What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

RTX 3090 / 3080 speculative specifications

Dwayne

Well-known member
Joined
Aug 17, 2008
Messages
1,258
Location
Courtenay, BC

I find it very interesting that there is a lot of CUDA core room between the 3080 and 3090. You could almost see a possible 3080 Ti slot in there around 4700 cores. Not too long to wait now. As for the power and heat, it seems the script is flipped once again. Seems to go in cycles between the fan boys out there. I recall that the Vega 56 and 64 were power hogs and hot as hell compared to the 10xx series. As a computer enthusiast, I just want goes fast... I can buy a bigger power supply, more fans, and pay more for my electricity.


RTX 3090RTX 3080RTX 3070
GPU7nm GA102-3007nm GA102-200?nn GA104-300
BoardPG132 SKU 30PG132 SKU 10PG142 SKU 10
CUDA Cores

5248


4352
TBC
Boost Clock

1695 MHz


1710 MHz
TBC
Memory

24GB G6X


10GB G6X


8GB G6
Memory Clock

19.5 Gbps


19 Gbps


16 Gbps
Memory Bus

384-bit


320-bit


256-bit
Bandwidth

936 GB/s


760 GB/s


512 GB/s
TGP

350W


320W


220W
 

Dwayne

Well-known member
Joined
Aug 17, 2008
Messages
1,258
Location
Courtenay, BC
Another leak posted to Videocardz. True, or just very well done graphics, who know? But here you go.


Basically a rumour/leak confirming another rumour/leak.
 

Shadowarez

Well-known member
Folding Team
Joined
Oct 4, 2013
Messages
1,849
Location
Arctic Canada
how long do you think before this new direct storage api from ms will be back ported to the 20 series. like the rtx voice? if it does work on older cards i can see the prices inflating again.
 

Sagath

Moderator
Staff member
Folding Team
Joined
Feb 7, 2009
Messages
4,741
Location
Edmonton, AB
how long do you think before this new direct storage api from ms will be back ported to the 20 series. like the rtx voice? if it does work on older cards i can see the prices inflating again.

I had to look in to this, because it wasnt clear at first glance...


This says "RTX gpus are capable" . Since its an API, it sounds like MS (DirectX) is the API, and nvidia is just the first to take advantage of it via RTX cores.

TLDR: It should be back ported, how far isnt clear.
 

Shadowarez

Well-known member
Folding Team
Joined
Oct 4, 2013
Messages
1,849
Location
Arctic Canada
yeah its funny though amd x570 and there zen 2 will be only one able to take the nvidia gpu for a nice lil run and fully test this pcie 4 spec. cant wait it see what this will do with pcie gen 4 ssd and the gpu,

and if i can accelerate this even further with primocache since this is all being read not written can choose to allocate ram to 100% read speed. should be interesting have ram speeds feeding the gpu+nvme drive.
 
Top