I doubt many developers are going to make games that only run on the latest, high-end consoles in the first place.
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
Hardware Unboxed video shows us what we already know - PC gaming is going to deal with I/O bottlenecks for a while.
- Thread starter Maple
- Start date
I doubt many developers are going to make games that only run on the latest, high-end consoles in the first place.
The problem when you dont test exotic solutions like virtual ram is it leads to bad conclusions as others pointed out.
Ms needs a better kernel in this area of features to maximize io available be it ssd or ram. This isnt a new idea either ram has been io king for a while but doesn't fully get reflected in windows.
These topics are a great to remind others how ms is lackluster in this area.
Ms needs a better kernel in this area of features to maximize io available be it ssd or ram. This isnt a new idea either ram has been io king for a while but doesn't fully get reflected in windows.
These topics are a great to remind others how ms is lackluster in this area.
I'd like to see PC Gaming wiggle its way out of this one
PC gaming easily produces the fastest loading times and best performance
Ah well, nevertheless.
PC gaming easily produces the fastest loading times and best performance
Ah well, nevertheless.
If you plan to keep your PC for many years, you'll want PCIE 4.0.
it has 8 GB of unified memory. Not all ram used for games is for assets displayed on screen.
Can we compare those RDR2 benchmarks on the PC to the Xbox Series X load times for it? That should tell us how the two system, both running unoptimized games, handle it and if its truely an IO issue or an optimization issue.
Edit:
looked it up, Here is a video of it: https://www.gamespot.com/videos/xbox-series-x-load-time-comparison-rdr2-control-ffxv/2300-6453867/
38 seconds to load on Xbox Seires X.
39 seconds on PC accoring to your benchmarks you posted.
Edit:
looked it up, Here is a video of it: https://www.gamespot.com/videos/xbox-series-x-load-time-comparison-rdr2-control-ffxv/2300-6453867/
38 seconds to load on Xbox Seires X.
39 seconds on PC accoring to your benchmarks you posted.
How is any of this different for consoles?
It is likely to take a bit longer on PC, but it's coming.
Last edited:
Why dont we look at this article:
www.eurogamer.net
Xbox Series X back-compat loading times: SSDs and HDDs compared
UPDATE: For the authoritative take on Microsoft's new flagship console, Digital Foundry's full Xbox Series X review is …
www.eurogamer.net
I swear I say this in every doom and gloom thread, but its the truth. Follow the $$$. Game makers won't ditch a sizable portion of their audience to "push the envelope" because it doesn't make them money. Sure, there will be flagship games that do, but that won't be the majority because if your flagship game can't be played by most, you won't recoup your investment.
You're reaching entirely the wrong conclusion from this.
All the data shows is that current games are not built to take advantage of high-performance drives.
That does not mean "major bottlenecks" are preventing NVMe drives from performing faster than SATA drives in a PC.
As an example:
If a game is built for an HDD, it might be considerably easier to build a single-threaded loading system.
That could be all you need with the HDD, as the data is coming in slow enough that there's no advantage to the additional complexity of building a multi-threaded loading system.
Maybe the single core that loading is running on is only hitting 33% utilization when the data is coming from an HDD.
Adding a SATA SSD will speed up that loading process by eliminating the disk bottleneck.
Load times in this example might speed up by a factor of three, which is a lot quicker than the HDD.
Replacing the SATA SSD with an NVMe SSD would only speed that up very slightly though. Why?
Well, the data transfer rate is much higher, but the bottleneck with the SATA SSD was already the single-threaded loading, not the data transfer rate.
The SATA SSD speeds alone were fast enough that the single-threaded loading is pushing that one CPU core to 100% to achieve the 3x speed-up.
Adding a faster drive doesn't affect how quickly the CPU can decompress that data.
A game built to take advantage of SSDs does not behave this way because the loading will be multi-threaded.
It may be more complex to build, but let's say you develop a multi-threaded loading system which scales linearly with the number of cores.
Now the limit on a 16-core CPU would be a 48x speed-up rather than 3x.
Switching from an HDD to a SATA SSD might be a 5x speed-up rather than 3x now, and switching to NVMe may reach that full 48x speed-up.
In that case, even the 16-core CPU might be holding back a really fast PCIe 4.0 NVMe drive; which is where new APIs and hardware decompression comes in. Instead of using all your CPU cores for the decompression, it's done in hardware which could reach speeds of several-hundred times without breaking a sweat - so you're back to being limited by the drive again.
Another factor in this is that NVMe drives are not always the big speed-up over SATA drives that is claimed.
Sure, large sequential reads might be able to hit 7 GB/s, but small random reads are only reaching 61 MB/s in this test.
That's why we need to move beyond NAND. Optane is a further 4–5x speed-up over typical NVMe drives, reaching almost 300 MB/s in the same test:
Optane DIMMs are many times faster than that too; but cost is prohibitively expensive for consumers right now. It's pro-grade hardware.
I don't know, if any game requires such a high bandwidth SSD to run smoothly, i can't see devs keeping out Direct storage, maybe in the 2000's they would, a lot of shit console to pc ports in those days.
I also believe it's going to be years before games really get designed around an SSD & i mean that in the sense of not just making the load times minimal like in Demons Souls & Ratchet & Clank, i don't think any PC gamers will give a shit if they have to wait 10-15 secs to load compared to 2-3 secs on PS5, no one has discussed PC's loading advantage that existed since Bill Clinton was President, to make out like loading is a massive deal now when future consoles get an advantage seems over hyped.
I also believe it's going to be years before games really get designed around an SSD & i mean that in the sense of not just making the load times minimal like in Demons Souls & Ratchet & Clank, i don't think any PC gamers will give a shit if they have to wait 10-15 secs to load compared to 2-3 secs on PS5, no one has discussed PC's loading advantage that existed since Bill Clinton was President, to make out like loading is a massive deal now when future consoles get an advantage seems over hyped.
No, it can't be independently tested as the API isn't released at this point.
The only real point of comparison we have is through Nvidia themselves and their "Marbles" tech demo loading.
1.62s (RTX I/O GPU decompression)
vs
5.25s (24core Threadripper CPU decompression)
Both utilizing the same NVMe Gen 4 drive (7GB/s)
So current games not designed with SSDs in mind don't utilise them as effectively as they should. This is the same with any platform including next gen consoles.
In one or two years, when games designed with with SSD and RTX I/O are releasing then we'll see the difference on PC and same as on consoles.
In one or two years, when games designed with with SSD and RTX I/O are releasing then we'll see the difference on PC and same as on consoles.
Like many others have said, the issue is mostly in software and considering all cross gen titles PCs will be fine. MS will come out with their new storage architecture and Nvidia with AMD will support it.
What this does illustrate is that PC gamers are fine with any decent SATA or NVME drives for next couple years at least,, and console games looking for BC storage are at same juncture.
If your motherboard supports it, buy NVME at 1 to 2TB vs SATA as pricing for cheaper NVME drives is similar.
Also, on PC, caching in RAM can easily be a solution as RAM is a fairly cheap upgrade.
Actually Linus did bunch of tests a while back for PC games and conclusion was that for PC gaming, whatever is the cheapest but still decent SSD is fine.
Also, you have a lot of professional applications that do utilize NVME properly so its definitely ain't a hardware IO issue.
What this does illustrate is that PC gamers are fine with any decent SATA or NVME drives for next couple years at least,, and console games looking for BC storage are at same juncture.
If your motherboard supports it, buy NVME at 1 to 2TB vs SATA as pricing for cheaper NVME drives is similar.
Also, on PC, caching in RAM can easily be a solution as RAM is a fairly cheap upgrade.
Actually Linus did bunch of tests a while back for PC games and conclusion was that for PC gaming, whatever is the cheapest but still decent SSD is fine.
Also, you have a lot of professional applications that do utilize NVME properly so its definitely ain't a hardware IO issue.
Using tests from software that hasn't been designed around faster storage solutions in order to further some weird take is genius.
OP
OP
I'm assuming you're talking about RTX I/O here. What about individuals using Radeon GPUs, or non 3000 series Nvidia GPUs? That gets back to the original point of the thead - some PCs can do it, but many can't. Cerny said that the PS5's custom decompressor equates to 9 Zen 2 cores. And then the dedicated DMA controller acts as yet another Zen 2 core to direct data coming off the SSD.
Even with a 16-core CPU, any PC that attempts that is going to become hamstrung. RTX I/O solves this, but again, this is not guaranteed tech inside every gaming PC. Years from now there will still be a ton of gaming PCs with 8-core CPUs, SATA SSDs, and non RTX GPUs. That means no NVME read speeds, and no RTX I/O decompression. DirectStorage helps, but there are still other limitations relative to consoles in this regard.
From the perspective of a developer, how do you approach this? If the consoles allow for your vision of a game, and some PCs with requisite hardware allow for your vision of a game, but many gaming PCs don't, what do you do? Do you cater to a minority of PC gamers, or change the structure of the game to accomodate the majority of gaming PCs?
I love how this is just being ignored because it doesn't fit the narrative.
These benchmarks are completely pointless. Check out the DF video on XSX BC. While there was an improvement on load times, any software not designed to take advantage of the faster transfer rate are bottlenecked by other factors. You also can't use the XSX BC video to claim the faster IO on the XSX doesn't make a difference. All those benchmarks show is that you won't see the advantage of faster transfer speeds until developers create software specifically tuned for those faster speeds.
Wont have to worry because all multi platform dev will have to build games that runs on old hardware, MS releasing 1st party games are also day 1 on pc have to work on older hardware. Only Sony and their 1st party don't need to take that into consideration and build games exclusively on ps5 where things will shine and you can only play them on ps. Cross gen games like horizon and spiderman will be an interesting showcase of what next gen can bring to show ppl why it is time to buy ps5.
I think we'll be fine lol. I'm gonna go out on a limb and say nvidia and amd know what devs need for future games.
I would assume you do it as always, make your game scalable to run on multiple hardware configurations. 🤷♂️
Gaming has handled different load times on different hardware in the past and will do it just fine in the future too. NVMe drives are awesome but I'm going to guess that it won't cause as much problem as you think, it's just noisier now because we can hear devs debating and users being concerned. There might be some clever new uses of loading masks but there is nothing earth-shattering happening now that the industry haven't already dealt with many times before. We've seen devs tackle moving from cartridges to cassette tapes, to floppies, to hdds, to cartridges, to cds, to blu-rays, to hdds, to ssds, to game cards, etc. I'm guessing they'll be able to tackle this fine as well.
eventually amd will have their own (likely before DirectStorage is officially out for PC), and DirectStorage will wrap around the hardware details allowing for agnostic hardware accelerated decompression as well as enhanced access to SSDs on Windows.
Arent they all slated for late next year? If that early estimate pan out, developer adoption will lag behind. I don't want to estimate when games using the new I/O technologies will be widely available for PC.
If you checked the thread, someone already did a comparison with gamespot's XSX test of current gen games to the PC test in the OP.
The difference between platforms for the game chosen (Red Dead Redemption 2) was 1 second.
It's not the CPU or the i/o.
It's that the random read and writes are the same. Or random reads for games specifically. So I don't agree with the OP, at least that's what I think.
All these SDDs perform more or less the same in games because random reads are way lower than sequential reads. PS5 and XSX touting 5gb/s are sequential reads, but it'll be interesting to see if people can test them for random reads? with high queue count and low thread rate that is more like a game.
Sequential read and writes
Random Read and Writes
from Tech Deals.
Anyways that's just my opinion. The difference is between SSDs and regular hard drives and how many people on PC will still be using the old hard drives and whether devs will cater for them.
Yeah this. Those 'load times' could be data decompressing, stuff processing in preparation for the game world to initialise. In other words - CPU stuff.
Does RTX IO /DirectStorage only benefit games or can the GPU also accelerate the IO system on generic workloads (loading other apps, compiling, copying files etc.)? I have only seen it mentioned in terms of games so far. I realise part of the appeal is getting data directly to the GPU memory but the being able to offload the workload of 'dozens of CPU cores' could be interesting across the board.
DirectStorage yes, not the base DX12. Will be cool to see the first games activelly using it. Although, without custom I/O hardware, PC will probably compensate with either GPU or cores on the CPU (and that could be where 12 cores CPU setup will start shining for gaming too?) to compensate the hardware part.
That wasn't the case also for Xbox 360/PS3 generation.
Every multiplaform game was better on PC.
I remember some benchmarks claiming that CoD2 on 360 was running better than the PC version, but they were wrong, since they were assuming that console version matched the PC maxed out, while it had several graphics features turned down.
The data presented by the OP does not match the conclusion as described in the topic's title and first post. I think that the OP lacks the necessary knowledge and understanding of the technology discussed.
The poster was being sarcastic.
The poster was being sarcastic.
I'm aware, I was being facetious.
But also, RAM is relatively cheap so maybe it's time PC gamers stopped settling for 16 GB.
Storage will catch up a lot faster than you think. Until then I believe RAM requirements will rise to compensate. You probably have at least 16gb of system ram plus whatever for VRAM while the next gen systems have 16gb all together