Friday, December 6th 2024

Intel Xe3 "Celestial" Architecture is Complete, Hardware Team Moves on to Xe4 "Druid" Design

We have already confirmed that Intel is continuing the development of Arc gaming GPUs beyond the current Xe2 "Battlemage" series, with the new Xe3 "Celestial" architecture in the works. However, thanks to PCWorld's The Full Nerd podcast, Tom Petersen of Intel confirmed that the Xe3 IP has been finished, and the hardware teams are already working on the next Xe4 "Druid" GPU IP. "Our architects are way ahead of us, and they are already working on not the next thing but the next thing after the next thing," said Petersen, adding: "The way I would like to comment is our IP that's kind of called Xe3, which is the one after Xe2, that's pretty much baked, right. And so the software teams have a lot of work to do on Xe3. The hardware teams are off on the next thing, right. That's our cadence, that we need to keep going."

The base IP of next-generation Xe3 "Celestial" GPUs is done. That means the basic media engines, Xe cores, XMX matrix engines, ray tracing engines, and other parts of the gaming GPU are already designed and most likely awaiting trial fabrication. The software to support this Xe3 is also being developed while Intel's team is working on enabling more optimizations for the Xe2 "Battlemage" architecture, which we previewed recently. We assume that Intel's Xe GPU will now follow a stricter cadence of releases, with SKUs getting updated much faster, given that a lot is prepared for the future.
Source: The Full Nerd Podcast (YouTube)
Add your own comment

36 Comments on Intel Xe3 "Celestial" Architecture is Complete, Hardware Team Moves on to Xe4 "Druid" Design

#26
Tek-Check
AusWolfSure, but bullshit sells products, let's be honest, because people are stupid.
You have low opinion of humanity
Posted on Reply
#27
tfp
Tek-CheckYou have low opinion of humanity
He has a realistic opinion of humanity.
Posted on Reply
#28
kondamin
Vya DomusWho the hell knows what "enthusiast" even means let alone what Intel thinks it means.
You need to be enthusiastic about exotic hardware that might be somewhat useful to purchase one of them?

an Other example why it’s not worth your time listening to rumour YouTubers
Posted on Reply
#29
Neo_Morpheus
Tek-CheckYou have low opinion of humanity
He is on the money.

Intel doesn’t deserve any mercy for all the illegal crap they pulled against AMD.

Their arrogant attitude towards their own customers.

They have lied over and over and yet, here we are, all their fanbois still giving them unwavering support.

Yeah, he is right.
Posted on Reply
#30
AusWolf
Tek-CheckYou have low opinion of humanity
I know there are decent individuals here and there, but in general, yes, I do.
Posted on Reply
#32
AggressiveCombo
JomaleReality:
Intel's 18A process, a node that is said to be a "turning point" for Intel Foundry, is now being said to feature only <10% yield rates.
And then Pat Gelsinger says he's very proud of the team that worked on it
Posted on Reply
#33
lexluthermiester
AggressiveComboAnd then Pat Gelsinger says he's very proud of the team that worked on it
That's because they've done something new for Intel and competitive to TSMC. Low yields are a fact of life for ANY new lithography process run. Everyone goes through that process, including TSMC. For example, TSMC's 3nm process started with a sub 5% yield. Yet they've improve as they refine and dial in the process. TADA!
Posted on Reply
#34
Tek-Check
lexluthermiesterThat's because they've done something new for Intel and competitive to TSMC. Low yields are a fact of life for ANY new lithography process run. Everyone goes through that process, including TSMC. For example, TSMC's 3nm process started with a sub 5% yield. Yet they've improve as they refine and dial in the process. TADA!
It seems that N2 trial run had already hit around 60% and then it was further boosted by 6% recently, while at the same time enabling smaller SRAM cells, as stated by Tom's. This suggests that their initial GAA nanosheet runs are far better than initial N3 attempts.

www.tomshardware.com/tech-industry/tsmc-engineer-boasts-of-recent-6-percent-boost-to-2nm-yields-passing-billions-in-savings-to-customers
Posted on Reply
#35
lexluthermiester
Tek-CheckIt seems that N2 trial run had already hit around 60% and then it was further boosted by 6% recently
Perhaps, but it didn't start at that yield level. 66% yield is still sub-par for production runs. Usable, but not optimal.
Posted on Reply
#36
Tek-Check
lexluthermiesterPerhaps, but it didn't start at that yield level. 66% yield is still sub-par for production runs. Usable, but not optimal.
True, but they have almost a year to fine tune it. Apple also used similar yields for initial M3 chips. Samsung's 8nm wasn't much better either, but it was cheap enough for Nvidia to bake Ampere on it.

I am sure that Apple, Nvidia, AMD and others would proceed with N2 even on cureent 66%. It will mature further. I'd expect at least 75% for larger chips and almost 80% for smaller chips by the moment of mass production.
Posted on Reply
Add your own comment
Dec 13th, 2024 01:54 EST change timezone

New Forum Posts

Popular Reviews

Controversial News Posts