Their potential alternative is to basically stop making Vega (outside its known markets like Apple etc that is) & sell Polaris where they can while dreaming of a brighter future.Its not like they have a choice, im sure they dont want to sell the 56 for Nvidia xx60 pricing but if the 2060 is as fast or faster than vega 56 they wont have a choice in the matter, they will have to sell it at whatever price its relative performance justifies or no one will buy it.
The TDP is for sure an issue. Not sure how they are going to tackle that one.
They don't need a new hardware. They just need the 17.320 batch.
Their potential alternative is to basically stop making Vega (outside its known markets like Apple etc that is) & sell Polaris where they can while dreaming of a brighter future.
I have a feeling you are right, but i hope for AMD's sake you are wrong. AMD is going to be in a world of hurt if they have nothing to counter volta till navi.
What do you mean ? Driver version ?
Data based on AMD Internal testing of a prototype RX Vega sample with a prototype branch of driver 17.320
https://forum.beyond3d.com/posts/1997699/Quick note on primitive shaders from my end: I had a chat with AMD PR a bit ago to clear up the earlier confusion. Primitive shaders are definitely, absolutely, 100% not enabled in any current public drivers.
The manual developer API is not ready, and the automatic feature to have the driver invoke them on its own is not enabled.
Zlatan. You need to tell us more
The ngg path is it in game engines or in the driver?
When can we expect to see it in action?
2 GHz is just the maximum clock. It may be more useful to not allow the full speed, and save some power.If HBM2 is fine why are clocks not 2 Ghz and why is pseudo channel missing.
2 GHz is just the maximum clock. It may be more useful to not allow the full speed, and save some power.
The pseudo channel mode is probably not stable with the current drivers.
Yep. But that point is beaten to death. Same for the memory. Come on.Pseudo channel mode is a HBM2 feature designed to improve bandwidth.
http://monitorinsider.com/HBM.html
Vega is struggling with real world bandwidth as a % of peak theoretical bandwidth.
https://techgage.com/article/a-look-at-amds-radeon-rx-vega-64-workstation-compute-performance/5/
I don't know if its a hardware or software issue. But the fact that its not working just proves how poorly the Vega launch has been executed.
Agree that they are always to early and to techhie. But lets see where vega lands when they finally enable ngg that is what makes vega interesting from a gamers perspective.Just get as far away from this [inferior] design and double down on Polaris and make that architecture great again for a big chip.
Nvidia has a compact design that uses very little power, AMD uses a more spread out design and need more power to use it, but since its also very complex they do end up losing performance.
Nvidia's design is a lot more simpler, focused on the needs of now, rather than in 2 years time. So their GPU's are best optimized for DX11, while engineered to do decent enough in DX12/Vulkan. Their next generation will most likely focus a lot more on Vulkan and DX12, which would be the now in about a year's time.
From the information I have, Geforce Volta runs in circles around Vega in DX12. So much difference that current Vega 64 DX12 benches are at GV106 level !!!Nvidia's design is a lot more simpler, focused on the needs of now, rather than in 2 years time. So their GPU's are best optimized for DX11, while engineered to do decent enough in DX12/Vulkan. Their next generation will most likely focus a lot more on Vulkan and DX12, which would be the now in about a year's time.
About the ngg path.
From amd whitepaper
http://radeon.com/_downloads/vega-whitepaper-11.6.17.pdf
"This next-generation geometry (NGG) path is much more flexible and programmable than before. To highlight one of the innovations in the new geometry engine, primitive shaders are a key element in its ability to achieve much higher polygon throughput per transistor. Previous hardware mapped quite closely to the standard Direct3D rendering pipeline, with several stages including input assembly, vertex shading, hull shading, tessellation, domain shading, and geometry shading. Given the wide variety of rendering technologies now being implemented by developers, however, including all of these stages isn’t always the most ecient way of doing things. Each stage has various restrictions on inputs and outputs that may have been necessary for earlier GPU designs, but such restrictions aren’t always needed on today’s more flexible hardware.
“Vega’s” new primitive shader support allows some parts of the geometry processing pipeline to be combined and replaced with a new, highly ecient shader type. These flexible, general-purpose shaders can be launched very quickly, enabling more than four times the peak primitive cull rate per clock cycle. In a typical scene, around half of the geometry will be discarded through various techniques such as frustum culling, back-face culling, and small-primitive culling. The faster these primitives are discarded, the faster the GPU can start rendering the visible geometry. Furthermore, traditional geometry pipelines discard primitives after vertex processing is completed, which can waste computing resources and create bottlenecks when storing a large batch of unnecessary attributes. Primitive shaders enable early culling to save those resources."
Well they better get that to work !
So basically, AMD designed a headline performance feature which would likely have made a huge difference in some scenarios, then launched the product with that headline performance feature disabled. The Software side foils AMD yet again.
I mean, it happened. GTX 480 vs GTX 580 was the prime example.It is hard to take a hot, power hungry, large die sized card that arrived a year late and expect AMD to suddenly alleviate at least one of these main drawbacks or pull a 1080ti competitor out of their hat. I understand why people conceive of this fairy tale, but it is simply not a grounded concept. It appears to me that AMD is now in the same spot versus NVIDIA that it was in versus Intel a decade ago. In my opinion, Vega is the Polaris to NVIDIA's Volta, arrived early. People just do not know that yet.
I mean, it happened. GTX 480 vs GTX 580 was the prime example.
The biggest difference didn't come from yields. 580 was the first time they used custom cell libraries in manufacturing, which still goes on today.Except this time around Vega was produced on a very mature node vs. the 480. The 580 was released within a six months of the 480 and roughly a year from initial small die 40nm gpus.
Vega should already be dialed in for yields, performance and power consumption at this point just from a hardware perspective, let alone the software. Vega is stillborn and I don't expect much if anything from AMD until Navi.
Except this time around Vega was produced on a very mature node vs. the 480. The 580 was released within a six months of the 480 and roughly a year from initial small die 40nm gpus.
Vega should already be dialed in for yields, performance and power consumption at this point just from a hardware perspective, let alone the software. Vega is stillborn and I don't expect much if anything from AMD until Navi.