AMD confirmed yesterday what was an open secret, that the Radeon RX 8000 will be a bet on mid-range graphics cards, and an effort to offer the best price-performance ratio possible. The company has adopted this strategy because it believes it is the one that can best serve to achieve a clear objective: to boost its market share to 40%-50%.
It is a goal very ambitious, and also quite complicatedespecially considering the dominance that NVIDIA has right now and the competition that Intel could offer with Arc Battlemage, but AMD is committed and willing to fight to achieve it, so we will have to wait a reasonable time to see what results it obtains with this new strategy.
New information about the Radeon RX 8000 has also been appearing over the past few weeks, so I think it’s a good time to post A new updated article with everything we know right now about this new generation of graphics cards. All the information I am going to share with you comes from fairly reliable sources, so it has a high credibility, but since it is not official it is not totally guaranteed that it will be fulfilled.
RDNA 4 architecture details
The RDNA 4 architecture will be the base on which the Radeon RX 8000 GPUs will be built, and if all the information I have been seeing is true, it will mean a very important evolution against RDNA 3.
The first thing to note is that this new architecture could mean abandoning the MCM design that we saw in the Radeon RX 6000 and RX 7000, which implies that AMD would return to a monolithic core design with the Radeon RX 8000, where there would be no outsourced components, and all the GPU elements would be on the same silicon die, including the L3 cache.
On the Radeon RX 6700 XT and above, and Radeon RX 7700 XT and above, the L3 cache is externalized to chiplets, which causes an increase in latency in the communication between the GPU and the cache. By integrating the L3 cache on the same silicon as the GPU, this problem is eliminated, and performance is improved.
In order to integrate the L3 cache on the same silicon chip, AMD will use the TSMC’s 4nm node, which reduces the size of transistors and improves efficiency, and will reduce its capacity below 96 MBThe most powerful models are likely to have 64 MB and the least powerful ones between 48 MB and 32 MB.
This architecture could Double the number of cores to accelerate ray tracing per compute unit, which would mean going from one to two, and will bring new technologies and features to improve ray tracing performance, including a dual ray intersection engine, improvements working with bounding volume hierarchies, and three-pair ray tracing optimization.
RDNA 4 will also offer performance improvements in rasterization and AI, thanks to the introduction of hardware for matrix acceleration and FP8 operations. It’s still unclear, but this hardware could be the basis for AMD to develop new intelligent upscaling technology that’s truly on par with NVIDIA DLSS.
AMD will launch four models within the Radeon RX 8000
And those four models will use two graphics cores with different configurations. Two models will use the core Navi 48which is the most powerful, and the other two will be based on the graphics core Navi 44To shape different graphics cards AMD will adjust the active computing units on each of those graphics cores.
The Radeon RX 8800 XT will be the most powerful of those four new graphics cards. Behind it will be the Radeon RX 8700 XT, and in the budget mid-range we will find the Radeon RX 8600 XT and Radeon RX 8500 XT. AMD may use a different nomenclature, with the two most powerful models coming identified as Radeon RX 8700 XT and Radeon RX 8700.
Possible Radeon RX 8800 XT specifications
- Navi 48 graphics core manufactured on TSMC’s 4nm node.
- 64 computing units.
- 4,096 shaders at a speed of over 3 GHz.
- 256 textured units.
- 128 raster units.
- 128 third-generation cores to accelerate ray tracing.
- 128 second-generation AI cores.
- Bus de 256 bits.
- 16 GB of GDDR6 memory at 20 Gbps (640 GB/s also bandwidth).
- 4 MB L2 cache.
- 64 MB of L3 cache.
Power and comparative performance
The Radeon RX 8800 XT would have a power of 49,15 TFLOPs en FP32which would place it slightly above the 48.74 TFLOPS reached by the GeForce RTX 4080In ray tracing its performance could be similar to that of a GeForce RTX 4070 Ti.
Possible Radeon RX 8700 XT specifications
- Navi 48 graphics core manufactured on TSMC’s 4nm node.
- 56 computing units.
- 3,584 shaders at 3 GHz speed.
- 224 textured units.
- 128 raster units.
- 112 third-generation cores to accelerate ray tracing.
- 112 second-generation AI cores.
- Bus de 256 bits.
- 16 GB of GDDR6 memory at 20 Gbps (640 GB/s also bandwidth).
- 4 MB L2 cache.
- 64 MB of L3 cache.
Power and comparative performance
This model will have a power of 43 TFLOPs en FP32enough to get high between the GeForce RTX 4070 SUPER and the GeForce RTX 4070 Ti. Its ray tracing performance could be similar to that of a GeForce RTX 4070.
Possible Radeon RX 8600 XT specifications
- Navi 44 graphics core manufactured on TSMC’s 4nm node.
- 40 computing units.
- 2,560 shaders at speeds of up to 3.2 GHz.
- 160 textured units.
- 64 raster units.
- 80 third-generation cores to accelerate ray tracing.
- 80 second-generation AI cores.
- Bus de 192 bits.
- 12 GB of GDDR6 memory at 18 Gbps (432 GB/s also bandwidth).
- 2 MB L2 cache.
- 48 MB of L3 cache.
Power and comparative performance
It could be the big star of the new AMD generation for price-performance. Its power in FP32 will be 30,72 TFLOPs, a figure that places it at the level of the GeForce RTX 4070. Its ray tracing performance should be closer to that of a GeForce RTX 4060 Ti.
Possible Radeon RX 8500 XT specifications
- Navi 44 graphics core manufactured on TSMC’s 4nm node.
- 32 computing units.
- 2,048 shaders at speeds of up to 3 GHz.
- 128 textured units.
- 64 raster units.
- 64 third-generation cores to accelerate ray tracing.
- 64 second-generation AI cores.
- Bus de 128 bits.
- 8 GB of GDDR6 memory at 18 Gbps (288 GB/s also bandwidth).
- 2 MB L2 cache.
- 32 MB of L3 cache.
Power and comparative performance
It will be the most economical of these four new graphics cards. It will have a power of 24,57 TFLOPs, a figure that would leave it a little above the GeForce RTX 4060 Ti. In ray tracing it should be more or less at the level of a GeForce RTX 4060.
Please note that TFLOPS allow us to make approximate performance estimates, but they are not absolute values when we talk about games, and therefore Performance in such applications may differ, as they depend on many other factors including things like resolution, bandwidth, and optimization.
Possible release date and price
There is no agreement yet, but the latest information says that Navi 48 will enter the testing phase. mass production during the fourth quarter this year, so we have two possibilities:
- That the Radeon RX 8800 XT and RX 8700 XT arrive before the end of 2024, with availability perhaps a little limited
- Both graphics cards are expected to be gone in early 2025.
In both cases it seems that the Radeon RX 8600 XT and Radeon RX 8500 XT will arrive later than the more powerful models. If this is true, they will not be available until first the second quarter of 2025.
Regarding the price, we know that AMD wants to fine-tune the cost-performance ratio of the Radeon RX 8000 as best as possible, and it is rumored that the most powerful model will cost between $499 and $599. This allows me make an estimate of the possible price of all models of this new generation:
- Radeon RX 8800 XT: between $499 and $599. Less than €649 at the current exchange rate.
- Radeon RX 8700 XT: between $399 and $499. Less than €549 at the current exchange rate.
- Radeon RX 8600 XT: between $299 and $349. Less than €400 at the current exchange rate.
- Radeon RX 8500 XT: between $199 and $249. Less than €300 at the exchange rate.
AI-generated images.
Source: www.muycomputer.com
