• 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.

AMD 7nm GPU News and Rumours 2018/2019

Man of Honour
Joined
13 Oct 2006
Posts
91,171
7nm process looks terrible. V20 with it's transistor count should be much smaller than it is.

Nothing wrong with the 7nm process - I think AMD has kind of jumped in the deep end with it though (definitely late to the party TSMC wise) and I'm getting the feeling the first round of releases might be not very optimised essentially indiscriminate "shrinks" from 12FF that don't really take advantage of the node.
 
Soldato
Joined
28 May 2007
Posts
18,262
Nothing wrong with the 7nm process - I think AMD has kind of jumped in the deep end with it though (definitely late to the party TSMC wise) and I'm getting the feeling the first round of releases might be not very optimised essentially indiscriminate "shrinks" from 12FF that don't really take advantage of the node.

Assumez, thinkz and feelz all in one post. Nice!
 
Associate
Joined
26 Mar 2016
Posts
150
Nothing wrong with the 7nm process - I think AMD has kind of jumped in the deep end with it though (definitely late to the party TSMC wise) and I'm getting the feeling the first round of releases might be not very optimised essentially indiscriminate "shrinks" from 12FF that don't really take advantage of the node.

It's not a simple shrink, as it has DP and additional DL stuff. I'm very sceptical.

https://www.amd.com/en/products/professional-graphics/instinct-mi60

20% performance increase by 7nm with 14,7 Tflops at 300W.
 
Man of Honour
Joined
13 Oct 2006
Posts
91,171
It's not a simple shrink, as it has DP and additional DL stuff. I'm very sceptical.

https://www.amd.com/en/products/professional-graphics/instinct-mi60

20% performance increase by 7nm with 14,7 Tflops at 300W.

Hence why I put shrink in quotation marks - in some cases a direct optical shrink isn't even possible - but it is still possible to design stuff out that you couldn't actually produce on a larger node that isn't optimised for a smaller one but is possible to produce on it.
 
Back
Top Bottom