NVIDIA GeForce RTX 4070 specs have (again) changed, now rumored with more cores and faster memory

Published: Aug 4th 2022, 07:04 GMT   Comments

Please note that this post is tagged as a rumor.

Rumor: RTX 4070 with 7680 CUDA cores and 12GB G6X 192b memory

Kopite7kimi has an update on RTX 4070 specs.

After a few disappointing rumors about the RTX 4070 getting 10GB and 160-bit memory configuration, NVIDIA may have changed their mind (again) and will launch a 12GB 192-bit model instead. Not only that, but this graphics card would also have faster memory than the previously mentioned 18 Gbps modules. A faster 21 Gbps memory is now rumored to be used by RTX 4070 SKU, same as alleged RTX 4090 and 4080. A wider bus and faster memory will enable 504 GB/s of maximum bandwidth, 36.5% higher than the previous RTX 4070 specs.

Furthermore, the PG141-SKU331 is rumored to feature 7680 CUDA cores, this is actually the full configuration of the AD104 die with 60 Streaming Multiprocessors. Such configuration was previously rumored for RTX 4070 Ti.

Kopite now claims that the 3DMark TimeSpy Extreme performance for such a configuration would increase from 10K points to 11K points, which now means it outclasses GPUs such as RTX 3090 Ti from the current generation.

Nevertheless, the specifications are not set in stone and are likely to change. This does not mean that current specs are wrong, on a contrary. A good example is the RTX 3050 model which had its first internal specs around December 2020. After receiving 3 major specs updates NVIDIA shelved the SKU till January this year. The point is, NVIDIA tends to shuffle the specs a lot and what Kopite is reporting are those frequent internal updates.

RUMORED NVIDIA GeForce RTX 40 Series Specs
VideoCardz.comGeForce RTX 4090GeForce RTX 4080(OLD)
GeForce RTX 4070
(NEW)
GeForce RTX 4070
ArchitectureAda (TSMC N4)Ada (TSMC N4)Ada (TSMC N4)Ada (TSMC N4)
GPUAD102-300AD103-300AD104-275AD104-400 (?)
Board NumberPG139-SKU330PG139-SKU360PG141-SKU341PG141-SKU331
SMs
 
128
 
80
 
56
 
60
CUDA Cores
 
16384
 
10240
 
7168
 
7680
Base Clock
 
2235 MHz
TBCTBCTBC
Boost Clock
 
2520 MHz
TBCTBCTBC
Memory
 
24 GB G6X
 
16 GB G6X
 
10GB G6
 
12GB G6X
Memory Bus
 
384-bit
 
256-bit
 
160-bit
 
192-bit
Memory Speed
 
21 Gbps
 
21 Gbps
 
18 Gbps
 
21 Gbps
Bandwidth
 
1008 GB/s
 
676 GB/s
 
360 GB/s
 
504 GB/s
TDP
 
~450W
 
~420W
 
~300W
 
~300W
3DMark TimeSpy Extreme
 
~19,000
 
~15,000
 
~10,000
 
~11,000
Launch DateOctober 2022TBCTBCTBC

Source: @kopite7kimi




Comment Policy
  1. Comments must be written in English and should not exceed 1000 characters.
  2. Comments deemed to be spam or solely promotional in nature will be deleted. Including a link to relevant content is permitted, but comments should be relevant to the post topic. Discussions about politics are not allowed on this website.
  3. Comments and usernames containing language or concepts that could be deemed offensive will be deleted.
  4. Comments complaining about the post subject or its source will be removed.
  5. A failure to comply with these rules will result in a warning and, in extreme cases, a ban. In addition, please note that comments that attack or harass an individual directly will result in a ban without warning.
  6. VideoCardz has never been sponsored by AMD, Intel, or NVIDIA. Users claiming otherwise will be banned.
  7. VideoCardz Moderating Team reserves the right to edit or delete any comments submitted to the site without notice.
  8. If you have any questions about the commenting policy, please let us know through the Contact Page.
Hide Comment Policy
Comments