Audiobooks can be had for less than $5 or free if you want to go that route. Or sing or play some counting game...
Audiobooks can be had for less than $5 or free if you want to go that route. Or sing or play some counting game...
Unless the new shield has better cooling than the old one this is unlikely. The old shield would throttle to the point where it was slower than the switch in docked mode (614 MHz for the old shield vs. 768 MHz for the Switch).
Given that the new Shield is significantly smaller than the old one, I would say it is very unlikely that it features better cooling.
NTMBK
Lifer
- Nov 14, 2011
- 10,456
- 5,843
- 136
I thought that most of the size reduction was due to ditching the hard drive bay?
Sure, and I never said it wasn't. Point is that the space savings from removing the bay apparently wasn't used on more cooling, but instead on shrinking the size of the console as a whole.
I would be interesting to see if this Shield TV GPU clock speed was true. The poster didn't provide any evidence that I've seen. He did say the CPU temps were in the 40-50's at load, so I find it hard to believe it's thermally throttling since the GPU is on the same package and usually runs close to the same temp. It could be a case of power limiting? I have a Shield, so I might have to take a look when I get the chance.
It could certainly be a case of power limit throttling, since he can actually get the GPU to run at the full 1 GHz if he downclocks the CPU (which would thus leave more headroom for the GPU).
The overall conclusion remains the same though, since I don't see why the new Shield would have a higher power usage limit than the old one.
Yeah I would bet that it's being power limited or "hard" limited. The Dolphin Dev states that rooting device allows the GPU to boost to 1GHz, this means that NVIDIA is doing it on purpose...Seems odd to me why they would do this? They have active cooling, so thermals would not be an issue. On the other hand, AT measured the device at 15w or so while gaming. That's kind of piggish...Not sure how much of that is CPU or GPU, but I would bet most of the power usage stems from the GPU.
I'm going to side load a CPU/GPU monitoring app and check it out myself as well.
poofyhairguy
Lifer
- Nov 20, 2005
- 14,612
- 318
- 126
Apple and Nintendo fans are very similar. That is why I loved the idea of Nintendo games on iOS and why I sometimes wish Apple would just buy Nintendo (could you imagine a Switch with an Apple SoCs inside? Yes please!!!).
I've been following the thread over at NEOGAF and from what I can gather it appears the X1 is power limited in the Shield TV. They seem to think it's thermally throttling, but there are no temp sensors on the GPU, just the CPU only. Temps are fine from what I can gather and we don't know what the thermal throttle point is the X1. In one of the posters(MDave) testing he noticed GPU clock speeds dip when both CPU/GPU are loaded significantly. Even when the CPU cores were forced to 1 GHz and stressed, the GPU would throttle down to 768 Mhz...which is exactly what the Switch supposed specs are. When the CPU is seeing low utilization, the GPU would boost to 1GHz and stay there. MDave was measure clock speeds directly from the Kernel.
From my own testing the CPU doesn't throttle at all while maxed at 100%. It will maintain 2014 Mhz on all 4 cores the entire time. I tested it for 10 min. The issue is the GPU. As the GPU uses a lot more power vs. the CPU when maxed. I'm still trying to get my DevCheck Pro App to work properly with the Shield, but I'm not having any luck.
On another note, I thought this post was an interesting one from NEOGAF.
Any techies here that know a good way of measuring memory bandwidth performance? If the Switch is operating at the clocks it is, I don't see why games won't be 1080p when docked. The Shield TV is pulling off 1080p 8xAA in Unity pretty comfortably! And this is using the Vulkan api too.
Using the same Vulkan API, same scene and same render quality settings (1080p 8x MSAA).
PC:
i5 4690K @ 4GHz, GTX 970 (3.5 TF):
284 FPS
Shield TV:
Clocked limited to 1GHZ CPU, GPU fluctuating between 614MHz and 1GHz, averaging about 768MHz most of the time:
44 FPS
Frame rate is barely fluctuating between 1 and 2 FPS on both platforms. Extrapolate that data between those platforms to get what it would perhaps be on the Xbox One and PS4, to see how far or how close the Switch might be? Hah!
http://www.neogaf.com/forum/showpost.php?p=228014696&postcount=9065
From my own testing the CPU doesn't throttle at all while maxed at 100%. It will maintain 2014 Mhz on all 4 cores the entire time. I tested it for 10 min. The issue is the GPU. As the GPU uses a lot more power vs. the CPU when maxed. I'm still trying to get my DevCheck Pro App to work properly with the Shield, but I'm not having any luck.
On another note, I thought this post was an interesting one from NEOGAF.
Any techies here that know a good way of measuring memory bandwidth performance? If the Switch is operating at the clocks it is, I don't see why games won't be 1080p when docked. The Shield TV is pulling off 1080p 8xAA in Unity pretty comfortably! And this is using the Vulkan api too.
Using the same Vulkan API, same scene and same render quality settings (1080p 8x MSAA).
PC:
i5 4690K @ 4GHz, GTX 970 (3.5 TF):
284 FPS
Shield TV:
Clocked limited to 1GHZ CPU, GPU fluctuating between 614MHz and 1GHz, averaging about 768MHz most of the time:
44 FPS
Frame rate is barely fluctuating between 1 and 2 FPS on both platforms. Extrapolate that data between those platforms to get what it would perhaps be on the Xbox One and PS4, to see how far or how close the Switch might be? Hah!
http://www.neogaf.com/forum/showpost.php?p=228014696&postcount=9065
Last edited:
We don't know anything about what's happening in that scene or what's happening in the GPU. That tells us nothing.
It appears it's the Unity 3rd Person Assets Demo
http://www.neogaf.com/forum/showpost.php?p=227905740&postcount=8844
I understand, but I think it's A neat comparison with both PC and Switch-like spec'd Shield running the same engine, and showing the performance of both.
He also used the 3Dmark benchmark earlier, which is arguably a better benchmark since it also loads the CPU, whereas the Unity demo only loads the GPU (which isn't really a realistic representation of an actual game). 3Dmark saw the same throttling down to 768 MHz, when the CPU was kept at stock clocks.
Also it might actually still be thermal throttling. Initially he reported the CPU only hitting 40-50 degrees with exhaust air at ~35 degrees, but with 3DMark he sees exhaust air at 48 degrees, which would suggest that the SoC is hitting 55-65, which would arguably be ok for a set top box, but might be too much for a handheld device like the switch where skin temperature matters (although with the Switch having a plastic shell it won't necessarily heat up that much).
With that being said though, it would certainly be nice if he could test it running some of the high end games the Shield TV has, like The Witness, Borderlands or Revengeance.
Last edited:
http://www.ign.com/articles/2017/01...ch-wii-u-versions-of-zelda-breath-of-the-wild
Switch: 900p @ 30fps while docked
WiiU: 720p @ 30fps
Switch: 900p @ 30fps while docked
WiiU: 720p @ 30fps
I wish there was a way to confirm it. I agree, I would also like to see how it performs in some of the newer and more graphical games.
On another note, It looks like the Switch has a vent area on the top. I wonder if it also has active cooling? I haven't heard anything about that?
Yup, that seems to the be the case. So it looks like in portable mode the Switch is the same as the Wii U. Reggie from Nintendo says it runs better on the Switch, but his response why was kind of vague. He said you don't see improvements so much in frame-rates, but in pop ins textures and overall smoothness. I'm looking forward to the Digital Foundary analysis video when it's released.
I'm personally okay with playing Zelda at 30fps. It's a more cinematic game anyway
Sonikku
Lifer
- Jun 23, 2005
- 15,901
- 4,927
- 136
For me a constant 30fps is more important than texture pop in or even 1080p over 900p. There's a big difference between a constant 30fps and 30fps some of the time 22fps other times. As a PC gamer 60 is my preferred so having to settle for half that is rough as it is. Dealing with games that dip even below that when things are their most intense is just the worst. This is really apparent in games like Dragon Age Inquisition when you can outright see a stutter at times, even during cut scenes.
Yeah I'm afraid it will have trouble keeping minimum of 30. I noticed other zelda titles lagging quite a bit while in combat, hope this one doesn't and is able to keep that solid 30.
Sonikku
Lifer
- Jun 23, 2005
- 15,901
- 4,927
- 136
It is a rather big issue. I just played Hyrule Warriors on my "New" 3DS and even it is obviously struggling to run the game well. It must be basically unplayable on a regular 3DS, never mind running it in 3D.
Judging by reviewers hands on time with the Switch and Zelda, it seems to drop below 30 FPS in certain scenes. It could be cleared up by the time of release, but I don't think it will.
Digital foundry did a frame by frame analysis, based on two 20 minute playthroughs, and they only saw a total of 3 framerate drops over those combined 40 minutes of play.
One of these happened during a big screen filling explosion, one happened when the DoF filter was turned on, and one happened when link runs through the shade of a nearby structure.
Apart from the drop from the big explosion, I would think that these are fixable, mainly because the DoF filter is turned on and off all the time and yet only produced a single drop once out of 40 minutes, and running through shadows obviously happen all the time as well, and yet also only resulted in a frame rate drop once.
It's also important to note that a lot of the reports of drops come from the earlier E3 Wii U demo.
Last edited:
Yeah I watched their video when it showed up in my YouTube feed. It's always nice to see their analysis, as they are always the ones to show show hard data. There were also a number of reviewers who had hands on time with the game at the reveal event and noticed it wasn't staying at 30fps while docked. Some said the grass also plays an issue with FPS drops as well.
I'm assuming the game cartridges were already in production by the time the reveal events were being held. Maybe we'll get a day one patch to rectify the issue? Even though, the issue doesn't seem to take away from the overall experience from what is saw. Though different areas could be worse or better? At any rate, I'm really excited for this game. Definitely could be a GOTY contender.
Interestingly enough Laura Kate Dale (probably the most prolific leaker of all for the switch), said that in her experience the frame rate drops was largely only an issue in the docked mode, which is a bit surprising (I would have imagined the other way around), and might further indicate that this is a fixable issue.
TRENDING THREADS
-
Discussion Zen 5 Speculation (EPYC Turin and Strix Point/Granite Ridge - Ryzen 9000)- Started by DisEnchantment
- Replies: 25K
-
Discussion Intel Meteor, Arrow, Lunar & Panther Lakes + WCL Discussion Threads
- Started by Tigerick
- Replies: 23K
-
Discussion Intel current and future Lakes & Rapids thread- Started by TheF34RChannel
- Replies: 23K
-
-
Facebook
Twitter