By using this site, you agree to our Privacy Policy and our Terms of Use. Close

Forums - Nintendo Discussion - Clarifying the 1.5TFLOPS of the SWITCH for those who just see the numbers.

vivster on 02 November 2016
OdinHades said:
When it comes to gaming, Teraflops are nothing more than a buzzword, FP16 or not. Just stop caring about that shit so much. It's getting worse than those MHz wars back in the days when AMD ran circles around Intel despite having less of dem MHz. Or how about the famous bit wars. What matters at the end of the day is what's on the screen. Even the Wii U with its 3.5 Nanoflops did some very impressive graphics after all. And then you have a game like Mafia 3 that looks like horseshit but will friggin' destroy your 10 Terfalop Titan X for no appearent reason.

Seriously, just stop it.

I think from all the specs that exist, FLOPS are the most descriptive for power comparison. Because they are a logical combination of the two most important factors in gaming graphics, cores and frequency. It's the best buzzword we have so far. Certainly better than MHz or bits or the worst one of all, RAM.

We should be happy that we're getting ever closer to real world performance.



If you demand respect or gratitude for your volunteer work, you're doing volunteering wrong.

Around the Network
walsufnir on 02 November 2016
BlkPaladin said:
Captain_Yuri said:
The thing that is really starting to urk me is that people are trying to spin it in a way to show that their console is more powerful than it really is. But it is soooo stupid cause FP16 can apply to all current gen hardware. So switch is 1.5TF, x1 is 2.6TF, ps4 is 3.6TF, Pro is 8.4TF and Scorpio is 12 TF and etc when you get PC into mind. So it ends up being the same in relative performance, just with bigger numbers.
Sighh

Actually not if you look in the post above you and my last post, it depends on the chip. You just cannot magically make a chip half percise to run faster. Depending on how they are made it may double the perfermance of FP16 instruction or it may run at the same speed. I use registers in my answer because that is how deep my knowledge goes about these things go, I'm sure there are other ways to speed of FP16 and FP32 instructions other ways. But a register for all intents and purposes of this explination can only run one instruction at a time. And depending on how the chip is made to run the FP32 instructions can influence if the chip experences a "speed boost" running thing at half-percision. For example some 32-bit instruction are run on two 16-bit registers. So if it is optimized to do so, if you put 16-bit instructions into this register you can put another instruction at the same time in the other register and thus "twice" the speed in this case. But there are 32-bit registers that will only do one instruction at a time no matter how small the instruction is. So just looking at terms of FLOPS and Full percision/Half percision doesn't tell the entire story.

FLOPS, like Hertz before it, is just a advertising go-to word that really has no real world inpact.

It's funny that back in the day people went crazy over higher bit consoles and now using less bits to store data is important :)



vivster on 02 November 2016
walsufnir said:
BlkPaladin said:

Actually not if you look in the post above you and my last post, it depends on the chip. You just cannot magically make a chip half percise to run faster. Depending on how they are made it may double the perfermance of FP16 instruction or it may run at the same speed. I use registers in my answer because that is how deep my knowledge goes about these things go, I'm sure there are other ways to speed of FP16 and FP32 instructions other ways. But a register for all intents and purposes of this explination can only run one instruction at a time. And depending on how the chip is made to run the FP32 instructions can influence if the chip experences a "speed boost" running thing at half-percision. For example some 32-bit instruction are run on two 16-bit registers. So if it is optimized to do so, if you put 16-bit instructions into this register you can put another instruction at the same time in the other register and thus "twice" the speed in this case. But there are 32-bit registers that will only do one instruction at a time no matter how small the instruction is. So just looking at terms of FLOPS and Full percision/Half percision doesn't tell the entire story.

FLOPS, like Hertz before it, is just a advertising go-to word that really has no real world inpact.

It's funny that back in the day people went crazy over higher bit consoles and now using less bits to store data is important :)

Not when it comes to OS.



If you demand respect or gratitude for your volunteer work, you're doing volunteering wrong.

fleischr on 02 November 2016

I do agree that this belief in FP16 magic is plain silly.

I don't count on 3rd parties to use FP16 to magically pull Mass Effect 4 and Red Dead 2 out of a hat onto the Switch.

I do think it's something Nintendo could utilize in their games though in a custom engine. But I doubt they'll do that.

Personally, I'm inclined to think that...
-The OS is significantly leaner because the system is ARM rather than x86
-The custom Parker chip is not the same Tegra X2 chip used in cars and actually is 1 teraflops in FP32. Given the lean OS and that nVidia punches above it's weight in performance / flop, Switch ends up being roughly around X1 in power level.
-nVidia's Optimus tech plays a role in the SCD meant to boost the Switch dock mode later in the future.



I predict NX launches in 2017 - not 2016

walsufnir on 02 November 2016
vivster said:
walsufnir said:

It's funny that back in the day people went crazy over higher bit consoles and now using less bits to store data is important :)

Not when it comes to OS.

What do you mean with OS?



Around the Network
vivster on 02 November 2016
walsufnir said:
vivster said:

Not when it comes to OS.

What do you mean with OS?

Operating systems. We all know that anyone who still uses a 32bit OS is garbage. #64bitOSmasterrace



If you demand respect or gratitude for your volunteer work, you're doing volunteering wrong.

Jizz_Beard_thePirate on 02 November 2016
BlkPaladin said:
Captain_Yuri said:
The thing that is really starting to urk me is that people are trying to spin it in a way to show that their console is more powerful than it really is. But it is soooo stupid cause FP16 can apply to all current gen hardware. So switch is 1.5TF, x1 is 2.6TF, ps4 is 3.6TF, Pro is 8.4TF and Scorpio is 12 TF and etc when you get PC into mind. So it ends up being the same in relative performance, just with bigger numbers.
Sighh

Actually not if you look in the post above you and my last post, it depends on the chip. You just cannot magically make a chip half percise to run faster. Depending on how they are made it may double the perfermance of FP16 instruction or it may run at the same speed. I use registers in my answer because that is how deep my knowledge goes about these things, I'm sure there are other ways to speed of FP16 instruction other ways. But a register for all intents and purposes of this explination can only run one instruction at a time. And depending on how the chip is made to run the FP32 instructions can influance if the chip experences as speed boost. For example some 32-bit instruction are run on two 16-bit registers. So if it is optimized to do so, if you put 16-bit instructions into this register you can put another instruction at the same time in the other register and thus "twice" the speed in this case. But there are 32-bit registers that will only do one instruction at a time no matter how small the instruction is. So just looking at terms of FLOPS and Full percision/Half percision doesn't tell the entire story.

FLOPS, like Hertz before it, is just a advertising go-to word that really has no real world inpact.

Yes, I know its not a perfect analogy but the point remains the same which is about relative performance. I have no doubt what you said is true but from what I can gather, using FP16 is very situational where as majority of the softwares are made around FP32. And I m sure that at a minimum, Switch, Pro and Scorpio will most likely have it. So I will edit the other two out



                  

PC Specs: CPU: 7800X3D || GPU: Strix 4090 || RAM: 32GB DDR5 6000 || Main SSD: WD 2TB SN850

vivster on 02 November 2016

PSA bump.



If you demand respect or gratitude for your volunteer work, you're doing volunteering wrong.

AZWification on 02 November 2016

Thank you for making this thread, sir Vivster.



                
       ---Member of the official Squeezol Fanclub---

pokoko on 02 November 2016

You know, I assumed this was an OnionBerry thread.