AMD Engineer confirms that RDNA 3 will make use of 5nm and 6nm silicon
AMD's high-end RDNA 3 products will use a hybrid architecture
Published: 4th February 2022 | Source: LinkedIn - Via @blueisviolet |
AMD Engineer confirms RDNA 3's hybrid lithography design
AMD has confirmed that new Radeon graphics cards based on the company's RDNA 3 architecture would be launching this year, succeeding their RDNA 2 product line with new enhancements and more performance. That said, AMD has said little about what RDNA 3 will offer, or how it improves upon RDNA 2's architectural achievements.
Thanks to an AMD engineer on LinkedIn (as spotted by @blueisviolet on Twitter), AMD's RDNA 3 based products will be built using a mix of 5nm and 6nm silicon. Three GPUs are listed, AMD's Navi 31, Navi 32 and Navi 33 silicon. Navi 31/32 appear to be built using 5nm and 6nm silicon, while Navi 33 appears to only use 6nm silicon. You can see this information on the screenshot below.
These data points suggest that a long-running rumour about AMD's RDNA 3 architecture is correct. It looks like AMD's high-end RDNA 3 products use a chiplet-based design, mixing silicon from various lithography nodes to create an ultra-high-end final product. It also looks like Navi 33 will use a monolithic design.
At this time, it is unknown how AMD will be using TSMC's 6nm and 5nm lithography nodes to create their RDNA 3 products. It looks like AMD will be using 6nm for their primary GPU silicon for Navi 33, with 5nm being used exclusively in their chiplet-based products. This suggests that 5nm chips will be used as a silicon bridge/inter-connect chip.
Currently, AMD has revealed nothing of substance about RDNA 3. That said, AMD knows the weaknesses of RDNA 2 (primarily ray tracing performance), and knows that a successfully implemented MCM gaming graphics card has the potential to deliver AMD staggering performance gains over rival silicon.
Using an MCM GPU design, AMD could allow two GPUs to act as a single, larger, chip, potentially enabling a near 2x performance boost over a single-chip design. If AMD is successful, their RDNA 3 graphics cards will deliver staggering performance levels.
You can join the discussion on the AMD engineer that blabbed about RDNA 3 on the OC3D Forums.
Most Recent Comments
|
Originally Posted by NeverBackDown
Don't care how it's done I just want comparable performance to Nvidia's next stuff (as they'll launch around the same time) in all areas of GPU performance that's around these days, rasterization, RT, AI upscaling, etc.
|
Navi was supposed to be "Scaleable" from day one. Meaning basically? Ryzen on a GPU. More than one core cluster. It wasn't. Nvidia will be doing it too, just seems a case of who will do it first now.
It's in everyone's interest. Small dies are much cheaper and far more successful when it comes to producing them by % so hopefully this will take the sting away from the stupid great tank dies that cost a fortune.Quote
|
Originally Posted by AlienALX
I think that may be why they are combining chips.
Navi was supposed to be "Scaleable" from day one. Meaning basically? Ryzen on a GPU. More than one core cluster. It wasn't. Nvidia will be doing it too, just seems a case of who will do it first now. It's in everyone's interest. Small dies are much cheaper and far more successful when it comes to producing them by % so hopefully this will take the sting away from the stupid great tank dies that cost a fortune. |
|
Originally Posted by AngryGoldfish
That's all well and good, but all the rumours point to the multi-chip design of RDNA3 being only for the super expensive stuff, namely Navi 31 and 32. The main benefits of Ryzen was that we could get 4 cores for the price of a 2 core, 6 core CPUs for the price of a 4 core, an 8 core for the price of a 6 core, a 12 core for the price of an 8 core, etc. RDNA3 however will be a 64 core at the price of a 48 core. The performance jump could be huge, but so will the price. We won't see the 'Ryzen of GPUs' for at least another generation. RDNA3 chiplet design is more like Threadripper, or even EPYC, than Ryzen.
|
This won't settle down properly until Intel have their own fabs able to make their own GPUs. What did they say? 2025? ICR.
The problem is that both AMD and Nvidia wait in a queue. And that queue has been of epic proportions.
I think it's high time Nvidia made their own fabs. AMD? don't have the capital. Nvidia however do.Quote
|
Originally Posted by AngryGoldfish
That's all well and good, but all the rumours point to the multi-chip design of RDNA3 being only for the super expensive stuff, namely Navi 31 and 32. The main benefits of Ryzen was that we could get 4 cores for the price of a 2 core, 6 core CPUs for the price of a 4 core, an 8 core for the price of a 6 core, a 12 core for the price of an 8 core, etc. RDNA3 however will be a 64 core at the price of a 48 core. The performance jump could be huge, but so will the price. We won't see the 'Ryzen of GPUs' for at least another generation. RDNA3 chiplet design is more like Threadripper, or even EPYC, than Ryzen.
|
I see no reason why we should compare CPU architecture to GPU architecture in terms of MCM design. Sure it's the only other available MCM design introduced in the high performance field, but they require quite different designs. For all we know they could just have Normal RDNA cores separated from the AI/RT cores connected by an Infinity Fabric, instead of everything on one die split multiple ways. That would be quite a change from Ryzen.Quote
CSS Valid