Recent content by Olikan

Discussion Speculation: Zen 4 (EPYC 4 "Genoa", Ryzen 7000, etc.)

humm... Zen4 L3 bandwidth is ~50% higher, userbenchmark sees a ~50% increase :smirk:
- Olikan
- Post #7,047
- Aug 5, 2022
- Forum: CPUs and Overclocking
Question Speculation: RDNA3 + CDNA2 Architectures Thread

At this point, i doubt the extra cost is more than $15, this is an old tech for AMD... they could easily made each chiplet larger, say 128-bit bus, the complexity would be halved, and would still had a good flexibility If anything, the tiny size of the chiplet means the savings of a yield...
- Olikan
- Post #1,242
- Jul 20, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

another interesting thing about these MCDs, is that AMD may reuse them for FGPAs
- Olikan
- Post #1,211
- Jul 17, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

What are the odds of something crazy like LPDDR5x? RDNA2 was surprising with its small bus. RDNA3 is rumored to have even bigger IF, more compression and clocks over 3GHz.... bandwidth shoundt be a problem, but latency might... a cheap, low latency and power efficient memory, might be good...
- Olikan
- Post #1,186
- Jul 14, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

Keep in mind, from the other patents: dual vector and super simd, both relie on auxiliar caches near the registry files
- Olikan
- Post #1,040
- Jun 28, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

Indeed, just the patent have weird ways to explain it... "a CU surrounded by L0s"
- Olikan
- Post #1,028
- Jun 28, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

UNIFORM CACHE SYSTEM FOR FAST DATA ACCESS https://www.freepatentsonline.com/y2022/0188232.html Someone at twitter found this patent some time ago, only now looking again at @Kepler_L2 diagram of RDNA3, i think i understood... Each SIMD will have it's own L0 cache, if a miss, it will look at...
- Olikan
- Post #1,026
- Jun 28, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

Pure expeculation: for next generation, AMD might improve the "dual vector issue" to be more flexible... Like 4 lanes of SIMD16, for a "quad vector issue"...do 1 wave 64, 2 wave32 and 4 wave16 per clock... Maybe even do unussual issues like wave48 + wave16
- Olikan
- Post #1,011
- Jun 22, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

It's a follow up of the patent "super-simd", RDNA3 can (aparently) chain mutiple instructions in theyr dual ALUs, pretty much bouncing the results of one alu to another... it's nothing alike the old gpus This last patent descibes that a compiler keeps track of any dependencys, and if it finds...
- Olikan
- Post #1,001
- Jun 20, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

Vliw compiler is back... but more advanced https://www.freepatentsonline.com/y2022/0188120.html
- Olikan
- Post #999
- Jun 20, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

I really can't remember another gpu that can do a wave64 in a single clock, very impressive.
- Olikan
- Post #996
- Jun 16, 2022
- Forum: Graphics Cards
Question AMD Phoenix/Zen 4 APU Speculation and Discussion

not a pro at this, but will do my best It alleviates one of the most important GPUs bottleneck: register file... (a major bottlenck of GCN gpus) doing a wave64 in 2 clocks means twice the time it needs to hold data... Problem shows up when the gpu needs to execute a critical wavefront, but the...
- Olikan
- Post #222
- May 25, 2022
- Forum: CPUs and Overclocking
Question Speculation: RDNA3 + CDNA2 Architectures Thread

Whats your theory on the double wavefront? Just saw at twitter an interesting speculation: the gpu launches 2 wave32 at the same time, emulating as 1 wave64... this would allow the gpu to launch 1 wave64 per clock
- Olikan
- Post #829
- May 14, 2022
- Forum: Graphics Cards
Question Speculation: RDNA3 + CDNA2 Architectures Thread

Nope, a wavefront will use as many shaders it can... I can see 2 uses of it: 1- Similar to hypertreading in CPUs, a way to use all resources avaible when possible. The drivers would need to be aware of it, will have diminushing returns If devs fine tune for it; 2- to Double the shaders...
- Olikan
- Post #822
- May 13, 2022
- Forum: Graphics Cards
Discussion RDNA 2 6nm Refresh

There was any talk about EUV improving Fmax?
- Olikan
- Post #16
- Dec 31, 2021
- Forum: Graphics Cards

RESOURCES

Top Bottom