Should you upgrade your GPU or CPU for faster gaming? We tested many hardware combos to find out

[bit_user]

It is not a good ROI if your metric is views per hour of work, but if most views come from search engine traffic I think you'll always struggle with that metric. Engaged regulars like interesting articles, but we make up a small proportion of viewers.

Yes. I was thinking about this some more, and it seems to me the internet has no shortage of news aggregators and clickbait peddlers. I don't see the need or value in having just one more.

I think it really gets at the site's mission. If it's just to maximize profit, then I can't necessarily say the clickbait approach is wrong. However, if the mission is broader, then I think articles like this serve a greater purpose. I'm tempted to borrow the logic of "prestige journalism" (to the extent I understand it), which is that you both use the "stuff that pays the bills" to enable the more consequential pieces, but they in turn elevate the site and help sustain a loyal & high-quality readership and help set the site apart from so many others out there.

Maybe these arguments will fall on deaf ears, but the two things that keep me coming back here are the forums and the more substantial pieces like deep dives and data-driven pieces like this. Unfortunately, I've noticed somewhat of a drop-off in the latter.

Maybe you should have titled this article "CPU vs GPU for gaming"

: D

 

[baboma]

>This is the whole point of stuff like social networking and YouTube. If you like this sort of article but don't feel inclined to share it, that's fine.

What I meant is that content promotion shouldn't be left to good will and volunteer work, but should be done by a community manager. Future Inc is not a small company, so surely it has employees for this essential task?

May I ask how traffic is broken down for direct, social media, and organic (Google)?

I see THW content links posted on FB, but as with here, there's no differentiation betwen filler and core content, so the filler just swamps the core pieces, as with here.

THW has a subreddit, but no links are posted there. Why not?

I note that you didn't address the elephant in the room, viz THW site layout, so I assume it's out of your hands. Perhaps direct traffic is only a small portion of your overall traffic? Regardless, THW is in sore need of a site makeover, but I get that simple changes aren't so simple when dealing with corporate bureaucracy.


>Now I have data to back that up that says, in essence, don't do hard articles — do the easier stuff. I'm a PC enthusiast and that's why I did this testing, because it's good to put concrete numbers out there.

Yes, it's evident from the sheer volume of data alone that it's a passion project. But I think you took on more than is prudent. There was no need to go back as far as 8700K for upgrade considerations. That 2017 PC would be 7 years old, which is well past its expire date in regards to upgrades.

Likewise, you could've dropped one of the GPUs from the list, either the 2080 or 3050. So, instead of dealing with 4^4=256 combos, 3^3=27 combos would've been much more manageable.

Other aspects of the piece limits its appeal. It's about upgrading older PCs, which screens out the majority of users wanting to build new PCs. All the PC configs have flagship CPUs, which further reduce the potential audience to a tiny portion.

It's also counter-intuitive: People who could afford flagship CPUs (and motherboards) in the first place, aren't the type who would settle for a compromised mixture of old+new tech. They would just get new PCs.

 

[JarredWaltonGPU]

The other article is titled "Product X vs Product Y", which is exactly the sort of phrase people will be searching for regularly. I'd argue that equalling that sort of article is actually pretty good. It is not a good ROI if your metric is views per hour of work, but if most views come from search engine traffic I think you'll always struggle with that metric. Engaged regulars like interesting articles, but we make up a small proportion of viewers. Unfortunately people coming from search engines are searching for phrases suitable for "pays the bills" articles and not interesting insights. That's why listicles are so popular. No effort required but wow do people like lists.

Maybe you should have titled this article "CPU vs GPU for gaming"

That last bit is actually probably the best advice. I'm not sure how we ended up at a headline with a freaking question mark at the end, because I hate those. (I mean, I know exactly how we ended up there, and it wasn't my doing.) I have updated the headline because saying "GPU versus CPU" or "CPU versus GPU" as the first thing is likely way more valuable for SEO.

 

[35below0]

There was no need to go back as far as 8700K for upgrade considerations. That 2017 PC would be 7 years old, which is well past its expire date in regards to upgrades.

I don't know about that. There are people upgrading from 6th or 4th generations. Those are a decade old but because they can still do the job well enough, people haven't felt the need to upgrade. Esp. if the computers were well built in the first place.

As for the GPUs, they were and are relevant. So dropping them would have made the article more useless, unfortunately.

 

[stonecarver]

But I think you took on more than is prudent. There was no need to go back as far as 8700K for upgrade considerations. That 2017 PC would be 7 years old, which is well past its expire date in regards to upgrades.

Without adding something like the 8700k as a base line this story would have been one of those all high end parts I can't afford to the larger amount of traffic Tom's gets. The article would have gotten opened but eyes rolled and never read.

JarredWaltonGPU again one of the most value able offering from Tom's staff that we need and the points you painstakingly provided with your time and hands on will be that current yard stick I will use.​

And it also shows a good point that an 7 year old part is still holding it's own.
The only thing that pushes still relevant parts to the recycle bin is marketing.

As to why not only this article and for that matter any thread with a simple question or harder repair question come down to readers all have low reader hits.

How we as Tom's regulars treat the new traffic that comes here for guidance and depending on how they ask and reply, there have been so many who would never give Tom's the time of day ever again being they were made to look stupid and never getting the help that they showed up on Tom's for help off a google search.

There are some of the best of the best PC guys on the forum who know there stuff but some times we forget there questions asked here because THERE NOT experts.

We need to get the traffic back up on Tom's.
Look at the "New post" page than the "Trending" post page.

Some of the same threads just sit and sit for sometimes a week and were still on page one. The views are lower than low for having millions all over the world come to Tom's.

Some days watching the threads reminds me of when you have to go a junkyard and the grumpy dirty cigar smoking guy behind the counter knows he is the only one that has what you need and uses that leverage to be a complete &ss.

You put up with his attitude but you never GO BACK.

 

[Jagar123]

Yes, it's evident from the sheer volume of data alone that it's a passion project. But I think you took on more than is prudent. There was no need to go back as far as 8700K for upgrade considerations. That 2017 PC would be 7 years old, which is well past its expire date in regards to upgrades.

This is a bad take imo. I have a friend on a 8700k and I myself am on a 2700x. We've been wanting to upgrade but haven't done so mostly due to the GPU market. This article gave me great detailed information that I could correlate with my system. Jarred picked out great hardware to represent multiple eras of CPUs and GPUs. If he can fine tune it so each permutation isn't as difficult to benchmark (less games tested as an example) then great, but what he provided was very useful to me.

 

[blargh4]

Yes, it's evident from the sheer volume of data alone that it's a passion project. But I think you took on more than is prudent. There was no need to go back as far as 8700K for upgrade considerations. That 2017 PC would be 7 years old, which is well past its expire date in regards to upgrades.

That is the most valuable part of this article. This isn't the heyday of Moore's law, people aren't compelled to upgrade their CPU every couple years. Including the 2600K or something might be a bridge too far but the 8700K is still a perfectly capable CPU that lots of people bought which is still happily running Windows 11 for me.

 

[JarredWaltonGPU]

That is the most valuable part of this article. This isn't the heyday of Moore's law, people aren't compelled to upgrade their CPU every couple years. Including the 2600K or something might be a bridge too far but the 8700K is still a perfectly capable CPU that lots of people bought which is still happily running Windows 11 for me.

FWIW, my daily driver — the PC I use for 99% of my work, outside of benchmarking GPUs — is running on Z390 with an i9-9900K. Actually, it's identical to the 8700K test PC, other than the CPU getting swapped out. And as long as I'm not trying to do something like video encoding on the CPU, or some other CPU intensive task, I never have cause for complaint. I have one PC on Win10 and the other on Win11... and though I still prefer quite a few things about Win10, there's not really a substantive difference these days.

 

[randomizer]

That is the most valuable part of this article. This isn't the heyday of Moore's law, people aren't compelled to upgrade their CPU every couple years. Including the 2600K or something might be a bridge too far but the 8700K is still a perfectly capable CPU that lots of people bought which is still happily running Windows 11 for me.

Until 2021 I was still running an X58 system with an i7 920 at stock speeds. I was playing games from 2020 (once I doubled my RAM to 12GB anyway). Sure, it wasn't a great experience, but it worked.

 

[Colif]

Yes, it's evident from the sheer volume of data alone that it's a passion project. But I think you took on more than is prudent. There was no need to go back as far as 8700K for upgrade considerations. That 2017 PC would be 7 years old, which is well past its expire date in regards to upgrades.

i know someone who is still using a 7700K and plays games on it .. It really depends what games you want to play and what resolution you happy with.
Not everyone buys a new PC every 5 years or so. Some people that post on here use way older.

 

[JarredWaltonGPU]

FYI, I've redirected the original article and URL to a new one that should likely draw more views over time. If you want to hop in the new thread, it's here: https://forums.tomshardware.com/thr...ost-your-gaming-performance-the-most.3847911/

The new URL: https://www.tomshardware.com/pc-components/gpus/cpu-vs-gpu-upgrade-benchmarks-testing

Here's hoping Google (and others) like the change and find it more often. Then I may see about doing some additional testing with AMD GPUs... but not until after my summer vacation and some other stuff gets finished.

 

[Colif]

AS per above, closing this thread. Please post in

CPU vs GPU: We tested 16 hardware combinations to show which upgrade will boost your gaming performance the most

We've tested four different CPUs and GPUs, spanning the past seven years of PC hardware, to show how different upgrade paths can perform. That's full benchmarks of 16 different hardware configurations to help you plan for your next gaming upgrade. CPU vs GPU: We tested 16 hardware combinations...

forums.tomshardware.com