• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Intel Serpent Lake processors with nVidia graphics planned for Q1 2028

M

Marees

Supreme [H]ardness
Joined
Sep 28, 2018
Messages
4,632
Intel x86 processors with NVIDIA RTX graphics reportedly planned for 2028
Today, 18:30 GMT • WhyCry •
Please note that this post is tagged as a rumor.

Intel roadmap reportedly lists NVIDIA RTX SoCs for early 2028

Source: VideoCardz.com
https://videocardz.com/newz/intel-x...idia-rtx-graphics-reportedly-planned-for-2028

Source: Erdi Ozuag twitter
https://x.com/fx57/status/2066467380625109491?s=20

Erdi Özüağ, a former editor at Turkish tech site DonanimHaber and now a YouTube-based tech reporter, has reached out to VideoCardz with exclusive information on Intel’s first client processors with NVIDIA graphics.

According to Erdi, Intel’s current roadmap lists the first x86 processors with NVIDIA RTX graphics for the first quarter of 2028. This would line up with a possible CES 2028 announcement, though Intel has not confirmed any product name, configuration, or launch date.

Serpent Canyon: Intel and NVIDIA RTX SoC in 2028​

The information appears to refer to the same client SoC project announced by Intel and NVIDIA in September 2025.
 
Which would mean big on die memory too? Or they hoping system ram will be good enough in 2 years time?

Intel APU…$1000 and 350W TDP
 
westrock2000 said:
Which would mean big on die memory too? Or they hoping system ram will be good enough in 2 years time?

Intel APU…$1000 and 350W TDP
Click to expand...
you mean on package like apple soc ?

High end APU doe snot yet put the memory on die, for performance they go on package like this:
55089-111839-lede-xl.jpg


The black rectangular chips at the right of it, yet to be able to do it on-die, die space is really precious and importantly the fab process that good to make memory is all different than the one that make logic, that why cache is so space innefficiant and does not scale well on cpu-gpu die.

As for the first launch on an intel APU, having to guess will be like for StrixHalo-N1X (with the latest best) on motherboard, Intel when they did try to do like Apple faced a lot of cost challenge:
https://www.tomshardware.com/pc-com...approach-for-future-cpus-due-to-margin-impact

Ability for oem to have some flexibility is also nice when you are not a pure vertical maker-seller like Apple.
 
Last edited:
Cactusj said:
Better hope that its not on die memory, or you will be stuck with 4/8gb...
Click to expand...
Not sure how you could reach that big of a number, look how much memory per mm you get from 3d vcache on giant epyc cpus... would probably be under 2GB even on a RTX 5090 sized die that has all its surface used for 3d cache. If you can stack them many layers with something new maybe reach that.
 
Captain Newmackwa said:
I wonder if this will be a one-off thing like what Intel and AMD did a while back.
Click to expand...
Anything can happen and 5 billions is not that much money for Nvidia anymore but still, that was a huge commitment and the initial plan is probably to try to make many generation of this.

Intel hammer lake roadmap seem to point toward Nvidia igpus:
https://www.notebookcheck.net/Intel...ifts-Titan-Lake-to-mobile-only.1118950.0.html

how well intel foundry work at making nvidia tile (if they use Intel foundry) being one element, how much of the deal had datacenter side of things and how much interest at an x86-Nvlink type of solution is there outthere and so on, but at least until they change socket I would expect using Nvidia igpu.
 
Captain Newmackwa said:
I wonder if this will be a one-off thing like what Intel and AMD did a while back.
Click to expand...

Doubt it with ARC spinning down (they'll still do low end iGPUs I believe I read/heard) - but all medium/high end are gonna be Nvidia for now IIRC - and then AMD for a future generation ala Kaby Lake G again maybe? Then go back and forth between those 2 for medium/high end as the contracts/relationships allow?
 
You must log in or register to reply here.
Back
Top