AMD Announces Radeon RX 5700 XT & RX 5700: The Next Gen of AMD Video Cards Starts on July 7th At $449/$379
by Ryan Smith on June 10, 2019 7:20 PM ESTA Quick Note on Architecture & Features
With pages upon pages of architectural documents still to get through in only a few hours, for today’s launch news I’m not going to have the time to go in depth on new features or the architecture. So I want to very briefly hit the high points on what the major features are, and also provide some answers to what are likely to be some common questions.
Starting with the architecture itself, one of the biggest changes for RDNA is the width of a wavefront, the fundamental group of work. GCN in all of its iterations was 64 threads wide, meaning 64 threads were bundled together into a single wavefront for execution. RDNA drops this to a native 32 threads wide. At the same time, AMD has expanded the width of their SIMDs from 16 slots to 32 (aka SIMD32), meaning the size of a wavefront now matches the SIMD size. This is one of AMD’s key architectural efficiency changes, as it helps them keep their SIMD slots occupied more often. It also means that a wavefront can be passed through the SIMDs in a single cycle, instead of over 4 cycles on GCN parts.
In terms of compute, there are not any notable feature changes here as far as gaming is concerned. How things work under the hood has changed dramatically at points, but from the perspective of a programmer, there aren’t really any new math operations here that are going to turn things on their head. RDNA of course supports Rapid Packed Math (Fast FP16), so programmers who make use of FP16 will get to enjoy those performance benefits.
With a single exception, there also aren’t any new graphics features. Navi does not include any hardware ray tracing support, nor does it support variable rate pixel shading. AMD is aware of the demands for these, and hardware support for ray tracing is in their roadmap for RDNA 2 (the architecture formally known as “Next Gen”). But none of that is present here.
The one exception to all of this is the primitive shader. Vega’s most infamous feature is back, and better still it’s enabled this time. The primitive shader is compiler controlled, and thanks to some hardware changes to make it more useful, it now makes sense for AMD to turn it on for gaming. Vega’s primitive shader, though fully hardware functional, was difficult to get a real-world performance boost from, and as a result AMD never exposed it on Vega.
Unique in consumer parts for the new 5700 series cards is support for PCI Express 4.0. Designed to go hand-in-hand with AMD’s Ryzen 3000 series CPUs, which are introducing support for the feature as well, PCIe 4.0 doubles the amount of bus bandwidth available to the card, rising from ~16GB/sec to ~32GB/sec. The real world performance implications of this are limited at this time, especially for a card in the 5700 series’ performance segment. But there are situations where it will be useful, particularly on the content creation side of matters.
Finally, AMD has partially updated their display controller. I say “partially” because while it’s technically an update, they aren’t bringing much new to the table. Notably, HDMI 2.1 support isn’t present – nor is more limited support for HDMI 2.1 Variable Rate Refresh. Instead, AMD’s display controller is a lot like Vega’s: DisplayPort 1.4 and HDMI 2.0b, including support for AMD’s proprietary Freesync-over-HDMI standard. So AMD does have variable rate capabilities for TVs, but it isn’t the HDMI standard’s own implementation.
The one notable change here is support for DisplayPort 1.4 Display Stream Compression. DSC, as implied by the name, compresses the image going out to the monitor to reduce the amount of bandwidth needed. This is important going forward for 4K@144Hz displays, as DP1.4 itself doesn’t provide enough bandwidth for them (leading to other workarounds such as NVIDIA’s 4:2:2 chroma subsampling on G-Sync HDR monitors). This is a feature we’ve talked off and on about for a while, and it’s taken some time for the tech to really get standardized and brought to a point where it’s viable in a consumer product.
Facebook
Twitter
RSS
326 Comments
View All Comments
Spunjji - Tuesday, June 11, 2019 - link
Underclocking / undervolting experiments have shown that GCN is actually quite competitive in terms of power/perf, *for a given level of performance*. Unfortunately for AMD, Nvidia have been consistently able to hit a higher absolute level of performance, forcing AMD to hot-clock their cards just to keep up.That is absolutely down to Nvidia hitting it out of the park with Maxwell - they nailed architectural efficiency in a way that has clearly taken AMD some time to catch up on, and they managed it with an architecture that scales up extremely well.
Meteor2 - Tuesday, June 11, 2019 - link
Define tangible. We’re seeing the same slow-down in performance increases that we’ve seen with CPUs.Pascal and Maxwell in particular were amazing; the GPU equivalent of Core or the Bridge series x86 cores. AMD has caught up on the CPU side, but not so much on the GPU.
4K 60+ FPS on max AAA settings is extremely hard to do. Nvidia have got there, just, but at what a price. AMD can’t build such a GPU; there’s not enough thermal capacity in a PC case for the amount of power such a Navi GPU would need.
eastcoast_pete - Monday, June 10, 2019 - link
Thanks Ryan! Are the 64 ROPs confirmed?xrror - Monday, June 10, 2019 - link
2nd this - 64 ROPs is something to get excited about.Ryan Smith - Monday, June 10, 2019 - link
Yes. It is confirmed.BenSkywalker - Monday, June 10, 2019 - link
Even if we take AMD slides as the gospel truth, 10% better performance than the 2060 for 8.6% more money but with no RTX hardware at all? These parts seem quite a bit overpriced, and that's if the rumors around the 'super' offerings from nVidia are wrong. If they are correct, or even half as good as the claims, these parts will lose in every metric. And that's assuming this isn't another 'overclocks like a dream' or 'poor Volta' moment.Fallen Kell - Monday, June 10, 2019 - link
Well, 3 years late to the party, sporting the new shiny coat of paint on last years performance... I just don't get it. AMD had every advantage for this card with a significant manufacturing process advantage over Nvidia, and yet, still can't beat what Nvidia had out ~3 years ago (1080ti). I can only believe that they have stopped trying. They most definitely didn't try the last 3 generations of cards. I mean, they didn't even make an attempt at a card that could perform in games as well as Nvidia's high end cards. It was sure good that AMD's cards could at least perform well for compute work, otherwise I have not idea how they have staying in the graphics card business the last 6 years.Look, I get it, the real sales are in the mid-range products for stand-alone cards, and in the low end products on the integrated side. But the thing that drives those sales are the high end. The consumers read, see, hear, all about how Nvidia is the fastest, highest performance, "best" card all over the place in the benchmarks, and all of a sudden Nvidia is equated with better products. Soon customers are buying that laptop or desktop because it has the Nvidia card in it, and there goes the sales on integrated and and low end systems from the big manufacturers. I doesn't matter that AMD might compete at price/performance, the brand itself is seen as second class because they havn't been able to compete with Nvidia on the high end in a decade or more now... I was REALLY hoping that this card/generation was going to be something different. 7nm vs 14nm should have been able to blow away Nvidia's performance, yet it can barely match Nvidia's 3rd or 4th fastest cards...
SaberKOG91 - Monday, June 10, 2019 - link
These are mid-range parts. Wait for Navi 20 with the full 64CU and 4096 shaders. Should fall between the 2080 and the 2080 Ti if I had to guess. Unless it is on 7nm+ which might net it some even better performance.AMD had every advantage...until they fell 6 months behind schedule due to a retape in October. We were supposed to have had Navi 10 in January and be getting Navi 20 now. Instead they are playing catch-up and licking their wounds. Stuff happens. Nvidia aren't really trying to get much faster, so AMD still have time to catch up. Closing the gap is a huge first step.
Phynaz - Tuesday, June 11, 2019 - link
You’re saying AMD has moved mid range to $500. And you’re happy about that.SaberKOG91 - Tuesday, June 11, 2019 - link
Nvidia moved it there. AMD are just not going to give up the margins this time around. I'm not happy about any of this. But clearly you are incapable of doing anything other than seeing someone who disagrees with you as a fanboy. Which is ironic because all of your petty little comments make you worse than any fanboy I have ever encountered.