Overclock.net - An Overclocking Community - Reply to Topic
Thread: [Tom's Hardware] Two Unknown Nvidia GPUs Spotted With Insane Core Counts Reply to Thread
Title:
Message:

Register Now

In order to be able to post messages on the Overclock.net - An Overclocking Community forums, you must first register.
Please enter your desired user name, your email address and other required details in the form below.
User Name:
If you do not want to register, fill this field only and the name will be used as user name for your post.
Password
Please enter a password for your user account. Note that passwords are case-sensitive.
Password:
Confirm Password:
Email Address
Please enter a valid email address for yourself.
Email Address:

Log-in


  Additional Options
Miscellaneous Options

  Topic Review (Newest First)
03-05-2020 06:46 PM
skupples slowing the cycle down would be dependent on how AMD performs. They may not have that luxury. Someone would then point out Wuhan slowing everything down.

cycle's are moving quicker since pascal dropped as well.
03-05-2020 06:33 PM
guttheslayer
Quote: Originally Posted by skupples View Post
they'll be priced the same, or higher.
Very likely, only the Ti variant will be priced $100-200 cheaper at most. Probably 899/999 for Non-FE/FE, but to justify the price it will be released much later maybe early 2021.


But with the high no of core counts, i think its safe to say the 3080 will be 25% faster and Ti to be 60% faster than 2080 Ti. It shouldn't be too hard given we seen almost doubling of cores. The boost speed might drop a little but the IPC improvement from micro-arch might compensate it. So we shall see...
03-05-2020 03:39 PM
skupples they'll be priced the same, or higher.
03-05-2020 02:20 PM
Slaughtahouse Excuse my ignroance, but is it incorrect to compare CUDA cores from generation to generation? What if the performance per CUDA core is lower than last generation. Or simply put, the boost frequency of the new architecture can't match older generations?

Seeing all these posts of new GPU obliterating old GPU's is just setting everyone up for disappointment. Especially considering we dont know how these cards will be priced.
03-05-2020 02:07 PM
EniGma1987
Quote: Originally Posted by bigjdubb View Post
Isn't Zen 2 considered an MCM design that uses infinity fabric?
I suppose it could be in a way, but normally when you talk about MCM the dies are right up against each other and are two full dies in one package.
Zen2 is a chiplet design, and each section while layed out in the same manner as an MCM, is its own separate die on a different part of the package and they do not contain the entire CPU within them. You have your front end in one die, and core sections in other dies. Ryzen 1 would probably be considered more of an MCM though.


This would be an MCM, where it was two, dual cores dies cut as a single unit on a package. You can see the seam between the die sections in the middle:



This was the next evolution and was a single die designed that way:


Then you have Zen 2, with its chiplet design which is kind of an evolution of MCM:





Likewise, the image shown by the other posters would be two GPU dies cut as a single in an MCM layout:
03-05-2020 12:40 PM
bigjdubb Isn't Zen 2 considered an MCM design that uses infinity fabric?
03-05-2020 11:45 AM
Roaches
Quote: Originally Posted by WannaBeOCer View Post
AMD already beat them to it with their Radeon Pro Vega II Duo.
Not really, AMD's current implementation on that is just a Infinity Fabric bridge like NV-link between 2 packages on a logic board. We're talking about a single integrated package solution...
03-05-2020 11:40 AM
WannaBeOCer
Quote: Originally Posted by Roaches View Post
Exotic packaging solutions are already happening I wouldn't be surprised to see Nvidia being the first to implement a GPU MCM...

AMD already beat them to it with their Radeon Pro Vega II Duo.
03-05-2020 11:27 AM
Roaches Exotic packaging solutions are already happening I wouldn't be surprised to see Nvidia being the first to implement a GPU MCM...

03-05-2020 10:26 AM
bigjdubb I don't know this benchmark at all, but it isn't uncommon at all for benchmarks to incorrectly read the clock speeds on unreleased hardware. Might be jumping the gun to assume that the maximum clock speed listed was the actual maximum clock speed.

Maybe I am just cynical after the 1/2 generation performance improvement of the 20 series. They will have to come out with a real monster to get the performance where it should be. The 3070 will need to absolutely clobber the 2080ti to get any sort of praise from me.
This thread has more than 10 replies. Click here to review the whole thread.

Posting Rules  
You may post new threads
You may 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