SemiAccurate Forums  

 
Go Back   SemiAccurate Forums > Main Category > CPUs

CPUs Talk about processors and related technology


Reply
 
Thread Tools Display Modes
  #131  
Old Today, 12:24 AM
aaronspink aaronspink is offline
10^2
 
Join Date: Feb 2010
Posts: 1,000
aaronspink is on a distinguished road
Default

Quote:
Originally Posted by poul View Post
Both GF and TSMC have fully developed TSV support, whish is called, Standard 2.5D. Intel does not support Standard 2.5D, but instead EMIB. The goal for EMIB, is to provide 'cheaper MCM interconnects'. Its not better than Standard 2.5D, neither does it have any advantages over Standard 2.5D, although the opposite might be argued.
EMIB actually does have significant advantages over TSV 2.5D. One of which is not having to do TSV. Second of which is not having to do yet another die that is larger than all the other dies put together. A third is not having an additional signal discontinuity for external signals. EMIB is basically all the advantages of TSV with basically none of the disadvantages.
__________________
speaking for myself inc.
Reply With Quote
  #132  
Old Today, 12:27 AM
aaronspink aaronspink is offline
10^2
 
Join Date: Feb 2010
Posts: 1,000
aaronspink is on a distinguished road
Default

Quote:
Originally Posted by toughluck View Post

IBM uses MCM for z books.
They use MCMs because they are already basically at their limit for monolithic die size.
__________________
speaking for myself inc.
Reply With Quote
  #133  
Old Today, 04:44 AM
juanrga's Avatar
juanrga juanrga is offline
>intel 4004
 
Join Date: Aug 2013
Posts: 4,615
juanrga will become famous soon enough
Default

Quote:
Originally Posted by raghu78 View Post
dude you just won't stop trolling, isn't it. 128w is instantaneous peak power draw supported by the AM4 socket specification and not TDP. Anyway here is how AMD calculates TDP.

<snip>
I am not trolling. You are regurgitating marketing BS, whereas ignoring all measurement data and technical stuff that contradicts you.

Before launch, we stated that the 1800X pulls 105W with turbo/XFR disabled. The real TDP for the stock settings was going to be higher than 105W. It was evident.

Reviews have shown that the 1800X pulls similar amounts of power than 125W FX-8350. whereas the 1700 pulls similar to 91W chips. I have given you many examples.

HFR measured 128W for the 1800X CPU on ordinary workloads as x264. CanardPC measured 90W for the 1700 CPU.

Updated HFR article states that the formula given by AMD, the one you repeat and I have snipped, is a marketing invention and that the real TDPs are 128W and 90W according to AMD:

Quote:
What results from this formula is not a TDP in the usual sense of the term [...] What are the TDPs, within the meaning of the consumption limit and therefore the maximum number of watts to be dissipated, of the Ryzen? AMD also communicates this value, less markedly: 128 watts for the 1800X / 1700X, and 90 watts for the 1700. These are the values that are most comparable with the TDP communicated by Intel.
We know the X399 socket is officially rated at 180W (real TDP >200W). All your fantasies about this platform have gone raghu, because it wasn't never 65W. And your claims about Naples will be soon officially debunked (they are unofficially).
__________________
Reply With Quote
  #134  
Old Today, 05:08 AM
Amiga500's Avatar
Amiga500 Amiga500 is offline
640k who needs more?
 
Join Date: Nov 2009
Posts: 737
Amiga500 is on a distinguished road
Default

Quote:
Originally Posted by aaronspink View Post
EMIB is basically all the advantages of TSV with basically none of the disadvantages.
Unless you need to connect to something on the far side of the die - in which case you are essentially replicating a TSV.

I suppose EMIB onto a net or ring that is dedicated to transferring data might work, but not simply adding core logic together like lego blocks.

Although you'd know better than me if there are ways to efficiently pass data through the separate blocks...?

Last edited by Amiga500; Today at 05:18 AM.
Reply With Quote
  #135  
Old Today, 05:55 AM
JeeBee JeeBee is offline
2^11
 
Join Date: Jan 2010
Location: UK
Posts: 3,209
JeeBee will become famous soon enough
Default

It depends on what the overall cost of embedding multiple smaller EMIB dies into the substrate with micrometer positioning accuracy (on multiple dies, in relation to each other too) is versus the single larger interposer is.

You'll still need TSVs for stacked stuff like HBM. However I could see a variant where the base logic die for the stack is also the EMIB (by being larger than the DRAM dies on top, and then underlying the die with the memory controller on it).
Reply With Quote
  #136  
Old Today, 07:25 AM
raghu78 raghu78 is offline
8-bit overflow
 
Join Date: Oct 2010
Posts: 465
raghu78 is on a distinguished road
Default

Quote:
Originally Posted by juanrga View Post
I am not trolling. You are regurgitating marketing BS, whereas ignoring all measurement data and technical stuff that contradicts you.

Before launch, we stated that the 1800X pulls 105W with turbo/XFR disabled. The real TDP for the stock settings was going to be higher than 105W. It was evident.

Reviews have shown that the 1800X pulls similar amounts of power than 125W FX-8350. whereas the 1700 pulls similar to 91W chips. I have given you many examples.

HFR measured 128W for the 1800X CPU on ordinary workloads as x264. CanardPC measured 90W for the 1700 CPU.

Updated HFR article states that the formula given by AMD, the one you repeat and I have snipped, is a marketing invention and that the real TDPs are 128W and 90W according to AMD:



We know the X399 socket is officially rated at 180W (real TDP >200W). All your fantasies about this platform have gone raghu, because it wasn't never 65W. And your claims about Naples will be soon officially debunked (they are unofficially).
Any review which shows data you do not like is BS according to you. I have shown multiple reviews of 1700 and 1700x which prove the actual power draw is true to rated TDP.

Talking of claims you are the one who trolled the forums with Zen predictions of 3 Ghz base and 3.5 Ghz boost at 95w. How did that end up for you R7 1700 at 3 Ghz base and 3.6 Ghz boost for 65w.

Right now its almost assured we will see AMD 16c/32t at 3.2 or 3.4 Ghz base for 180w TDP. For Naples 32c/64t we can expect 2.4 -2.5 Ghz at 180w TDP. Do you want to make some predictions

btw what were my so called fantasies about the x399 platform. I would like to hear from troll master

Last edited by raghu78; Today at 07:32 AM.
Reply With Quote
  #137  
Old Today, 07:29 AM
Amiga500's Avatar
Amiga500 Amiga500 is offline
640k who needs more?
 
Join Date: Nov 2009
Posts: 737
Amiga500 is on a distinguished road
Default

Please please please don't quote! It renders the ignore functionality useless.
Reply With Quote
  #138  
Old Today, 07:40 AM
schroeder schroeder is offline
640k who needs more?
 
Join Date: Jan 2013
Location: Portugal
Posts: 652
schroeder is on a distinguished road
Default

I'll second that.

Stop feeding the TROLL, if you people want peace of mind, just put him in the ignore list!
Reply With Quote
Reply


Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


Forum Jump


All times are GMT -5. The time now is 08:22 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2017, vBulletin Solutions, Inc.
SemiAccurate is a division of Stone Arch Networking Services, Inc. Copyright 2009 Stone Arch Networking Services, Inc, all rights reserved.