Now that 2021 is approaching, we thought it would be a good idea to do a quick review of the rumors about AMD not for the short term, but for the medium-long term, all based on its use of node 6 nm of TSMC. In order to get an idea of what the rival of Intel and NVIDIA can offer us, we have compiled and organized the information that has been leaked in recent months.
AMD is now in a very strong moment, but especially it has a close relationship with TSMC after the decision of some leading companies to go with other providers such as Samsung, as NVIDIA and Qualcomm have done. On the other hand with Apple having made the jump to 5nm, this leaves TSMC’s 6nm node fully available to AMD.
What is the TSMC 6nm Node?
TSMC’s 6nm fabrication process, known as N6, was announced by the Taiwanese mega-foundry in April 2019 as an easy transition node to using EUV technology.
Unlike the N7 + node, the N6 node allows manufacturers to use the designs for the standard 7nm manufacturing process, this means that AMD can redeploy its portfolio of 7nm engineered cores under the new node.
TSMC has not given any advantage regarding speed and consumption in order to use the new node, but in terms of density where node N6 is 18% denser than N7 , which means being able to remove a greater number of chips per wafer or be able to add additional elements to create more powerful variants of known CPUs or GPUs.
AMD plans go through 6 nm
Although it is not official, in certain cases the leaks are controlled by the manufacturers themselves and we have evidence through a version of Zen 3 under TSMC’s 6 nm node
This version is called Zen 3+. We do not know how it differs from the standard Zen 3 version or if AMD has taken the opportunity to do some last minute tweaks that increase the IPC a bit, what we do know is that we are going to see a new series of AMD CPUs and SoCs making use of of the Zen 3+ core.
On the one hand we have the Rembrandt-H SoCs, which should replace the Cezanne-H that will be presented in a few days, on the other hand Warhol, which would be the replacement of the current Ryzen 5000 desktop.
Is 2001 a transition year for AMD?
A couple of years ago, AMD and GLOBALFOUNDRIES reached an agreement to manufacture the IOD, which includes the Northbridge and Southbridge of the Zen 2 architecture and which has also been used in the Zen 3. It is precisely in this year 2021 that the contract between AMD and Global Foundries will come to an end.
If Warhol processors are going to be compatible with both DDR5 and PCIe 5 memory, then a new IOD is necessary that is different from Global Foundries, this leads us to the construction of a new Zen processor that we could see for sale at late 2021 or early 2022.
This means that AMD instead of opting for a direct change of generation to adopt DDR5 in its processors, Alder Lake, AMD taking advantage of the chiplet configuration of some of its AMD Ryzen models can easily build variants compatible with DDR5 without making a drastic change in the chip.
AMD RDNA 2+ at 6nm by 2021?
One of the things that has leaked into the roadmap is SoCs with RDNA 2 integrated GPUs, which shouldn’t surprise us considering how easy it is to port designs for TSMC node N7 to node N6.
In the same way that AMD apparently has ported Zen 3 to said node, it can also do so with RDNA 2 and take advantage of the 18% extra density to launch a new generation of AMD graphics cards. After all, an annual launch of new graphics cards is not a crazy idea, not always under the same architecture but it is what has traditionally always been done.
But is a configuration beyond the Navi 21 possible? Yes, it is possible and we have the answer in Navi 14 used in the RX 5500 and RX 5300 that had a configuration of 12 Compute Units per Shader Array instead of 10, if the same configuration was applied in a RX 5900 XT then we would have a 96 Compute Units GPU.
The other possibility is AMD RDNA 3
From the presentation they made last October, they talked about RDNA 3 under an advanced manufacturing node more advanced than that of 7 nn, the problem is that the assertion of the most advanced node can include the 6 nm and the time map of the slide falls within the product launch period under node N6 by AMD.
There are a number of improvements that AMD can integrate into RDNA 2 in order to improve the performance of the architecture and be able to beat NVIDIA on all fronts, not only in rasterization, but also in Ray Tracing as well as Artificial Intelligence.
An RDNA 3 for 2021 is not impossible, we know that Navi 31 exists since it was filtered in the drivers a month ago, but without going into the impossible, what we can expect from the third generation of RDNA 3 are things like the following:
- The Ray Accelerator Unit would have been improved, now instead of performing 4 box-intersections and 1 triangle-intersection it can perform 4 box-intersections and 2 triangle-intersections.
- AMD would have integrated CDNA “Tensor Cores” into this RDNA 2 enhancement.
- GDDR6X instead of GDDR6, but with densities of 2 GB per chip.
What we have more doubts is about doubling the number of ALUs per Compute Unit, this implies much more profound changes in the GPU that require a much longer development time, but AMD, from improving certain points of RDNA 2, can deploy a third generation of your RDNA 3s in less than a year.
