It's not a legit slide deck from AMD until you find embarrassing typos.
-
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.
Discussion RDNA 5 / UDNA (CDNA Next) speculation
Page 15 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
It's not a legit slide deck from AMD until you find embarrassing typos.
Ah fair Verona!
https://wccftech.com/next-gen-hbm-a...gb-capacity-per-24-hi-stack-embedded-cooling/
Looking at the details for HBM4, the memory aims to offer a data rate of 8 Gbps with a 2048-bit IO, 2.0 TB/s of memory bandwidth per stack, 24 Gb capacity per die for up to 36-48 GB memory capacity, and a per-stack power package of 75W. HBM4 goes with the standard direct-to-chip (D2C) Liquid cooling and employs a custom HBM-based die (HBM-LPDDR).
Interestingly, AMD's Instinct MI400, which launches next year too, takes things up a notch versus Rubin and is aiming to offer 432 GB of HBM4 capacity with memory bandwidth of up to 19.6 TB/s.
Looking at the details for HBM4, the memory aims to offer a data rate of 8 Gbps with a 2048-bit IO, 2.0 TB/s of memory bandwidth per stack, 24 Gb capacity per die for up to 36-48 GB memory capacity, and a per-stack power package of 75W. HBM4 goes with the standard direct-to-chip (D2C) Liquid cooling and employs a custom HBM-based die (HBM-LPDDR).
Interestingly, AMD's Instinct MI400, which launches next year too, takes things up a notch versus Rubin and is aiming to offer 432 GB of HBM4 capacity with memory bandwidth of up to 19.6 TB/s.
Attachments
Last edited:
Embedded cooling, aka DARPA ICEcool.
branch_suggestion
Senior member
You know, AMD could change their client GPU roadmap to something like 2 mainstream parts every 2 years or so, and 1 halo tier part every 3-4 years.
Surely they can occasionally pull people off the hellscape that is GPGPU.
Surely they can occasionally pull people off the hellscape that is GPGPU.
adroc_thurston
Diamond Member
Lmao no, yearly shader cores in GPGPU (new or derived) await.
ToTTenTranz
Golden Member
I don't get this.
On one side we have leakers claiming AMD's mobile APUs will start watering down their iGPUs because the market isn't interesting enough. On the other side we see Microsoft claiming their next gaming consoles are basically Windows handhelds running on present and future mobile APUs from AMD.
Which one is it?
adroc_thurston
Diamond Member
They aren't.
I can explain this. Semicustom parts with high volume vs off the shelf parts with low volume.
You misunderstand the announcement.
poke01
Diamond Member
Can Intel match this?
I mean, console wise they already were.
The last Xbox with nVidia gfx was the OG Xbox, and the last Xbox with a non AMD/x86 CPU was Xbox 360.
Even without next gen or the rest of this gen that's already 12 years of consistent partnership on consoles, 2.5 generations worth when you count the XB1 mid cycle.
Honestly I'm surprised that they aren't more in sync on the Surface side of things.
marees
Platinum Member
Surface is like a company of its own.
They partner only with intel/Qualcomm
The danger for AMD was that it was the surface team which was working on the first party handheld
Now that Verge's tom warren says that the surface handheld is essentially cancelled, AMD can breathe easy
marees
Platinum Member
Both scheduled for H2 next year
SolidQ
Golden Member
Xbox Series X was released 10th november.
6800XT(RDNA2 uarch) - 18th november
So i assume RDNA5/Xbox next gonna be same route
Actually that's pretty par for the course.
The first PC APU with RDNA2 (Rembrandt IIRC) was announced in Jan 2022, over a year after PS5 and XSX/XSS launched in 2020.
Then Steam Deck with Van Gogh/Aerith was announced just a month later in Feb 2022.
Oh ye RDNA5/UDNA/woteva dGPUs will likely launch pretty close to at least one console.
Be interesting to see where Valve lands in terms of what semi custom stuff they have next on their roadmap with AMD.
adroc_thurston
Diamond Member
None.
Valve got told to pound sand.
For what reason?
Does Steam Deck not sell well enough to justify the VG production and AMD/TSMC are bitter about this?
Or is Valve delinquent on paying chip royalties?
adroc_thurston
Diamond Member
Pathetic volumes.
Van Gogh was made for that dual screen Surface dingus that got canned.
'eck used it as a salvage.
poke01
Diamond Member
Steam Deck in terms of volume is sad. Even the pathetic Wii u sold more
Update on HIP RT/RDNA4 performance, I scoured the GitHub repo and found this:
github.com
Evidently they are likely updating it to the nu HIP 7, with optimisations likely to be included along the way.
Question about current RDNA4 performance · Issue #35 · GPUOpen-LibrariesAndSDKs/HIPRT
Hello, we did some tests on blender 4.4 on w7900 and 9070xt card. In barbershop scene, w7900 got 80 seconds and 9070xt got 87 seconds. So can this result reflect actual hardware performance? Or HIP...
github.com
Evidently they are likely updating it to the nu HIP 7, with optimisations likely to be included along the way.
Last edited:
marees
Platinum Member
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 25K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 24K
-
-
