Reader Mail for August Loose Ends

high tom, any news about MRDNA (samsung collaboration) and what do you think will be the future of AMD in the ARM space.

How likely do you think it is that Ethereum pushes back the transition to Proof of Stake?

Will Arc Xe-HPG include a software stack for Machine Learning and how it will compare the performance of the hardware compared to NVIDIA (in the machine learning field)

Valko Milev

Tom, have you heard anything about Infinity Fabric on Zen 4? With DDR5 on the way, I wonder if they will finally increase the physical links and run at 2:1, because the frequency of IF hasn't improved much gen over gen.

Stu

Tom , I really do love what you do and enjoy your channel. It's why I subscribe as everyone else should on Patreon. You recently made a comment about angry members who say and I quote "me angry...." . I think what you are hearing is frustration regarding the state of the PC market place, supply and demand regarding GPU'S and the absolutely bonkers pricing. It doesn't help when tech tuners like Linus says things like high GPU prices are our(consumer's) fault.

Dr Forbin

Which quarter do you think they will release Nvidia GPUs next year?

Falto

Alder Lake platform has been stated to support the 12V.O. standard, will people will be able to build 12V.O. systems at launch or do you expect delays on the new standards?

Rootknight

Hi Tom. People assume that upcoming Intel GPUs will be cheap, or at least cheaper compared to current AMD and Nvidia prices, but can we be really sure of that? Intel was never afraid to use overly high prices even when having worse products next to competition, and we also still have some components shortages and AIBs ready to jack up the price as high as the market allows them to. Do you think there's anything that can "guarantee" that Intel GPUs will actually be cheap and have stock that will match the demand?

Tom, when will we have GDDR7?

Stu

Do you think that future lower power variant cpu's will primarily focus more on the effiecny of the little core in intel's Big.Little? As in the big cores will still be highly clocked (max clock) and the little cores will be the focus, and or be manipulated in the clock speeds. Also, will this sort of power effiecny carry over to gaming laptops, making them far more competitive as daily drivers for work?

nathan moes

I'm wondering if Dan thinks his GPU will last him for 3 to 4 years before needing an upgrade?

matthew saltzberg

I know you've been heavy on Intel lately, with desktop alder lake hitting soon, do you see laptop alder lake & Rembrandt hitting at CES or do you think Rembrandt will come later in the year?

B. Fish

Now Nvidia (And AMD) have got away with extremely high prices for the last 2 years along with the pricing rumours for next gen gpu's is it possible nvidia especially will never go back? Instead they follow the Apple business practice-can't afford a £2000 card from us? Too bad you penniless peasant Keep up the good work, best few £ I spend a month being a patreon

Alex

Takeaways from this year's Hot Chips? What impressed you most, or least? Do you feel the frequency of "Moore's law isn't dead" statements has decreased, year over year? Or maybe, what was the buzzword of the Hot Chips 33?

KarbinCry

First question: amd drop today came and went, no luck in 30mins (tho 6600xt wasnt there to begin with)... What will september, october look, both in actual products and news cycle perspective(referring to mention of august being heavy in news content)?

Timo H

As a college student who just got released from covid jail I've gotta ask, what was your favorite alcoholic beverage when you were in college, any recommendations for someone on a budget?

Chance of multi GPU coming back with Intel in the market now or is the 300-600w GPUs replacing multi GPU? Any update on the node (Samsung or TSMC) Nvidia will be using for the RTX 4000? I have a bit of a hard time seeing Nvidia going back to TSMC unless they absolutely have to, given how many times they have burned TSMC throughout the years. Will Nvidia be using HMB to get the necessary bandwidth or doubling down on G6X?

Benjamin Cannon