• 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.

News Intel GPUs - we've given up on B770, where's Celestial already

Page 56 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Iris XE - 3200 Mhz vs. 4266 Mhz - RAM difference



These iGPUs need faster RAM.




Looks like DG2 (supposedly) supports AV1 encode upto 8192x8192 res.


Also the DG2-128 variant apparently.

 
Iris XE - 3200 Mhz vs. 4266 Mhz - RAM difference



These iGPUs need faster RAM.

Yea... no.

The linear to superlinear scaling flies in the face of proper tests such as from Techspot. It suggest it's bound 100% by memory bandwidth. Random Youtube reviews can't be taken seriously.

One is also TGL-H and the other is TGL-U.

This is proper testing: https://www.techspot.com/review/2235-amd-ryzen-5800u/

And it shows in some games the big difference in 15W configurations between two RAM setups almost disappear at 28W. And it doesn't go much above 20%.

The 28W DDR4 setup outperforms 15W LPDDR4x setup in all tests. How would that be possible if Tigerlake is 100% bound by memory bandwidth?
 
Yea... no.

The linear to superlinear scaling flies in the face of proper tests such as from Techspot. It suggest it's bound 100% by memory bandwidth. Random Youtube reviews can't be taken seriously.


The test is fine, he is using a soldered RAM DDR4 device which is quite a bit slower than a two slot DDR4 x8 device.

This is proper testing: https://www.techspot.com/review/2235-amd-ryzen-5800u/

And it shows in some games the big difference in 15W configurations between two RAM setups almost disappear at 28W. And it doesn't go much above 20%.

The 28W DDR4 setup outperforms 15W LPDDR4x setup in all tests. How would that be possible if Tigerlake is 100% bound by memory bandwidth?


There is no real device overview, what Tigerlake devices and DDR4 configuration do they use? I don't see it. A proper test should do better in this regards. I mean if it's a dual rank DDR4 device (32GB) it will be faster than usual. Dual rank is quite a bit faster than single rank, the gap to LPDDR4x will be much lower than usual in this case. It requires 32GB though, not many devices have this amount.

Only 3 games by the way and GTA V 25W looks odd, the gap to ICL-U and CML-U is oddly low compared to TGL/Cezanne. This is a very limited iGPU test. Reviews from big tech sites are not always great or even trustworthy, check out the newest DDR5 test from anandtech, it's pathethic.
 
Acer Swift X 16-inch is the first ARC announcement.

Intel says 50 design wins. Death Stranding supports XeSS plus few other companies announced support for it.

Variant of Alienware X17 and Lenovo Yoga 7i will be among the high end with ARC.
 
Last edited:
Seems sort of funny people were saying “no way intel makes 1st quarter”, then intel does not meet the goal and people are surprised/pissed about it.
Whatever, earlier would be great but regardless it is more inventory for the market.
 
DG1 is horrible. Not the performance as much as they didn't put work into making a vBIOS chip for it. It's also an opportunity lost for them as people would have took it and used it for other systems.
 
Surely mean creatures?
Undocumented creatures with an "anthill inside":
AFAIK the forum does now do timestamps, but anyway the "Ant Hill Inside" and "Undocumented Creatures" gag starts at 10:50.
That's a DOS game?!

Holly Molly, I can't believe I missed out on something this cool when I was young. I did try to run a *ahem* altered version of Discworld 2 with videos and voices stripped out but it didn't run for some reason (maybe expanded memory related). Good thing they released it on PS1. ePSXe here I come! 😀
 
Back
Top