Discussion ARM Cortex/Neoverse IP + SoCs (no custom cores) Discussion

Page 57 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

mvprod123

Senior member
Jun 22, 2024
399
456
96
And it's out... heck, the X4 is interesting. But again Google has to choose an inferior GPU this time.

Why Google who can choose something good, makes the screwup to make an awul desicion?

Meanwhile... no announcements of future cores.
There is a suspicion that a driver supporting Vulkan 1.4 does not yet exist or is still in development. The Imagination website states that support for DXTP-48-1536 is only planned for 2025. 1756137586853.png
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136

Seeing this.... this might be the next GPU from Google after Tensor G5:
View attachment 129237
They must really be impressed with IMG TEC's numbers if they are that set on shifting away from Mali when it's finally doing well against Adreno.

As to "OGL 4.6 via Zink" I'll believe it when I see it - you need a fair chunk of VK extensions to make Zink work somewhat performantly.

Given they aren't even advertising VK 1.4 support nearly 8 months into the year they promised to deliver it I have my doubts.
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136
heck, the X4 is interesting
2+ years old at this point tho.
Meanwhile... no announcements of future cores.
There is an "ARM unlocked" event on September 10th, and an AI Infrastructure event on Sept 9th-11th, either of which might be the IP announcement bonanza.

I'd be very surprised if they leave it any longer and risk Qualcomm announcing Oryon3 first.
 
  • Like
Reactions: 511

DZero

Golden Member
Jun 20, 2024
1,625
629
96
They must really be impressed with IMG TEC's numbers if they are that set on shifting away from Mali when it's finally doing well against Adreno.

As to "OGL 4.6 via Zink" I'll believe it when I see it - you need a fair chunk of VK extensions to make Zink work somewhat performantly.

Given they aren't even advertising VK 1.4 support nearly 8 months into the year they promised to deliver it I have my doubts.
Unless Google is paying a LOT to both GPUs and allows to Imagination to update their drivers.

Mali might step up now since there will be support with Linux in a future and that could open the way to see a driver for Windows. Something Imagination lacks
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136
Mali might step up now since there will be support with Linux in a future
PanVK already has 1.4 compliance for Mali v10+ (v10 = G710) according to mesamatrix.

API standard compliance =/= optimisation though, so I wouldn't expect great performance for the time being.

The better VK driver efforts seem to have their own special compiler like ACO for RADV and NAK for NVK, so hopefully we might see a similar effort for PanVK and/or Panfrost in the future.

Panfrost OGL support is much patchier, especially compared to Asahi (Apple M1/2 GPU) oddly enough.

It doesn't even yet have full OGL 3.2 compliance (geometry shader support still lacking), though it does have compliance for all 3.3 extensions, and for 4.0+:

4.0 = 48%%, 4.1 = 50%, 4.2 = 83%, 4.3 = 91%, 4.4 = 93%, 4.5 = 58% and 4.6 = 36%.

If the Panfrost effort had put as much time in as for Asahi we might already have 4.6 compliance, but as it is sadly not.

Zink will give some relief for OGL problems, if not very performantly for the time being until PanVK gets more optimisation TLC.

There is another driver effort called Tyr underway built on Rust language code, but I think that is still pretty early in its development.
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136
and they have confirmed if anyone had doubts that MediaTek is doing the CPU chiplet
I would assume probably to get their wifi/bt/5G modems.

Unless they are really that bad at non GPU silicon design even when most of the work is already done for them by ARM and the fab.

Qualcomm did flail somewhat with SD 810, so perhaps even with all that help there is still some assembly required that MTK is simply better suited to do, especially when multiple different radio PHYs are added to the equation requiring extra expertise.
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136
RIP Google. They went away from Exynos and now Exynos is recovering big time.
TSMC won't fab Exynos, even though Samsung have asked.

This puts it at a significant disadvantage unless Sammy can finally match TSMC again.
 
  • Like
Reactions: 511

DZero

Golden Member
Jun 20, 2024
1,625
629
96
TSMC won't fab Exynos, even though Samsung have asked.

This puts it at a significant disadvantage unless Sammy can finally match TSMC again.
I mean to Google. Google jumped from Samsung since it wasn't good.
Now Samsung is comming back and Google might start to regret that desicion.
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136
I mean to Google. Google jumped from Samsung since it wasn't good.
Now Samsung is comming back and Google might start to regret that desicion.
That's a mischaracterisation.

They didn't jump from Samsung so much as shift from a semi custom design collaboration with Samsung to a fully in house custom design.

Given the tools at their disposal have served them pretty well (allegedly at least) for their TPU and VCU (Argos) hardware I definitely would not bet against them in the long term, even if they happen to stumble a bit in the near term for their first fully custom SoC designs.

The decision to go with X4 rather than X925 tells me that they have chosen to be cautious for this generation, but going forward they may well shift to a more hi end cutting edge product.
 

soresu

Diamond Member
Dec 19, 2014
4,105
3,566
136
If ARM doesn't announce their 2025 CPU/GPU IP this week I'm going to start getting worried.
 

DZero

Golden Member
Jun 20, 2024
1,625
629
96
If ARM doesn't announce their 2025 CPU/GPU IP this week I'm going to start getting worried.
I'll take this month, the issue are that MTK and other are ready to announce their new processors bassed on unannounced cores.
 
  • Like
Reactions: soresu

gdansk

Diamond Member
Feb 8, 2011
4,568
7,682
136
The C1-SME2 unit is implemented inside a C1-DSU cluster. It is a shared unit between all the cores and is accessible through the C1-DynamIQ Shared Unit (DSU), that behaves as a full interconnect with shared L3 support and full coherency between the cores and the C1-SME2 unit.
Lame. So that big GB6 increase ARM is advertising is not representative of the C1U(X930?) itself.
 
Last edited: