50- 100 budget cpu benchmarks: amd ryzen r5 1600 af vs. Intel i3-9100f, athlon 3000g
video description
Date: 2020-05-06
Related videos
Comments and reviews: 10
mapesdhs
3:50 - That isn't gouging, it's called a free market. If popularity pushes demand beyond supply, then pricing will rise. An astute buyer will then see that value shifts more towards a 2600X, 3600, old XEON or other options. Nobody is forcing anyone to buy at a higher price. Buying a CPU is a voluntary act. I bought a 1600 AF yesterday for 100 UKP (approx. 125) from CCL, the best price I could find. This is obviously more than the 85 you mentioned, but then I'm not in the US (pricing is always higher in the UK, the proverbial cross-the-Pond markup) and other CPUs were a lot more expensive anyway, the 2700X having risen since January from 150 UKP to closer to 200 UKP (eg. 189 from Amazon, but delivery lead times are terrible) and the 3600 around 160 UKP. Allowing prices to rise when demand increases is actually very healthy. The media-encouraged obsession with preventing this is why hoarding and black markets emerge, because stores don't raise prices when they should; state/media interference prevents consumers from responding in a sensible manner. It's up to the buyer to decide whether the price is suitable, and up to them whether they make the effort to research the basis for their decision sensibly or not. As for the data, it's academically interesting but of dubious practical value for someone looking for a budget gaming CPU since nobody in this market segment is going to be pairing one with a 2080 Ti. Fact is, the way these budget CPUs compare can be very different when tested with a more budget realistic GPU like a 5600 XT, 5700 or used RX 570/580, GTX 970/1070, etc. Namely, performance gaps can narrow or be rather different for AMD vs. NVIDIA. RandomingGamingHD, PhilsComputerLab and Tech Yes City often test budget CPUs with lesser GPUs, as do I. Indeed, for a budget gamer looking to push out the boat somewhat, a 5700 XT or used Vega 56, 1080 or 1080 Ti is a more likely ceiling for purchase consideration.
reply
3:50 - That isn't gouging, it's called a free market. If popularity pushes demand beyond supply, then pricing will rise. An astute buyer will then see that value shifts more towards a 2600X, 3600, old XEON or other options. Nobody is forcing anyone to buy at a higher price. Buying a CPU is a voluntary act. I bought a 1600 AF yesterday for 100 UKP (approx. 125) from CCL, the best price I could find. This is obviously more than the 85 you mentioned, but then I'm not in the US (pricing is always higher in the UK, the proverbial cross-the-Pond markup) and other CPUs were a lot more expensive anyway, the 2700X having risen since January from 150 UKP to closer to 200 UKP (eg. 189 from Amazon, but delivery lead times are terrible) and the 3600 around 160 UKP. Allowing prices to rise when demand increases is actually very healthy. The media-encouraged obsession with preventing this is why hoarding and black markets emerge, because stores don't raise prices when they should; state/media interference prevents consumers from responding in a sensible manner. It's up to the buyer to decide whether the price is suitable, and up to them whether they make the effort to research the basis for their decision sensibly or not. As for the data, it's academically interesting but of dubious practical value for someone looking for a budget gaming CPU since nobody in this market segment is going to be pairing one with a 2080 Ti. Fact is, the way these budget CPUs compare can be very different when tested with a more budget realistic GPU like a 5600 XT, 5700 or used RX 570/580, GTX 970/1070, etc. Namely, performance gaps can narrow or be rather different for AMD vs. NVIDIA. RandomingGamingHD, PhilsComputerLab and Tech Yes City often test budget CPUs with lesser GPUs, as do I. Indeed, for a budget gamer looking to push out the boat somewhat, a 5700 XT or used Vega 56, 1080 or 1080 Ti is a more likely ceiling for purchase consideration.
reply
Blahorga
The Division 2 is really easy to have totally GPU limited. When it was still relatively new I compared benchmark numbers with a friend. He had just built a new machine close to the real high end with an intel i9 9900K, 32GB of fast ram, a very fast NVMe drive and so on. The only thing that wasn't really top shelf was the Nvidia RTX 2070 Super, but it was still respectable. I had just upgraded my machine with a 2070 Super from another manufacturer, but other than that it was running a Intel i7 6700 and 16GB of memory. When benching TD2 1440p, ultra settings my machine actually managed to squeeze out a higher average FPS than my friends. We re-ran the benchmark at least three times and this was actually very consistent. Now it wasn't more than 1 - 1.5 fps difference, but it was there every time we ran the benchmark. After scratching our heads a while we cam to the conclusion that the slight factory overclock on my GPU probably was the reason my machine got that slight edge. But the story certainly doesn't end there. Looking at the GPU and CPU load averages there was a significant difference. My GPU load average was listed as 104%, as silly as it sounds, while my friends machine reported something like 99%. CPU load however was another beast. On my machine it was reported as something like 98 or 99 % while my friends machine was logging less than 30%... I have no doubt that the frame times was pretty ugly on my machine when compared to my friends, but that's the sad thing about most in game benchmark, they don't really have any way to tell you. In the end I dialed down the quality settings and with a lot of testing ended up with something that looks good to me while being a bit easier on the CPU and GPU which allows for more a more stable frame rate.
reply
The Division 2 is really easy to have totally GPU limited. When it was still relatively new I compared benchmark numbers with a friend. He had just built a new machine close to the real high end with an intel i9 9900K, 32GB of fast ram, a very fast NVMe drive and so on. The only thing that wasn't really top shelf was the Nvidia RTX 2070 Super, but it was still respectable. I had just upgraded my machine with a 2070 Super from another manufacturer, but other than that it was running a Intel i7 6700 and 16GB of memory. When benching TD2 1440p, ultra settings my machine actually managed to squeeze out a higher average FPS than my friends. We re-ran the benchmark at least three times and this was actually very consistent. Now it wasn't more than 1 - 1.5 fps difference, but it was there every time we ran the benchmark. After scratching our heads a while we cam to the conclusion that the slight factory overclock on my GPU probably was the reason my machine got that slight edge. But the story certainly doesn't end there. Looking at the GPU and CPU load averages there was a significant difference. My GPU load average was listed as 104%, as silly as it sounds, while my friends machine reported something like 99%. CPU load however was another beast. On my machine it was reported as something like 98 or 99 % while my friends machine was logging less than 30%... I have no doubt that the frame times was pretty ugly on my machine when compared to my friends, but that's the sad thing about most in game benchmark, they don't really have any way to tell you. In the end I dialed down the quality settings and with a lot of testing ended up with something that looks good to me while being a bit easier on the CPU and GPU which allows for more a more stable frame rate.
reply
Sander169
I have a little critique. 20:12 I think power consumption figures should be always put into context of the specific workload, because of CPU utilization. The 1600 AF consumes almost duble in Blender, but it also performs much better (22:29). On the other hand, the power consumption difference in gaming doesnt have to as big. It could be, that the 1600 AF isnt even fully utilized (due to some engine or architectual restrictions maybe) and hence drawing less power, whereas the 9100F could be wery well sitting at 100%, reducing the gap in power consumption in gaming vs heavy multithreaded applications. To be clear, Im not defending AMD or accusing GN of not being aware of that, I just see normal users and even enthusiasts often neglecting that fact and assuming a 9900k for example would be always drawing 180, 200 or 250 watts. My critique is: that statement at 20:12 - 20:19 could perpetuate this misconception, because of the link drawn between the consumption figures in Blender and the FPS performance in gaming. It could have been more specific IMO.
reply
I have a little critique. 20:12 I think power consumption figures should be always put into context of the specific workload, because of CPU utilization. The 1600 AF consumes almost duble in Blender, but it also performs much better (22:29). On the other hand, the power consumption difference in gaming doesnt have to as big. It could be, that the 1600 AF isnt even fully utilized (due to some engine or architectual restrictions maybe) and hence drawing less power, whereas the 9100F could be wery well sitting at 100%, reducing the gap in power consumption in gaming vs heavy multithreaded applications. To be clear, Im not defending AMD or accusing GN of not being aware of that, I just see normal users and even enthusiasts often neglecting that fact and assuming a 9900k for example would be always drawing 180, 200 or 250 watts. My critique is: that statement at 20:12 - 20:19 could perpetuate this misconception, because of the link drawn between the consumption figures in Blender and the FPS performance in gaming. It could have been more specific IMO.
reply
Doge
I love the fact that i can trust Gamer Nexus blindly, that you put so much time and effort to have perfect data and trustable testing methodology. But you should put like a disclaimer in the videos like CPU BOTTLENECK IS TESTED AT 1080p with a 2080TI and doesn't really matter in a medium high GPU END SCENARIO Because dummy dumms can't read datas or charts. Reddit is full of brainwashed scare people that suggest an upgrade to users from 2600 to 3600, for gaming in the same budget, instead of the upgrade from 1060 to 2060 super (user said the budget allowed it or even a 2070), because of the CPU BOTTLENECK. I mean alot of people suggested that and supported those post. It drives me crazy, if you don't have a 1200 GPU you can't be bottlenecked so hard, or you will not be at all (like at 1440p resolution), but mainly, you are giving up 40%+ more performance for having 4-10% more cpu performance that in game will give you around 0, with the same low end GPU.
reply
I love the fact that i can trust Gamer Nexus blindly, that you put so much time and effort to have perfect data and trustable testing methodology. But you should put like a disclaimer in the videos like CPU BOTTLENECK IS TESTED AT 1080p with a 2080TI and doesn't really matter in a medium high GPU END SCENARIO Because dummy dumms can't read datas or charts. Reddit is full of brainwashed scare people that suggest an upgrade to users from 2600 to 3600, for gaming in the same budget, instead of the upgrade from 1060 to 2060 super (user said the budget allowed it or even a 2070), because of the CPU BOTTLENECK. I mean alot of people suggested that and supported those post. It drives me crazy, if you don't have a 1200 GPU you can't be bottlenecked so hard, or you will not be at all (like at 1440p resolution), but mainly, you are giving up 40%+ more performance for having 4-10% more cpu performance that in game will give you around 0, with the same low end GPU.
reply
ChrisXCrozz
Is the Upcoming Ryzen 3 3100 and 3300x comparable against the ryzen 5 2600? (When it releases of course) Im planning to buy either one for my next CPU upgrade and cant wait to see how it performs. Also i live in the Philippines and we dont have the Ryzen 5 1600 AF in a local or online store, Only amazon im finding it. So please for the people who are gonna recommend to me getting the 1600 AF. Answers NO. Are you people willing to spend 150 on the ship fee alone getting it half way across the world? at 235, id rather spend that on the ryzen 5 3600 with an AIO as well.
reply
Is the Upcoming Ryzen 3 3100 and 3300x comparable against the ryzen 5 2600? (When it releases of course) Im planning to buy either one for my next CPU upgrade and cant wait to see how it performs. Also i live in the Philippines and we dont have the Ryzen 5 1600 AF in a local or online store, Only amazon im finding it. So please for the people who are gonna recommend to me getting the 1600 AF. Answers NO. Are you people willing to spend 150 on the ship fee alone getting it half way across the world? at 235, id rather spend that on the ryzen 5 3600 with an AIO as well.
reply
Tudor
I own an intel i3-9100F because, at least in Romania, it's the lowest bang for the buck you can get. The AMD 1600 AF is much more expensive, somewhere close to 100-120 US Dollars, compared to the 75-85 US Dollars for the i3. It is a very potent CPU, even though it's a locked quad-core. I do light gaming with it and paired with a GTX 970 OC, I can easily get 100+, even 200+ on mainstream games such as CS:GO, StarCraft 2, PUBG etc., all at High. It can also play Battlefield V, but there will be few frame dips here and there. And yes... Crysis will run also. :-)
reply
I own an intel i3-9100F because, at least in Romania, it's the lowest bang for the buck you can get. The AMD 1600 AF is much more expensive, somewhere close to 100-120 US Dollars, compared to the 75-85 US Dollars for the i3. It is a very potent CPU, even though it's a locked quad-core. I do light gaming with it and paired with a GTX 970 OC, I can easily get 100+, even 200+ on mainstream games such as CS:GO, StarCraft 2, PUBG etc., all at High. It can also play Battlefield V, but there will be few frame dips here and there. And yes... Crysis will run also. :-)
reply
starshock2002
It frustrates me so much that people are price gouging something like the 1600af. It only works because people read headlines but don't seem to adsorb the context. The 1600af is a fantastic CPU for the price, the more it costs the worse it is. I've seen instances of them being sold side by side with the 2600 for the same price by people on amazon, at that point the 2600 is always the better choice, they basically identical but the 2600 has slightly higher stock clocks... but people buy the 1600af because they have heard its good :(
reply
It frustrates me so much that people are price gouging something like the 1600af. It only works because people read headlines but don't seem to adsorb the context. The 1600af is a fantastic CPU for the price, the more it costs the worse it is. I've seen instances of them being sold side by side with the 2600 for the same price by people on amazon, at that point the 2600 is always the better choice, they basically identical but the 2600 has slightly higher stock clocks... but people buy the 1600af because they have heard its good :(
reply
Florian
Attention: misleading Power Consumption Figures! The charts showing Blender Power Consumption DO NOT show or represent Efficiency. In fact the higher Power Consumption of the 1600AF and the higher Performance/Time Saving almost exactly equal out to the same Efficiency (total power consumed for same render job done). Which means the 1600AF PC will actually be more efficient since the rest of the system will not consume an equally higher amount of power while rendering and finish faster.
reply
Attention: misleading Power Consumption Figures! The charts showing Blender Power Consumption DO NOT show or represent Efficiency. In fact the higher Power Consumption of the 1600AF and the higher Performance/Time Saving almost exactly equal out to the same Efficiency (total power consumed for same render job done). Which means the 1600AF PC will actually be more efficient since the rest of the system will not consume an equally higher amount of power while rendering and finish faster.
reply
Hofnaerrchen
Tests like this show that intel's platform is the limiting factor and because intel doesn't seem to care about enthusiasts (especially those on the other end) much or at all. There is no cost-efficient environment for people who like pushing low-spec products to their limits without the need to buy high-end motherbords - at least if the cpu allows for OC at all, which is another fail by intel but is also totally legit if you make most of the money in other markets. TL;DR: So sad!
reply
Tests like this show that intel's platform is the limiting factor and because intel doesn't seem to care about enthusiasts (especially those on the other end) much or at all. There is no cost-efficient environment for people who like pushing low-spec products to their limits without the need to buy high-end motherbords - at least if the cpu allows for OC at all, which is another fail by intel but is also totally legit if you make most of the money in other markets. TL;DR: So sad!
reply
Kevin
i only think 1600's were 85 dollars for a limited amount of units from a used vendor, nothing to keep repeating. I cheaped out and got an i3 in canada because 1600/2600s are literally 250+ vs 100 dollar i3s. In USA, i3 9100fs are 74.99 all day on amazon, but still treated as if they cost the same or more than 1600afs? and 9400f's are 149.99 all day compared to 175 3600's. Of course they will perform worse but are decent value
reply
i only think 1600's were 85 dollars for a limited amount of units from a used vendor, nothing to keep repeating. I cheaped out and got an i3 in canada because 1600/2600s are literally 250+ vs 100 dollar i3s. In USA, i3 9100fs are 74.99 all day on amazon, but still treated as if they cost the same or more than 1600afs? and 9400f's are 149.99 all day compared to 175 3600's. Of course they will perform worse but are decent value
reply
Add a review, comment
Other channel videos
