Budget CPU Battle Royale: Intel i3-13100F CPU Review & Benchmarks
video description
Date: 2023-01-25
Related videos
Comments and reviews: 14
butterfingersman
love this video, and i love the extensive comparisons, even managing to squeeze a 3700x into those charts! im just now wrapping up a build for a friend who has been stuck on a xeon from 2012 for 6+ years. its fared well enough (for a sub 300 budget build 6+ years ago) but new games requiring AVX has pushed him into the world of a CPU upgrade! i wanted to go 5600 all the way, but availability has been miserable. new b550 motherboards have been pricey too lately, with decent sales but not sub 100 for decent motherboards really. i ended up snagging a 3800x & b550 motherboard with great VRM heatsinks for 220 secondhand locally. imagine my delight when there's a new budget comparison video for me to directly compare to! i know it isnt as good as a 5600 would have been, or maybe even a 12100f, but i feel like i wouldve been chasing prices for weeks or settling on a 160 processor and 130 motherboard, pushing the budget way higher than i intended. plus, this way i get a nice bump in productivity and multithreaded workloads, which i valued pretty highly. the price was super fair for the performance i can expect from the build, even though it's definitely not the direction most people have been going in. this budget direction would probably make the most sense for somebody hosting servers or programming / video editing / 3d art / compression / game engine workloads. which, in our case, doesnt make money, but is fun to mess around with and setup, and sometimes even crucial (decompression). i think we're going to be very happy with the performance uplift for the price!
reply
love this video, and i love the extensive comparisons, even managing to squeeze a 3700x into those charts! im just now wrapping up a build for a friend who has been stuck on a xeon from 2012 for 6+ years. its fared well enough (for a sub 300 budget build 6+ years ago) but new games requiring AVX has pushed him into the world of a CPU upgrade! i wanted to go 5600 all the way, but availability has been miserable. new b550 motherboards have been pricey too lately, with decent sales but not sub 100 for decent motherboards really. i ended up snagging a 3800x & b550 motherboard with great VRM heatsinks for 220 secondhand locally. imagine my delight when there's a new budget comparison video for me to directly compare to! i know it isnt as good as a 5600 would have been, or maybe even a 12100f, but i feel like i wouldve been chasing prices for weeks or settling on a 160 processor and 130 motherboard, pushing the budget way higher than i intended. plus, this way i get a nice bump in productivity and multithreaded workloads, which i valued pretty highly. the price was super fair for the performance i can expect from the build, even though it's definitely not the direction most people have been going in. this budget direction would probably make the most sense for somebody hosting servers or programming / video editing / 3d art / compression / game engine workloads. which, in our case, doesnt make money, but is fun to mess around with and setup, and sometimes even crucial (decompression). i think we're going to be very happy with the performance uplift for the price!
reply
Byte
i must say, that performance per dollar is an extremely difficult topic. Since these are budget CPUs, they are bound to be paired up with budget GPUs. The real performance per dollar then is the combined price of the CPU platform cost (MB, RAM, CPU) and the GPU.
For example if you are CPU bottlenecked with a 6600xt (like in csgo low FHD) the 40 dollar more expensive CPU suddenly has a way better performance per dollar. And if your System is GPU bottlenecked, (for example 4k rise of the tombraider on a 6600xt), the opposite holds true. Even if the cost of the CPU alone would suggest otherwise in the charts.
This is not saying that GamersNexus should be calculating Performance per dollar that way, because that is a tremendous undertaking, to test every CPU with every GPU and the gains will become obsolete as soon as prices change. If you were to undertake such madness tho, you'd have to make a interactive version on your website, that uses the performance numbers you calculate with a automatic price checker in order to keep the chart up to date, because even tho the prices change constantly, the performance does not. One could also limit oneself to just configs of reasonably matched cpu/gpu combos, (no 4090 x 13600k or below) but as already stated, waaaay to much work.
reply
i must say, that performance per dollar is an extremely difficult topic. Since these are budget CPUs, they are bound to be paired up with budget GPUs. The real performance per dollar then is the combined price of the CPU platform cost (MB, RAM, CPU) and the GPU.
For example if you are CPU bottlenecked with a 6600xt (like in csgo low FHD) the 40 dollar more expensive CPU suddenly has a way better performance per dollar. And if your System is GPU bottlenecked, (for example 4k rise of the tombraider on a 6600xt), the opposite holds true. Even if the cost of the CPU alone would suggest otherwise in the charts.
This is not saying that GamersNexus should be calculating Performance per dollar that way, because that is a tremendous undertaking, to test every CPU with every GPU and the gains will become obsolete as soon as prices change. If you were to undertake such madness tho, you'd have to make a interactive version on your website, that uses the performance numbers you calculate with a automatic price checker in order to keep the chart up to date, because even tho the prices change constantly, the performance does not. One could also limit oneself to just configs of reasonably matched cpu/gpu combos, (no 4090 x 13600k or below) but as already stated, waaaay to much work.
reply
Cybernetic
The i3 13100 is basically super Skylake i7. It's a hyperthreaded 4-core intel which is 30-50% faster than the 6700k, depending on use case. I think framed like that, it's a monster for 125, which is what it costs right now. You also get the new chipset which supports all the newest bells and whistles, including DDR5. If you want to use DDR4 and the rest of your existing components, you can get the CPU and mobo together for about the cost of an i5-13400 or Ryzen 7600 CPU WITHOUT their motherboards.
So for a person on an aging setup, say a 6700k or so with DDR4, this means they can spend 225 and be modernized and raise the CPU bottleneck. For someone looking to game at 1080 who is starting from scratch, an entire build with 32gb DDR5, a 1TB 970 M.2, a 600W gold PSU, a case, and a 3060 12GB or RX 6700 is playing every game out right now for under 1000. That's way more palatable than a year ago an equivalent rig costing 1500.
reply
The i3 13100 is basically super Skylake i7. It's a hyperthreaded 4-core intel which is 30-50% faster than the 6700k, depending on use case. I think framed like that, it's a monster for 125, which is what it costs right now. You also get the new chipset which supports all the newest bells and whistles, including DDR5. If you want to use DDR4 and the rest of your existing components, you can get the CPU and mobo together for about the cost of an i5-13400 or Ryzen 7600 CPU WITHOUT their motherboards.
So for a person on an aging setup, say a 6700k or so with DDR4, this means they can spend 225 and be modernized and raise the CPU bottleneck. For someone looking to game at 1080 who is starting from scratch, an entire build with 32gb DDR5, a 1TB 970 M.2, a 600W gold PSU, a case, and a 3060 12GB or RX 6700 is playing every game out right now for under 1000. That's way more palatable than a year ago an equivalent rig costing 1500.
reply
Jesper
I bought the 12100F and I am VERY happy with it for 1080p gaming and general media consumption... it's low power and it is FAST...
I have had Ryzen 1700 - 3600 - 5800X - Intel 11400 and 12100 and graphics cards have been RX 580, Vega 56, RTX 3070TI - 3060 and 3060TI...
currently I run with the 12100 and RTX 3060 TI... it's low power when I don't game, but kicks ass when I game... It doesn't win any charts but it is more than plenty for my needs. Only got the 3060TI over the planned 3060 because it was 10 bucks more, so the choice was obvious. Most people doesn't NEED more than this but some WANT bigger and better but don't really utilize it since they still run 1080p or play games that don't need hundreds of fps... and I could see on my powerbill when I changed from 5800X-3070TI to a 11400-3060 and I didn't lose anything gaming wise.
reply
I bought the 12100F and I am VERY happy with it for 1080p gaming and general media consumption... it's low power and it is FAST...
I have had Ryzen 1700 - 3600 - 5800X - Intel 11400 and 12100 and graphics cards have been RX 580, Vega 56, RTX 3070TI - 3060 and 3060TI...
currently I run with the 12100 and RTX 3060 TI... it's low power when I don't game, but kicks ass when I game... It doesn't win any charts but it is more than plenty for my needs. Only got the 3060TI over the planned 3060 because it was 10 bucks more, so the choice was obvious. Most people doesn't NEED more than this but some WANT bigger and better but don't really utilize it since they still run 1080p or play games that don't need hundreds of fps... and I could see on my powerbill when I changed from 5800X-3070TI to a 11400-3060 and I didn't lose anything gaming wise.
reply
ThysiosX
Is there a global CPU ranking chart on your website or somewhere that shows every CPU you've tested? That way if the CPU someone wanted to compare to isn't in the list they could just look at the global chart that gets updated every new CPU release.
FPS per dollar feels a bit... usless to me. Well to me personally anyway. I tend to buy the best CPU I can afford. FIgure out my budget, then figure out which CPU will give me the best performance while staying under that budget.
Knowing a cheaper/weaker CPU is better value isn't going to make me buy it as I'd still be losing performance. The 13100F could be 100 x better value than any other CPU, but it's meaninless as I know always want a CPU more powerful than that. I suppose it's an interesting metric to know, but not something I'd ever use for making a purchase decision.
reply
Is there a global CPU ranking chart on your website or somewhere that shows every CPU you've tested? That way if the CPU someone wanted to compare to isn't in the list they could just look at the global chart that gets updated every new CPU release.
FPS per dollar feels a bit... usless to me. Well to me personally anyway. I tend to buy the best CPU I can afford. FIgure out my budget, then figure out which CPU will give me the best performance while staying under that budget.
Knowing a cheaper/weaker CPU is better value isn't going to make me buy it as I'd still be losing performance. The 13100F could be 100 x better value than any other CPU, but it's meaninless as I know always want a CPU more powerful than that. I suppose it's an interesting metric to know, but not something I'd ever use for making a purchase decision.
reply
Michael
In regards to the stated disadvantages of using a frames per dollar chart, I would suggest you combine both the FPS and FPS/ on the same graph.
The easiest way to do this would be to just use an FPS bar chart (for each benchmark) but then annotate each of the bars with the FPS/ metric. Perhaps a colour gradient could be used for each bar to help indicate the FPS/ as well.
A better way to visualise this might be to have two bars per piece of hardware (one bar for FPS and another for FPS/ ) and then use two different axis keys on the same axis; this would make it easier to visually compare the FPS/ . I would still order the hardware by FPS though as I would expect most people care about the hardware's performance but you could consider ordering by FPS/ if you were making a video specifically to compare hardware value.
reply
In regards to the stated disadvantages of using a frames per dollar chart, I would suggest you combine both the FPS and FPS/ on the same graph.
The easiest way to do this would be to just use an FPS bar chart (for each benchmark) but then annotate each of the bars with the FPS/ metric. Perhaps a colour gradient could be used for each bar to help indicate the FPS/ as well.
A better way to visualise this might be to have two bars per piece of hardware (one bar for FPS and another for FPS/ ) and then use two different axis keys on the same axis; this would make it easier to visually compare the FPS/ . I would still order the hardware by FPS though as I would expect most people care about the hardware's performance but you could consider ordering by FPS/ if you were making a video specifically to compare hardware value.
reply
Lame
20:59 But they are not the same, Steve. The 5600X is 12.4 FPS faster than the 5600. I dont care that it is only 2.2%, the actual FPS is what matters. People will laugh at me at a LAN party when they see that I only get 530.3 FPS in Siege, and I will be made to sit with the peasants. 540 FPS gets you into the VVIP area, and that is obviously where you want to be. Also, you cant compete with 530 FPS against people with 542 FPS in an online competitive shooter. To make my argument stick, at 542 FPS you have an average frame time of 1.84ms. At 530 FPS that increases dramatically to 1.88ms. That is 0.04ms of increased latency, and my 10 Microsoft Wireless mouse will definitely feel a lot more sluggish, making me miss all of my shots.
reply
20:59 But they are not the same, Steve. The 5600X is 12.4 FPS faster than the 5600. I dont care that it is only 2.2%, the actual FPS is what matters. People will laugh at me at a LAN party when they see that I only get 530.3 FPS in Siege, and I will be made to sit with the peasants. 540 FPS gets you into the VVIP area, and that is obviously where you want to be. Also, you cant compete with 530 FPS against people with 542 FPS in an online competitive shooter. To make my argument stick, at 542 FPS you have an average frame time of 1.84ms. At 530 FPS that increases dramatically to 1.88ms. That is 0.04ms of increased latency, and my 10 Microsoft Wireless mouse will definitely feel a lot more sluggish, making me miss all of my shots.
reply
nineEngine
Im sure you'll get many comments about this, but two things I love about Hardware Unboxed's version of price to performance charts are A) They show the price they used for calculation, which is important for a viewer to know so they can compare to the actual prices they themselves can find at the moment, which may be different for many reasons, and B) they show FPS on the same chart, so you get a really condensed view of FPS, price, and price/performance all in one. Maybe there's reasons not to do things this way, but personally I find that they bring a lot of quick, summarized information to the table, which as you mentioned is the main advantage of those kinds of charts.
reply
Im sure you'll get many comments about this, but two things I love about Hardware Unboxed's version of price to performance charts are A) They show the price they used for calculation, which is important for a viewer to know so they can compare to the actual prices they themselves can find at the moment, which may be different for many reasons, and B) they show FPS on the same chart, so you get a really condensed view of FPS, price, and price/performance all in one. Maybe there's reasons not to do things this way, but personally I find that they bring a lot of quick, summarized information to the table, which as you mentioned is the main advantage of those kinds of charts.
reply
nleiilly
Would have been nice to see some performance at 4K max settings using the 4090 across the selection of CPU's.. As resolution goes up the CPU requirements go down, however where is the sweet-spot for say Cyberpunk.. 4, 6, 8 or more cores and 4ghz or 6ghz?
I'm using an ancient E5-2695 v3 (Turbo on All 14 cores unlocked & Resizeable BAR support inserted into x99m Fatal1ty bios) -- MW2 1440p ultra is 120fps.. MW2's built in benchmark says it's 50/50 bottleneck between the CPU and my 4090!
E5-2695 v3's can be had for around 50, x99 motherboards are 100'ish and 64gb ECC memory is uber cheap.. that's get you a proper 'budget & performance' build!
reply
Would have been nice to see some performance at 4K max settings using the 4090 across the selection of CPU's.. As resolution goes up the CPU requirements go down, however where is the sweet-spot for say Cyberpunk.. 4, 6, 8 or more cores and 4ghz or 6ghz?
I'm using an ancient E5-2695 v3 (Turbo on All 14 cores unlocked & Resizeable BAR support inserted into x99m Fatal1ty bios) -- MW2 1440p ultra is 120fps.. MW2's built in benchmark says it's 50/50 bottleneck between the CPU and my 4090!
E5-2695 v3's can be had for around 50, x99 motherboards are 100'ish and 64gb ECC memory is uber cheap.. that's get you a proper 'budget & performance' build!
reply
Maggi
One big problem with FPS/Dollar is also the difference in price in different regions, for me in germany the 5600 is a complete no brainer compared to the 13100F (same price or even 5 cheaper) and compared to the 12100F it is a 10-15 difference so you get good scaling when it comes to the money you pay more there. At the end of the day the most important thing for me in these reviews is the absolute performance and efficency of the different parts. People should be able to compare prices in their specific market and the time you buy. Who knows how much these CPUs will cost in a month or half a year. Price changes performance stays.
reply
One big problem with FPS/Dollar is also the difference in price in different regions, for me in germany the 5600 is a complete no brainer compared to the 13100F (same price or even 5 cheaper) and compared to the 12100F it is a 10-15 difference so you get good scaling when it comes to the money you pay more there. At the end of the day the most important thing for me in these reviews is the absolute performance and efficency of the different parts. People should be able to compare prices in their specific market and the time you buy. Who knows how much these CPUs will cost in a month or half a year. Price changes performance stays.
reply
reynardt
Here is my current view on CPU's and I am not one to distinguish between AMD and Intel. (I currently own 2 AMD's and 1 Intel). Currently the high-end AMD CPU's are great but when you hit the lower market where I usually shop Intel is just better at the moment. For the same reason that I bought the AMD 3600 I bought an Intel 12400 at the time of purchase IT just is better than its competition. (my metric is simple I allocate how much I am willing to spend and then get as much information on all the CPU's around that bracket. AMD use to be king here but lately it seems they are more concerned about beating intel at the high end)
reply
Here is my current view on CPU's and I am not one to distinguish between AMD and Intel. (I currently own 2 AMD's and 1 Intel). Currently the high-end AMD CPU's are great but when you hit the lower market where I usually shop Intel is just better at the moment. For the same reason that I bought the AMD 3600 I bought an Intel 12400 at the time of purchase IT just is better than its competition. (my metric is simple I allocate how much I am willing to spend and then get as much information on all the CPU's around that bracket. AMD use to be king here but lately it seems they are more concerned about beating intel at the high end)
reply
TomcaT
Something that I think would also benefit in budget CPUs reviews like this is a chart or two where the CPUs are paired with one or two budget class / mid tier GPUs as well. Maybe something like the 6600 as kind of a reality check against what an actual budget build with the CPU would perform like. The point being that gaming performance between these CPUs may well flatten out with a lower end GPU which could change the value equation. If a CPU could technically run a game 30% faster, but only when paired w/ a 4090, then you'd be paying a premium for a performance difference you're never going to see. Just a thought.
reply
Something that I think would also benefit in budget CPUs reviews like this is a chart or two where the CPUs are paired with one or two budget class / mid tier GPUs as well. Maybe something like the 6600 as kind of a reality check against what an actual budget build with the CPU would perform like. The point being that gaming performance between these CPUs may well flatten out with a lower end GPU which could change the value equation. If a CPU could technically run a game 30% faster, but only when paired w/ a 4090, then you'd be paying a premium for a performance difference you're never going to see. Just a thought.
reply
MlemDota
One thing I would like to see in CPU reviews is workloads that actually look like what a lot of gamers do. A lot of people don't just have their game open and nothing else; they'll be on discord, have 10 chrome tabs open, music on, maybe be updating a game in the background, maybe be streaming. It would be nice to see how CPUs fair in realistic multitasking workloads, rather than just how it does in a single task. Do intel's E-cores give a bigger advantage than single-program tasks show in these cases? Does having a higher core count help in general? Does it not matter? That's what I'd like to know.
reply
One thing I would like to see in CPU reviews is workloads that actually look like what a lot of gamers do. A lot of people don't just have their game open and nothing else; they'll be on discord, have 10 chrome tabs open, music on, maybe be updating a game in the background, maybe be streaming. It would be nice to see how CPUs fair in realistic multitasking workloads, rather than just how it does in a single task. Do intel's E-cores give a bigger advantage than single-program tasks show in these cases? Does having a higher core count help in general? Does it not matter? That's what I'd like to know.
reply
AirSafe
Personally, I love the fps/ charts especially on budget parts, but i have one specific suggestion. Since prices are always changing, I think it would be really important to have the pricing for every specific part on the chart screen itself. That way, people can multiply the chart number with the price onscreen, and divide by the price they are potentially being shown when they are buying so they themselves can evaluate if something is a good value, or even the best value. Just my 2 cents. Keep up with the incredible work and thanks for keeping manufacturers humble and exposing their BS when needed.
reply
Personally, I love the fps/ charts especially on budget parts, but i have one specific suggestion. Since prices are always changing, I think it would be really important to have the pricing for every specific part on the chart screen itself. That way, people can multiply the chart number with the price onscreen, and divide by the price they are potentially being shown when they are buying so they themselves can evaluate if something is a good value, or even the best value. Just my 2 cents. Keep up with the incredible work and thanks for keeping manufacturers humble and exposing their BS when needed.
reply
Add a review, comment
Other channel videos
