GTX285 MARS REVIEW

however the GTX 260 was also 65nm process prior to to the release of the GTX 295.
Sure it's from the same tree, as was the 8800GTS to the GTX...but it's not the same.
 
PlatinumPC said:
however the GTX 260 was also 65nm process prior to to the release of the GTX 295.
Sure it's from the same tree, as was the 8800GTS to the GTX...but it's not the same.
Click to expand...


they released the 55nm version at the same time as the gtx 295 and gtx 285.. but even saying that.. means absolutely nothing because the 65nm core isnt any different.. the 55nm core is a die shrink of the same 65nm core.. same architecture.. same performance..
 
PlatinumPC said:
however the GTX 260 was also 65nm process prior to to the release of the GTX 295.
Sure it's from the same tree, as was the 8800GTS to the GTX...but it's not the same.
Click to expand...
the gtx260 going to 55nm is what allowed Nvidia to do a gtx295 in the first place. the gtx295 IS 2 gtx260 cards with the last cluster unlocked. besides that unlocked cluster a gtx260 is identical to the gpus used on the gtx295. the gtx275 is just a gtx260 with a the same unlocked rop/tmu cluster, better specced memory and faster clocks. they are all basically the same gpu so stop making an issue over something so silly.
 
Fact remains, the GTX 275 is closer speced to the GTX 295 than the gtx 260 is.

Just wanted to point that out....don't see what all the fuss is about.

I think other will agree when I say that TMU's and Stream Processors are a little more important than a change in clock speed.
 
PlatinumPC said:
Fact remains, the GTX 275 is closer speced to the GTX 295 than the gtx 260 is.

Just wanted to point that out....don't see what all the fuss is about.

I think other will agree when I say that TMU's and Stream Processors are a little more important than a change in clock speed.
Click to expand...
its semantics. before the gtx275 EVER came out the gtx295 took two gtx260 cards unlocked the last cluster and put them together. the 55nm gtx260 is identical in every other way except for that unlocked cluster. then later Nvidia took a gtx260 unlocked the last cluster, put on better specced memory and raised the clocks and called it a gtx275. of course you could also say they took one of the gpus form the gtx295 and put the better memory on it and raised the clocks to make a gtx275 but its the SAME end result. they are ALL the same basic gpu that started life even back when the gtx260 was 65nm and had two clusters disabled.
 
Last edited:
MurderCityDevil said:
seriously what was asus thinking
Click to expand...

ASUS was thinking "Man, there are some geeks out there that will spend insane amounts of money on high end graphics, even if they don't need it. We can make a card that is a tiny bit more powerful than what else is out there, and those people will pay a stupid premium for it."

That is what it comes down to. The thing is like $1000. They are making a massive premium on it, not to mention the advertising it gets them.

That's also why they are doing the "limited edition" thing. They know 99.9999% of people will never buy one, so it is going to be a small production run anyhow. They they make it "limited" to seem cooler and justify it more to the people that pick one up.

Over all, a smart idea. Any money they spent developing the thing (which probably wasn't much, nVidia supplies the cards, they just made a thermal solution) they made back on the massive premiums, and it gets people talking as is happening here.

Some people just to have something that's moar bettar than what everyone else has, this is for them. It isn't actually useful in any realistic term.
 
all three cards live in unison though ...as long as people aren't kidding themsleves in thinking that they perform the same..then that way of looking at things is fine. I would suggest it's easier however to call a spade for a spade though and see the current specs for what they are i.e. GTX 295 = 2 x GTX 275's not 2 x special unlocked GTX 260's.
 
PlatinumPC said:
all three cards live in unison though ...as long as people aren't kidding themsleves in thinking that they perform the same..then that way of looking at things is fine. I would suggest it's easier however to call a spade for a spade though and see the current specs for what they are i.e. GTX 295 = 2 x GTX 275's not 2 x special unlocked GTX 260's.
Click to expand...
and what do you think the gtx275 is? oh thats right, a gtx260 with the last cluster unlocked. for the third time when they made the gtx295, they took two gtx260 cards and unlocked the last cluster on each. so again they are all based off the SAME thing.
 
cannondale06 said:
and what do you think the gtx275 is? oh thats right, a gtx260 with the last cluster unlocked. for the third time when they made the gtx295, they took two gtx260 cards and unlocked the last cluster on each. so again they are all based off the SAME thing.
Click to expand...

based on that, I could argue all GPU of one generation are just "unlocked" down.

They (nVidia/ATI) typically release the (nearly) highest end version first, then a few low end cards (locked = never put on the die), then unlocked/OCed the final bit out just for the uber high end card.

Until you can reliably convert a GTX260 into GTX275, they aren't "locked," just merely low end.

(I'm sorry if this looks like I am attacking you!)
 
jeremyshaw said:
based on that, I could argue all GPU of one generation are just "unlocked" down.

They (nVidia/ATI) typically release the (nearly) highest end version first, then a few low end cards (locked = never put on the die), then unlocked/OCed the final bit out just for the uber high end card.

Until you can reliably convert a GTX260 into GTX275, they aren't "locked," just merely low end.

(I'm sorry if this looks like I am attacking you!)
Click to expand...
for the FOURTH time to make the gtx295 they took two gtx260s and unlocked the the last cluster. everything else is the same. this is exactly how the gtx295 was described when it was introduced. now that the gtx275 is out you can say that the gtx295 has two of those with lower clocks. either way its the same darn thing.
 
jeremyshaw said:
based on that, I could argue all GPU of one generation are just "unlocked" down.

They (nVidia/ATI) typically release the (nearly) highest end version first, then a few low end cards (locked = never put on the die), then unlocked/OCed the final bit out just for the uber high end card.

Until you can reliably convert a GTX260 into GTX275, they aren't "locked," just merely low end.

(I'm sorry if this looks like I am attacking you!)
Click to expand...
You're right in saying that they aren't locked. However, they are disabled, and they are located on the die. They are simply disconnected physically from the rest of the GPU during manufacturing (probably with a laser). The GTX260, GTX275, GTX280, GTX285, and GTX295 all use the same GT200 core, just with various parts of it enabled or disabled depending on the specific model. That is an entirely different story compared to for example ATI's current generation, where the 4350, 4650/70, 4770, 4850/70, and 4890 all use different cores which differ in terms of what is actually physically included in the die when it is created. The 4650 isn't just a 4870 with certain parts disconnected, it uses a completely different core which has fewer physical components.
 
Zero82z said:
You're right in saying that they aren't locked. However, they are disabled, and they are located on the die. They are simply disconnected physically from the rest of the GPU during manufacturing (probably with a laser). The GTX260, GTX275, GTX280, GTX285, and GTX295 all use the same GT200 core, just with various parts of it enabled or disabled depending on the specific model. That is an entirely different story compared to for example ATI's current generation, where the 4350, 4650/70, 4770, 4850/70, and 4890 all use different cores which differ in terms of what is actually physically included in the die when it is created. The 4650 isn't just a 4870 with certain parts disconnected, it uses a completely different core which has fewer physical components.
Click to expand...

Oops... sorry!

At least it explains why ATI takes so long for their low-end line (after the high end) to come out! (and why they use less power, too?)
 
You must log in or register to reply here.
Back
Top