• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Discussion Intel Nova Lake in H2-2026: Discussion Threads

Page 4 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
I am mostly curious, how many boards will have integrated 10Gb LAN? Hopefully that is a more common thing soon.

There is this new cheaper 10 GbE chip from Realtek coming soon:


Hopefully that means it’ll become more common on motherboards going forward.
 
There is this new cheaper 10 GbE chip from Realtek coming soon:


Hopefully that means it’ll become more common on motherboards going forward.
intel boards are usually bundled with Intel Ethernet controllers
 
There is this new cheaper 10 GbE chip from Realtek coming soon:


Hopefully that means it’ll become more common on motherboards going forward.

That's good news but that ain't real good stuff outside Windows and (maybe) MacOS.
Realtek support under linux e.g. Debian.

Still nice 'cos that will drag down the price of 10GbE.
 
They likely couldn't physically fit two CCDs on the same package if one of them was a large cache version. Current rumor is that the large cache isn't "stacked" but on the die with the cores.
 
They likely couldn't physically fit two CCDs on the same package if one of them was a large cache version. Current rumor is that the large cache isn't "stacked" but on the die with the cores.
They are using foveros and the physical limit for Foveros is Full Wafer. It's impossible they couldn't fit two 100mm2 plus few other dies of around 30-40mm2 on a single wafer.

Also yes the larger cache is just brute forced die at least 30-35mm2 larger.
 
For latest reliable info about bLLC, Exist50 on reddit seems so sure that bLLC won't be on NovaLake, maybe until RazerLake.
He also doesn't seem very sure. I guess it's up to Intel how serious they are competing with ZenX3D chips. If they are able to put out at least an 8+16 core SKU with bLLC several months before Zen6X3D, they could take the gamer market by storm.
 
It's ridiculous that the SOC and GPU tiles are made with a high cost of 18A
At least the GPU tile, which has a lot of XE cores for mobile use, may use the most advanced nodes. I don't think a mere desktop GPU tile will use such a cutting edge process.
I think I'll probably use Intel 3
 
It's ridiculous that the SOC and GPU tiles are made with a high cost of 18A
It's not tbh if you are talking about tiles 60mm2 large and iGPU Tile will be roughly 20-30mm2 at such a small size it's not costly
They gotta eat the cost if they want to learn the ropes making these things in-house.
Intel Advanced Accounting™.
Even if it is made in-houseYou don't need to go out of your way to create a GPUTile for desktop use with cutting edge 18A
I think Intel 3 from a generation ago is enough
It's not like that they are reusing design from the Xe3P iGPU lol they won't have to make the iGPU on separate node.
 
Even if it is made in-houseYou don't need to go out of your way to create a GPUTile for desktop use with cutting edge 18A
I think Intel 3 from a generation ago is enough
I feel that it is important for Intel to do it for their NVL-HX chips. Intel does not have any new discrete GPU for laptops--they need something with good performance but low power. Plus, a tiny tile is perfect for breaking in a new node.

Then once the mobile chip has it, why redesign for the desktop chips which probably will be sold at a lower margin and lower volume?
 
Sorry, I'm not saying I'm designing the architecture from its location.
What I wanted to say is, regardless of architecture It was an opinion that there is no need to go out of your way to manufacture an IGPU for desktop with Intel 18A.
 
Sorry, I'm not saying I'm designing the architecture from its location.
What I wanted to say is, regardless of architecture It was an opinion that there is no need to go out of your way to manufacture an IGPU for desktop with Intel 18A.
They never went out of their way they shared so many things on 18A/18AP IP wise.
 
I've never heard of XE3P being an architecture (IP) dedicated to Intel 18A. That information is interesting
It comes down to a Raichu rumor. Then people extrapolate it to be 18A. The timing makes sense especially if 18A is not being as utilized as Intel hoped by 3rd parties. 18A would have 29% fewer masks and if the equipment is underutilized this is a win-win for salvaging the underutilized equipment and personnel. I suspect Intel would rather sell that equipment time, but this is their backup option.
 
Last edited:
Back
Top