Soldato
Doesn't the size reference refer to the actual size of the transistor itself.
Just saying...
Last edited:
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
Nvidia Geforce 'Maxwell' Thread
- Thread starter Boomstick777
- Start date
Doesn't the size reference refer to the actual size of the transistor itself.
Just saying...
Not looked at the context but could be in reference to i.e. 16nm FF which is just 20nm modified for finfets.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
I have/had no idea what he was "just saying...." so I ignored it.
The only point I was making was if you had 32nm gpu's or not, it has zero bearing on the 28nm process, how many transistors can fit into a certain size or what configurations of the architecture work or don't. THe 7970 was around 350mm^2, it will be as big as they were going to go with a 384bit bus, and if they went with a 256bit bus they would have gone smaller.
They weren't planning, or going to do a 450mm^2 core on launch because they knew you couldn't do it with suitable yields, Nvidia couldn't do it either. You can't just make a 400mm^2 die because you want to, it has to be balanced Nvidia couldn't just make something inbetween the gk104/gk110, the later had poor yields even pushing for a later than usual launch. Gk104 was as big as they would go on 256bit bus.
A 32nm cayman would have made absolutely no difference, not even one transistor difference, to 28nm.
They don't make a chip on a new process a specific size based on the previous processes, it's an architecture that has balanced points. 7970/680gtx was what could be made at launch, you can't magically up performance by 30% because a previous gen card was faster.
The only point I was making was if you had 32nm gpu's or not, it has zero bearing on the 28nm process, how many transistors can fit into a certain size or what configurations of the architecture work or don't. THe 7970 was around 350mm^2, it will be as big as they were going to go with a 384bit bus, and if they went with a 256bit bus they would have gone smaller.
They weren't planning, or going to do a 450mm^2 core on launch because they knew you couldn't do it with suitable yields, Nvidia couldn't do it either. You can't just make a 400mm^2 die because you want to, it has to be balanced Nvidia couldn't just make something inbetween the gk104/gk110, the later had poor yields even pushing for a later than usual launch. Gk104 was as big as they would go on 256bit bus.
A 32nm cayman would have made absolutely no difference, not even one transistor difference, to 28nm.
They don't make a chip on a new process a specific size based on the previous processes, it's an architecture that has balanced points. 7970/680gtx was what could be made at launch, you can't magically up performance by 30% because a previous gen card was faster.
Soldato
You said 32nm wouldn't make 28nm transistors smaller. That doesn't make any sense
You can ignore it, you're still babbling though. Throw in some wafer sizes so everyone thinks you know what you're on about. Mission complete
You can ignore it, you're still babbling though. Throw in some wafer sizes so everyone thinks you know what you're on about. Mission complete
Last edited:
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Well done, you've posted twice and still don't quite comprehend something as simple as posting something to make a point to the guy who thinks a 6970 on 32nm would have any bearing on the 7970/680gtx in the slightest.You said 32nm wouldn't make 28nm transistors smaller. That doesn't make any sense
You can ignore it, you're still babbling though. Throw in some wafer sizes so everyone thinks you know what you're on about. Mission complete
I'm also not sure why you think it doesn't make any sense.... do you think 32nm DOES make 28nm transistors smaller... that would be something that doesn't make sense. I was highlighting the point for the other guy.
But once again you pick out something I say, randomly quote and for want of a better word, call me out over it... for no apparent reason except to have a go at me... which is exactly your posting style.
I still have no idea what you're trying to say with your post, because your words
actually don't make sense because it had literally no relevance to what I was saying. It's really really simple, I was stating something most assume obvious, because it makes his idea, frankly, silly. You can't just make a 7970 faster because you will it so. Bigger cores weren't going to yield well or make a sensible product, AMD went as big as they thought they could go on a immature process for a complex chip, nothing on any other process effects what that point is. How fast or slow the 6970 is has no effect on what a sensible size chip to make on 28nm was.
Last edited:
Soldato
...a 32nm transistor doesn't make a 28nm transistor smaller.
Ok you're right it does make sense. If you're a simpleton making a really obvious point.
I think maybe you just got your terminology mixed up DM. No need to dig a hole chap
Ok you're right it does make sense. If you're a simpleton making a really obvious point.
I think maybe you just got your terminology mixed up DM. No need to dig a hole chap
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
Regardless of the context is great, until you realise that we weren't talking about shrinking anything at all, making your point .... utterly irrelevant.
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
...a 32nm transistor doesn't make a 28nm transistor smaller.
Ok you're right it does make sense. If you're a simpleton making a really obvious point.
I think maybe you just got your terminology mixed up DM. No need to dig a hole chap
No thanks, you got it wrong and now you're doing the usual thing of trying to make it about someone else being wrong. You posted one thing, either you wanted to have a go or you had no idea what you were talking about. Then two people wonder wtf you're on about, and you go defensive mode... standard.
Now you've even said... wait, it does make sense. That was the ENTIRE POINT, sometimes in a discussion you say something patently obvious to make a point you think the other person doesn't comprehend. There is no mixed terminolgy, I was stating something simple, and what I believed was obvious because he doesn't seem to get it.
His stance... 6970 on 32nm would magically make AMD?Nvidia make their 7970/680gtx faster... defying laws of physics. My stance... you can't change the laws of physics, regardless of the 6970 being on 32nm, 40nm, 768nm..... 28nm is the size it is and that makes the sizes of a specific architecture relatively set in stone on a given process.
Keep digging though, make it about me, change my own point because it suits you, when you're in this deep, you've just got to keep pretending you had a point to start with.
Soldato
** edited **
If your entire point was that bigger transistors don't make smaller ones smaller, then that's fantastic
If your entire point was that bigger transistors don't make smaller ones smaller, then that's fantastic
Last edited:
Caporegime
- Joined
- 18 Oct 2002
- Posts
- 33,188
** edited **
Can't except you had zero idea what you were talking about. My point was clear.. 32nm has no effect on 28nm... you thought I was saying something else entirely... you posted something incoherent, followed by claiming what I said made no sense, followed by claiming it DID make sense but that I didn't understand why I posted it.... followed by the inability to point out how it didn't make sense in my original post.
Can't except you had zero idea what you were talking about. My point was clear.. 32nm has no effect on 28nm... you thought I was saying something else entirely... you posted something incoherent, followed by claiming what I said made no sense, followed by claiming it DID make sense but that I didn't understand why I posted it.... followed by the inability to point out how it didn't make sense in my original post.
Yes it may of been irrelevant to what your were talking about ( AMD 6970's or something or other) but as a statement what I said is accurate.
But all of that is irrelevant to this thread which is supposed to be about Nvidia Maxwell, not AMD anything.
Only because the comment it was in reply to didn't either...
Davedree said "32nm would have made 40nm transistors smaller" which makes no sense, but I can interpret it to mean "if they had used the 32nm process they would've had a shrink available". DM was replying to Engrish.
Last edited:
Soldato
32nm doesn't make smaller transistors smaller.
I don't have to keep saying it do I? 32nm what exactly? You've agreed you meant 32nm transistors don't make 28nm ones smaller to make a point. Why not just say that then. All you said was 32nm doesn't make 28nm transistors smaller? What the **** does that matter?
32nm what? The nm refers to the size of the transistor or more specifically the space between the gate. So saying 32nm doesn't make 28nm transistors smaller, doesn't teach anyone anything, does it buddy.
I just wondered what you meant, but apparently you were being obviously sarcastic. Maybe I forgot who I was talking to
I don't have to keep saying it do I? 32nm what exactly? You've agreed you meant 32nm transistors don't make 28nm ones smaller to make a point. Why not just say that then. All you said was 32nm doesn't make 28nm transistors smaller? What the **** does that matter?
32nm what? The nm refers to the size of the transistor or more specifically the space between the gate. So saying 32nm doesn't make 28nm transistors smaller, doesn't teach anyone anything, does it buddy.
I just wondered what you meant, but apparently you were being obviously sarcastic. Maybe I forgot who I was talking to
Back on topic.
NVIDIA will reportedly launch its GeForce GTX 990 in early 2015
NVIDIA will reportedly launch its GeForce GTX 990 in early 2015
Last edited:
Associate
Wait, GTX 990? B-but the 800 series hasn't even been announced yet? And the 990 would be a dual card which usually comes late into a series...
Soldato
Why does everyone keep saying NV are skipping 20nm completely. I still think a lot of these sites are getting their wires crossed with that
Associate
- Joined
- 20 Mar 2014
- Posts
- 2,361
For the next high end it's gonna be 20nm, DirectX12 and have an ARM core.
Time for an upgrade, gonna buy 3 of these to go with the 8 core Haswell E CPU and 32GB DDR4. Raid 0 SSDs etc.
I have not had an upgrade for a long time so it's gonna be a big un.
Time for an upgrade, gonna buy 3 of these to go with the 8 core Haswell E CPU and 32GB DDR4. Raid 0 SSDs etc.
I have not had an upgrade for a long time so it's gonna be a big un.