ฝัง
- เผยแพร่เมื่อ 27 มิ.ย. 2024
- Welcome to today's mind-blowing episode where I uncover the hidden truths behind the Steam Deck OLED's performance at lower TDPs! 🕹️💡 While it's easy to assume that lower TDP means weaker performance, the reality is far more fascinating and complex.
🧐 Here's what you'll learn:
- 💻 Steam Deck OLED at Lower TDPs: Why the performance metrics don't tell the whole story.
- 🔋 Efficiency Over Power: How Valve has optimized the Steam Deck OLED for a balance between battery life and performance.
- 🎮 Real-World Gaming Tests: Demonstrating the device's capabilities with popular titles at different TDP settings.
- 🛠️ Tweaking for Maximum Efficiency: Tips and tricks to get the most out of your Steam Deck OLED.
👀 Prepare to have your mind blown as I unravel the mysteries behind the Steam Deck OLED's performance.
📌 Don't forget to hit that like button, subscribe for more deep dives into gaming tech, and ring the bell for notifications!
👇 Drop your questions and thoughts in the comments below - let's discuss!
#SteamDeckOLED #Valve #steamdeck
Chapters
00:00 - SteamDeck OLED is WORSE?!
03:20 - SteamDeck LCD Vs. OLED Benchmarks
11:11 - Uncore Power Matters, Forcing 400Mhz on RAM
13:56 - Why Does The Steam Deck OLED Have "Worse" Performance at 5 Watt
15:00 - 5 WATT TDP CAP - Steam Deck LCD CPU Workload and Gaming Workload
16:17 - 5 WATT TDP CAP - Steam Deck OLED CPU Workload and Gaming Workload
18:10 - 15 WATT TDP CAP - Steam Deck LCD CPU Workload and Gaming Workload
22:32 - 15 WATT TDP CAP - Steam Deck OLED CPU Workload and Gaming Workload
24:47 - How TDP Works In Different Scenarios
26:08 - Best Case Battery Life on Steam Deck LCD and OLED
29:20 - Worst Case Battery Life on Steam Deck LCD and OLED - เกม
Thank you so much for this-- as a recent entrant to the hobby, I'm so often confused by all the technical aspects of these handhelds, and videos like this just breaking down how some of these basic concepts *actually* work are tremendously helpful
I'd say this is above basic, but yes
I would love for Phawx to do a series educational videos on the meaning and significance of all the specs, acronyms, etc.
@@bipbop3121 the *video* goes beyond basic, but I meant that "TDP" is a "basic concept," not that the video itself was basic
Yep, its good there are someone exposed Steam deck issues instead of trash tuber who shilling for valve and steam.
I admire your restraint by avoiding phrasing it as "Sephiroth kills Aerith" [when it comes to efficiency]
WOW how did i not think of this lol
I just wanna say thank you for such detailed videos. As a recent handheld fan, your channel is by far the most interesting on TH-cam for me.
Id much rather have a measurement of total system power because it seems easier to calculate battery life.
Well that’s what the overlay is giving you and what it’s using to calculate battery life
I think another good way to get this point across be a line chart of FPS in a game against total system power (at its associated TDP settings). This should show the point at which the OLED steam deck should hit the same performance at a lower total system power(/its associated TDP setting).
I think that visualisation may be a nice way to aid in explaining how these better power characteristics help with batter life.
Lol Steam Deck oled is already a flop just like Switch oled. Only blinded Valve or steam fanboy overhyping this piece of low quality turd. At least we still have other much better PC handheld from other vendor like Asus.
@@runninginthe90s75 your brain dead take aside, I commented an idea of another way to do data visualisation because it's something that would help me pick a handheld (including from other vendors 😂) where in that you got that I was a fanboy is beyond me especially as this is the steam deck compared to itself 😂
P.S. when they sell over half a million of these (regardless of mine and your opinions about the device) in the next year come back tell me it's still a flop especially considering they've sold close to 3 million "inferior LCD ones" 😂
@@ignotassalna8958 He's just trolling the comments. He repeated the same two comments twice each as replies to the first four comments. He's either a fanboy himself of one of the other handhelds or just trolling for the sake of trolling.
I won't use a TDP cap . if the game is too demanding I cap the fps at 30 or 40 (depending on the game and settings) and when it's lighter I cap it at 60 . So I let the APU use whatever amount of power to keep the framerate locked at any demanding or non demanding moment of the game .
in almost all circumstances, that is the best way to go
Agreed. Some games, like Hades, use more power than they have any right to and capping TDP has no effect on performance
Thanks for the video. I discovered your channel when searching for information on the new OLED Steam Deck and your videos have been incredibly helpful. Can't wait to get my Deck in to start tinkering.
Thanks for the time dedicated to make this video, you are definitely my go to when it comes to handhelds, it's hard to trust anyone else :)
This was your most impressive thumbnail effect so far. I was laugjing about you wathing the Benchmark results❤😂
This almost seems like an extra cap they put in to keep the Sephiroth performance at the same target as the Aerith performance but didn't realize that the method would not match at lower TDP settings.
In my mind Valve would want you to be able to get the same performance on the same settings on both Deck versions so that when a friend or youtuber recommends you some settings for a game, you can just use them yourself without thinking. Only the lower TDP settings now don't match up. I wonder if they'd want to adress that in firmware.
Agreed.
It seems that they're actually going to apply the new TDP profile to LCD version as well in the next OS update
I hope they can lift this in an update instead of take down the LCD with it. Would be cool to have that extra bit of headroom
so that means battery life will increase on LCD?@@didjidks
the steam deck never favored performance over consistency, so id rather have the stable power draw instead of extra headroom. this new cap should help battery life in the LCD model too, which especially makes sense because Valve still sells the LCD model as the new base model.@@ThompYT
Highly recommend showing a weighted average of power usage for this kind of data. Around 24:00 you identify the big weakness for a pie chart to represent this data, where not all >16W readings are equal. Weighted averages would probably give a rough picture of how much additional energy the LCD is using at the same TDP as the OLED and give an idea of how many additional W of TDP cap a user would need to give the OLED to get roughly equivalent power usage at the CPU/GPU level. For instance, I get 14.6 average W for OLED GPU bench based on your pie chart and 15.1 for LCD (results approximate because I guessed for unmarked
For the actual average wattage in a gaming workload =
15w cap on OLED averages = 14.7389881 watts
15w cap on LCD averages = 15.27805393 watts
So you were very close with your approximation! The thing that is interesting is even though we see we are "averaging" around the correct TDP this has huge implications on total system power used. between LCD and OLED we see a ~10% efficiency gain on total system power on OLED. Meaning, if LCD uses 5.5w, OLED will use 5w. Or if LCD uses 24w, OLED will use 22w total system power.
Thanks for the suggestion on the weighted average, my goal in this sense was to show the ratio split is the same between LCD/OLED, I didn't think of using a weighted average at all
very clear explanation, thank you Phawx for another awesome video!
As someone that put a TDP cap an nearly every single game. This was very helpful, so thanks for doing this work. I'll have to adjust all my games but I'm still real excited for the oled model.
Thanks Phawx. Your channel continues to be like school for handheld tech. I learn new things every time I watch.
I think this is the first time I've come across your channel and holy shit, that intro was great. Transitioning from the thumbnail to the start of the video. Just because of that, I'm subscribing
Thank you so much for all of the work to thoroughly explain something that seems pretty simple, and for giving us information that bigger outlets neglect to test. I think you should increase font size on those pie charts if possible to make it easier to read for phone users. Keep it up! ❤
This video was very informative; I didn't know how TDP works, so this was very interesting, especially the pie graphs showing the ratios.
Thanks for the explanation, I was a bit worried as I play LOADS of games at around 3 ~ 6W settings 😂 (mostly lightweight indie games on my deck)
No reason to worry, just up it a couple watts and you'll effectively be using the same power.
Just up it to 7 watts and you'll be using basically the same power as you would on 5 watts on the old steam deck
Lol Steam Deck oled is already a flop just like Switch oled. Only blinded Valve or steam fanboy overhyping this piece of low quality turd. At least we still have other much better PC handheld from other vendor like Asus.
@@runninginthe90s75 Historically speaking, higher specced devices in the portable sector have rarely done that well going back to the early 90's, mainly because it's a balancing act from performance and battery life, and even though rivals offer better performance, you're having to throw more juice at it to get that, and even when not doing that, the performance uplift over the Deck ends up being quite small of around 15%
But don't get me wrong, I would love more performance, but not if it means having worse battery life, which for me, the original Deck was already pushing the limits on what they can get away with in portable form, the rivals to it don't cut it where it counts for a lot of gamers, whereas the OLED have improved in the hours you can play whiles pushing hard, that matters a lot, far more then having better specs for a lot of gamers.
Nintendo is another company that understands that, it's not about performance but portability, so even thought I like what's going on with the ROG and Legion Go, it's not going to interest a lot of gamers that want the entire package and is mostly aimed at higher end gamers, which is fine, but it's always going to have limited market share compared to what the Steam Deck can have, that matters because developers go on market share, in other words, the Steam Deck wins out by being an overall better package, even thought it's the weakest hardware of the bunch.
Besides, who the heck wants Windows on a portable device? That's a massive selling point in favour of the Deck over these rivals, which is ironic how a Linux OS in SteamOS could become a better option for playing Windows games, that shows how far SteamOS has gone for it to be a none issue for most gamers.
A little advice to the rivals, if they really want to gain any traction, offering a SteamOS option would go a long way, especially if it's well integrated, also offering a cut down version with the core specs being the same but cuts in other areas to reduce the price, as it is, the price point of these rivals is too high to really make a dent, whereas Valve have got it right in offering a $400 version, that is where the game changer is.
To be blunt, the Steam Deck isn't targeting PC elites like you, it's targeting the core gaming audience from PC and consoles, they matter far more than people like you, hence why most games target consoles and games that do target the PC, usually target modest hardware, hence why Valve are playing this game smart compared to the rivals.
There is a silver lining in all this, the rivals do have the advantage over laptops, which cost much more for that level of performance, but that's because laptops are way overpriced and need to come down, hence there are many buying the ROG or Legion Go as a laptop replacement as it's a better deal then other laptops on the market, but again, that's not the market Valve is targeting and that won't help them if the laptop market starts competing again and reducing the price point, which is an area Valve doesn't have to worry about but Asus and Lenovo do.
@@runninginthe90s75ok thanks for your opinion that nobody asked for
Cary that was an excellent video, very informative and wonderfully presented.
Thanks!
Amazing analysis as always. I have much better understanding of how to squeeze battery life on UPMCs thanks to your videos.
Really interesting - thanks! Would be really interested to see whether there is any more undervolting potential with the OLED, given that the SOC is a slightly smaller process.
Really good video man, well done
Im liking the way your doing your thumbnails!
Yo, great vid! Looks like it's placed normally in the search of steam deck oled ^^ (Unlike the old steam deck videos which had the odd discoverability issues.)
Thanks for letting me know!
The phawx coming in with the facts again. Thanks!
Great video as always, great overview of steam deck power consumption!
11hr battery life is crazy!
Been waiting for this one 🍿
This is a unique video about steam deck oled battery and watt usage. As a deck LCD user with 10tdp cap all the time, I wonder how 10tdp lcd vs 8tdp oled battery would be? Great video, keep up the good work
Great job
0:04 love the intros. 😂
Thank you so much, this was exactly what I was wondering about when looking at some other reviews! It did take me a few moments to fully realise what it was you were saying. Maybe it would help to overlay the GPU/CPU power draw graphs with different colors so its immediately obvious?
Now i'm interested how it splits up power in that lower tdp limit
I think this is why taki udon did his teating with the overall battery drain metric, and not straight TDP limit
Aerith and Sephiroth...nice touch!
Great intro bit haha
Very very interesting. So the TDP cap is not a constraint on the maximum of the distribution but a constraint on the average (a target really since you can't 100% guarantee where the average will be). That makes sense and gives a lot of optimization headroom. One thing I didn't get though is why is the OLED more efficient? Is it because the wattage distribution is shifted to the left? But that would only be more efficient if they both achieved the same performance right? It seems like the hard TDP cap makes it easier to better control the average TDP but it also sacrifices some performance (at a lower power) on the way?
6nm is slightly improved over 7nm. And for most games oled display itself should use less power.
"But that would only be more efficient if they both achieved the same performance right?"
They did, look at the performance tests. On 15w OLED is either the same or even better compared to LCD, meanwhile also consuming less power. On lower TDPs the performance is worse because LCD deck just consumes unproportionally more power. This is why it was also "better" on tests compared to other handhelds, but it was a cheat. It wasn't some super efficient optimization at all, the opposite actually.
To be precise, I cannot really tell if those "same or better" results are due to TDP optimizations since OLED also comes with overclocked RAM which probably contributed more to the performance. So that's not really a fair comparison
@@didjidks ahh, got it. At 15W the OLED is actually both more performant and total power draw is shifted to the left, hence more efficient. At 5W the LCD Deck can't actually hit the 5W TDP target and actually uses more power, hence the cheat. Got it, got it. Thanks for the explanation!
Generally speaking from my limited testing is 10w LCD vs 10w OLED will produce roughly the same performance. And when all settings are matched between both decks we are looking at around ~10% better total system power on the OLED.
@@ThePhawx got it. That would be the equal performance scenario and also shifted to the left/ more efficient 💯. Would be interesting to see how the OLED Deck performs in the 5-15W range vs the Z1 Extremes now that it better honors the TDP target
Hey love your videos you are the man I go to when it comes to handhelds etc would love to see some testing with input latency with this new screen as I notice a huge increase in latency when I limit my deck to 30fps almost unplayable for me, I know oled tvs/monitors tend to have less input latency I was wondering if this is the case with the new screen?
We should maybe stop calling it a TDP "cap" on Steam Deck and call it like a "target average" instead. The results here show it's not treated as any kind of limit, especially for the GPU.
Professor Phawx always teachin' us something
Thank you my man 🙏
Thank you so much! Really appreciate that
I'm just trying to figure out what this means for me as a consumer, cause even after the full explanation it still sounds like worse performance. Even at the end showing the last of us, we see the arrow pointing to the wattage being more efficient on OLED, using 21.5w instead of 24.5 on the lcd, BUT you can see the LCD gets a 5fps bump in performance due to using more power. So why should I care? I'm getting less fps. How do I as an OLED owner get that 35fps in the last of us? How can I boost and use more power like the LCD does to get that? Am I just capped and gonna see less power potential due to these "efficiencies"? This is what I'm trying to figure out. I just care about the end result, I play plugged in and maxed out.
This is indeed strange since on his graphs OLED had better performance
This also makes the comparison not so valid because obviously OG deck would have worked longer on 30 fps too
to dumb it down... "TDP" setting is not comparable across platforms. 5 TDP on the SD OLED is not the same as 5 TDP on SD LCD. THe SD LCD will draw more power and eat battery up faster at that same settings, hence it will seem to run faster because its technically drawing more power. This means setting 5 TDP on a SD OLED is an even lower setting at lower power draw than is possible on the SD LCD. Run your SD OLED at a setting of 7 or 8 might be similar to setting SD LCD to 5. Someone would have to do the math and benchmarks to check what would be closest in power draw.
@@AyaWettsbut when tdp is uncapped and the device is free to try as hard as it can the LCD is shown with higher tdp and higher fps while the OLED has less tdp and fps. So it seems the ceiling is lower on OLED. At least in the last of us example at the end
@@mrexpendable1232 There is no uncapped. They have it currently capped at 15... because they are focusing on efficiency and power, not speed... and in most cases its just as fast. The OLED SD is running lower power even when "maxed" out. A little overclocking and it might be nice to see what that APU can do, when not behind Valve's conservative limits.
tbh i've always set the limit on the frame rate rather than mess with the TDP, the battery estimates remain the same, but it just remains a very consistent amazing experiance. if it needs power for a bit it just draws it for a second without creating a dip
Your intros are always hilarious. 😂
I hope you can explain the power efficiency of the Win Mini against the other GPD 7840u handhelds once the official IGG backer version is ready.
I discovered this when I ran BOTW on Cemu capped at 3W TDP but it was actually using around 11.
Even by your own high standards that was a great video start today.
No one can match ThePhawx's thumbnail game.
Wow thanks!
Great way to do lunch today 😊
One suggestion for future videos, can the font of the benchmarks be increased a bit? It's a tad difficult to watch on a phone
Agreed it's a good lunch break video. Did you notice he was going in and out of focus and when the charts were added it seems like his audio is slightly out of sync?
Part way through watching this I found myself wondering about analysis vs overanalyses. But I am glad that I watched it to the end as there was a conclusion that summed up TDP for me. Just like HDR and other industry terms, they are full of ambiguity, wholly misused or misrepresented (what we normally cite as lying to children) and not worth the storage that 'standards' are written on. 1024 (accurate) vs 1000 is a classic industry misuse of basic science to make a product seem better than it actually is. So TDP is, rather like an Antutu benchmark, a top level guide to give you a very basic idea of the possibility of performance, even though the real world use will not always live up to that for varying reasons.
Short answer: Valve is gimping the new chipset for battery life?
It still would've been nice if Valve asked for a 780m, they already got an APU on a new lithography so it shouldn't have been too much more for them to ask for what would've given them relative parity with the Ally and other Z1 based handhelds.
So this means we really need more battery life tests attached to the 7840u vs. 6800u benchmarks.
And I assume "TDP" management can change per-handheld, this makes benchmarking a huge pain, I wonder if "TDP improvements" in the firmware of some handhelds isn't just lifting power draw restrictions. Maybe we're all in an arms race to have TDP ratings that lie the most?
Since max power consumption of Deck Oled is 24.5W could a 25W PD powerbank be enough to keep it going without discharge the internal battery and actually charging it very slowly? Cause those powerbanks are much smaller and lighter to carry around or even mount to the back of Deck
This makes me curious what the new tdp efficiency sweet spot is for the OLED Deck. LCD sweet spot was 9-11W.
So the new OLED model boosts less in power hungry situations at lower TDPs? I wonder how the TDP curves are stacking up against the Z1 Extreme and 7840u chips, since the steam deck used to be the low TDP king.
Also, I wonder how much different the performance will be in 2d indie games
its still the low TDP king... he's trying to explain that the TDP number isn't comparable across platforms. like a setting of 7 TDP on the SD OLED could be the same as a 5 TDP on SD LCD, which means you can basically just run even lower power settings on the SD OLED than you could ever get on SD LCD.
The total weight of power distribution in a device is key. Chipset TDP is most of, but not all of the weight.
Is this why GPD 's 7840u always some magical improvement over other brands ' when at the same TDP?
Hi phawx do you plan to do a z1 not extreme analysis ?
not sure it's already asked but is there any tool or way to change the hard cap ourselves, is it even possible to update it or is it something hard wired and can't be changed? I understand valve/AMD may not provide option OOTB but what about something like hacked BIOS.
Hmmm so to keep it more or less performant compared to the LCD version and to increase battery life, Valve upgraded the speed of the ram to compensate for the new power limit imposed on the OLED model.
It might be a little premature to assume this is how it's always going to be, Valve can tweak a lot of stuff in OS and BIOS updates. This might be the new behavior for all models, or it might be a pre-launch bug, who knows.
it doesn't matter, everyone with working brain will accept that this is how it works between lcd vs oled *in current state*. if there are new improvements, people can just test them out again.
@@noname76787 I doubt it'll get re-tested though and even if it is, this old info will keep getting passed around. Always happens in these niche circles, I just recently spoke to someone who insisted SD has a plastic screen on the 64 & 256 GB models... speculate on something once and it never goes away. I'm just saying he could qualify his statements better, the world is full of dummies that need to be spoon-fed.
@@AFistfulOf4Kit doesn't matter who gonna re test, this way of thinking is just too narrow. and you are worrying too much. some random content creator will do it eventually even if this content creator won't. also my first point on previous reply already says everything: everyone with the right mind understand that this is a comparison at that exact point of time, and you mentioned as well that Valve might tweak a lot of stuff. If you can figure that out I'm pretty sure whole internet which is like you and me can figure that out as well. Pretty standard
Is this OLED screen DC dimming or PWM low frequency dimming, like the switch OLED screen? Thanks😅
pwm you can si the screen flick when the screen is at minimum
Hey Phawx can you tell us what was the battery life on Steam Deck OLED for the Dead Cells "best case scenario" testing?? You noted 7h 45min for the LCD model, but got sidetracked with panel power usage and never said the Steam Deck OLED battery test results. :)
oh sorry! 10 hours and 45 mins
Brightness at 50%, wifi/BT were off
If you yanked the battery but plugged the device in, could you measure the inputted power?
Suprised at the oled temperature at the lowest wattage, I get 36 degrees on the lowest wattage on my ally which is like 7 watts.
TDP doesn't mean much nowadays anyways. Both Intel and AMD calculate TDP differently. It would be wiser to use an actual power cap.
That's incorrect.
Whatever you need to say to make the deck better for you ❤
@@X862goagreed
Lol Steam Deck oled is already a flop just like Switch oled. Only blinded Valve or steam fanboy overhyping this piece of low quality turd. At least we still have other much better PC handheld from other vendor like Asus.
@@runninginthe90s75 lols calling out valve when 90 percent of pc users get their games from steam 🤦♂️ your funding them it’s amazing thanks for your money. Bro if the legion go was that good digital foundry would have said something but they have said nothing 😂 only think they have said is steamdeck oled is the best handheld on the market and it’s true
Histograms are the superior chart :)
So why does cyberpunk run worse at 15tdp on the OLED? How come other videos show it runs better? Is it uncapped tdp or something?
The lows are better on the oled tho. Cyberpunk is also hard to get a like for like scenario since it’s open world with npcs everywhere and whatnot
Because 15 TDP on the LCD Deck and 15 TDP on the OLED aren't the same thing. LCD Deck is always drawing more power than the limit says it is, OLED is way more accurate to whats actually being drawn.
@@WhiskeyAfterHours oled is not running worse at 15wats
@@alex1212456 I didnt say it was try to reread what I typed.
I feel like this isn't surprising really. TDP is Thermal Design Power. It's a metric meant more for judging what the cooling solution should be to mitigate heat generated. IIRC, the TDP measurement is supposed to be a rating of how much heat needs to be dissipated over time(I'm not sure what that window of time is).
The way I understand it is that, watts are the same but thermal energy is slower to transfer so you can fluctuate electrical watts a lot more and the resulting thermal energy ends up being an averaging. All that matters in that case is that the heatsink and fan(or whatever cooling solution is used) can dissipate it all fast enough that it doesn't keep building.
With special appearances from Anbernic Arc
Is this the same for the low performance in the z1 extreme / 7840u in 5w tdp?
Remember that the 7840u and Z1E family of chips have double the cpu cores and more gpu cores to power, and will always be power starved as it costs power/energy just to have those extra cores. In theory if you could disable/park half of the cpu cores then more power would be available for the remaining cores & gpu. But even then there’s a rather large physical size difference between those apus and power gating/core parking would only go so far. I would estimate that the 7840u family of chips would be able to match or exceed the steam deck at 10w if there was a way to power gate portions of the apu (cpu disabling/core parking).
Really nice video. Also solved a worry I had: some light games already work great at 3W TDP on the LCD model, but will draw more if unrestricted.
On the OLED, with the APU being more efficient, I was thinking that not being able to easily set it lower than that would be a waste of power, but it seems you answered that. Just set it to 3W again and it is actually lower. Fantastic 😀
I am very pleased that once again there is proof that TDP is a useless term.
i guess i will put a lot of black /dark scenes on my game to optimize it for oled devices (to use less energy saving it for when it need more performance =P)
Any word on Cryoutilities with the OLED?
How did you make sure that the faster RAMs in the OLED doesn't play a part in better performance?
Really you just shouldn't TDP cap, it literally makes no sense to do so, you should FPS cap, then you get what you need for performance, you're not going to limit to 5W and play with 10-15fps, you should just cap to 30fps
26:47 10h20min of Dead Cells, really impressive 👏🏻 😳 . Does anyone know the battery life of the Nintendo Switch in the same game?
So does this mean the oled at max tdp 15 isn’t as performant as the lcd at 15?
Sometimes its more, others is less. But its drawing less power than the lcd model, even with 15w tdp in both
No it means the LCD steam decks power draw isnt as accurate as the OLED steam decks
Please use a larger font on the charts
Is it possible to overclock the new steamdeck to match the power draw of the steamdeck OG and compare the perf gain!
It would be dope.
Our problem is officially we have a 15w TDP cap. If you remove this cap, then yes you can, but we'll be spending more power to do so
The title of this video is very confusing, so after watching this video twice now, is the oled version less performance or not? Yes/No?
In not that complicated to understand actually, simple... The new Steam Deck is more optimized it has a less energy consumption compared to the old one, what it is 10TDP for the old Steam Deck, on the new Deck will be around 11 or 12 TDP because remember that the componentes of the New Deck is consuming less power at any TDP compared to the old one.
I really hope someone can give us some YUZU stats before launch. A 3-7 FPS gain would bring those games that dip into the mid 20s into the 30s. Hoping someone does this before launch. Stuck between this and a Legion GO.
Bro I tell you, I feel the Legion is better for all around performance..the 144hz refresh rate is huge especially on a 8.8 inch display. Just wait as these updates will be coming to Legion soon to fix some things it's not even a whole month since Legion been out and the deck had way more issues then the Legion at this point of launch
Alright I get the whole lower tdp performance, but if this is without a lock, why does it have 5 fps less than the LCD version, as seen here 29:59? also even drops to 29 fps.
is that without a lock? he isn't clear... he said while trying to make them run similar, so it might have a lock on it
@@AyaWetts Even if that's true, why does it drop below 30? Or is that just how it's locked?
@@kidgaming1245 I don't think he gives enough info for us to know here. I've seen other videos of people benchmarking the same game and showing it faster on OLED. I also know on my LCD Deck its supposed to have a TDP of 15 watts, but if I hard set the setting to say 15 watts, games run slightly slower than if I leave it as no limit, so the power profile setting may work slightly different.
Arkham Knight uses an extremely modified Unreal 3 engine, it’s possibly the least useful representative if you’re trying to compare engines. It’s like Using Call of Duty to compare the Quake 3 engine to other engines.
Isn't the recent mass effect legendary on unreal 3? Like a modified unreal 3. That's probably an interesting game to benchmark in that bracket. But if Arkham Knight works then I don't see any reason to. Just a suggestion.
wow
did not expect it is so comlicated
Nothing some updates can't fix
It starts 😂
I don't go below 10W, so even if were objectively correct, no harm no foul.
Please dude bigger fonts on charts :/
Wouldnt the oled version use less power so at 15wtdp the oled is using less apu power
Also mind you that the OLED version of the Steam deck is using slightly faster LPDDR5x RAM at 6400 vs 5600 aka they are running the RAM chips at their native 6400 speed rather than underclocking it because the OG deck couldn't handle 6400 mt/s RAM for whatever the reasoning(can be any number of things). Which might effect the 1% and 0.2% framerates.
nah it's could handle it it's a cost saving mesure The steam use two different ram supplier micron and samsung .The samsung module are 6400mt/s and the micron ones are 5500mt/s
@@potatorigs2155 underclocking ram saves money how?
@@clankfish The OG steamdeck use samsung and micron ram chips the micron one are 5500mt/s and the samsung one are 6400mt/s dual sourcing is often used to same money. Also there was a chips shortage not to long ago.
@@potatorigs2155 isn't the Micron ram the one that is rated at a higher max frequency and not the Samsung ram?
Honestly, for indies I always go 8 W or 10 W. Never try 5 W.
WQould be nice if you can show the actual steam decks.
TDP is not useful to predict CPU power consumption, as it's the maximum amount of heat generated by a computer chip or component that the cooling system in a computer is designed to dissipate under any workload.
If the 15w tdp in the Oled model is using less power than the same tdp in the lcd model, they should increase the tdp limit to 16 or 17w in the oled, to match the lcd. It make no sense that cyberpunk is getting worse performance in the new one, because its power limited when comparing to the old one
@@alistairsmith2396 it depends on the game, and I'm gessing the settings used also. These results by the Phawx show a decrease in performance at 15w with the SD preset and native resolution. Digital foundry always tests with a lot of upscaling, I'm guessing that changes results.
Every test from Phawx here shows OLED is generally better or same on 15w compared to LCD. Only Cyberpunk and RDR2 have slightly worse average fps (-2), but 1% and 0.2% are still better
After i noticed your shirt: This is a hint, that GPD are planning to make a PSP Go-like handheld?
Btw, thx for this video.
And i need a little advice: What do you guys think? buy now a 'refurbished 512 og' steamdeck, or buy the 'oled 512gb' this summer? i mean, the price difference is almost 200€, and that is a lot in my country. (I mean, i know, the real answer is to buy og now, and buy the oled too later next year when it goes to sale.)
3:19 video/audio out off *SYNC*
So the Final word, TDP Cap is not TDP Cap.
I don’t understand why Valve would do this. They should have kept the same tdp behavior. Its just confusing
Not exactly, the OG Steam Deck was drawing way much more power using the same TDP cap, so the overall efficiency was not that good.