Hardware reviews: Can we increase the focus on required software?

[R-T-B]

Cool, so there's an entire two of you.

I mean, I'm here too.

 

[looniam]

Do you work at being this cynical?

i can assure you that it comes with age.

you'll see.

 

[Assimilator]

Do you work at being this cynical?

It's not cynicism, it's realism. The fact that a handful of people, including you, care about desktop Linux does not detract from the fact that a handful is not a viable economic proposition for any company - including TPU - to expend time and money on. Again, Valve is only putting money into Proton because (a) they have more cash than they know what to do with (b) they want to build their own hardware and software ecosystem, including Proton and Steam, that will allow them to lock users entirely into said ecosystem.

 

[chrcoluk]

Its only bloatware if you have to use it and causes performance drains.

Vaild points are being made here and should be discussed in the review. Worth a article at some point installing all the software and seeing what the dropoff is. I think the days of 1-4 cpus and 4GB of ram has convinced people all optional software is bloatware.

Isn't Intel Atom (E-cores) exactly for this? Taking care of small background tasks.

Perhaps a compromise is have a seperate review just for the software, its own article comparing all the vendors, as I think there is no need to test software every time a motherboard is released that would be ridiculous. Maybe redo the software review once a year or something when the vendors make new versions. Could include ryzen master and intel tuning app in this as well.

 

[Mussels]

I spent nearly two weeks violently ill and forgot this thread, but heres a great example that reminded me of my irritation with required software:

That top one? The moment i disconnect my blue yeti mic, it drops to 70MB.
I'm losing 3GB of RAM to having discord open and my microphone plugged in. Any user on 8GB or even 16GB, is going to feel that.

It seems to be related to Nvidia broadcast being integrated into iCue and Logitech Ghub these days, the high RAM usage is with them disabled - it gets worse when its enabled.

Ghub runs my mouse, microphone and speakers.
iCue controls my fans (static or via temperature sensors), but at present NO RGB is controlled via it.


Meanwhile razer synapse which has a bad reputation from its older versions runs my keyboards RGB, mousepads RGB, my entire systems RGB (Chrome ARGB controller), syncs my RGB with phillips hue - and uses almost nothing

 

[Sybaris_Caesar]

I think the current status quo is fine. DVDs containing softwares aren’t provided these days and people rarely bother to download them from the maze that is most manufacturers support page.

And most things you mentioned don't need software running since they ought to have onboard memory. Logitech mice has onboard memory for decades. Quick googling tells me same for Blue Yeti.

Unless your Corsair fans are connected to a iCue controller, your motherboard's fan controller is fine. No need for software since you can access it from BIOS.

Same for Razer products. But I don’t think a RGB mousepad might have onboard memory.

If any software is consuming more RAM than you think it should, chances are it’s an Electron app. Discord definitely is. I'd wager Logitech G Hub is as well, cause I don’t remember Logitech Gaming Software being that thirsty.
And RAM consumption is usually dynamic iirc. If there’s more RAM, chromium instances (like chrome based browsers and Electron softwares) will claim more as insurance.

 

[mama]

Dunno, but I always, and I think we always tell people to troubleshoot their frametime/fps woes by returning to stock.

Its common sense. Run lighter, means run faster. The fact is that commerce (and as a result: popular opinion) clouds the view of common sense. Maybe we should review our own psychology

I keep telling people around me much of the same idea: the more crap you run in and around the house, the more you'll worry about keeping it all going. When you own a house, you start noticing. Every bit of comfort brings a bit of maintenance. And what if you can't fully do that maintenance the exact way you want it? Then you've bought into something that causes more trouble than it solves. The perfect commercial story: create a need where there wasn't one. Note: every single smart home appliance is firmly in this category. And those are remarkably similar to PC accessories like RGB or numerous feature buttons that go beyond keyboard functions.

RGB and all that other additional crap we use in and around our PCs is exactly the same. If it doesn't cover bare needs, you don't need it and it probably won't help you. And if you can't fully control it and change it yourself, you've been enslaved to it - while paying to get there.

So rather than making reviews even more complicated, how about getting that general idea ingrained in everything we do? Otherwise where do you stop?

I've heard RGB makes it go faster...

 

[ir_cow]

I spent nearly two weeks violently ill and forgot this thread, but heres a great example that reminded me of my irritation with required software:

View attachment 255331

That top one? The moment i disconnect my blue yeti mic, it drops to 70MB.
I'm losing 3GB of RAM to having discord open and my microphone plugged in. Any user on 8GB or even 16GB, is going to feel that.

It seems to be related to Nvidia broadcast being integrated into iCue and Logitech Ghub these days, the high RAM usage is with them disabled - it gets worse when its enabled.

Ghub runs my mouse, microphone and speakers.
iCue controls my fans (static or via temperature sensors), but at present NO RGB is controlled via it.


Meanwhile razer synapse which has a bad reputation from its older versions runs my keyboards RGB, mousepads RGB, my entire systems RGB (Chrome ARGB controller), syncs my RGB with phillips hue - and uses almost nothing
View attachment 255332

I haven't checked in a Looong time but Razer (use to?) require internet to use a f-ing mouse. I had the deathadder and gave it away because it becomes a $5 mouse without the software. 800dpi, no marcos, etc.

Not to mention the software (was/is) complete bloatware.

Edit: I've been fighting off covid again so I feel your pain of being sick.

 

[Mussels]

I think the current status quo is fine. DVDs containing softwares aren’t provided these days and people rarely bother to download them from the maze that is most manufacturers support page.

And most things you mentioned don't need software running since they ought to have onboard memory. Logitech mice has onboard memory for decades. Quick googling tells me same for Blue Yeti.

Unless your Corsair fans are connected to a iCue controller, your motherboard's fan controller is fine. No need for software since you can access it from BIOS.

Same for Razer products. But I don’t think a RGB mousepad might have onboard memory.

If any software is consuming more RAM than you think it should, chances are it’s an Electron app. Discord definitely is. I'd wager Logitech G Hub is as well, cause I don’t remember Logitech Gaming Software being that thirsty.
And RAM consumption is usually dynamic iirc. If there’s more RAM, chromium instances (like chrome based browsers and Electron softwares) will claim more as insurance.

So you missed the post right above yours, where 2.6GB of RAM is used for one microphone?

The 500MB for icue is almost excusable, but it's bloated for running a single device that has internal hardware memory - i literally open it to adjust those hardware fan and lighting settings and close it, but if you want any temperature based fan control or use devices without hardware memory, you NEED to keep it open.

This is kinda my point tho:
As users, we need to know if devices *require* software or it can be removed after setup (Device has internal memory), and either way - what kind of performance impact does it have?


I thought about listing all the features i lose by uninstalling icue and LGH, but honestly the list is too damn long. Anything other than default rainbow vomit lighting, stereo sound and default button mapping is lost.


You buy an aorus 3080, which has an LCD panel that lets you install custom GIF's or show hardware stats, yay!
But then you find out the software makes your system run like ass and no reviews mentioned it, or requires an entire bundle of software to use (main control app to install the secondary apps that have firmware updates, then another secondary app to control the LCD that bundles RGB control, etc) - or its outright incompatible with required software from that own company (aorus/gigabyte is REALLY bad for this - the software for their monitors only supports the exact model it's for, so if you dare to use mismatched screens you're in for hell)

 

[Sybaris_Caesar]

This is kinda my point tho:
As users, we need to know if devices *require* software or it can be removed after setup (Device has internal memory), and either way - what kind of performance impact does it have?

Isn't that too much variable for a primarily hardware review site?

Just from my anecdote, MSI's B450 chipset boards are fucked in a sense that no third-party RGB control software (SignalRGB, OpenRGB) support them because of bricking hazard.
And there's no RGB control or off-switch in the BIOS. So software is the only option.
If I bought any MSI B450 boards recently, the support page lists a shitty bloated UWP software suite called MSI Center.
But as I bought the board in 2019, I know MSI listed individual software for individual tasks. Mystic Light 3.0 for RGB control. Another one for fan control. Another for CPU OCing. You get the gist. I use Mystic Light 3.0 that's pretty lightweight. 2-3MB last I checked.

That's already two variable for one board.

 

[Mussels]

Isn't that too much variable for a primarily hardware review site?

Just from my anecdote, MSI's B450 chipset boards are fucked in a sense that no third-party RGB control software (SignalRGB, OpenRGB) support them because of bricking hazard.
And there's no RGB control or off-switch in the BIOS. So software is the only option.
If I bought any MSI B450 boards recently, the support page lists a shitty bloated UWP software suite called MSI Center.
But as I bought the board in 2019, I know MSI listed individual software for individual tasks. Mystic Light 3.0 for RGB control. Another one for fan control. Another for CPU OCing. You get the gist. I use Mystic Light 3.0 that's pretty lightweight. 2-3MB last I checked.

That's already two variable for one board.

Yes, which is why it's been ignored almost universally

But again, if the software is MANDATORY and end users cant avoid using it - reviewers can at least give a snapshot view of what it's like at the time of review.

+ Great motherboard, awesome OLED display, overclocks like a champ
- OLED display cost 6% in cinebench


The key to my view here is, nothing short of reviews drawing attention to this will compel manufacturers to do anything about it

iCue was broken on ryzen CPU's for over a year with no temperature readings for 3000/4000/5000 series CPU's, leading to users having no temperature controlled cooling (and broken boost clocks, lowering overall performance) - If a single major reviewer had brought that up it would have got fixed a lot sooner



Edit: Figured out how to word part of it better
To be clear I don't mean testing every option in the software in all its various configurations, but just the default config - the bare minimum, out of the box experience to get the advertised features working.

I've got motherboards here with Killer networking - they can run just fine with the driver alone, with *optional* extra software - but if that software was required to run the ethernet at all, it'd have to be part of the testing and review process. ARGB software, OLED displays, icue/synapse/Ghub shouldn't be exceptions to this.


Reviewers review the product, as it is as the time of review.
We still get revision changes, we still get part-swaps (The SSD scandals). We still get driver changes and firmware updates.

Nothing stops them doing their normal review of the hardware, and then installing all the required software, recording the drive spaced used, RAM used at idle, and re-running the performance benchmarks and noting in the conclusion if the software had a positive, negative or neutral impact.

Will software updates change that in time? Sure. But so will firmware updates to a motherboard or GPU, or drivers from nvidia and AMD.

 

[ir_cow]

Nothing stops them doing their normal review of the hardware, and then installing all the required software, recording the drive spaced used, RAM used at idle, and re-running the performance benchmarks and noting in the conclusion if the software had a positive, negative or neutral impact.

This double or triples the review time. All so the software gets update next month and time was wasted. Say it is bloatware - the review says so. A update comes out and it fixes ALL the problems (magically). But the review isn't update. Now what?

 

[Sybaris_Caesar]

Yes, which is why it's been ignored almost universally

But again, if the software is MANDATORY and end users cant avoid using it - reviewers can at least give a snapshot view of what it's like at the time of review.

+ Great motherboard, awesome OLED display, overclocks like a champ
- OLED display cost 6% in cinebench


The key to my view here is, nothing short of reviews drawing attention to this will compel manufacturers to do anything about it

iCue was broken on ryzen CPU's for over a year with no temperature readings for 3000/4000/5000 series CPU's, leading to users having no temperature controlled cooling (and broken boost clocks, lowering overall performance) - If a single major reviewer had brought that up it would have got fixed a lot sooner



Edit: Figured out how to word part of it better
To be clear I don't mean testing every option in the software in all its various configurations, but just the default config - the bare minimum, out of the box experience to get the advertised features working.

I've got motherboards here with Killer networking - they can run just fine with the driver alone, with *optional* extra software - but if that software was required to run the ethernet at all, it'd have to be part of the testing and review process. ARGB software, OLED displays, icue/synapse/Ghub shouldn't be exceptions to this.


Reviewers review the product, as it is as the time of review.
We still get revision changes, we still get part-swaps (The SSD scandals). We still get driver changes and firmware updates.

Nothing stops them doing their normal review of the hardware, and then installing all the required software, recording the drive spaced used, RAM used at idle, and re-running the performance benchmarks and noting in the conclusion if the software had a positive, negative or neutral impact.

Will software updates change that in time? Sure. But so will firmware updates to a motherboard or GPU, or drivers from nvidia and AMD.

I got the gist of it. And I thought reviewers could list software version used like BIOS version and GPU driver version.

But then another thought came to me that ultimately leads the topic in a circle. The software you mentioned are ultimately complimentary. The hardware for the most part are plug-n-play. They're not unusable without the software like GPU driver.
Sure the default of RGB peripherals is rainbow vomit but isn’t the goal of review sites like TPU to review products as-is that come default?

Disregarding the time inflation because of complimentary software review, wouldn’t it make sandardisation a headache. Gigabyte RTX 3080 released at start of the year with RGB software might have different performance loss versus a SKU released at the end of the year with different version of RGB software.

Best option imo if someone at TPU do decide to take on the challenge would be to do a annual or bi-annual feature article on vampiric loss of these softwares like TPU does for PCIe saturation. Maybe we could compile a bunch of scenarios and softwares too. I usually open thw browser in Steam overlay or an Edge (basically Chrome) instance on games without overlay for music listening or walkthrough. I have qBitTorrent always running in the background as well. I'd like to know how much performance I'm sacrificing.

 

[Vayra86]

This is kinda my point tho:
As users, we need to know if devices *require* software or it can be removed after setup (Device has internal memory), and either way - what kind of performance impact does it have?

This is called common sense & experience. Users are consumers that need to apply due diligence before they jump on every product they see.

TPU doesn't need to be a babysitting club, that's a race to the bottom more so than anything else. You cater to stupid, you get stupid flock. Its not a bad thing to expect some level of intelligence and normal thought process.

Most products even market their software as the next best thing. If you miss that, you're an idiot. And products that run plug & play ALSO market that as a unique selling point, like my Xtrfy mouse... This world isn't overly complicated. Consumers are, they simply don't read the whole box before they click buy, or they use shitty sites like Ebay and Amazon for a good deal and forget the fine print.

You have to also consider that gamurs 'want' the flashy features and take the app for granted. The only fix here is experience and noticing how crap it is. And in the category of good advice, I think the best comment on this whole affair is always 'Run your PC with as little as possible for a good experience'. Less is more, really.

As for manufacturers doing something about it?
Again, we WANT this. There is a part of the market that buys this, consciously, and buys into all those extra features - even on monitors the features creep in on all segments. Most of them don't have performance hits or its obvious they do, as per your example of an OLED screen (you're driving an extra screen with data, duh?). But RGB: its an actual purchase that comes with software to control it. And for most of these addons and peripherals, again, the market is diverse, you can buy plug and play, or you can buy turbo ultra mice with 18601 buttons/macro options/settings/flashy lights. To each their own. The market supports both. You say software irritates you, and yet you keep running your Deathadder that requires a whopping... 8?! processes to run it proper and here's the kicker: its not a better mouse than others that run without software. Something tells me this problem is PEBKAC more than anything else...

And as for firmware/software updates over time... We have the internet. If your stuff suddenly runs like ass, you search for it and you find it. If that is found in a review, great, but if it happens 1-2 years post release, user base will definitely speak up anyway.

 

[ir_cow]

For you guys its simple question. Would you the readers like a MB review every two months or every two weeks. Because going into detail about each piece of software, running benchmarks with and without multiple times (lower margin of error) is extremely time consuming.

My personal computer. I install nothing extra and leave the rainbow RGB default. All OC is done in BIOS. I use to close all extra apps when playing games until one day I forgot to. Yeah I was techincally "losing" fps, but I have g-sync on and playing at max settings. Im already GPU bound for the most part. In my use case, it doesnt matter if a MB or graphics cards comes with software because I don't install it.

Edit: Now if software is requires. Aka OLED screen for user control, yes it would be good to note that in the review, but is that not just common sense? How else are you going to control it... Like comon, lets not hold people hands here. Do I really need to write a paragraph each time explaining to control certain functions require software?

 

[AlwaysHope]

For you guys its simple question. Would you the readers like a MB review every two months or every two weeks. Because going into detail about each piece of software, running benchmarks with and without multiple times (lower margin of error) is extremely time consuming.

My personal computer. I install nothing extra and leave the rainbow RGB default. All OC is done in BIOS. I use to close all extra apps when playing games until one day I forgot to. Yeah I was techincally "losing" fps, but I have g-sync on and playing at max settings. Im already GPU bound for the most part. In my use case, it doesnt matter if a MB or graphics cards comes with software because I don't install it.

Edit: Now if software is requires. Aka OLED screen for user control, yes it would be good to note that in the review, but is that not just common sense? How else are you going to control it... Like comon, lets not hold people hands here. Do I really need to write a paragraph each time explaining to control certain functions require software?

I can't speak for every other member on here of course but my problem is reviewers should benchmark systems with an OS that is "off the shelf" so to speak. That means disabling things like windows defender for example skews the results for plain everyday computing. So why can't they just leave the OS as it is from the manufacturer so regular consumers like myself can clearly see data that is unbiased for daily computing needs & thus make our minds up on purchase decisions based on that type of honest data?

 

[dgianstefani]

I spent nearly two weeks violently ill and forgot this thread, but heres a great example that reminded me of my irritation with required software:

View attachment 255331

That top one? The moment i disconnect my blue yeti mic, it drops to 70MB.
I'm losing 3GB of RAM to having discord open and my microphone plugged in. Any user on 8GB or even 16GB, is going to feel that.

It seems to be related to Nvidia broadcast being integrated into iCue and Logitech Ghub these days, the high RAM usage is with them disabled - it gets worse when its enabled.

Ghub runs my mouse, microphone and speakers.
iCue controls my fans (static or via temperature sensors), but at present NO RGB is controlled via it.


Meanwhile razer synapse which has a bad reputation from its older versions runs my keyboards RGB, mousepads RGB, my entire systems RGB (Chrome ARGB controller), syncs my RGB with phillips hue - and uses almost nothing
View attachment 255332

Synapse is good software. I don't know why it gets so much flak, I've literally never had an issue with it, and it's why all my peripherals and future lighting will be chroma compatible.

 

[Deleted member 24505]

I can't speak for every other member on here of course but my problem is reviewers should benchmark systems with an OS that is "off the shelf" so to speak. That means disabling things like windows defender for example skews the results for plain everyday computing. So why can't they just leave the OS as it is from the manufacturer so regular consumers like myself can clearly see data that is unbiased for daily computing needs & thus make our minds up on purchase decisions based on that type of honest data?

I agree.

Most avg users will not disable defender or anything else. The hardware review should be relevant to every user and not just us.

 

[Sybaris_Caesar]

Synapse is good software. I don't know why it gets so much flak, I've literally never had an issue with it, and it's why all my peripherals and future lighting will be chroma compatible.

Synapse is good now. But could be dogshit later like it wasn’t for the longest time. Software development works like that.

There was a brief period in time when Razer thought it was a good idea to exclude onboard memory on its peripherals AND require Razer account for the software. It still requires now but after backlash they added offline guest account circumvention. So if you transferred your peripheral to another PC and had no internet, tough luck. Your settings saved conveniently in the cloud is stuck there until you facilitate internet.

 

[ir_cow]

I can't speak for every other member on here of course but my problem is reviewers should benchmark systems with an OS that is "off the shelf" so to speak. That means disabling things like windows defender for example skews the results for plain everyday computing. So why can't they just leave the OS as it is from the manufacturer so regular consumers like myself can clearly see data that is unbiased for daily computing needs & thus make our minds up on purchase decisions based on that type of honest data?

Having a stipped down os gives better accurate results. Leave defender on and one video card might be 20 fps lower just because Windows is doing its own thing in the background. The impact of defender idle is about the same as it not being enabled. So really all it does is create more work to ensure the results are accurate.

 

[Mussels]

This double or triples the review time. All so the software gets update next month and time was wasted. Say it is bloatware - the review says so. A update comes out and it fixes ALL the problems (magically). But the review isn't update. Now what?

Going too far certainly could - but that's not what i've asked for or suggested.

Finish the normal part of a review, install the basic software on it's default settings (which is ALREADY DONE for reviews to get all the screenshots of the software, mind you) and re-run specific, repeateable tests - one or two single threaded and multi threaded tests, and if its gaming related one repeatable 3D benchmark thats not always GPU bound

Considering that all the software and tools are already on the test system at this point, it's being overly dramatic to say it'll triple the test time.

As for the second part:
It's almost like reviews mention a software version, as well as time of review and have comments sections.
We've had TPU reviews literally do this in the past, as well - it may take me time to find examples but TPU reviews have had pre-release hardware and launch day hardware with buggy software, and they've literally just stated the at-the-time situation, said they've contacted the manufacturer and then users commenting in the forum threads provided the various updates as time went on

I'll try and find one of them for you, but it's not like it'll show up in google search easily


This isn't personally aimed at you, but at hardware reviews in general - and while it's lost knowledge now thanks to 3Dchipset going under (the owner literally vanished overnight, the website soon after) I used to BE a hardware reviewer. This was my view on how I did reviews myself.

If you aren't confirming the advertised features work, you aren't reviewing the hardware. You're repeating the marketing.
You cant claim a motherboard is a great overclocker because the marketing says so - you need to test it.
You can't say one product is better or worse than another, without actually comparing the entire product as a whole - back then, fan noise was an often ignored example of this. People only cared about fastest and coldest.
You cant repeat claims and stake your websites reputation on specific features of the hardware working and how they function, without actually testing them - and you need to repeat the same experience end users will have as neutrally as possible, or it's exactly the same as running cherry picked hardware.

You cant review a corsair USB or wireless headset without iCue running, since you drop from 7.1 virtual surround sound to stereo sound, lose the RGB and equalizers and it literally runs different drivers - It's not exactly expected to throw a 3D games suite into a headset review, but you're certainly expected to actually play games while wearing them and nothing stops you from running a quick test and noting if the software features came at a cost or not


(This is why TPU's method of a debloated OS is fine despite it not being how end users run their system - it's removing variables that have nothing to do with the hardware being reviewed. You can't assume what antivirus an end user will use, so use none.)

Synapse is good now. But could be dogshit later like it wasn’t for the longest time. Software development works like that.

Old synapse is where logitech and corsair are at right now, with resource heavy bloat.
They took the criticisms and fixed their software, as well as increasing the amount of their devices with hardware profile storage




Edit 9000: This post has had a billion edits, but heres the perfect example of what i mean how this DOES happen in reviews including TPU reviews, but erratically.
No one expects performance benchmarks in a mouse review, but we DO want it acknowledged if the software has issues or is resource heavy.

This is not triple the level of effort in a review, but it's exactly what users need to know - you lose control of the lighting if you don't use the resource heavy software, and you can understand that's going to affect battery life negatively.

 

[chrcoluk]

I will agree with mussels on the specific point of never taking a manufacturers word for it, I dont like reading a review praising a product feature that hasnt been personally tested by the reviewer. If you dont have time to confirm such a feature then dont mention it in the review (at least not as a working feature).

The one example I can think off (and I have no idea if TPU made this mistake) is my b450 pro 4 board has 2 PCIE slots fed from the CPU directly, but the reviews I read on it said the second one was chipset routed, I assume they got it from asrock media rep's who fed them wrong information and then didnt personally verify it.

Also if something doesnt work properly avoid comments like "we assume they will fix it in next bios/software/driver update", it always feels like the reviewer in that case is making too much effort to be diplomatic to the vendor, when the reality is they got no idea if it will be fixed and have to tell it how it is.

 

[ir_cow]

This isn't personally aimed at you, but at hardware reviews in general - and while it's lost knowledge now thanks to 3Dchipset going under (the owner literally vanished overnight, the website soon after) I used to BE a hardware reviewer. This was my view on how I did reviews myself.

I'm sorry to crush your dreams on this subject but unless Wiz specifically says I need to do this re-testing with software and commentary, it just isn't going to happen from me. I understand where you are coming from and wanting a more rounded review that includes software "testing", but it really does add a lot of time to a review (at least for MBs).

checking out the software for say a computer mouse makes perfect sense. All the features are locked in the software. Same goes for Keyboards and some other peripherals. But it doesn't for MBs, Memory, GPUs, CPU, Cases (expect for Corsair ique). Basically everything that can be fully used without the need for software to be installed. OR a 3rd party software that does it better. Like why install any other GPU software when MSI Afterburner is the best already?

 

[neatfeatguy]

I have 16GB of RAM on my system and I don't notice any issues. I don't really run anything exciting in the background when I'm gaming - I close out of Chrome and that pretty much leaves iCue running and Steam and/or Galaxy 2.0 running. I'm not trying to stream anything, I don't have multiples of other RGB controlled software running and I don't use Discord (sometimes Team Speak is open).

If I had a bunch of stuff that was clearly impacting the performance of my system I'd streamline things to remove what isn't needed or I'd add more system RAM to help offset the RAM hogs I'm running in the background. I don't see it being a necessary need for TPU to run these programs for benchmarking. Most of the time these programs are negligible and if you're feeling the pinch, then you need to do something about it on your end to remedy the issue.

If anything, perhaps an individual review can be done for some of the more widely used software; such as iCue or Synapse or Discord and in there they could reference the amount of memory usage they see from these programs as they run. Even then, I don't really see this as being very beneficial use of time for TPU unless W1zzard thought it would be a good idea.

Synapse is good software. I don't know why it gets so much flak, I've literally never had an issue with it, and it's why all my peripherals and future lighting will be chroma compatible.

I found the software, when I used it, to be downright awful. This was back in 2014/2015. Synapse caused my GPU to run 3D clocks all the time and caused conflicts with my Razer mouse. I actually had to scrub my registry of the synapse software to resolve the high forced clocks for my GPU. Just to make sure it was a Synapse issue or not, I re-installed it and the GPU clocks ran high with the software (I did try multiple GPU drivers, too, but it made no difference)....so another scrubbing of the registry. After many reported issues to Razer about my mouse and the software, things never improved and I moved away from Razer. If they've fixed their software to run better, good for them, but it was atrocious software to use years ago.

 

[Vario]

Is there any research on the effect of Nvidia Geforce Experience versus only driver install? I've been doing only driver install on the premise that the additional features of Geforce Experience must add some amount of bloat.