Reader Mail for End of 2024 Loose Ends!
Added 2025-01-02 20:20:21 +0000 UTCWe've just gotten back from Christmas break...and WOW there's been A LOT of news dropping over the past couple weeks!!! As such, Tom will be conducting a livestream going over all of the recent news (which will include NEW info received from sources over break), and of course he will also answer your questions LIVE!
Yes, there will be A LOT of RDNA 4 & Blackwell info divulged/confirmed, but please feel free to give input and ask about ANYTHING that comes to mind from recent news...or upcoming launches! You have at least 12 hours to submit! (EDIT: Livestream will be Friday Evening)
Comments
Since the 5000 series will use a more efficient node instead of jumping to 3nm, do you think the 6000 series will use 3nm or leap to 2nm?
Stoff
2025-01-04 00:56:31 +0000 UTCDid you happen to hear any news of the Switch 2 release timing? It's clear sometime this year, but very open in the air given still no official announcement.
Trevor Renfro
2025-01-04 00:23:33 +0000 UTCAny news on FSR4 and whether it will be ready for launch? Is it still software based or will RDNA4 have some hardware it can use.
Sad XTX 999
2025-01-03 23:44:55 +0000 UTCHi Tom it’s been a year, hope all is well. Curious if you have any sales info on Intels latest and greatest CPU’s!
Sad XTX 999
2025-01-03 23:42:44 +0000 UTCJust give us the XTXTXT (add as many X's and T's as you think proper) a few months down the line with 3GB ram modules.
Lucas
2025-01-03 22:16:33 +0000 UTCHappy new year Tom! Obviously asking before the new leaks but is there a chance that AMD release RDNA 4 before NVIDIA release their cards? It feels like AMD always has been waiting to see what NVIDIA does first before they make a move. If AMD truly does have a killer mid-range product, then they should release it first with conviction regardless of competition.
MultiNati
2025-01-03 13:14:19 +0000 UTCHi Tom, any updates on Navi 48 with more than 16GB of VRAM? In my opinion a 24GB or 32GB Flagship would make sense in 2025, even if it's just to win over AI/compute prosumers.
KingKoro
2025-01-03 06:48:57 +0000 UTCAre there any info or estimates on compute performance for full Navi 48 compared to the 4090? Like, for scientific compute
Allen Chen
2025-01-03 02:24:50 +0000 UTCAre there final confirmations for if AM5 will have dual v-cache CCD Zen 5 options? How about all v-cache CCD Zen 5 options for TRX50
Allen Chen
2025-01-03 02:22:31 +0000 UTCWill AM5 become like AM4, or will it not support as many generations?
Mark Smith
2025-01-03 01:45:00 +0000 UTCAre you expecting International Pricing outside the US to be as usual (aka direct conversion + sales tax + 5% or so), or do you think there will be a larger difference than usual due to America having Tariffs, if so will we see international pricing increase, or be lower?
TechnoLadz
2025-01-03 01:10:53 +0000 UTCRTX 4090 owners who have a ATX 3.0 power supply with native 12VHPWR cable, should we be upgrading to ATX 3.1 for the RTX 5090?
RisingMiles
2025-01-02 23:58:56 +0000 UTCWith all the Switch 2 news in such a short time have you received any new information or confirmations of recent leaks?
Planet_9
2025-01-02 23:33:23 +0000 UTCWhat are you hearing about strix Halo adoption? The Asus leak made it look like only the z13 was getting it and there's been one leaked HP model. Given that strix point and zen 5 x3d are both out, id have thought this would have the most buzz.
B. Fish
2025-01-02 23:32:34 +0000 UTCIs Intel foundry getting any orders from the likes of Amazon/Meta etc?
Sakosha
2025-01-02 22:53:34 +0000 UTCI am really curious if you still have Intel contacts that have any info on what they plan to do with their smaller CPUs. My favorite post-12 Gen product is hands down the N100/N200 chip, but despite there already being TWO of the new Core Ultra generations out and still no sign of replacements for those.
Wolfie
2025-01-02 22:51:32 +0000 UTCDo you plan on buying into RTX 50 or RDNA 4 or is it too soon for an upgrade for you?
coladict
2025-01-02 22:26:03 +0000 UTCHey Tom. Who will make an apu first that can achieve effective partity with 8 core and 7900XTX in ultra settings gaming? AMD or Nvidia?
BaconHouse Collective
2025-01-02 21:59:16 +0000 UTCDo you see leading edge process nodes getting too crowded or expensive that nvidia and amd hold consumer chips back on older nodes ?
shea
2025-01-02 21:34:17 +0000 UTCHi Tom, hope you and the team had a good holiday season. Based on current information, do you think 4090s will hold their value on the used market relative to the other 40 series cards? Depending on 5090 reviews, I may want to upgrade, so when would be the best time to sell my 4090? I may hand it down to the kids, but just wanting your take.
Gob
2025-01-02 21:28:34 +0000 UTCIf it is true that RDNA 4 adopts the 9070 name I think that is a very good thing. The exaggerated names of the 7000 series damaged the brand. A 70 class name should restore confidence in the naming scheme.
ProfNexGen
2025-01-02 21:19:37 +0000 UTCFinal performance estimates on RDNA 4? Will the (9?)800xt still be around 7900xt performance? Or has it gotten closer to 7900xtx?
FloridaMan
2025-01-02 21:04:58 +0000 UTCHi Tom! I am looking to get the RTX 5080 Founder's Edition. The leaked specs show the same dimensions, or at least length, as the 4080. Is this just a placeholder value or is it indeed the same?
Craig W.
2025-01-02 20:54:16 +0000 UTCDo you know how large the 5090 founders edition will be? Will it be roughly the same as a 4090 or will it be longer and/or take up more slots?
Frobozz
2025-01-02 20:47:57 +0000 UTCIs there anything to indicate that Nvidia is concerned at all for the low end and/or mid range GPUs due to RDNA 4 and Battlemage?
CompressedAIBlocks
2025-01-02 20:46:34 +0000 UTCDo you think it was a mistake that AMD did not use AI in their GPU naming scheme? /s
CompressedAIBlocks
2025-01-02 20:43:42 +0000 UTCDo you think gamers will accept 8GB VRAM in 2025? I think it's finally hit the mainstream gamer that 8GB VRAM is no longer acceptable but we're still getting 8GB 60 class cards from Nvidia and AMD this generation and presumably next generation (unless AMD goes full in with 16GB on 128 bit GDDR6 for N44).
CompressedAIBlocks
2025-01-02 20:42:21 +0000 UTCWhy did AMD think it was a good idea to completely overhaul their GPU numbering scheme for this generation? It would've made sense for their first UDNA generation, but it's only slightly easier for normies to compare against Nvidia.
Cleansweep
2025-01-02 20:36:23 +0000 UTCI've been saying since Lunar Lake was first announced that it would be perfect for handheld PCs. It's the first Intel lineup in a long time that has genuinely interested and excited me. The review embargo for the 2025 MSI Claw with Lunar Lake just lifted and independent reviews have shown huge increases in performance and efficiency over not only Meteor Lake, but also Hawk Point and Phoenix in these power-constrained handheld PCs. The graphics is competitive with the 890M in Strix, but it has fewer and less-powerful CPU cores so the GPU is never starving for power. Does AMD have anything releasing in 2025 that should take a similar design approach to Lunar Lake for handhelds and ultra thin laptops? Strix is a great lineup, but it definitely focuses a lot more on CPU performance than it does balancing out the CPU and GPU performance for lower power budgets, and it doesn't perform extremely well when you're only giving it 10-30 watts, while Lunar Lake handles that without breaking a sweat.
Question Generated By JensenGPT
2025-01-02 20:35:48 +0000 UTCGood day Tom, have you seen the review for the MSI Claw 8 AI + from the Phawx. As device it will be likely the best Handheld Windows Gaming device if there weren't the Battlemage Driver problems in mostly DX11 games. Will it force AMD to optimize there Drivers further for Handheld Gaming devices?
AI-Soldier789
2025-01-02 20:25:32 +0000 UTCAny updates on RDNA4 pricing? Is it possible that AMD lowers their previously planned price because of Intel Battlemage?
dietzi96
2025-01-02 20:24:08 +0000 UTCHello Tom. How did Intel's fixes manage to worsen the performance of arrow lake even more? Maybe Intel has some to-do list of things it can still fuck up. But from management standpoint how does this even happen?
QuickJumper
2025-01-02 20:23:41 +0000 UTC
