And still no clue as to what is at the $300 end of the price range mentioned at Computex.
New models.
AMD Announces 470 and 460 along with 480 launching June 29th
http://venturebeat.com/2016/06/13/a...w-cost-polaris-graphics-chips-that-can-run-vr
Looks like RX 480 is full P10, RX 470 the cut down version and RX 460 is Polaris 11 for desktops. OP updated, will post all news there as well from now on.
Maybe this simply mean anyone with a budget of $100-$300 will seek Polaris. If 1070 is $379 minimum, it would be true even if the 8GB 480 was $230-$250... I guess.
There's more uses for VR than gaming. Most likely the masses will be purchasing VR for the other uses in the end anyways.
Have you tried any other content such as movies, digital tours, etc? If so how taxing is it compared to gaming on your rig?
LOS ANGELES—AMD has unveiled two new graphics cards based on its latest Polaris architecture, the RX 470 and the RX 460. The company says the RX 470 is ideal for 1080p, 60FPS full HD gaming, while the RX 460 is for e-sports players looking for a small low-power graphics card (just 75W) that offers high frame rates in games like Overwatch and League of Legends.
AMD didn't go into technical details on either card, but it's a good bet that the RX 470 will feature a slimmer version of the chip inside the RX 480, a mid-range graphics card designed for VR and 1440p gaming that's due out on June 29. AMD hasn't priced either the RX 470 or RX 460, or announced a release date, but given that the RX 480 is set to retail for $199 (for a 4GB version), they'll likely be light on the wallet.
AMD's didn't go into further details on the Polaris architecture behind the two new graphics cards either, but we do know that Polaris is new 14nm FinFET manufacturing process that's allowing AMD to shrink down its GCN architecture to a far more efficient level. Polaris isn't designed to take on the likes of Nvidia's GTX 1080 and GTX 1070, and is instead aimed at the mainstream, with the top-of-the-range RX 480 offering VR-ready performance. The RX 480 will feature 36 compute units, along with some fast GDDR5 memory attached to a 256-bit memory bus for 256GB/s of bandwidth.
Going after a mainstream audience is potentially a very smart move for AMD. It has struggled to compete with Nvidia at the high-end of the market, despite some decent efforts in the form of the R9 Fury range. Be reaching volume at cheaper price points, the company can hope to claw back some of the market share it has lost to Nvidia, which has traditionally not fared so well at lower price points.
Seems like rx480 is either full chip or that full chip is all going to apple.
This man gets it.
AMD's Command Processor is inefficient in DX11 because it was designed to work alongside ACEs which DX11 cannot access, ie. Graphics Queues via CP, Compute/Copy queues via ACEs.
It also lacks a deep queue buffer, so if the CPU is hammered, it stalls. NV's Gigathread has a deep buffer and it receives the queues from the software (driver) scheduler that's multi-threaded on the CPU, so it rarely gets stalled. In DX12, it loses it's advantage, and even suffers slightly via software scheduling overhead.
When GCN is driven under DX12, the CP can be fed with multi-core CPUs and so it's operating at a higher efficiency. Add ACEs to distribute compute queues and it offloads some of the work from the CP, as it was designed to work together.
This is why we see DX12 performance is reflective of the Tflops of GCN.
Polaris gets a lot of improvements to enhance it's performance in DX11. A new CP, Pre-Fetch, Cache system are all things that address the critical flaws by removing bottlenecks in DX11. I don't know whether this would improve it's DX12, but given how well current GCN runs in DX12, I have to wonder if there's room left to improve.
What it may mean is that if it's 5.5 TFlops, in DX12, it's going to be on-par with the 390X, but in DX11, it will be ahead. The biggest unknown though is how effective the Discard Accelerator will function. That may really skew the performance edge to Polaris.
I know this wasn't the true release for Polaris but, why does AMD trot out cards just to say literally zero about them? The Zen being tested in the lab was pretty cool, and was appropriate for a product still 6 months away, but they couldn't show ANY metrics about these new cards?
This man gets it.
AMD's Command Processor is inefficient in DX11 because it was designed to work alongside ACEs which DX11 cannot access, ie. Graphics Queues via CP, Compute/Copy queues via ACEs.
It also lacks a deep queue buffer, so if the CPU is hammered, it stalls. NV's Gigathread has a deep buffer and it receives the queues from the software (driver) scheduler that's multi-threaded on the CPU, so it rarely gets stalled. In DX12, it loses it's advantage, and even suffers slightly via software scheduling overhead.
When GCN is driven under DX12, the CP can be fed with multi-core CPUs and so it's operating at a higher efficiency. Add ACEs to distribute compute queues and it offloads some of the work from the CP, as it was designed to work together.
This is why we see DX12 performance is reflective of the Tflops of GCN.
Polaris gets a lot of improvements to enhance it's performance in DX11. A new CP, Pre-Fetch, Cache system are all things that address the critical flaws by removing bottlenecks in DX11. I don't know whether this would improve it's DX12, but given how well current GCN runs in DX12, I have to wonder if there's room left to improve.
What it may mean is that if it's 5.5 TFlops, in DX12, it's going to be on-par with the 390X, but in DX11, it will be ahead. The biggest unknown though is how effective the Discard Accelerator will function. That may really skew the performance edge to Polaris.
I know this wasn't the true release for Polaris but, why does AMD trot out cards just to say literally zero about them? The Zen being tested in the lab was pretty cool, and was appropriate for a product still 6 months away, but they couldn't show ANY metrics about these new cards?
And still no clue as to what is at the $300 end of the price range mentioned at Computex.
So if they are anouncing two new models that go on sale at the same day could this mean that globalfoundries finally doesn't have fab problems at 14nm?
Seems like rx480 is either full chip or that full chip is all going to apple.
Maybe the longer they wait for specific performance, the more likely Nvidia will pick their clock rates for 1060 and 1050 by the time they find out? Launching late has its benefits, especially when your tech is potentially better and you can easily have it be 5-10% above the competition (see 680 against launch, 925MHz 7970). But you can change clocks up until pretty late so I'm not sure the secrecy is helping.
The big problem of nVIDIA is that they can't perfectly scale down since they have an IPC regression. It will hurt them hard.