In a recent hardware deep-dive, we took a seek for at how CPU cores and cache impacted gaming efficiency. To produce that, we archaic three Intel Core processors and compared their core scaling efficiency, revealing that in this present day’s games quite a bit of the efficiency gains you look when going from a Core i5-10600K to a Core i7-10700K and even Core i9-10900K are largely in consequence of the elevate L3 cache capacity.
This became a enchanting prognosis on story of most other folk who upgrade from an older generation Intel half reminiscent of the Core i7-8700K to the more moderen Core i9-10900K and saw a gain efficiency uplift in games, tend to imagine this comes as the of the 67% elevate in cores, however for basically the most half it’s if fact be told in consequence of the 67% elevate in L3 cache — in any case that is the case for this present day’s most tense games.
Coming some distance from that making an are attempting out, many of you wished to take hold of how mighty distinction the L3 cache capacity makes with shiny 4 cores active, wondering if the margins might perchance well be even increased. So we’ve long past support and retested a 4-core, 8-thread configuration while adding three extra games to the checklist along with a fourth processor, the 4C/8T Core i3-10105F.
For the rationale that Core i3 items are locked, we’re unable to make utilize of a 4.5 GHz clock frequency, as a change the 10105F ran at 4.2 GHz which is the spec all-core frequency, so it’s going to be working at a 7% decrease frequency than the K-SKU aspects. The Core i3-10325 would had been extra appropriate sort for this take a look at, however we had been unable to gain one in time for this take a look at. Even when the Core i3 half is working at a disadvantage, it must composed reside a enchanting addition given the mighty extra puny 6MB L3 cache.
The fastest quad-core CPU that Intel has produced so some distance is the Core i7-7700K, or one in all the higher clocked Comet Lake Core i3s, reminiscent of the Core i3-10325, both of which characteristic an 8MB L3 cache, or 2MB better than the 10105F. So it’s going to be enchanting to peer what extra or much less gains might perchance well be that you might perchance well per chance also think the usage of a quad-core CPU with 20MB of L3 cache, something we are in a position to produce with the 10900K by disabling half a dozen cores on that half.
To place this take a look at together we archaic the Gigabyte Z590 Aorus Xtreme motherboard, clocking the three Intel K-SKU CPUs at 4.5 GHz with a 45x multiplier for the ring bus and archaic DDR4-3200 CL14 dual-abominable, dual-channel memory with all predominant, secondary and tertiary timings manually configured. The Core i3-10105F archaic the same spec memory.
The bulk of the making an are attempting out became ran with the Radeon RX 6900 XT as it’s the fastest 1080p gaming graphics card you might perchance well per chance also expend, despite the indisputable fact that we maintain included some results with the RTX 3090 for a seek for at Nvidia’s overhead, which we suspect will maintain an be conscious on the quad-core configurations.
We’ll initiate with Rainbow Six Siege the place we previously saw the most essential efficiency variations between cache capacities. As an example, with all CPUs locked to 6 cores, we saw an enormous 18% elevate from 12MB to 20 MB of L3 cache.
Then with handiest 4 cores enabled, the margin from the 10600K to the 10900K is generally reduced to 13%, fairly just a tiny no longer up to the margin seen with 6 cores enabled. This offers us a flee that the higher L3 cache turns into much less effective at boosting efficiency with fewer cores accessible to originate basically the most of it.
Right here is simplest illustrated by the Core i3-10105F, which became shiny 9% slower than the 10600K, and be wide awake it is some distance clocked 7% decrease, so presumably in any case half that margin is down to the distinction in clock toddle. It be composed enchanting to existing that it became that you might perchance well per chance also think to enhance efficiency from the 4-core Core i3 by 24%, when the usage of the 10900K with shiny 4 cores active. That’s an enormous elevate given both CPUs utilize the same Comet Lake architecture with the final notice distinction being a 7% clock frequency variation and the L3 cache capacity, which is over 3x higher for the i9 half.
The takeaway right here is that no longer all cores are equal, despite the indisputable fact that the cores are bodily the same, a distinction in cache capacity can originate your total distinction. That and cramming extra cache correct into a CPU is composed important with shiny 4 cores active, however it absolutely’s much less effective than what we saw with 6 cores.
Subsequent up we maintain now Assassin’s Creed Valhalla which is a heavily GPU-sure title, in particular when the usage of the ‘Extremely Excessive’ quality preset, even with a Radeon RX 6900 XT at 1080p. Which capability the benchmark relies nearly fully on the GPU and handiest the Core i3 sees a runt dip in 1% low efficiency, suggesting that frame time efficiency isn’t as just with the 6MB L3 cache, however no longer critically worse both.
A brand unique addition to this making an are attempting out is Battlefield V. Right here we are in a position to look that the 20 MB L3 cache of the 10900K is terribly important, even when puny to 6 cores. Dialing down to 4 cores creates a CPU bottleneck that can’t be solved with extra cache. 1% low efficiency became improved by 9% when going from 16 to 20 MB which isn’t nothing, however it absolutely’s no longer up to the 13% assemble we saw with 6 cores enabled.
Now, with handiest 4 cores enabled, the 10900K became much like the 10600K stock. Truly, the frame time efficiency became higher, bettering 1% lows by a 12% margin. When it got right here to 1% low efficiency the Core i3 struggled a tiny, dipping down to 71 fps making it 19% slower than the 10600K and an enormous 30% slower than the 10900K. All yet again that’s an amazing distinction that’s nearly fully down to the distinction in L3 cache capacity.
The F1 2020 results are much like what we’ve seen so some distance, the place the L3 cache paucity made extra of a distinction with 6 cores, though we produce look some distinction with 4 cores active. The 10700K became shiny 3% slower than the 10900K, while the 10600K became a additional 3% slower than the 10700K. So shiny consistent scaling there. The Core i3 half though became 8% slower than the 10600K, or 11% slower if we seek for on the 1% low results.
We can estimate that at most ~5% of that margin shall be in consequence of the distinction in clock toddle, and if that is the case, scaling is composed roughly inline with that we saw for the K-SKUs.
Hitman is a CPU intensive title, and we are in a position to look a 9% elevate in efficiency when going from a 12MB to 20MB L3 cache with 6 cores enabled. That margin became reduced with shiny 4 cores enabled, this time to 6%. Then we saw a 9% drop in efficiency from the 10600K to the 10105F.
Horizon Zero Morning time is plenty like Assassin’s Creed Valhalla in the sense that it’s primarily a GPU sure title, even at 1080p with a 6900 XT. Which capability, the 4-core configurations weren’t an infinite deal slower than what we saw with 6, 8 and 10 cores enabled. The Core i3 half also performed smartly relative to the higher-stop Core i5, i7 and i9 items.
Cyberpunk 2077 is a if fact be told CPU intensive recreation and you’re going to also look an infinite merit when upgrading to a contemporary 6-core processor, reminiscent of the Core i5-10600K. That mentioned, had the 10600K been armed with a 20MB L3 cache it’d be 13% faster in this title when the 1% low efficiency. That margin remained rather in line with shiny 4 cores enabled as the 10900K configuration became 12% faster than the 10600K.
Curiously, the 4-core 10600K configuration became shiny 6% faster than the Core i3-10105F on life like, however 19% faster when comparing the 1% low efficiency and right here’s the place the Core i3 if fact be told struggled with shiny 64 fps.
When the Core i3 half we look an enormous 56% efficiency disparity between the 1% low and life like frame price, whereas the 10600K sees shiny a 39% margin and this suggests mighty extra consistent frame time efficiency.
It’s rate noting that the 10900K with shiny 4 cores active played thoroughly, providing relaxed and consistent efficiency, without reference to being 25% slower than the 6, 8 and 10-core configurations. The 10105F though suffered from noticeable stuttering and again that became reflected in the mighty weaker 1% low efficiency. So while a 10th-gen quad-core with a elephantine L3 cache can play the game shiny ravishing, it is some distance composed mighty slower than a 6-core an analogous, while going to eight cores offers no development in life like frame price efficiency or frame time efficiency.
Subsequent up we maintain now Shadow of the Tomb Raider which is one other CPU tense recreation however again, with shiny 4 cores enabled, the 10th-gen CPUs aren’t in a state to prefer merit of that additional L3 cache like they’re with 6-cores. It’s no longer till the L3 cache is place to 6MB with the Core i3 half that efficiency begins to descend away, losing the 1% low efficiency by 16% when compared with the 10600K.
Checking out with the RTX 3090
As we had been wrapping up the quad-core making an are attempting out, we thought it is some distance also enchanting to re-trot about a of these the usage of the GeForce RTX 3090. With 6-cores enabled in SoTR, the 10600K became 13% slower than the 10900K the usage of the Radeon and 15% slower with the GeForce, so no longer an infinite trade there.
Then with 4 cores enabled, the 10600K became 4% slower than the 10900K with the Radeon, while we look radically assorted results with the RTX 3090. Right here the 10600K is 14% slower than the 10900K for the frequent frame price and 13% slower for the 1% low. Right here is the of Nvidia’s added overhead by the usage of the CPU for plenty of its GPU scheduling.
Nevertheless, it’s the Core i3 half that is if fact be told crippled by the utility scheduling, tanking 1% low efficiency to shiny 54 fps, making it 28% slower than the 10600K, whereas it became 16% slower earlier than. Had we archaic the RTX 3090 for all making an are attempting out, the 4-core ends in Hitman, Cyberpunk 2077, Battlefield V, etc can also want been mighty extra essential.
We also examined Survey Dogs Legion with both Radeon and GeForce high-stop GPUs. Before with the Radeon we didn’t look mighty of a distinction with the K-SKU aspects, with 6 cores enabled the implications had been mighty the same and then 8 cores handiest offered a runt boost. Nevertheless, 4 cores drop efficiency fairly critically. The 10600K, for instance, became 17% slower with 4 cores enabled, and we saw an analogous margins with the 10700K and 10900K. That mentioned, efficiency between the a total lot of 4-core configurations is expounded and even the Core i3 half manages to hold in there.
The utilize of the RTX 3090 doesn’t produce mighty to the 4-core results. The 10600K became 19% slower with shiny 4 cores enabled when compared with its stock 6-core configuration. Impulsively though, the margins with 6 and 8 cores enabled are fairly assorted and curiously the cache performs the next characteristic with the RTX 3090 installed, presumably for the rationale that CPU is having to produce extra work.
It’s also enchanting how we don’t look the same scaling with 4 cores enabled, however it absolutely does seem as if right here’s too few cores to prefer merit of the increased L3 cache capacity.
What We Realized
This became a enchanting seek for at CPU efficiency however doubtlessly no longer what many of you had been looking ahead to. We imagine the expectation became that with fewer cores, the higher L3 cache of the 10900K would play an splendid increased characteristic, however for basically the most half that doesn’t seem like the case.
It’s clear that for mid to high-stop gaming, quad-cores are formally out. We’ve identified this for some time, which is why AMD and Intel maintain stopped producing mid-fluctuate quad-core processors. For decrease-stop systems nonetheless, quad-cores composed work smartly, though it ought to be clarified that after we’re announcing quad cores, we mean 4-core/8-thread processors that produce enhance simultaneous multi-threading.
Frame time efficiency can also just moreover be seen suffering in tense titles reminiscent of Battlefield V, Shadow of the Tomb Raider and Cyberpunk 2077, for instance, however in most games if you’re the usage of a funds GPU, reminiscent of the Radeon RX 5500 XT, GeForce GTX 1650 Neat, or anything slower, a just quad-core will permit an appropriate level of efficiency.
Frame capping to 60 fps will lend a hand relaxed our frame rates in a title like Cyberpunk 2077 as it reduces CPU load as smartly as frame to frame variance. So if you’re working into stuttering considerations, are attempting capping the frame price to something extra sustainable as that will lend a hand.