HITMAN Lead Dev: DX12 Gains Will Take Time, But They're Possible After Ditching DX11

HITMAN Lead Dev: DX12 Gains Volition Have Time, Only They're Possible After Ditching DX11

HITMAN (or rather, its prologue) came out but over a month ago and was received pretty well overall, Wccftech review included. The game was also ane of the first to introduce DirectX 12 support on PC, but the gains were minimal on AMD and non-real on NVIDIA.

We've had the hazard to enquire virtually this specific topic with IO Interactive'southward Atomic number 82 Developer Jonas Meyer, who mentioned that the fabled xx% CPU performance and l% GPU performance boosts advertised by Microsoft are possible, but only in time and afterward dropping DirectX 11 support entirely. HITMAN itself was more of a straight port to DirectX 12, for case.

HITMAN on DirectX 12 simply offers approximately a 10% operation boost on AMD cards. Are y'all planning to improve DirectX 12 functioning and/or add any further DX12 features via patch and if so, which ones?
We don't accept any DX12 specific improvements planned for our subsequent releases. While we have x% improvement when completely GPU bound, I think there are gains above that when you're CPU bound.

NVIDIA cards, on the other hand, have basically the same performance under DX11 and DX12. Y'all've mentioned during the GDC 2022 talk that you're working with NVIDIA to improve their Async Compute implementation - how are things looking on that end and exercise yous take whatever ETA?
I don't have any news on that forepart, sorry.

Async Compute in item has received a lot of attention from PC enthusiasts, specifically in regards to NVIDIA GPUs lacking hardware support for it. However, in the GDC 2022 talk y'all said that fifty-fifty AMD cards only got a 5-10% boost and furthermore, you described Async Compute as "super hard" to melody considering too much work can make information technology a penalty. Is it fair to say that the importance of Async Compute has been perhaps overstated in comparison to other factors that determine performance? Do you think NVIDIA may be in trouble if Pascal doesn't implement a hardware solution for Async Compute?
The main reason it's hard is that every GPU ideally needs custom tweaking – the bandwidth to compute ration is dissimilar for each GPU, ideally requiring tweaking the amount of async work for each one. I don't recollect information technology'due south overstated, merely obviously YMMW (your mileage may vary). In the current state, Async compute is a nice & piece of cake operation win. In the long run information technology will be interesting to run across if GPU's get better at running parallel work, since we could potentially get even better wins.

Several DirectX 12 games are out now, but the start outlook isn't nearly as positive every bit Microsoft stated (upward to 20% more than operation from the GPU and upwards to 50% more operation from the CPU). Practice you think that information technology'south simply a matter of fourth dimension before developers learn how to employ the new API, or maybe the functioning benefits accept been somewhat overestimated?
I think it will have a bit of time, and the drivers & games need to mature and exercise the right things. Merely reaching parity with DX11 is a lot of piece of work. 50% performance from CPU is possible, simply it depends a lot on your game, the driver, and how well they work together. Improving performance by 20% when GPU bound will be very hard, especially when you have a DX11 driver team trying to meliorate performance on platform likewise. It's worth mentioning nosotros did only a straight port, once we starting time using some of the new features of dx12, information technology volition open upwardly a lot of new possibilities – then the gains will definitely be possible. Nosotros probably won't offset on those features until we can ditch DX11, since a lot of them require fundamental changes to our render code.

Do you believe HITMAN's DirectX 12 performance could have been better if the game had been built entirely on information technology, rather than having DX11 as min spec?
Yes, information technology obviously would take. DX12 includes new hardware features, which I think over time will brand it possible for us to brand games run fifty-fifty improve.

Another low level API has been recently released: Vulkan. What do you recall of it in terms of performance and features? Do you have any plans to add Vulkan support to HITMAN?
Vulkan is a graphics programmer'due south wet dream: A high performance API, like d3d12, for all platforms. With that said, we don't have any plans to add together Vulkan support to Hitman. Also, unfortunately it looks like Vulkan won't exist supported on all platforms.

Finally, as a developer, what are your thoughts on the Universal Windows Platform?
I'one thousand not qualified to talk almost UWP – I don't know annihilation about it. I can say, I really similar the Win32 API, it'due south been around forever, is well documented, and it works.

Our take

While many take been disappointed by the functioning results of the showtime DirectX 12 compatible games, information technology is worth to add to Meyer's words that we have seen the same verbal situation in the past. DirectX 11 took a very long fourth dimension to be properly implemented in games; given the faster adoption of Windows 10, information technology's likely that DirectX 12 will get there faster but information technology clearly still needs time, mostly because developers have to sympathise how to best utilize its new capabilities.

Dropping DirectX xi, which might happen sooner rather than later on (DICE's Johan Anderson wished DX12 just games for Vacation 2022 well before Windows 10 came out), should let programmers to explore these possibilities. Of course, hither at Wccftech we'll keep you updated on everything concerning DirectX 12 & Vulkan, so stay tuned.