Question Zen 6 Speculation Thread

Page 82 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

Ghostsonplanets

Senior member
Mar 1, 2024
773
1,227
96
Would there be enough mem bandwidth from DDR5? It's unlikely that iGPU will have lots of "infinity cache"
RDNA 4 is more efficient perf/Gbps. But, also, 8CU/512 lanes are easily feed on DDR5

And the Zen 6 rumors also state memory controller will be vastly improved, with support for higher frequencies at 1:1 fclk.
 

Win2012R2

Senior member
Dec 5, 2024
792
795
96
RDNA 4 is more efficient perf/Gbps
With 64 MB cache and 645 GB/s from GDDR6 - and DDR5 will give what, 100+ GB/sec from two channels and cache size 16 MB at best?

It's certainly a LOT better than current iGPU, but would it be enough for 1080p gaming?
 

marees

Senior member
Apr 28, 2024
940
1,252
96
With 64 MB cache and 645 GB/s from GDDR6 - and DDR5 will give what, 100+ GB/sec from two channels and cache size 16 MB at best?

It's certainly a LOT better than current iGPU, but would it be enough for 1080p gaming?
Probably yes, with FSR 4 upscaling ?
 

OneEng2

Senior member
Sep 19, 2022
461
692
106
With 64 MB cache and 645 GB/s from GDDR6 - and DDR5 will give what, 100+ GB/sec from two channels and cache size 16 MB at best?

It's certainly a LOT better than current iGPU, but would it be enough for 1080p gaming?
Also, it isn't like all the bandwidth can go to the GPU either. Zen 6 is upping the core counts considerably, and likely each core will be capable of moving more memory than Zen 5.

I have worked in Automotive for decades. Combustion engines are really just air movers. If you want more power, you move more air through an engine. "Hopping up" an engine is done by a combination of pushing more air INTO the engine, Making the engine have more displacement (either more cylinders, or bigger ones), and opening up the exhaust.

If you think about it, processor architecture isn't that different.

If you raise the processing power, you have to feed it and get rid of the exhaust at the same rate is the execution engines use them or you become bottlenecked.

Sharing bandwidth with so many cores and a bigger GPU is going to be a big strain on DDR5 dual channel IMO. I suspect it will be the limiting factor to Zen 6 performance in many situations.
 

Win2012R2

Senior member
Dec 5, 2024
792
795
96
Probably yes, with FSR 4 upscaling ?
That would mean less compute needed per pixel, but you'd still have to move same data, unless you are dropping textures quality massively, probably geometry too - with smaller cache you'd need to fit more of them to compensate for much slower "v"RAM. With slow memory like this you need to have bigger caches to compensate, not much smaller.

I suspect it will be the limiting factor to Zen 6 performance in many situations.
If this iGPU is involved then for sure - it will suck all mem bandwidth easy.
 

eek2121

Diamond Member
Aug 2, 2005
3,274
4,820
136
You guys all make me laugh. I've never known someone that does DIY and doesn't game, and I know a bunch of DYIers. Thanks to Kepler for giving a few sources. It seems like there is a narrative being pushed that says "PC gaming is small, irrelevant, doesn't matter, and will disappear" when the opposite is true. A similar narrative about ARM vs. x86 is also constantly coming up (x86 has been dead since announced decades ago if some parties were to be believed. There are actual magazine covers claiming all this nonsense at least since the 90s, possibly before). AMD's constant peddling to gamers is pretty big evidence. X3D is a hot product, and so is the RX 9070 series. They can't make enough! It doesn't sound like much of a dying market to me! Maybe just chill on this type of talk until we see evidence. Shoot, "desktop is dying" is also a meme that his going on forever, yet I bet many of you have a desktop PC. What do you think folks are doing with all of those millions of GPUs sold each generation, putting them on Christmas trees? 🤣

At any rate (and back to the topic at hand) I don't care who makes the IO portion of the chip, I just hope AMD improves memory performance, and the next time we do a socket change, we NEED more PCIE lanes. Shoot, boards are so far ahead with bandwidth, I'd rather AMD stick with PCIE 5.0 for a great deal longer and give more lanes, vs. trying to move to PCIE6. GPUs don't scale much beyond 2.0, and NVME drives don't see much of a noticeable uplift. Just my thoughts.
 
Jul 27, 2020
23,462
16,510
146
I've never known someone that does DIY and doesn't game, and I know a bunch of DYIers.
Then maybe you don't know a lot of old people?

I would call myself a "game explorer" instead of a gamer coz I like to fire up games sometimes and look around and marvel at the graphics and environments. The last real game I played on PC was, oh my gosh, can't even remember properly coz work sucks all the energy out of me

Looking to get back in if I get my annual vacation (that is actually more than 10 days). Lost my last two annual vacations trying to be a bonafide C++ coder in the French 42 bootcamp (learned some interesting stuff. loved their examshell. but overall, I discovered that the French are really cruel and have no sympathy for people who can code but can't code up to the standard THEY think is required).
 

Markfw

Moderator Emeritus, Elite Member
May 16, 2002
26,800
15,810
136
Then maybe you don't know a lot of old people?

I would call myself a "game explorer" instead of a gamer coz I like to fire up games sometimes and look around and marvel at the graphics and environments. The last real game I played on PC was, oh my gosh, can't even remember properly coz work sucks all the energy out of me

Looking to get back in if I get my annual vacation (that is actually more than 10 days). Lost my last two annual vacations trying to be a bonafide C++ coder in the French 42 bootcamp (learned some interesting stuff. loved their examshell. but overall, I discovered that the French are really cruel and have no sympathy for people who can code but can't code up to the standard THEY think is required).
I game everyday, Dune 2000 from approx that year. But I don't call myself a gamer, as I have not played any of the current or ever semi-current games. C&C type games is all I have done in the last 25 years.
 
Jul 27, 2020
23,462
16,510
146
I game everyday, Dune 2000 from approx that year.
I don't have the IQ and patience for that kind of game

I don't do well with thinking patiently

The longest I've played a strategy game was this one: https://www.gog.com/en/game/theme_hospital

Coz it's so darn funny!

By the way, give Uplink a try.


May enjoy the first half hour (coz after that it gets too difficult.

And Portal 2, also a modern thinking game.

 
Jul 27, 2020
23,462
16,510
146
Only for interpreted languages!
I think they know something we don't. Maybe operating systems will increasingly become just a "support system" to launch a browser? I can certainly believe that because at home and at work, almost 90% of my clicks is clicking on browser tabs!
 

reaperrr3

Member
May 31, 2024
80
261
86
Also, it isn't like all the bandwidth can go to the GPU either. Zen 6 is upping the core counts considerably, and likely each core will be capable of moving more memory than Zen 5.
(...)
Sharing bandwidth with so many cores and a bigger GPU is going to be a big strain on DDR5 dual channel IMO. I suspect it will be the limiting factor to Zen 6 performance in many situations.
Not saying that bandwidth won't be an issue with Zen6 at all, but using bandwidth as argument to depict a bigger iGPU as something potentially negative feels misleading at best.

How often would both the iGPU and all CPU cores be under full load and bandwidth pressure at the same time?
In games, the iGPU's (lack of) raw power will be the bottleneck and the CPU part will idle around a lot.
Outside of games, the iGPU will likely do little work apart from accelerating video encode/decode.

Feel free to share any realistic workload scenarios where making the iGPU "bigger" (or in this case, less of a joke) would hurt CPU performance because it takes away its bandwidth. I'm genuinely curious.

but would it be enough for 1080p gaming?
Sounds an awful lot like "but can it run Crysis?"

It'll always massively depends on the specific games a person wants to play on this.
Some older games may run fine and even look decent enough, while some newer games may look and run horrible, no matter how much you lower settings.


Not that I think gaming is the primary reason for this upgrade.

If I were at AMD, I'd mostly
a) be worried MS may add some fancy useless shader effects to Win12 desktop that 2 CUs can't handle fluidly
b) want to finally take away that one remaining checkbox desktop feature advantage Intel had over the non-G desktop models from AMD.


...Although, maybe the new IOD was just pad-limited on the new process due to all the IO, so adding more CUs was virtually free because the alternative would've been some unused whitespace, who knows.
 

fastandfurious6

Senior member
Jun 1, 2024
429
583
96
You guys all make me laugh. I've never known someone that does DIY and doesn't game, and I know a bunch of DYIers. Thanks to Kepler for giving a few sources. It seems like there is a narrative being pushed that says "PC gaming is small, irrelevant, doesn't matter, and will disappear" when the opposite is true. A similar narrative about ARM vs. x86 is also constantly coming up (x86 has been dead since announced decades ago if some parties were to be believed. There are actual magazine covers claiming all this nonsense at least since the 90s, possibly before). AMD's constant peddling to gamers is pretty big evidence. X3D is a hot product, and so is the RX 9070 series. They can't make enough! It doesn't sound like much of a dying market to me! Maybe just chill on this type of talk until we see evidence. Shoot, "desktop is dying" is also a meme that his going on forever, yet I bet many of you have a desktop PC. What do you think folks are doing with all of those millions of GPUs sold each generation, putting them on Christmas trees? 🤣

At any rate (and back to the topic at hand) I don't care who makes the IO portion of the chip, I just hope AMD improves memory performance, and the next time we do a socket change, we NEED more PCIE lanes. Shoot, boards are so far ahead with bandwidth, I'd rather AMD stick with PCIE 5.0 for a great deal longer and give more lanes, vs. trying to move to PCIE6. GPUs don't scale much beyond 2.0, and NVME drives don't see much of a noticeable uplift. Just my thoughts.

correct me if I'm wrong but perf demand from games is plateauing isn't it

there has to be something radically new in games tech to demand a lot more perf

i.e. zen 6 X3D + RDNA 5 dgpu will likely handle all games ultra until 2030+ at least
 

soresu

Diamond Member
Dec 19, 2014
3,688
3,025
136
correct me if I'm wrong but perf demand from games is plateauing isn't it
In terms of the fact that 8K is pointless and there really isn't much to stress a GPU beyond path tracing then yes.

Point of fact it will be contracting as faster and higher quality perf hacks on real time PT are introduced - we're not out of low hanging fruit territory on the software side of things there yet (relative to raster which was fairly played out with deferred lighting years ago).
 

GTracing

Senior member
Aug 6, 2021
440
1,033
106
correct me if I'm wrong but perf demand from games is plateauing isn't it

there has to be something radically new in games tech to demand a lot more perf

i.e. zen 6 X3D + RDNA 5 dgpu will likely handle all games ultra until 2030+ at least
From what I've noticed, it's pretty normal for graphical demands to increase a lot in the first half a console generation, and then plateau in the second half (there are some notable exceptions of course). Games will get much more demanding after the PS6 comes out.
 
Reactions: Tlh97 and Kepler_L2

CakeMonster

Golden Member
Nov 22, 2012
1,600
779
136
Semi OT:
The next console generation both excites me and worries me. If we presume Z6 that will be a very welcome IPC increase, I'm not really worried about that. Hopefully more than 8c/16t as well, I'm counting on 12c/24t minimum. On the GPU side I feel a lot more iffy, whether tech currently quite far in the pipeline will be able to meet the AI/ML/RT needs of 2027-2034, can those GPU cores be beefy enough given AMD's current situation (or NV for that matter, just iteration since Ada)? I also fear for the (V)RAM capacity, it should be able to hold future AI models in memory along with improved textures, and there will probably be a lot of AI breakthroughs in that time period, and we just can't imagine all the potential use cases right now, so I fear being hobbled by 16GB or 24GB, hopefully they go for 32, 48, or even 64GB. History tells us that memory has been one of the most vulnerable specs.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |