GTX 260 55nm images (Uploaded on imageshack)

[zlojack]

I've been thinking they might do that with the RAM. If you look at the inside of EVGA's stock heatsinks they have twice the number of ram blocks in there and 1/2 of them were doing nothing. That might've been a good idea to do on the full cover waterblocks. This does present an issue with aftermarket cooling solutions. There may be alot of GPU blocks with RAM sinks again.

Dangit, I could order one of these to SLI and get one like this ona 280 and even the Unisink wouldn't work. I am getting a stadarnd non-fullcover block though. Might have to stepup and then go SLI later.

Watercooling these things does help them tremndously. They have alot of core and memory clock in them. I've found my max clocks already on the 280, but can't leave it there becasue of thermal runaway. I need more cooling. Watercooling these is really a great way to get solid max clocks from them.

Definitely. Talonman got some great clocks out of his GTX 280 with the EVGA FC block on it.

I was planning on grabbing three of the 260 Core 216 cards for now and then going through step-up when the 55nm GTX 280s came out. But if the layout is different, that throws a wrench in my plan as I'd be stuck with three useless blocks.

Maybe I'll just go for one card now and then step that up and sell the block and buy two more of the 55nm 280s later.

[jtf2]

in looking at the dtek unisink,each ram chip that makes contact with the dtek block,seems to have a "raised"part of the sink,to make contact with the chip.
i wonder if it would work to just get small pieces of copper the size of each chip to act as a shim bewteen the unisink and the memchips.
i have on order the dtek block and unisink,and if i can`t get this idea to work i am going to have to just purchase a current revision of the 260.

[T_Flight]

Definitely. Talonman got some great clocks out of his GTX 280 with the EVGA FC block on it.

I was planning on grabbing three of the 260 Core 216 cards for now and then going through step-up when the 55nm GTX 280s came out. But if the layout is different, that throws a wrench in my plan as I'd be stuck with three useless blocks.

Maybe I'll just go for one card now and then step that up and sell the block and buy two more of the 55nm 280s later.

Yep, after I saw that thread I was convinced more than ever. After I saw how much these cards had in them and the Furmark tests I did I was *sold*. There is no way I can run my card at max clocks for any length of time...the temps run away and I have to back out of it. You can actually see it in Furmark when you get what I call "the flickers". Those blocky artifacts. It doesn't do it too bad, but it'll tell you "Hey! Back out! I gotta cool down." hehe

[Blacky]

I hope that blue pcb is only from zotac, since its the only brand who puts weirds colors to their pcbs (orange)...by other hand the toned-down power circuitry overall, with 4 phase vGPU and 2 phase vMem would limit the OCing compared to the current one no?

[spursindonesia]

Hmm, if it just quietly replace the current line up of GTX series card, the reference clock won't see any change or improvement, correct ? Just some TDP reduction. But atleast nVidia does their customers a good deed by modifying its PCB design, so the buyers will atleast be able to tell the difference beetween the old card and the new one.

[RPGWiZaRD]

This version certainly looks less cluttered and cheaper to produce. Let's just hope it's gonna boost overclocks a bit further than original GTX 260 and won't cost anymore either (than Core 216 version). They should be cheaper to produce so I'd be disappointed if the cost would be higher just cuz it's a "new" product.

[Stevethegreat]

This version certainly looks less cluttered and cheaper to produce. Let's just hope it's gonna boost overclocks a bit further than original GTX 260 and won't cost anymore either (than Core 216 version). They should be cheaper to produce so I'd be disappointed if the cost would be higher just cuz it's a "new" product.

If g92b is of any indication, the 260s will gain a 6%-8% overclocking headroom, which is nice if it is accompanied by lower prices (since GT200b is cheaper to produce).

[munim]

Can we tell that it's been shopped by looking at the pixels and having seen many shops in our time?

[jas420221]

The naming convention is changing yes?

The sticker says GTX260 2162

[highoctane]

I really don't see the price going down unless ATI changes the pricing on their lineup, if anything the price will go up with the revision bringing a premium.

I've been waiting for the 55nm shrink to upgrade my tried and true 8800gtx.

[NapalmV5]

no more one pcb for all..


if this 55nm 260 is for real then:

216 shader gtx260
240 shader gtx280
2x216 shader gtx295

same old nvidia

but how do the gtx 270/290 come into play then? missinfo?

or is gtx 270/290 higher clocked versions of 260/280?:

216 shader gtx260
216 shader gtx270 higher clocks
240 shader gtx280
240 shader gtx290 higher clocks
2x216 shader gtx295

or are the supposed gtx 270/290 new chips?:

216 shader gtx260
240 shader gtx280
264 shader gtx270
288 shader gtx290
2x216 shader gtx295

dont make sense..

one more.. or:

216 shader gtx260
240 shader gtx270
264 shader gtx280
288 shader gtx290
2x216 shader gtx295


eenee meenee minee mo eenee meenee minee mo..

none?!? ahhh!!

[moogle]

No ram on the back, so no need for a cooling plate on the other side.

That plate was to cool ram on the reverse side? I kinda liked that plate covering the whole card up. I hope they keep it with the newer gen cards.

[VOID WARRANTY]

the blue pcb would go good with my gigabyte......55nm+removing the his+water cooling...will equal some great temps...im stoked

[fireice2]

GTX 260x2 216 anyone?

[VOID WARRANTY]

GTX 260x2 216 anyone?

yea if they price it right...

[Blacky]

no more one pcb for all..


if this 55nm 260 is for real then:

216 shader gtx260
240 shader gtx280
2x216 shader gtx295

same old nvidia

but how do the gtx 270/290 come into play then? missinfo?

or is gtx 270/290 higher clocked versions of 260/280?:

216 shader gtx260
216 shader gtx270 higher clocks
240 shader gtx280
240 shader gtx290 higher clocks
2x216 shader gtx295

or are the supposed gtx 270/290 new chips?:

216 shader gtx260
240 shader gtx280
264 shader gtx270
288 shader gtx290
2x216 shader gtx295

dont make sense..

one more.. or:

216 shader gtx260
240 shader gtx270
264 shader gtx280
288 shader gtx290
2x216 shader gtx295


eenee meenee minee mo eenee meenee minee mo..

none?!? ahhh!!

the 270 and 290 was made by the stupid inq with that BS article few months ago...nvidia never announced them either in the first place.

[labs23]

Good! Benchmarks for these card will surface out soon. A bit sooner i hope. Well, I won't expect much from a die-shrink.

[NapalmV5]

http://www.techpowerup.com/img/08-12-05/106f.jpg

lol GTX 260² isnt it more like GTX 2603 ?

this is the 3rd 260.. whats next.. probably a 40nm 260 in the works

wouldnt GTX 260² be more fitting for the 2x260 gtx 295?

whatta mess..

the 270 and 290 was made by the stupid inq with that BS article few months ago...nvidia never announced them either in the first place.

i dont think they made 270/290 up..

55nm 216/240 shader gtx 270/290 makes sense

not that im defending inq

but here we have a 3rd 260.. why name 2x260 gtx 295?

[Ket]

TWO PCIE cables needed even with 55nm, longer than the Titanic and a blast furnace for a GPU? No thanks..

[Blacky]

lol GTX 260² isnt it more like GTX 2603 ?

this is the 3rd 260.. whats next.. probably a 40nm 260 in the works

wouldnt GTX 260² be more fitting for the 2x260 gtx 295?

whatta mess..




i dont think they made 270/290 up..

55nm 216/240 shader gtx 270/290 makes sense

not that im defending inq

but here we have a 3rd 260.. why name 2x260 gtx 295?

well there isn't a press realease by nvidia saying those changing name won't make sense since its the same chip just with a die shrink just like 9800GTX > 9800GTX+ also imo nvidia doesnt want to screw up again with a large product line, they want it to have a simple line just like ATI is doing with their new line of products that is...just my 2 cents

[VOID WARRANTY]

TWO PCIE cables needed even with 55nm, longer than the Titanic and a blast furnace for a GPU? No thanks..

a blast furnace..your saying they run hot right?..so how long have you had your 55nm gtx260?...you must have some serious connections...a 55nm before they even hit the shelves..tell me what were your temps...

[nosboost300]

a blast furnace..your saying they run hot right?..so how long have you had your 55nm gtx260?...you must have some serious connections...a 55nm before they even hit the shelves..tell me what were your temps...

even then so.. my gtx260 core 216 only runs 44 idle... and thats sitting directly touching the pcb to another gtx260!!!!

people really shouldnt judge a product if they don't own it

[LordEC911]

people really shouldnt judge a product if they don't own it

..so how long have you had your 55nm gtx260?...you must have some serious connections...a 55nm before they even hit the shelves..tell me what were your temps...

LoL.

...55nm+removing the his+water cooling...will equal some great temps...im stoked

Whoops...

[VOID WARRANTY]

the blue pcb would go good with my gigabyte......55nm+removing the his+water cooling...will equal some great temps...im stoked

how can it not have lower temps...nvidia is saying the 55nm will use less power...i forget ware the article was...but there saying it will use something like 100watts less then the 65nm version...a decrease in power consumption should make it a cooler running chip....but hey it could be BS its hard to find reliable sources...it very well could run hotter then the 65nm....but not likely...

[LordEC911]

how can it not have lower temps...nvidia is saying the 55nm will use less power...i forget ware the article was...but there saying it will use almost 100watts less then the 65nm version...hey it could be BS though its hard to find reliable sources

That wasn't Nvidia. Nvidia hasn't come out with any numbers or scheduled any events because there is little to no improvements.

The numbers you are quoting are realworld power consumption vs TDP.