NVIDIA is filling in what it presumes to be holes in its next-generation GPU lineup, adding the 40nm G210M, GT 230M, GT 240M and GTS 250M, with GDDR3 memory ranging from 512MB to 1GB, to its existing GTX 280M, GTX 260M and GTS 160M laptop graphics cards. Apparently the new cards sport "double the performance" and "half the power consumption" over the last generation of discrete GPUs they're replacing. The cards are SLI, HybridPower, CUDA, Windows 7 and DirectX 10.1 compatible, and all support PhysX other than the low-end G210M. Of course, with integrated graphics like the 9400M starting to obviate discrete graphics in the mid range -- even including Apple's
latest low-end 15-inch MacBook Pro -- we're not sure what we'll do with eight different GPU options, but we suppose NVIDIA's yet-to-be-announced price sheet for these cards will make it all clear in time.
posted Jun 15th 2009 10:22AM
All I have to say is kudos!
They're not really new ... they're just higher clocked, die shrinks of 3 or 4 generation old chips - it's the same with all of NVIDIA's laptop range.
i came.
old meme is old.
Let's hope their driver support doesn't suck like the French this time around.
These would have been damn nice with those new Macbook Pros. Doh!
Is this a joke? 1GB of memory on a 128bit card? They can put GDDR10 on it and it's still going to be useless...
Wrong.
The GTS250M (128-bit, 1GB GDDR5) has the same performance as the old 8800M GTX. It's basically the 8800M GTX in 40nm - same clockspeeds (500/1250/1600), architecture and number of shader cores (96). The only difference is the 128-bit interface, which is compensated for by the GDDR5 memory.
It has a 28W TDP, compared to 65W of the 8800M GTX.
Oranges and Apples man, 2007 vs 2009 tech, different cores, and even with the GDDR5 it wont beat the 8800m GTX, so FAIL!
I owned a 8600m GT the worst card i ever seen... I wont touch 128 bit ever again! Loads of memory are useless, unless you have enough bus width to use it...
Wrong again. They are not different cores, the architecture is the same, both are G92-based.
GDDR5 has approx twice the memory bandwidth of GDDR3, so it pretty much exactly compensates for the 128-bit memory bus difference. This (GTS 250M) is effecitvely a 40nm 8800M GTX (or 9800M GT), no more, no less.
GTS260M is a little faster still.
Sure... Lets use medicine a dropper to fuel a ferrari too, oh wait, nvidia only sells videocards... There is no excuse to make a 1GB card with 128-bit interface, period!
Memory bandwidth, the value which represents how fast data can be transferred between the processor and the memory, is calculated by:
Memory Bandwidth = Bus Width * Memory Clock * Data Rate
GDDR3 has Double Data Rate (2) and GDDR5 has Quad Data Rate (4) so therefore:
8800M GTX / 9800M GT = 256 (bus width) * 800 (base memory clock) * 2 (data rate) = 51.2 GB/s (memory bandwidth)
GTS 250M = 128 (bus width) * 800 (base memory clock) * 4 (data rate) = 51.2 GB/s (memory bandwidth)
Therefore, as far as memory performance is concerned, they are equivalent. And then if you compare the other specifications (shaders, core clock, shader clock, and gigaflops) they are also equivalent. So thus, the GTS 250M is essentially a 40nm 8800M GTX / 9800M GT with more than half the TDP. Now if 1GB of VRAM is really helpful with 51.2 GB/s memory bandwidth is a different question (the performance gain is neglible and usually only seen at high resolutions), but a 128-bit bus with GDDR5 is equivalent to a 256-bit bus with GDDR3. Also, the GTS 250M is essentially a die-shrink of the 9800M GT with a 128-bit bus and GDDR5 instead of a 256-bit bus and GDDR3 (to lower TDP and manufacturing costs), and the 9800M GT is essentially a rebrand of the 8800M GTX.
So is my 9800m super obsolete already?
I think will pass on nvidia video cards for long time, the last fiasco nvidia laptop chips overheating made spent almost the price of the Dell Vostro Laptop I paid for. So long nvidia. Fortunately dell sold me a refurbished 8600M GT and I replaced by myself after letting specialists in Europe trying to find out unsuccessfully about the problem, the problem was I bought a 999 laptop vostro 17" fully loaded as a gift for my dad and shipped it overseas to Switzerland. At the end, I never sent it back to Dell and just replaced it with a new video card and problems solved.
I hear ya there. I also got nailed by the defective parts, and lost my out of warranty Thinkpad. For the at least $300 it would have cost me to replace the motherboard, I figured buying a more recent machine would have been a better choice. It's kinda made me leery to go anywhere near Nvidia, out of spite. It's definitely a problem they should have had the decency to recall... though I suppose that would have killed their business, considering it affected pretty much every mobile video card they released last generation.
Every OEM for your case offers out of warranty free replacement of the motherboard in your laptop if you are affected by this issue.
Why release these now? Why not wait for Direct X 11 and include support for it? Or are they planning on holding back DX11 support for a while, arbitrarily? Seems like a very short and unnecessary stopgap to me.
The answer is simple.
It's just more....money!
Glad to see that it's done using a 40nm process; this adds validity to my hopes of seeing 40nm desktop GPUs from Nvidia (since ATI already says that's what they're using).
The only question is.. Will they melt?
the question for me is when to make a good investment on laptop~ since technology is changing everyday, and it's endless ~~
Ridiculously pseudo high tech article image... Reminds me of Hackers but not as ridiculous.
"Of course, with integrated graphics like the 9400M starting to obviate discrete graphics in the mid range -- even including Apple's latest low-end 15-inch MacBook Pro "
Thats because the idiots at Engadget think a 9400M is a good card ! it's better than Intel Extreme but thats about it.
too bad none of the stats will mean anything to me. Why don't reviewers ever compare mobile gpus and desktop gpus straight up?
The GeForce GTS 260M has about the same performance of a desktop GeForce 9600 GSO based on the specs (real world performance may vary).
I want to see ATI's 40nm 48xx series!
I gotta say, the latest macbooks kinda are obsolete. I say they are weak in graphics now. I mean, come on, if these were in the Macbooks that would kicka**!