1 API, and some feature the new GDDR5 memory interface that doubles effective bandwidth. Enter GeForce GTS 260M, GTS 250M, GT 240M, GT 230M, and G 210M.
The G 210M succeeds the GeForce 9400, at least as far as the manufacturing technology and clock speeds go. It holds 16 shader processors, a 64-bit GDDR3 memory interface, and 512 MB of memory. The GT 230/240M hold 48 shaders, 1 GB of 128-bit GDDR3 memory, and clock-speeds that make up each variant. The GTS 260/250M are especially new, as they feature a 128-bit GDDR5 memory interface. 96 shaders and 1 GB of memory make the rest of their specs., while their clock speeds make for the variants. The table below spreads out the specifications.
The peculiar thing about these GPUs is their timing of entry into the market. This is the fag-end of the DirectX 10(.1) generation of GPUs, with no more high-end GPUs in sight. The company is facing stiff competition from AMD, and needs to cut manufacturing costs, while also decorating specs sheets. The use of GDDR5 allows manufacturers to narrow memory bus width, and reduce the number of memory chips, in turn board footprint, while maintaining the same levels of performance as GDDR3 with higher bus width. The use of 40 nm manufacturing process is certainly a step in the right direction, although we wonder how the company is able to get large yields of these chips, considering AMD isn't able to keep up with demand for its 40 nm GPUs, and is developing alternative SKUs at the same price points.
These GPUs will only be available to the mobile GPU board segment to begin with. There is no indication as to when the company makes desktop graphics cards based on these. AMD launched the RV740 in a similar fashion, with a mGPU board being released weeks ahead of the desktop board (Radeon HD 4770).
The G 210M succeeds the GeForce 9400, at least as far as the manufacturing technology and clock speeds go. It holds 16 shader processors, a 64-bit GDDR3 memory interface, and 512 MB of memory. The GT 230/240M hold 48 shaders, 1 GB of 128-bit GDDR3 memory, and clock-speeds that make up each variant. The GTS 260/250M are especially new, as they feature a 128-bit GDDR5 memory interface. 96 shaders and 1 GB of memory make the rest of their specs., while their clock speeds make for the variants. The table below spreads out the specifications.
The peculiar thing about these GPUs is their timing of entry into the market. This is the fag-end of the DirectX 10(.1) generation of GPUs, with no more high-end GPUs in sight. The company is facing stiff competition from AMD, and needs to cut manufacturing costs, while also decorating specs sheets. The use of GDDR5 allows manufacturers to narrow memory bus width, and reduce the number of memory chips, in turn board footprint, while maintaining the same levels of performance as GDDR3 with higher bus width. The use of 40 nm manufacturing process is certainly a step in the right direction, although we wonder how the company is able to get large yields of these chips, considering AMD isn't able to keep up with demand for its 40 nm GPUs, and is developing alternative SKUs at the same price points.
| GeForce G 210M | GeForce GT 230M | GeForce GT 240M | GeForce GTS 250M | GeForce GTS 260M | |
| Shader Units | 16 | 48 | 48 | 96 | 96 |
| Memory Size | 512 MB | 1024 MB | 1024 MB | 1024 MB | 1024 MB |
| Memory Bus Width/Type | 64 bit GDDR3 | 128 bit GDDR3 | 128 bit GDDR3 | 128 bit GDDR5 | 128 bit GDDR5 |
| Core/Shader/Memory Clock | 625/1500/800 MHz | 500/1100/800 MHz | 550/1210/800 MHz | 500/1250/800 (3.2 GT/s) MHz | 550/1375/900 (3.6 GT/s) MHz |
User comments
Terrible timing to enter the market when DX10.1 is a standard which will be soon superseded. Oh well at least they rebadge them :rolleyes:
by: TheLaughingManThe GPU does the calculations, and native PPUS are just extremely weak GPUs, but without output for display.
lets clear this up. The ATI system is completely different in architecture and implementation than the Nvidia system. This is comparing apples and oranges again. ATI has always had an unusually large number of processor cores compared to Nvidia. In this case more does not equal more power. There is a lot of other stuff going on the chip level than processor core count.
Good example is a ATI GPU core does physics better than the Nvidia core of the same class; BUT the Nvidia card will have an Aegis physics chip that it offloads the physics to which gives it a huge boost in physics. It definitely has a huge number of PPU's that don't get counted when shader specs are listed.
Hey hey im a ati fanboi as much as the next ati fanboi, but when it comes to laptops, its whoever gets to the 40nm medium sized, fairly slim laptop first thats the fasted.... is what il think about getting.
by: ShogoXTlower fab process doesnt always mean lower temps & power usage.
Hey hey im a ati fanboi as much as the next ati fanboi, but when it comes to laptops, its whoever gets to the 40nm medium sized, fairly slim laptop first thats the fasted.... is what il think about getting.
But it does mean they can eventualy stuff more in there :rockout:
by: EastCoasthandleThank you.
Source
It's always great fun reading this douchebag bs and failed attempts to look smart and be funny when we all know just how much of a douchebag he is.
Gotta love Charlie the douchebag, he is so stupid and pathetic that he's kinda funny in a way. :nutkick:
:roll:
Am I the only one bothered by the fact that there is no 10.1 for desktop users? WTF, there's way more desktop gamers than there are laptop gamers... :/
They're all mobile chips. :|
I suppose that means don't expect a mobile NVIDIA DX11 chip any time soon. They wouldn't be releasing DX10.1 mobile processors if DX11 was almost ready.
I suppose that means don't expect a mobile NVIDIA DX11 chip any time soon. They wouldn't be releasing DX10.1 mobile processors if DX11 was almost ready.
by: TheLaughingManPeople who correct me often amuse me.
lets clear this up. The ATI system is completely different in architecture and implementation than the Nvidia system. This is comparing apples and oranges again. ATI has always had an unusually large number of processor cores compared to Nvidia. In this case more does not equal more power. There is a lot of other stuff going on the chip level than processor core count.
Good example is a ATI GPU core does physics better than the Nvidia core of the same class; BUT the Nvidia card will have an Aegis physics chip that it offloads the physics to which gives it a huge boost in physics. It definitely has a huge number of PPU's that don't get counted when shader specs are listed.
The reason that 240 beat 800 is quite simple. We're talking about a Superscalar architecture. Every 5 ALU's only have a single functioning 'brain'. Each of the 240 are SP's and have their own 'brain', as the architecture is Scalar, which is the next level upwards, just as Superscalar is to Vec4 (Vector - 4 ALU/'brain').
ATI doesn't do physics at all, actually. But, if they did, the performance would only be relative in the exact same way as real-world performance. Or, if nothing else, nVidia would have the upper hand. As I've said from the very beginning of G80 - Scalar is better and Superscalar is out-dated. It's not that ATI has 320 and nVidia has 128, or 800 and 240 and whatever comes next. It's that 1600 fail to beat 480.
by: PP MguireWhich is exactly my point - why don't they?
But it does mean they can eventualy stuff more in there :rockout:
I remember back in 2006. They had a fully-functional G71M that was identical in almost every way to G71 - the desktop variant and greatest GPU on earth, until R580, of course. R580 still is my favorite GPU.
at last nvidia do what ATI do before a year
Just in time....
I suppose they are moving to DX11 and just decided to show that.
I suppose they are moving to DX11 and just decided to show that.
by: WeerInteresting, cause I don't remember correcting anyone. I merely pointed out that the design of the chips is different. And that Nvidia, historically speaking has never used as many shaders as ATI.
People who correct me often amuse me.
by: tkpenaltyPPU's (Physics processing unit) are just calculation based number crunching units. They don't generate images which is why it will never have a display. It basically calculates random numbers faster and more higher precision than GPU's and CPU's. That is all it does. It is nothing like a GPU core.
Terrible timing to enter the market when DX10.1 is a standard which will be soon superseded. Oh well at least they rebadge them :rolleyes:
The GPU does the calculations, and native PPUS are just extremely weak GPUs, but without output for display.
Recently, there has been a move to create GPGPU (General-purpose computing on graphics processing units) that can been used for both as needed.
by: TheLaughingManNo they don't. Can't find it on the PCB. It's done at the driver level. Just like how pre physx 8800GT's etc can do physx.
Nvidia has had onboard PhysX chip since they bought out Aegis. Every Nvidia board from the 8 series and up will have one.
http://www.nvidia.com/object/physx_new.html
Where is this PhysX chip at exactly? Oh, that's right, there isn't one.
by: ShadowFoldUgh red pcb :shadedshu
http://www.techpowerup.com/reviews/Palit/GeForce_GTX_260_Sonic_216_SP/images/front.jpg
Where is this PhysX chip at exactly? Oh, that's right, there isn't one.
by: ShadowFold
http://www.techpowerup.com/reviews/Palit/GeForce_GTX_260_Sonic_216_SP/images/front.jpg
Where is this PhysX chip at exactly? Oh, that's right, there isn't one.
Ok, ok. I yeild on that note. Bad example on my part. There is no dedicated chip like I thought.
What is the green chip on the card for? I didn't see that on any ATI board I have seen.
by: WeerMy XFX 7950GT was the bomb digity.
Which is exactly my point - why don't they?
I remember back in 2006. They had a fully-functional G71M that was identical in almost every way to G71 - the desktop variant and greatest GPU on earth, until R580, of course. R580 still is my favorite GPU.
by: PP MguireSalvay, brotha.
My XFX 7950GT was the bomb digity.
7900 GT with 512MB of RAM. That, and the 7900 GTO - same as GTX but with lesser RAM.
Both.. honorable choices. Wish I had one.
by: ShadowFoldShadow, I'm going to eat your Xbox 360 camera! :D
http://www.techpowerup.com/reviews/Palit/GeForce_GTX_260_Sonic_216_SP/images/front.jpg
Where is this PhysX chip at exactly? Oh, that's right, there isn't one.
The chip exists. Doubt your suspicions. Everything can be proven.
And if not, my 8400 GS can run it. It's healthy enough.
by: TheLaughingManYou seemed to take a superior tone and explain to me the reason why ATI's 800 can't beat nVidia's 240. That is, without actually knowing why. I understand you're new. No hard feelings towards those that learn.
Interesting, cause I don't remember correcting anyone. I merely pointed out that the design of the chips is different. And that Nvidia, historically speaking has never used as many shaders as ATI.
by: TheLaughingManIts a NVIO chip I believe used for outputting sound through HDMI or something of that sort.
Ok, ok. I yeild on that note. Bad example on my part. There is no dedicated chip like I thought.
What is the green chip on the card for? I didn't see that on any ATI board I have seen.
I got a 7900GT right now. It is still awesome for anything pre-Crysis. If only it had 512MB ram...
by: hatBeing the self admitted resident cheapskate of tpu how much would you pay for a gpu maximum ? (Trying not to sound offensive)
I got a 7900GT right now. It is still awesome for anything pre-Crysis. If only it had 512MB ram...
by: DrPepper
Its a NVIO chip I believe used for outputting sound through HDMI or something of that sort.
That's a pretty big chip (not to mention hot) for handling a menial task. :eek: