First Look of 55nm GeForce GTX260
Since we unveiled the first image of 55nm GT200, we thought it would take longer for concrete products getting prepared. Nobody is sure whether it has NDA problem. As we estimated, GeForce GTX 260 and GTX295 based on 55nm GT200 will not be available until January next year.
New Zotac GeForce 260 Graphic Card based on NVIDIA P654 PCB design
Not all the manufacturers are capable of producing VGA cards like GeForce GTX260 of such complicated structure. Most of them tend to purchase reference cards by NVIDIA, and then put on labels of their own logo. However, friends from Zotac claims they have been prepared to manufacture the latest version of GeForce GTX260 on their own, and send us the pictures below to prove their saying. Except for the PCB is changed to be blue color, we don’t find any difference on it from reference card. So I am going to introduce them together.
The above pictures are both NVIDIA reference cards
Based on brand-new P654 reference design, the latest version of GeForce GTX260 Graphic Card utilizes 26.5CM PCB. All the 14 video memories are put on the front side of PCB, with 10 layers of PCB design. Previous P651 PCB shared by GeForce GTX260/280 was 14 layers. Previous version of GTX260’s features 3+2 phase power module (The 3 CSP sealed Volterra VT1195SF MULTIPHASE chip would be most eye-catching).
Remark by one of AIC: the difficulty in purchasing Volterra has resulted in limits to mass production.
New P654 power modules has upgraded to 4+2 phase, equipping solid-state capacitor, magnetism shield inductance and MOSFET. Where the two 6Pin are connected features a large input inductance separately. Therefore, the reduced difficulties in design will bring down the cost.
The GT200b of GeForce GTX260 will be renamed to be “G200-103-B2” instead of 65nm “G200-103-A2”. The “B” means newest 55nm processing is used. The number of its stream processors remains as 216. The video memory uses Samsung K4J52324QH-HJ1A chip. The specs are 896MB/448bit, with default frequency of 575/2000MHz and memory bandwidth of 112GB/s.
Updated GeForce 260GTX continues to use all-in-one cooler by CoolerMaster, in order to cool the chip, video memory and power module effectively. The cooler is equipped with four copper heat sinks and 80mm concave turbofan with temperature controller. The heat sinks and fins are connected together by jointing. Due to the change of circuit, design of cooler has been adjusted accordingly.
As it will not be available until Jan next year, we’re not sure about its retail price. Its frequency stays the same as 65nm’s, so there won’t be much difference in performance. But we can expect its overclocking ability and power consumption will be improved thanks to 55nm GT200. BTW: This significant AMP product from Zotac will upgrade to 0.8ns for higher default frequency.
December 5th, 2008 at 7:45 pm
55nm GTX260 finally here.I need one of this baby.
December 5th, 2008 at 10:26 pm
Mmmm… don’t understand Why Nv use DDR3 ?
December 6th, 2008 at 12:13 am
Because GDDR3 is cheap and their memory controller hasn’t been updated. I think they should have gone with GDDR4 at least. They would also have to redesign the PCB traces/power circuitry and BIOS among other tweaks.
Lower power + higher bandwidth would be a wonderful thing and the die-shrink would be a good time to implement it. Oh well, GTX300 40nm is where the next memory tech will be.
December 6th, 2008 at 1:09 am
good job!!
December 6th, 2008 at 2:33 am
ddr3 + 512bit bus or ddr5 and on a 256bit bus is performance wise just the same. its marketing for ati; another “futureproof” statement to convince average joe…
December 6th, 2008 at 5:08 am
But don’t forget 256bit means a smaller gpu die, half mem chips and simpler pcb design than 512bit.
yeah, that’s a perfect maketing and business for ATI.
December 6th, 2008 at 6:38 am
I wonder why there is not a DisplyPort on it, or am i wrong ?
Jean
December 6th, 2008 at 12:08 pm
Paralou: because this is essentially just a die shrink revision, supporting Display Port would require more significant hardware changes so the GPU would be able to output the proper signals. Perhaps whenever the new series (G3xx unless nv decides to change it’s naming scheme again) comes out display port will be added, but where to put it on single slot cards will be a question since it would require either dropping the 2nd DVI port or the TVout connector.
December 7th, 2008 at 2:05 am
Why don’t these beefy cards have a HDMI output? Ridiculous.
December 7th, 2008 at 7:36 am
Maybe if you took the 5 seconds it takes to google it you can put on an HDMI adapter and send a SPDIF cord to your videocard and have it output both sound and video.
Why don’t people get their head out of their ass and google things? Ridiculous.
December 8th, 2008 at 5:54 am
Actually he meant why don’t they have a built in full HDMI output. What you’ve sugested is to use separate HW for the audio.
Not really an nVidia fan but they sure did to good job on the reconfiguration… the memory modules on the bake side was a joke… a not so cool joke.
One premium partner that’s worthy of its place!
December 8th, 2008 at 2:48 pm
Because it’s sill to waste PCB space fro audio processing, when this is a GRAPHICS board.
December 8th, 2008 at 4:04 pm
Actually HDMI is DVI + sound.
So video part is are covered.
December 8th, 2008 at 4:13 pm
@headfoot
Why not put some cheep soundcard on VGA (it`s only digital audio) and just select that output in your player?
90% ppl even don`t know where is digital out from their motherboards and most of them are back on your computer where your keyboard and mouse are. I`m personally using that output already when i playing 5.1 games. What will be if someone kid want to watch cartoon on your Plazma and you want to play a game at same time (because with ATI that is perfectly doable). I don`t care if this card have 5%-10% more fps than ATI. If you ask me, ATI is 3 years in advance in this “HD” stuff.
December 8th, 2008 at 9:44 pm
Few years ago i’m Nvida fan, but now I love ATI Radeon, but again i can think about Nvida. but 55nm GTX 260 good choice??? Still use DDR3, no built in HDMI. My 42″ lcd tv support HDMI…………..
December 9th, 2008 at 2:01 am
Thilina, yea mean full-HDMI…
Bob, it’s not processing… it’s rerouting! Most of the sound processing is done by the southbridge, a dedicated on-board sound chip or a separate sound card and it’s rerouted through the HDMI port through the PCIe interface.
The part about it being faster is questionable since comparing the image quality in a few cases has led me to conclude that the Big Bang gain is nothing but history repeating itself.
This design is actually more useful to the overclockers who use water-cooling, just my opinion.
December 10th, 2008 at 3:29 am
the integration of the sound card onto the video card is a terrible idea, i give you three reasons:
1.) the card would require more power draw.
2.) it wouldn’t be any better than using your on-board sound.
3.) the GT200 cards are already freakin huge, and this would just make the card bigger.
personally, i’d much rather see dedicated PhysX Calculaion cores built into the GPU.
December 13th, 2008 at 3:19 am
@pyroxide
GeForce cards already have PhysX Calculation if you have the newest driver and enable it in the NVIDIA Control Panel.
January 28th, 2009 at 7:52 pm
will it have lower power requirements ?
Will be priced lower than the current 65nm ?