Support NeoGAF

[SlimySnake]

Why is it that every PlayStation console generation has had multiple Tech demo's apart from the PS5 which only had one. (UE5) don't think we can blame Covid, is it just a case UE5 was more than enough to showcase the consoles potential.

The PS5 marketing strategy has been piss poor imo. I was just thinking about how they are hiding Cerny like a protected species. No interviews with the press, no videos, no tech demos. The one time they did let him out of his cage, the man showed up with a powerpoint presentation instead of a tech demo. Just do another duck demo to show the GPU power of the PS5. The amazing spiderman ssd demo was shown behind closed doors for some reason. Surely they shouldve had something like that for the Road to PS5.

Their reliance on trailers instead of actual gameplay demos for next gen games is also bizarre. I think Ratchet is the only next gen game that got a demo and as amazing as it was, it's not enough. Revealing GoW with a title instead of gameplay footage or even cutscene footage is also bizarre.

 

[LiquidRex]

The PS5 marketing strategy has been piss poor imo. I was just thinking about how they are hiding Cerny like a protected species. No interviews with the press, no videos, no tech demos. The one time they did let him out of his cage, the man showed up with a powerpoint presentation instead of a tech demo. Just do another duck demo to show the GPU power of the PS5. The amazing spiderman ssd demo was shown behind closed doors for some reason. Surely they shouldve had something like that for the Road to PS5.

Their reliance on trailers instead of actual gameplay demos for next gen games is also bizarre. I think Ratchet is the only next gen game that got a demo and as amazing as it was, it's not enough. Revealing GoW with a title instead of gameplay footage or even cutscene footage is also bizarre.

Yeah I remember the the spiderman SSD demo which had to be leaked, there's also the Gran Turismo 8K presentation and Ray Tracing demo which was also behind closed doors.

 

[kyliethicc]

My guess is Microsoft want to turn it into their version of Naughty Dog lol

I highly doubt they want Ninja Theory to have like 400 employees but idk. They seem much smaller.

Hellblade is a cool game but its 5 hours long. Its AA. Lets see gameplay for the sequel before assuming too much.

Other than Rockstar, no one else is close to Naughty Dog. Naughty Dog are the crown jewel of Sony's WWS and the most acclaimed developer of the last 10-12 years. They put out a lot of games and they sell tens of millions. 7 big hit AAA games in the last 15 years. Probably like 60-65 million units sold.

 

[kyliethicc]

The PS5 marketing strategy has been piss poor imo. I was just thinking about how they are hiding Cerny like a protected species. No interviews with the press, no videos, no tech demos. The one time they did let him out of his cage, the man showed up with a powerpoint presentation instead of a tech demo. Just do another duck demo to show the GPU power of the PS5. The amazing spiderman ssd demo was shown behind closed doors for some reason. Surely they shouldve had something like that for the Road to PS5.

Their reliance on trailers instead of actual gameplay demos for next gen games is also bizarre. I think Ratchet is the only next gen game that got a demo and as amazing as it was, it's not enough. Revealing GoW with a title instead of gameplay footage or even cutscene footage is also bizarre.

We have GT7 gameplay.

 

[Imtjnotu]

This MGS2 reboot looking

 

[kyliethicc]

Fucking beautiful. Masterpiece.

 

[FranXico]

Their reliance on trailers instead of actual gameplay demos for next gen games is also bizarre. I think Ratchet is the only next gen game that got a demo and as amazing as it was, it's not enough. Revealing GoW with a title instead of gameplay footage or even cutscene footage is also bizarre.

They've mostly shown gameplay footage for most of their first party games. God of War and Horizon FW are the notable exceptions.

 

[TheThreadsThatBindUs]

They've mostly shown gameplay footage for most of their first party games. God of War and Horizon FW are the notable exceptions.

Even the in-engine trailers are better than the CGI target renders MGS have mostly put out.

 

[HoofHearted]

Dad never came home today, You bastard.

 

[BatSu]

 

[TheRedRiders]

I highly doubt they want Ninja Theory to have like 400 employees but idk. They seem much smaller.

Hellblade is a cool game but its 5 hours long. Its AA. Lets see gameplay for the sequel before assuming too much.

Other than Rockstar, no one else is close to Naughty Dog. Naughty Dog are the crown jewel of Sony's WWS and the most acclaimed developer of the last 10-12 years. They put out a lot of games and they sell tens of millions. 7 big hit AAA games in the last 15 years. Probably like 60-65 million units sold.

Agreed. I highly doubt they will reach the level of Naughty Dog. The Sony first parties have significantly more experience and they've been outputting games which have become hits on a very consistent basis. The infrastructure and the culture of the first parties is just something that cannot be replicated by throwing around money and resources. It's why I've been saying for a while now that by the end of this generation, the best looking games will be from the Sony first parties.

 

[SlimySnake]

Agreed. I highly doubt they will reach the level of Naughty Dog. The Sony first parties have significantly more experience and they've been outputting games which have become hits on a very consistent basis. The infrastructure and the culture of the first parties is just something that cannot be replicated by throwing around money and resources. It's why I've been saying for a while now that by the end of this generation, the best looking games will be from the Sony first parties.

Heavenly Sword was on par with Uncharted Drake Fortune when it came out. At least when it comes to visuals and production values. They have regressed since then though. Especially since DmC which was a great looking game with fantastic combat despite what DMC fanboys might tell you. Hellblade to me was a massive disappointment. I dont care if 20 people worked on that game or 200, it had the Ninja theory game behind it and it felt like a glorified indie title. I have no wtf they were trying to do with Bleeding edge but ND since UDF has made U2, U3, TLOU2, U4, ULL and TLOU2. Four of those games won GOTY awards. NJ had so much potential.

If MS gives them the budget I think they have the talent to be where ND was 10 years ago and push out a U2 quality title.

 

[kyliethicc]

Agreed. I highly doubt they will reach the level of Naughty Dog. The Sony first parties have significantly more experience and they've been outputting games which have become hits on a very consistent basis. The infrastructure and the culture of the first parties is just something that cannot be replicated by throwing around money and resources. It's why I've been saying for a while now that by the end of this generation, the best looking games will be from the Sony first parties.

And to be fair to other studios, almost no one can reach the level of Naughty Dog. Their nonstop success for the last 15 years, both commercial and critical has made them essentially the top dog studio in all of gaming.

Of course it's debatable, but they're automatically in that debate with a few other studios, like Rockstar.

Its like saying Sony's Bend Studio will become the next Naughty Dog.. its not fair to expect that from them. Its basically impossible.

 

[assurdum]

 

[carloseduardocd]

"It's an error" - Alex Battaglia.

 

[Lysandros]

"It's an error" - Alex Battaglia.

Must be a glitch in the Matrix.

 

[kyliethicc]

what the fuck is going on with the XSX version (on the right) at 4:33 in that video? looks fucking bad as the camera pans up, there's like flickering or something

 

[Superayate]

what the fuck is going on with the XSX version (on the right) at 4:33 in that video? looks fucking bad as the camera pans up, there's like flickering or something

Simply think that they have divided by two the framerate in the video, as we can see how fast she runs just after

 

[kyliethicc]

Simply think that they have divided by two the framerate in the video, as we can see how fast she runs just after

Sure they're playing the footage back in slow mo but it looks weird from like 4:33 to 4:50 on XSX, but its smoother on the left on XSS.

Its like a weird flickering in the footage as the camera pans over the desk.. goes away when they cut to the next clip of her running.

 

[assurdum]

"It's an error" - Alex Battaglia.

"We already contacted Remedy. They said they are looking to it.".
Hitman 1800p on ps5.
"The 4K advantage Xbox owners are waiting for, people accept the reality , 18% of more GPU land to a 44% of more pixels, it's normal"

 

[SlimySnake]

what the fuck is going on with the XSX version (on the right) at 4:33 in that video? looks fucking bad as the camera pans up, there's like flickering or something

XSS version is 60 fps, the XSX graphics mode is 30 fps.

Thats how bad any 30 fps mode looks in comparison to 60 fps.

 

[SlimySnake]

"We already contacted Remedy. They said they are looking to it.".
Hitman 1800p on ps5.
"The 4K advantage Xbox owners are waiting for, people accept the reality , 18% of more GPU land to a 44% of more pixels, it's normal"

What I dont understand is why didnt Alex ask Remedy about the XSX graphics settings being the same as the PS5 settings which are mostly set to Low. Surely an extra 2 tflops at the same resolution shouldve allowed them to set at least some of these settings to Medium. RT contact shadows actually have a very low impact on the GPU, so that couldve been enabled. Otherwise, wtf is the point of having 44% extra CUs if you arent able to take advantage of the RT hardware in them?

I have watched both videos and they dont even mention the settings on the Xbox aside from a quick mention on the written review.

They have access to devs and dont bother asking these fairly common sense questions.

 

[ethomaz]

What I dont understand is why didnt Alex ask Remedy about the XSX graphics settings being the same as the PS5 settings which are mostly set to Low. Surely an extra 2 tflops at the same resolution shouldve allowed them to set at least some of these settings to Medium. RT contact shadows actually have a very low impact on the GPU, so that couldve been enabled. Otherwise, wtf is the point of having 44% extra CUs if you arent able to take advantage of the RT hardware in them?

I have watched both videos and they dont even mention the settings on the Xbox aside from a quick mention on the written review.

They have access to devs and dont bother asking these fairly common sense questions.

Welcome do reality... the TFs advantage doesn't matter that much.

 

[kyliethicc]

What I dont understand is why didnt Alex ask Remedy about the XSX graphics settings being the same as the PS5 settings which are mostly set to Low. Surely an extra 2 tflops at the same resolution shouldve allowed them to set at least some of these settings to Medium. RT contact shadows actually have a very low impact on the GPU, so that couldve been enabled. Otherwise, wtf is the point of having 44% extra CUs if you arent able to take advantage of the RT hardware in them?

I have watched both videos and they dont even mention the settings on the Xbox aside from a quick mention on the written review.

They have access to devs and dont bother asking these fairly common sense questions.

He also never asked how these PS5 settings compare to the PS4 Pro version.

I bet they just took the PS4 Pro version, used the same settings, raised res to 1440p from 1080p, and locked it at 60 Hz. Then added some RT and capped it to 30 FPS for the graphics mode.

 

[FrankWza]

What I dont understand is why didnt Alex ask Remedy about the XSX graphics settings being the same as the PS5 settings which are mostly set to Low.

He also never asked how these PS5 settings compare to the PS4 Pro version.

Have any of them mentioned the s and how it is or isn’t detrimental to overall development? I know they mentioned cross gen in at least a couple of videos but I have not seen series s mentioned.

 

[Insane Metal]

Another multiplat where 10.2 TF performs pretty much on par with 12.1TF. Hmm.

 

[SlimySnake]

Another multiplat where 10.2 TF performs pretty much on par with 12.1TF. Hmm.

 

[kyliethicc]

Another multiplat where 10.2 TF performs pretty much on par with 12.1TF. Hmm.

"Its dangerous to rely on tittyflops as an absolute measure of performance."

- Mark "Father of Knack" Cerny

 

[assurdum]

What I dont understand is why didnt Alex ask Remedy about the XSX graphics settings being the same as the PS5 settings which are mostly set to Low. Surely an extra 2 tflops at the same resolution shouldve allowed them to set at least some of these settings to Medium. RT contact shadows actually have a very low impact on the GPU, so that couldve been enabled. Otherwise, wtf is the point of having 44% extra CUs if you arent able to take advantage of the RT hardware in them?

I have watched both videos and they dont even mention the settings on the Xbox aside from a quick mention on the written review.

They have access to devs and dont bother asking these fairly common sense questions.

Alex confirmed both console have the same graphic setting on Resetera. Funny stuff: Alex when analised the one X version he claimed everything was medium but he was (obviously) completely wrong looking to the data provided by Remedy about the next generation of console

 

[DJ12]

Another multiplat where 12.1 TF performs pretty much on par with 10.2TF. Hmm.

Fixed it for you.

Xbox not quite there for parity yet, even when performance is being left on the table it just cannot keep up.

 

[TheThreadsThatBindUs]

Heavenly Sword was on par with Uncharted Drake Fortune when it came out. At least when it comes to visuals and production values. They have regressed since then though. Especially since DmC which was a great looking game with fantastic combat despite what DMC fanboys might tell you. Hellblade to me was a massive disappointment. I dont care if 20 people worked on that game or 200, it had the Ninja theory game behind it and it felt like a glorified indie title. I have no wtf they were trying to do with Bleeding edge but ND since UDF has made U2, U3, TLOU2, U4, ULL and TLOU2. Four of those games won GOTY awards. NJ had so much potential.

If MS gives them the budget I think they have the talent to be where ND was 10 years ago and push out a U2 quality title.

Ninja Theory has potential to be a great studio. Not ND level (very few devs are), but they are very talented.

Imho, Heavenly Sword was their best game in terms of graphical presentation, overall production values etc.. That said, the engine and all the tech belonged to Sony, so when they parted ways, NT had to move to UE3, which is why Enslaved their next game was such a big step down.

I thought Enslaved was pretty mediocre myself. It was entertaining, but I just couldn't get past the horrible "UE3 look" of the graphics that like 99% of third party games had during the PS3/X360 gen. I also kinda felt the riff on "Journey to the West" was pretty old and played out, so that didn't help. Also, Andy Serkis as muscle-bound gripper didn't really work.

 

[SlimySnake]

Welcome do reality... the TFs advantage doesn't matter that much.

I dont know if thats true. More tflops has traditionally offered more performance. we saw this with the PS4 and Xbox one then again with the PS4 Pro and X1X. We see this in PC all the time. A 5700xt will never perform like a 5700. My rtx 2080 will always perform better than a 2070, and not because of some magical secret sauce in it. But because it has more tflops or more shader processing power.

What's happening here simply doesnt make much sense. Maybe the PS5 is overperforming because of the insanely high GPU clock, but mostly because there is something seriously wrong the xsx. I dont know if there is a bottleneck in the GPU design or the split ram bandwidth is causing issues. Maybe it IS the tools, but an extra 20% raw horsepower and an extra 44% more RT hardware shouldve allowed them to push either high resolutions or higher graphical settings.

If this wasn't happening in pretty much every multiplatform game, I would think Sony was paying Remedy for parity. Hopefully, devs and MS can sort this out and we can start to see better settings in XSX versions of games for the sake of Xbox owners. People bought that system thinking they were getting a more powerful console and getting an identical version of the game with no improvements is simply disheartening.

 

[TheThreadsThatBindUs]

I dont know if thats true. More tflops has traditionally offered more performance. we saw this with the PS4 and Xbox one then again with the PS4 Pro and X1X. We see this in PC all the time. A 5700xt will never perform like a 5700. My rtx 2080 will always perform better than a 2070, and not because of some magical secret sauce in it. But because it has more tflops or more shader processing power.

What's happening here simply doesnt make much sense. Maybe the PS5 is overperforming because of the insanely high GPU clock, but mostly because there is something seriously wrong the xsx. I dont know if there is a bottleneck in the GPU design or the split ram bandwidth is causing issues. Maybe it IS the tools, but an extra 20% raw horsepower and an extra 44% more RT hardware shouldve allowed them to push either high resolutions or higher graphical settings.

If this wasn't happening in pretty much every multiplatform game, I would think Sony was paying Remedy for parity. Hopefully, devs and MS can sort this out and we can start to see better settings in XSX versions of games for the sake of Xbox owners. People bought that system thinking they were getting a more powerful console and getting an identical version of the game with no improvements is simply disheartening.

The problem with your thesis is that with PS4 and XB1 and with PS4 Pro and XB1X there was more than a 40% gulf in raw theoretical shading performance between them.

This time it's a much smaller difference, i.e. 18%, and don't forget we're talking about maximum theoretical performance. Actual performance in real game will be lower than this and will be largely governed by the efficiency of the design.

PS5 has a faster GPU front end and render back end. Faster internal cache bandwidth. It has cache scrubbers that increase further effective cache bandwidth and a single unified pool of RAM instead of spilt, interleaved solution XSX went for. It also has double the SSD bandwidth and more potent decompression hardware.

The smaller the difference in theoretical flops, the bigger the difference these other features of the design will make in real game performance. To the point, like we're seeing, where games are consistently performing better on the system with the lower TFLOPs.

TL;DR, there's more to GPU performance than tittyflops (thank you kyliethicc , I'm borrowing this), because these consoles are doing much more than just shading pixels all day.

 

[ethomaz]

Alex confirmed both console have the same graphic setting on Resetera. Funny stuff: Alex when analised the one X version he claimed everything was medium but he was (obviously) completely wrong looking to the data provided by Remedy about the next generation of console

When Alex analyzed the last gen versions? Because that is a new version... not the old Control.

PC is a new version too.

 

[ethomaz]

I dont know if thats true. More tflops has traditionally offered more performance. we saw this with the PS4 and Xbox one then again with the PS4 Pro and X1X. We see this in PC all the time. A 5700xt will never perform like a 5700. My rtx 2080 will always perform better than a 2070, and not because of some magical secret sauce in it. But because it has more tflops or more shader processing power.

What's happening here simply doesnt make much sense. Maybe the PS5 is overperforming because of the insanely high GPU clock, but mostly because there is something seriously wrong the xsx. I dont know if there is a bottleneck in the GPU design or the split ram bandwidth is causing issues. Maybe it IS the tools, but an extra 20% raw horsepower and an extra 44% more RT hardware shouldve allowed them to push either high resolutions or higher graphical settings.

If this wasn't happening in pretty much every multiplatform game, I would think Sony was paying Remedy for parity. Hopefully, devs and MS can sort this out and we can start to see better settings in XSX versions of games for the sake of Xbox owners. People bought that system thinking they were getting a more powerful console and getting an identical version of the game with no improvements is simply disheartening.

RX 5700 XT performs better than Vega 64.
Just to show one example.

 

[HoofHearted]

"Its dangerous to rely on tittyflops as an absolute measure of performance."

- Mark "Father of Knack" Cerny

You do not want to see what I found when googling "titty flops"...

 

[MasterCornholio]

You do not want to see what I found when googling "titty flops"...

I've seen them in episodes of botched.

 

[TheRedRiders]

The problem with your thesis is that with PS4 and XB1 and with PS4 Pro and XB1X there was more than a 40% gulf in raw theoretical shading performance between them.

This time it's a much smaller difference, i.e. 18%, and don't forget we're talking about maximum theoretical performance. Actual performance in real game will be lower than this and will be largely governed by the efficiency of the design.

PS5 has a faster GPU front end and render back end. Faster internal cache bandwidth. It has cache scrubbers that increase further effective cache bandwidth and a single unified pool of RAM instead of spilt, interleaved solution XSX went for. It also has double the SSD bandwidth and more potent decompression hardware.

The smaller the difference in theoretical flops, the bigger the difference these other features of the design will make in real game performance. To the point, like we're seeing, where games are consistently performing better on the system will lower TFLOPs.

TL;DR, there's more to GPU performance than tittyflops (thank you kyliethicc , I'm borrowing this), because these consoles are doing much more than just shading pixels all day.

Mark Cerny stressed cache coherency and latency in the Road to PS5 talk.

I remember LeviathanGamer2 (hardware engineer) stated that PS5 GPU was designed in a way so that developers could target a high cache hit rate which is important for performance, likewise Matt Hargett the Playstation engineer stated that we'll see better performance from the PS5 overtime as developers learn to hit the caches more.

 

[kyliethicc]

Funny stuff: Alex when analised

surely

 

[Superayate]

The problem with your thesis is that with PS4 and XB1 and with PS4 Pro and XB1X there was more than a 40% gulf in raw theoretical shading performance between them.

This time it's a much smaller difference, i.e. 18%, and don't forget we're talking about maximum theoretical performance. Actual performance in real game will be lower than this and will be largely governed by the efficiency of the design.

PS5 has a faster GPU front end and render back end. Faster internal cache bandwidth. It has cache scrubbers that increase further effective cache bandwidth and a single unified pool of RAM instead of spilt, interleaved solution XSX went for. It also has double the SSD bandwidth and more potent decompression hardware.

The smaller the difference in theoretical flops, the bigger the difference these other features of the design will make in real game performance. To the point, like we're seeing, where games are consistently performing better on the system will lower TFLOPs.

TL;DR, there's more to GPU performance than tittyflops (thank you kyliethicc , I'm borrowing this), because these consoles are doing much more than just shading pixels all day.

On the paper, yes it's obvious that the the differences at specs level was lower between XsX and PS5 than what we had with the previous gen, where PS4 was clearly stronger in all GPU specs over XO and memory part. For the midgen, XOX was also clearly stronger on most GPU specs. PS5 vs XsX it's clearly different, with some specs, that's PS5 that has the lead, and on others that's the series X. So it's not surprising that overall they offer similar performances. I thought that the differences could be higher depending on the games (when the PS5 has le lead or the XsX). Honestly, first time we have so close consoles.

 

[SlimySnake]

RX 5700 XT performs better than Vega 64.
Just to show one example.

different architectures. vega 56 doesnt outperform vega 64 or vega 7 for example.

This time it's a much smaller difference, i.e. 18%, and don't forget we're talking about maximum theoretical performance. Actual performance in real game will be lower than this and will be largely governed by the efficiency of the design.

But again, the difference between 70 series Nvidia cards and 80 series cards is usually 18-20% and that always translates into 18-20% better performance. Same goes for the 5700 and 5700xt and 6800 and 6800xt.

I am not saying I disagree with you because I do believe the PS5's higher clocks, unified ram pool and custom APU is mitigating the tflops gap. Hitman of course not withstanding.

 

[IntentionalPun]

The question about XSX is whether the extra CUs are currently being utilized, or if they aren't, if they ever will be.

I like Cerny, but you guys hang on every word like he isn't someone marketing his product first and foremost.

 

[Tchu-Espresso]

I've seen them in episodes of botched.

...why do that to yourself?

 

[MasterCornholio]

The question about XSX is whether the extra CUs are currently being utilized, or if they aren't, if they ever will be.

I like Cerny, but you guys hang on every word like he isn't someone marketing his product first and foremost.

I don't think it's a question of those CUs being used. I think it's a question if they are being used efficiently.

 

[MasterCornholio]

...why do that to yourself?

It was basically a video of a plastic surgeon reacting to those videos. Still saw all the nasty stuff though.

 

[kyliethicc]

The question about XSX is whether the extra CUs are currently being utilized, or if they aren't, if they ever will be.

I like Cerny, but you guys hang on every word like he isn't someone marketing his product first and foremost.

Thing is there's more to it than that.

PS4 Pro is better than PS4 in every way. So games should always run best on Pro. If a game was not running better on Pro, then it'd make sense to conclude its not being used fully.

Similar with PS4 Pro and Xbox One X. The One X is the better spec system in basically every way.
If a game was not running better on OneX, then it'd make sense to conclude its not being used fully. (Like Hitman 3.)

But the PS5 and XSX are not like that. PS5 has some legit advantages and therefore when a game is running same on both, or better on PS5 or XSX, its not logical to assume the XSX or PS5 are being under-utilized. It could just be thats legit the best the systems can do.

Neither system is clearly better spec overall, despite 12 being a larger number than 10.

 

[IntentionalPun]

I don't think it's a question of those CUs being used. I think it's a question if they are being used efficiently.

Cerny specifically talked about how their narrower design is easier to design for as CU occupancy (how many CUs are used at any given time) is less of an issue, and that the CU's of the PS5 will run faster.

If a next-gen game is occupying the same amount of CUs on either system, the faster CUs on the PS5 will perform better.

 

[IntentionalPun]

But the PS5 and XSX are not like that. PS5 has some legit advantages and therefore when a game is running same on both, or better on PS5 or XSX, its not logical to assume the XSX or PS5 are being under-utilized. It could just be thats legit the best the systems can do.

Well I'm not assuming that; hence why I said "whether they are currently being utilized."

That's really the room XSX would have to grow it's performance over the PS5; if they currently are being utilized and the system is still only on-Par with PS5 then it doesn't have that room to grow.

PS5 of course likely has room to grow performance w/ it's more custom features; but we also don't know for sure how much of that is currently being used or not, or for which games. Same with XSX; we just don't have the details.

 

[MasterCornholio]

Cerny specifically talked about how their narrower design is easier to design for as CU occupancy (how many CUs are used at any given time) is less of an issue, and that the CU's of the PS5 will run faster.

If a next-gen game is occupying the same amount of CUs on either system, the faster CUs on the PS5 will perform better.

I honestly don't believe the developers are not using those additional CUs. Maybe they are just struggling to keep them fed. Maybe more shader arrays and the same CU count can fix the issue?

That's what AMD did with their desktop GPUs. Instead of increasing the CU count per shader array they increased the shader arrays.

 

[IntentionalPun]

I honestly don't believe the developers are not using those additional CUs. Maybe they are just struggling to keep them fed. Maybe more shader arrays and the same CU count can fix the issue?

I don't believe or disbelieve personally.

Nobody but developers who have done profiling of their own games across both systems knows.