Last Minute Telegrams: Coreteks!

Considering Suzanne Plummer's "Zen team" working on RDNA 2.0, how efficient do you think this architecture will be compared to RDNA 1.0? Would we be seeing RTG's Zen 2 moment this year or would this be similar to a hypothetical Zen 3 improvement over Zen?

The Immortal Cameraman

Hopefully not too late, but in his most recent video (posted today) Coreteks seemed to have a much more optimistic take on Intel's GPUs than many other tech tubers/bloggers. I'd just be curious to hear a bit more of his insight/take on Intel's future for discrete graphics.

MyNameIsCrease

With AMD seemingly starting to properly challenge PhysX with their own GPU real-time physics API, do you think they’d pull a GSync and actually bring it into a more open status rather than keeping it as an underused black box that performs terribly on other hardware?

Nils

Is it ever worth it to consider ECC RAM for a workstation computer?

Tommy

Aside from the CCX redesign, what other new features do you think will be in Zen 3?

Andrew Martin

Why are AMD’s APU’s called APU’s chip (CPU and GPU on one package) while Intel chips are called CPU’s even though there are graphics cores and central processing cores on the die? Is it down to marketing? I never could find a clear answer on the difference is on a chip level.

John DeLuca

I can see it being an add-in PCIE card, just like back in the days where you needed two graphics cards, one for 2D and one for 3D

John DeLuca

Is there a possibility of sound raytracing on the ps5? If so what would the benefits of sound raytracing be and what would be the computational cost? I would assume with the route Sony is going on for the next gen it would be much more likely they do that then Microsoft.

John DeLuca

AMD GPUs with it's driver issues

DeusNightFire

How likely would you think are dedicated ray tracing accelerator chips for gaming PCs, in addition to, say, a normal graphics card. Similar to ray tracing cores on Turing, but on a dedicated card/chip. I am not certain which factor is stronger: More integration and lower cost vs the performance requirements of ray tracing and the opportunity with multi-chip designs.

Have you ever wondered how we will look back on computing technology in, say, 10-15 years? For example, we look back at the early 2000s and the performance, size, graphical fidelity, etc. seems laughably (and obviously) bad from a retrospective point of view. What might we say in the future when looking back on Zen 2, Navi, Turing, (Ampere?) Sorry for the long-windedness.

Kaden Picknell

Q3: Do you actually think Intel can get it together for 11th gen, or will the 11900k be 12 cores at 500 watts?

Durmij

Q2: What are the odds the 4600 beats the 9900k in some gaming scenarios, seeing as the 3600 is within 5% in some titles?

Durmij

When the chains of the ATX form-factor are finally broken, will we see a future where Add-in Memory is placed closer to the processor; perhaps, as a downward extension of the CPU socket?

Bollocks

Q1: In your last video you mentioned Zen 3 is coming early 2021. Was that referring to the whole line up, or can we still expect the lower end skus late 2020?

Durmij

What do you guys think about the ML ASICs coming this year? Specifically the batshit, wafer scale integrated Cerebras chip? Vapourware or is Nvidia in trouble?

Edward Huff

You can speculate on the possibility of a mixed HBM/GDDR GPU, with say 8GB HBM and some GDDR used similarly to HBCC segment. As we think big RDNA will have both HBM and GDDR memory controllers, why not... use both? And thus work around the two tradeoffs - power/bandwidth and price/capacity.

KarbinCry