Intel i5 new: Intel’s new $196 16-thread Core i5 is an awesome reason to upgrade from an older CPU

12 марта, 2023

Intel’s new $196 16-thread Core i5 is an awesome reason to upgrade from an older CPU

Skip to main content

When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works.

(Image credit: Future)

Audio player loading…

The new Intel Core i5 13400F looks to be a mighty processor for the money. Coming with 10 cores and 16 threads, I’m actually wondering why I should spend any more on my next CPU upgrade than the relatively paltry $196 Intel’s asking for with this chip. From the specs announced at CES 2023 alone, this could be a budget gamer’s dream.

Let’s get right into this chip’s more than impressive spec. Like I mentioned, the Core i5 13400 comes with 10 cores, and those 10 cores are made up of six Performance-cores (P-cores) and four Efficient-cores (E-cores).

Those six P-cores are of a speedy ilk designed to accelerate your gaming experience. They’re capable of running up to 4.6GHz Max Turbo on the Core i5 13400F, which, while low next to say the Core i9 13900K’s 5. 8GHz Max Turbo speed, is absolutely quick enough for fast-paced gaming. 

That’s also a big P-core clock speed bump over the Core i5 12400F from the last generation. The 12th Gen chip operates at a more modest 4.4GHz, though admittedly even that sort of speed was enough to win Dave over in his Core i5 12400 review .

The Core i5 13400F’s four E-cores are more focused on background tasks and taking the strain off of the P-cores. There’s only a handful here, but their mere inclusion is cause for celebration—the Core i5 12400F didn’t come with any.

That makes for a 16-thread chip overall in the Core i5 13400. That’s a whole lot for this sort of cash, and should be handy for even more CPU intensive workloads beyond gaming.

In terms of cache, the Core i5 13400F comes with 20MB of L3 and 9.5MB of L2—another improvement over its 12th Gen counterpart. It supports both DDR4 and DDR5 across both 600-series and 700-series motherboards, which means you should be able to pick up a cheap board and get off to a good start. You could even go as cheap as ASRock’s B660M Pro RS, which is only $120 on Newegg right now.

What that all means is the Core i5 13400F will arrive with a similar recommended price tag to the Core i5 12400F but with four extra cores, all the benefits of Intel’s hybrid architecture, and run up to 200MHz faster. Not to mention the other Raptor Lake-based upgrades I’ve spoken about in more depth in my Core i9 13900K review . Only downside is it consumes a bit more power.

The ‘F’ at the end of the Core i5 13400F denotes this chip’s lack of integrated graphics, however. You won’t really miss the Intel UHD Graphics 730 chip you’ll find on the Core i5 13400, though if that processor is more readily available then it’s only a touch more expensive at $221.

Just bear in mind, these are Intel’s recommended customer prices. While we tend to see chips float around these sorts of prices once the initial launch day fervour has settled, right now the 13th Gen is a bit pricier than we’d like. These chips may take a little while to filter out into the market, too, so just keep an eye out for the best price as they come in.

The Core i5 12400 scored 95% in our review. (Image credit: Future)

Neither the Core i5 13400 or 13400F are overclockable the traditional way. There’s a slim chance some motherboard manufacturers may allow some sneaky BCLK overclocking on the side with these locked chips, as this was the case with some of Intel’s 12th Gen processors (and we loved the Core i5 12400 for that). Yet so far it appears as though BCLK overclocking in this way doesn’t play nicely with the 13th Gen, so we’ll have to wait and test for ourselves to be sure either way.

Swipe to scroll horizontally

13th Gen Core i5 versus 12th Gen Core i5
Processor Cores (P+E) Threads Max P-core Turbo (GHz) Processor Base Power/Max Turbo Power (W) Price
Core i5 13400 6+4 16 4. 6 65/148 $221
Core i5 13400F 6+4 16 4.6 65/148 $196
Core i5 12400 6+0 12 4.4 65/117 $211–221
Core i5 12400F 6+0 12 4.4 65/117 $184–194

Your next upgrade

(Image credit: Future)

Best CPU for gaming : The top chips from Intel and AMD
Best gaming motherboard : The right boards
Best graphics card : Your perfect pixel-pusher awaits
Best SSD for gaming : Get into the game ahead of the rest

For any PC gamer still rocking an Intel Core i7 6700K or a chip of a similar four-core design (I have a build with one such chip in it still to this day), it’s tough to deny the Core i5 13400F as a huge upgrade. Heck, it’s a pretty awesome upgrade even if you’ve got a more recent CPU than that, but then again I’m not a big proponent of needlessly changing CPUs for the sake of it. 

There’s no denying that CPUs really have come on leaps and bounds as of late, and it’s that firm competition between Intel and AMD that’s fuelling the fire. There’s no better example of that in my mind than the Core i5 13400F’s impressive spec—a big improvement on what was already a mighty mid-range offering in the Core i5 12400. 

In terms of competition, the Core i5 13400F should be quite a bit cheaper than AMD’s latest Ryzen 5 7600X, which sells for around $299, meaning there’s no real opposite number from the red team—yet. AMD does have a keynote at CES 2023 tomorrow, however, and rumour has it we could hear more about its more entry-level desktop CPU plans then or some time this month.

In the meantime, I’m going to do my best to get this chip in the PC Gamer test bench and see what it’s really able to deliver. Though I don’t foresee it being anything other than awesome.

Sign up to get the best content of the week, and great gaming deals, as picked by the editors.

Contact me with news and offers from other Future brandsReceive email from us on behalf of our trusted partners or sponsors

Jacob earned his first byline writing for his own tech blog from his hometown in Wales in 2017. From there, he graduated to professionally breaking things as hardware writer at PCGamesN, where he would later win command of the kit cupboard as hardware editor. Nowadays, as senior hardware editor at PC Gamer, he spends his days reporting on the latest developments in the technology and gaming industry. When he’s not writing about GPUs and CPUs, however, you’ll find him trying to get as far away from the modern world as possible by wild camping.

PC Gamer is part of Future plc, an international media group and leading digital publisher. Visit our corporate site .

©
Future Publishing Limited Quay House, The Ambury,
Bath
BA1 1UA. All rights reserved. England and Wales company registration number 2008885.

Intel Core i5-13500 Review | TechSpot

The Core i5-13500 is Intel’s new $250 offering, packing 14 cores in total, 20 threads and 35.5 MB cache. This is an interesting almost hybrid-like CPU as it packs the larger L3 cache capacity of the Raptor Lake i5’s while using the smaller L2 cache of Alder Lake, so it’s not merely a locked version of the impressive 13600K, so we’re very keen to see how it performs.

The 13500 is an interesting processor. When compared to the 13600K you get the same 6 P-Cores with Hyper-Threading support for 12 threads, but where the 13600K cores clock as high as 5.1 GHz, the 13500 is capped to 4.8 GHz.

The big change is to the P-cores, where the L2 cache has been downgraded from 2 MB to 1.25 MB, which is the same capacity as Alder Lake CPUs, while the K-SKU Raptor Lake parts all received a 2 MB upgrade which does help boost performance.

The 13500 does retain all 8 E-cores, for the same core count as the 13600K and twice that of the 12600K. This is a big deal as 12th-gen locked Core i5’s such as the 12500 didn’t feature any E-cores at all, so we’re going from zero to eight in a single generation. However, like the P-cores, the E-cores are receiving the 12th-gen cache capacity, so unlike the 13600K E-cores which pack a 4 MB cache capacity per cluster, the 13500 has been downgraded to 2 MB per cluster.

Despite these changes, the L3 cache capacity remains 24 MB, the same as the 13600K and 4 MB more than what you get with the 12600K and 6 MB more than the 12500.

At $250 the Core i5-13500 could be well positioned as the best value all rounder for budget oriented builds and its main competition here would be the Ryzen 5 7600.

The 13500 has the advantage of working on cheaper B660 boards using DDR4 memory, so we’ll be looking at that along with a DDR5 configuration. For this we’ll be using G.Skill Ripjaws V 32GB DDR4-3600 CL16 memory as it can currently be had for $115, it’s a good quality kit that offers a nice balance of price to performance.

Please note that the locked Intel processors can’t run above 3466 if you want to use the Gear 1 mode, and ideally you’d want to use Gear 1 as the memory has to run well above 4000 before Gear 2 can match Gear 1. The reason for this limitation is due to the System Agent (or SA voltage) which is locked on Non-K SKU processors, and this limits the memory controller’s ability to support high frequency memory using Gear 1 mode. That being the case, we’ve tested the 12100 and 13100 using the Ripjaws memory adjusted to 3466, as this is the optimal configuration for these processors short of going and manually tuning the sub-timings, which we’re not doing for this review.

The key competitor for the Core i5-13500 is the Ryzen 5 7600, so this will be the comparison we focus on. However, we’ve also included other processors for reference, but please note the data isn’t always an apples to apples comparison in terms of memory configurations. The AM4 processors were tested using DDR4-3200 CL14 memory (in the past we have found this comparable to DDR4-3600 CL16), but we still wanted to point out that difference. Let’s get into the data…

Clock Speeds

Here’s a look at how the Core i5-13500 operates using the supplied box cooler, which we won’t be using for our testing as it’s not very good and you should definitely replace it when buying this processor, a cheap $20 tower style cooler will do the trick.

With the box cooler installed thermal throttling kicked in after roughly 25 seconds. Initially the 13500 P-cores ran at 4.5 GHz with the E-cores at 3.5 GHz, after an hour the P-cores were running at 4.2 GHz and the E-cores at 3.4 GHz, so a 7% reduction on the P-cores and just a 3% drop for the E-cores.

Those aren’t massive performance dips but it’s worth noting that the core temperature ran at 92c with a CPU package power of 100c, and this is while throttling.

So as usual we ditched the terrible box cooler for the testing, and instead used the be quiet! Pure Loop 2 FX 360mm AIO. With the liquid cooler the i5-13500 maintained a P-core frequency of 4.5 GHz and an E-core frequency of 3. 5 GHz, so after an hour of load we saw no dip in the clock speeds and that’s because the 13500 ran at just 62c, which is well below the thermal threshold.

Application Benchmarks

Starting with Cinebench R23 multi-core results we see that the Core i5-13500 is very impressive for core-heavy productivity workloads that can utilize the E-cores well.

When compared to the Ryzen 7600, we’re looking at a 44% increase in performance, a massive gain that sees the locked Core i5 even outpacing the 7700X. It was also just 12% slower than the 13600K and that difference in this test is purely down to clock speed as cache plays a very small role here.

The single core performance is similar to the Ryzen 7600 and slightly less than the 12600K for the simple fact that the 12th-gen part clocks slightly higher. Still, when compared to the 13600K we’re talking about a small 6% reduction in performance.

The 7-Zip compression results are heavily influenced by memory performance. The Core i5-13500 using an optimal DDR4 configuration is only ~10% faster than the 5600X and 16% slower than the 7600. However, when armed with DDR5 memory the 13500 is 14% faster than the 7600, though it did still trail the 13600K by a 14% margin. Still, overall very good performance for the price.

DRAM performance has almost no influence on decompression performance, so here the 13500 does well using either DDR4 or DDR5 memory, delivering comparable performance to the 5800X, making it up to 10% faster than the 7600.

The Blender Open Data results are very impressive, here the E-cores are working well and as a result we’re looking at Ryzen 7 7700-like performance, making the 13500 at least 28% faster than the 7600 and just 12% slower than the 13600K.

It’s a similar story in the Corona benchmark, here the Core i5-13500 is 13% slower than the 13600K, but 37% faster than the 7600, and performance was much the same using either DDR4 or DDR5 memory.

Next up we have Premiere Pro 2022, and here we see when using DDR4 memory that the 13500 is slower than the Ryzen 5800X, as well as the 7600, trailing by a 9% margin.

But when paired with DDR5 memory the locked Core i5 enjoys a massive 21% performance bump, allowing it to beat the 7600 by an 11% margin to come in just 4% slower than the 13600K.

We’re also seeing solid performance gains when upgrading from DDR4 to DDR5 in Photoshop, with DDR5 providing an additional 17% performance. Even so the E-cores are virtually useless in this lightly threaded application, meaning the Ryzen 7600 is still slightly faster here.

In Adobe After Effects the Core i5-13500 managed to beat the Ryzen 7600, even when using DDR4 memory, though performance was virtually identical. Armed with DDR5 though the 13500 was 9% faster than the 7600 and just a percent slower than the 13600K.

The last productivity benchmark we have is code compilation and here the Core i5-13500 again performed very well, trailing the 13600K by a mere 3% margin or 9% when using DDR4 memory. Either configuration saw it beat the 7600 and it was 17% faster when using DDR5.

Gaming Benchmarks

Time for some gaming benchmarks and we’ll start with Factorio. Here the Core i5-13500 performed much the same using either DDR4 or DDR5 memory and when using DDR5 it was just 3% slower than the Ryzen 7600 and just 5% slower than the i5-13600K. Overall a great result for the locked Core i5 part.

We’re going to focus on the 1080p data but we will show you the 1440p average data as well for the 10 games we tested.

In Watch Dogs Legion, we see that the i5-13500 performed much better using DDR5 memory, delivering almost 20% greater performance. When limited to DDR4 we’re looking at similar performance to the Ryzen 5800X, but with DDR5 the 13500 was able to match the 7600, making it 9% slower than the 13600K.

The Rainbow Six Extraction results are very interesting as here the 13500 appears to struggle quite a bit, though the results make sense. The E-cores are of course useless here, so when it comes to gaming performance the Core i5-13500 is going to be similar to the 12600K, it’s clocked slightly lower, but it packs 20% more L3 cache.

We see that when armed with the same DDR5 memory that the 13500 is 5% faster than the 12600K, but 14% slower than the 13600K and a massive 23% slower than the 7600. The Zen 4 processors perform very well in this title, though to be fair the 1% lows for the 13600K and 7600 are comparable.

The Hitman 3 performance, while very good overall, is a bit lackluster when compared to the Ryzen 7600, which matched the 13600K, at least for the average frame rate. Again, DDR5 provided the best results, boosting the average frame rate by 13% to 205 fps, the same performance delivered by the 5800X3D, and this saw the 13500 trailing the 7600 by an 8% margin.

The Core i5-13500 managed 110 fps in Spider-Man Remastered when using DDR5 and 99 fps with DDR4 making the DDR5 configuration 11% faster.

That also meant that the 13500 was just 4% slower than the 7600 and 7% slower than the 13600K which it beat the 12600K by a 7% margin, so overall solid performance and right where we’d expect this part to land.

Here we’re looking at a rather substantial 20% performance uplift for the Core i5-13500 using DDR5 opposed to DDR4, pumping out 228 fps to make it 5% slower than the Ryzen 7600 and 10% slower than the 13600K. So again where we’d expect it to be given the specs, around 7% faster than the 12600K.

Like Rainbow Six Extraction, Horizon Zero Dawn heavily favors Zen 4 CPUs, and in this instance quite heavily. DDR5 memory also doesn’t help the 13500 here and although it was just 8% slower than the 13600K, that made it a massive 26% slower than the 7600.

The benefits of DDR5 memory are also apparent in Cyberpunk 2077 where the Core i5-13500 is up to 22% faster thanks to the additional bandwidth. Even so with 181 fps on average that meant it was slightly slower than the Ryzen 7600, just a 4% margin in this example.

ACC benefits more from L3 cache performance than it does DRAM performance, but even so DDR5 is offering a mild 5% jump for the 13500, which saw it come in 10% slower than the 13600K and 7% slower than the 7600.

In The Riftbreaker the 13500 outperforms the 7600, delivering 8% greater performance when armed with DDR5 memory. That also made it just 5% slower than the 13600K, a great result for Intel’s new locked Core i5 processor.

Last up we have Counter-Strike: Global Offensive and this is another title where AMD Zen 4 processors perform really well. With 414 fps on average, the i5-13500 was 11% faster than the 12600K, just 6% slower than the 13600K and 11% slower than the Ryzen 7600. A good result relative to the 13600K, but a weaker than typical result compared to the mainstream Ryzen part.

10 Game Average

Across the games tested, the Core i5-13500 with DDR5 memory was on average 9% slower than the Ryzen 7600.

If we were to remove games that heavily favored the Zen 4 processors such as CSGO, Rainbow Six Extraction, and in particular Horizon Zero Dawn, these two processors would be more evenly matched. However, we’ve found this sample of games generally reflects what we see across massive 40-50 game benchmarks, at least within a few percent points.

The Core i5-13500 does stack up well to previous generation CPUs when using DDR4 memory, offering 7% greater performance than the Ryzen 5600X, which we estimate would make it around 12% faster than the 12400.

Also here are the 1440p results for those of you interested, also using the RTX 4090.

The Ryzen 7600 was 9% faster than the 13500 using DDR5 memory, or 16% faster when limiting the i5 to DDR4. Using DDR5 the 13500 was also 5% faster than the 12600K, but 7% slower than the 13600K.

Power Consumption

When it comes to power consumption for an all-core workload, the Core i5-13500 is relatively power efficient, pushing total system usage to 231 watts using DDR4 and 266 watts with DDR5.

That’s comparable power usage to the Ryzen 7700X for similar performance, the i5-13500 was just 5% slower in this test.

Having said that, the 7700X isn’t exactly an efficient product, the non-X version delivered basically the same level of performance while using considerably less power. Still, in this example, the 13500 matched the 7700 when using DDR4 memory and here we see that power usage for the entire system is just 9% higher, so not too bad.

Cost vs. Performance

Running a cost per frame analysis, for the 13500 we’ve got two prices, one with the $115 G.Skill DDR4-3600 CL16 kit and another using the DDR5-6400 kit. The DDR5 configuration is based on a $160 Z690 board, while the DDR4 configuration is based on a decent $130 B660 board.

The Zen 4 processors are priced with the DDR5-6000 CL30 memory they were tested with along with a $180 B650 board. We know there is one Gigabyte B650 board selling for $160, but most are around $180 and up as this increases your options to three B650 boards.

Based on those prices the Ryzen 7600 comes out to a cost of $2.53 per frame and technically with a little overclocking you can match the 7600X which would reduce the cost per frame of the non-X part to $2.43, but we’re not including that data here as were focusing on out of the box performance.

When it comes to gaming performance the Core i5-13500 is a little disappointing in terms of value, with the DDR4 version coming in at a cost of $2.71 per frame or $2.77 for the DDR5 configuration, making it 9% more costly per frame when compared to the 7600.

Even if we ignore the performance data, the cost of the 13500 on a DDR5 motherboard is equal to that of the 7600, while opting for DDR4 will save you a mere $65, and will without question result in less gaming performance.

What We Learned

The Core i5-13500 is a very interesting offering and certainly the best of the locked 13th-gen CPUs we’ve tested so far. With its 8 E-cores, this is a genuinely strong productivity CPU, power consumption is very reasonable and thermals are easy to keep in check with a decent cooler. Gaming performance is also solid, though it needs to go DDR5 to compete with the likes of the Ryzen 5 7600.

When it comes to productivity it can edge out the Ryzen 7 7700X, though depending on the workload it’s not always a great deal faster than the 7600X like what we saw in our code compilation benchmark, and it can end up slower for lightly threaded applications such as Photoshop. Overall though, the 13500 is the superior productivity CPU.

Gaming is a lot less clear cut and we think things swing the other way with the 7600 being the superior choice here, even when factoring in value. The Ryzen 7600 can also be easily overclocked to produce 7600X levels of gaming performance, and while that doesn’t change things drastically, it does further advantage the Ryzen processor.

Sadly, Intel has completely locked down BCLK overclocking for 13th-gen Non-K processors, so even if you have a motherboard sporting an external clock generator, it won’t be able to do anything with the 13500, which really sucks and is a big blow to this part for enthusiasts.

The Intel box cooler also sucks, it’s loud and sees the i5-13500 suffer thermal throttling issues in well under a minute whereas the Wraith Stealth that comes with the Ryzen 7600 avoids any throttling and is also relatively quiet. You can also achieve the performance shown in this review with the AMD box cooler, but that’s not always possible with Intel’s cooler.

But in our opinion the single biggest blow to the Core i5-13500 is the LGA 1700 platform, which is now at the end of the road with no future prospects.

Meanwhile, AM5 will be supported for years to come and will support future CPU generations. Zen 5 is expected to be a radical update, so having the ability to support future processors is a massive win, and it’s why we’d buy the Ryzen 7600 over the i5-13500.

Had the Intel CPU offered more value for gamers, then it would make the choice a lot harder, but the fact that you need DDR5 memory to compete with the 7600, and DDR4 doesn’t help much with the value equation anyway, unless you opt for a 16GB kit, but we think at this point new system builders will want to invest in a 32GB kit. You can also purchase 16GB DDR5-4800 CL38 kits for around $80 and while that might be around twice the price of a comparable DDR4 kit, a $40 saving is peanuts once you factor in the cost of the motherboard and CPU.

For productivity we could see us investing in the Core i5-13500, though we’ve got to admit for just $80 more the Ryzen 7 7700 would tempt us, as performance overall is similar and you get that upgrade path. But we could just keep going, the non-K 13700 can be had for $410 or the Ryzen 9 7900 for $430, and at that point we’re talking about a very different budget.

Shopping Shortcuts:
  • Intel Core i3-13500 on Amazon
  • AMD Ryzen 7 7600 on Amazon
  • AMD Ryzen 7 7700 on Amazon
  • AMD Ryzen 7 7700X on Amazon
  • Intel Core i5-13600K on Amazon
  • Intel Core i7-13700K on Amazon
  • AMD Ryzen 7 5800X3D on Amazon

Testing an unusual Intel Core i5-12400F. What does scalping, polishing and pressure frame give? — i2HARD

Petr alykard__

August 8, 2022

We are conducting clinical trials with the Intel Core i5-12400F processor to reduce the temperature regime.

Intel’s 12th generation Alder Lake family of processors brought many major changes, making the new product announcement a welcome milestone. Among the innovations, first of all, we note the transition to the LGA1700 socket with support for the PCI Express 5. 0 interface and, of course, the new DDR5 RAM standard, the hybrid architecture with the breakdown of cores into the so-called P- and E-cores. The changes affected the power management scheme, the introduction of a microcontroller and Thread Director scheduler technology, Intel Z69 system logic0 and many more.

The competitive struggle between the two giants (AMD and Intel) leads to the fact that manufacturers are constantly striving to develop modern technologies and expand capabilities, giving rise to revolutionary and even evolutionary processes. But the emergence of new solutions does not always go smoothly, so if at first Intel successfully eliminated some features of the Aldel Lake processors with firmware updates, then users fix other problems themselves. For example, too «soft» socket LGA1700, deformation of the heat-distributing cover — problems that the blue giant officially denies.
nine0003

Using the popular Intel Core i5-12400F processor purchased from a Chinese site for home use as an example, we will try to understand some of the nuances of its operation and perform optimizations, as well as check popular methods for eliminating them. The project WORKSHOP by i2hard will help us with this.

Learn more about Core i5-12400F


Core i5-12400F
Core i5-12400
Platform LGA1700 LGA1700
Microarchitecture Alder Lake Alder Lake
Technical process, nm 10 10
Cores/Threads 6P+0E/12 6P+0E/12
Frequency (nominal/turbo), GHz
2.5 — 4.4 2.5 — 4.4
Full load frequency, GHz
4.0 4.0
L2 cache, KMB
6×1.25 6 × 1.25
L3 cache, MB eighteen eighteen
nine0013 PL1/PL2 limits, W 65/117 65/117
Memory
DDR4-3200
DDR5-4800 		DDR4-3200
DDR5-4800
Integrated graphics Not UHD 730

You can find a complete review of the Intel Core i5-12400 processor at the link «Intel Core i5-12400 test. Comparison with Core i5-11400 and Ryzen 5 5600X».
nine0003

The Core i5-12400F processor continues the glorious dynasty of the «people’s» chosen six-core Core i5-10400 and Core i5-11400, being the golden mean for gaming. The novelty is deprived of energy-efficient E-cores, for the sake of cost, hybridity and Thread Director technology also went under the knife.

In nature, there are two steppings of the Intel Core i5-12400 (F) processor: C0 — a rejection from older processors with an area of ​​215 mm 2 (which have an 8 + 8 layout) and H0 — a pure 6-core processor with an area of ​​162 mm 2 , which physically lacks E-nuclei. Each model has its own advantages and disadvantages, for example, the H0 version will have slightly lower inter-core delays due to the lack of blocked non-working cores, the C0 version will have a slightly better temperature regime due to the larger crystal area. However, these results will be very low and are at the level of errors.

Test bench configuration

  • Motherboard: Gigabyte B660I Aorus Pro DDR4; nine0182
  • Processor: Intel Core i5-12400F;
  • Cooling system: Noctua NH-L9i-17xx chromax.black;
  • Thermal interface: Noctua NH-h2;
  • RAM: Team Group T-Force Xtreem 8Pack DDR4-4500 2×8 GB;
  • Video card: Radeon RX 6400;
  • Drive: Western Digital Black SN750, 500 GB;
  • Power supply: be quiet! Straight Power 11 850W;
  • Monitor: AOC U2790PQU;
  • Housing: open stand.

    • nine0201

    First test results

    For a compact gaming PC based on the Intel Core i5-12400F processor paired with the Noctua NH-L9i-17xx chromax.black cooling system, an interesting project is being prepared in the original case, which you will get acquainted with in the near future on the i2hard portal. In its stock state, the Intel Core i5-12400F processor warms up to 70-80 ° C, depending on the load and fan speed, which cannot be called a critical value. However, a logical question has arisen: what if we optimize the operation of the processor and improve its cooling in some popular ways, without resorting to replacing the cooler? And to begin with, it was decided to study the reference state of affairs.
    nine0003

    At the first acquaintance with the mount, the fact of the curved distribution of thermal paste was noted (the popular method is used with applying one drop in the middle). A characteristic hump has appeared on the left side of the heat-distributing cover of the Core i5-12400F, and a crescent moon mark has appeared on the right side. There are multiple tests confirming that the standard LGA1700 mount presses harder on the side where the lock lever is located. Even with reference fasteners, there are several recommendations for eliminating them, for example, artificially shifting the frame to the left when closing. However, our goals go further.
    nine0003

    Unfortunately, we did not manage to get a picture of the exact fit of the cooler base to the heat-distributing cover. But you can see how unevenly distributed thermal paste.

    It is also necessary to say a few words about the tested Intel Core i5-12400F processor: this is the C0 stepping, bought on Aliexpress. Lucky so lucky.

    Pressure plate Thermalright LGA1700-BCF

    One of the popular Intel LGA1700 soft socket solutions is the Thermalright LGA1700-BCF pressure plate, which replaces the original. Thermalright has released four color frames (black, grey, red and blue) that enthusiasts will love. For our projects WORKSHOP we purchased black and red plates.

    Thermalright LGA1700-BCF comes in a small box with installation instructions, the frame itself and a hex key. The manual describes the installation mechanism in detail and gives appropriate recommendations.

    The original mounting frame on the example of the Gigabyte B660I Aorus Pro DDR4 motherboard.

    After removing the four screws, the side posts are removed, revealing a bare socket with spring-loaded 1700 pins. The processor is installed in the socket, Thermalright LGA1700-BCF is mounted on top. For clarity, we chose the red color, carefully twist the reference screws diagonally in half a turn.
    nine0003

    The Thermalright LGA1700-BCF plate itself is not involved in cooling. Its main task is to evenly redistribute the load on the socket and eliminate the bending of the processor. There is no need to press it all the way, otherwise we will also get a negative effect and bending of the motherboard textolite. Therefore, we recommend loosening the mount by 1/4.

    The final result of installing the Thermalright LGA1700-BCF is pleasing to the eye. The mounting frame is designed in such a way that the Intel 12th processors are perfectly fixed around the perimeter.
    nine0003

    Let’s look at the imprint of thermal paste on the base of the Noctua NH-L9i-17xx chromax.black cooling system. You can call it much better than what we saw earlier.

    And the footprint on the Intel Core i5-12400F’s heat-spreading cover is nearly flawless. From this we conclude that there is an effect from using the Thermalright LGA1700-BCF. But we will get acquainted with the detailed results after testing.

    Scalping Core i5-12400F by WORKSHOP

    The next step is familiar to enthusiasts and has become widespread since the 7th generation Intel Core processors Kaby Lake — a method of scalping the processor and then replacing the thermal interface with liquid metal. Despite the use of solder in the new 12th generation Intel Core line, replacing it also has a positive effect.

    The WORKSHOP by i2HARD gaming computer workshop provides a scalping service for Intel 12th Alder Lake processors of any complexity. Scalping the low-end Core i5-12400F processor is overkill, but we wanted to explore all sorts of ways.
    nine0003

    The scalping technique hasn’t changed in a few years: lateral shift of the heat-distributing cover with fixation of the processor platform. The main difference is that the 12th generation Intel Core processors use solder with a soldering method. We were interested in looking at the scalped Core i5-12400F C0 stepping processor; the die area is identical to the Core i9-12900K version.

    Due to the wide profile and rich experience, we have noticed that the sealant has become tougher. The solder itself is also difficult to remove, but freelance liquid metal also helps in partial removal.
    nine0003

    The reference Intel Core i5-12400F processor without a heat-spreading cover looks absolutely terrible! But we will wash it, clean it, varnish the small SMD capacitors and bring them into a divine form.

    Nearly cleaned lid with stock Intel solder removed. Next, the processor is going back.

    Polishing Core i5-12400F

    Another solution to reduce the temperature of the Intel Core i5-12400F is to polish the heat spreader cover. There are many examples of the curvature of its surface on the web. In our case, the deformation was most severe on the left side, which we clearly tried to capture in a schematic image. nine0003

    It took several steps to align the cover with a dremel, and then with #2000 sandpaper.

    However, we did not take into account that even on the finest sandpaper, litter can get caught, which, like the smallest sand, will scratch the heat-distributing cover.

    Using polishing paste can achieve a mirror effect, but will not remove deep scratches. They are not felt tactilely, and visually live they do not look as depressing as the camera transmits.
    nine0003

    Using a dremel with a soft nozzle and GOI No. 3 paste, we grind the cover.

    Again we resort to the help of polishing paste and after half an hour of the total time we achieve the desired result!

    Downvolting Core i5-12400F

    Separately, it is worth mentioning the method that is the safest and does not require mechanical action on the processor, unnecessary expenses and nerves — manually lowering the voltage in the BIOS or downvolting. Modern AMD and Intel processors have become directly dependent on the operating voltage, so its change is possible in a short range, then this will lead to a decrease in performance, and with a strong decrease — loss of stability. The method is safe in that if problems arise, you can always reset the BIOS settings and try everything again.
    nine0003

    For 12th generation Intel Core processors on a dozen units from i3-12100F to i9-12900K, the average downvolting value ranges from 0.05 V to 0.1 V (or 50-100 mV). When the threshold value is exceeded by 0.025 V, there is a performance drop of 3-7%, another 0.025-0.050 V — unstable operation (BSOD, errors in LinX, etc.).

    For our copy of the Intel Core i5-12400F, the negative Offset of 0.125 V became the decisive value! This means that in a simple operating voltage and load, the system reduces by 125 mV, which directly affects power consumption and, as a result, temperature. This is consistent with formula P = I * V (in its rough manifestation for the processor).

    Core i5-12400F Processor Test

    The C0 stepping Intel Core i5-12400F was tested on a Gigabyte B660I Aorus Pro DDR4 motherboard with the latest F4b BIOS. The cooling system was a low-profile Noctua NH-L9i-17xx chromax.black cooler with a fixed rotation speed of ~2200 rpm.

    First, let’s look at the performance of the memory subsystem with the stock JEDEC DDR4-2133 profile.
    nine0003

    When XMP is activated with DDR4-4500 parameters, the available clock frequency is set to 4533 MHz due to a change in the multiplier. The timings are increased by one cycle, which is a Gigabyte BIOS feature for a possible increase in stability. The Gear-2 mode with a frequency of 4533 MHz turns out to be more productive than the Gear-1 DDR4-3600, so we will leave the current settings. Let’s move on to the most interesting!

    Testing the Intel Core i5-12400F processor in several modes showed interesting system behavior. At idle, the Thermalright LGA1700-BCF clamping frame produces a temperature decrease of a couple of degrees Celsius. Scalping allows the gap to widen slightly.
    nine0003

    A significant effect and temperature delta is visible with a high load of the Cinebench R23 test on all cores. Activating the XMP profile leads to a temperature increase of 4°C due to increased voltage to the memory controller and other CPU nodes. The Thermalright LGA1700-BCF pressure plate is not involved in cooling, but it evens out the surface of the processor, which allows you to bring down three degrees Celsius. If additional scalping is performed, we will get a decrease in heating by 5°C, polishing — by another minus 2°C. The effect of downvolting is also noticeable. As a result, we manage to reduce the temperature of the Core i5-12400F by 13°C in total!
    nine0003

    Under extreme load with Prime95, we see an identical temperature pattern depending on the methods used. The end result is that the Intel Core i5-12400F is up to 15°C cooler when fully optimized! It is important to know that if downvolting is used alone, without polishing, scalping and clamping frame, the result is also good: from 72°C we have achieved a decrease in temperature to 66°C.

    Not paradoxical, but true! As the temperature decreases, the performance of the latest generations of processors increases. An increase of +2% is observed with an increase in the clock frequency of DDR4 memory, but our methods also allowed us to win back +2.2%.
    nine0003

    On the older Intel Core i9-12900K processor, after scalping, we were able to achieve a temperature drop of 11-15°C, and more interestingly, the processor consumption also decreased by 10 watts from 289 watts to 279 watts. Lower temperature — lower leakage currents.

    We decided to check and measure such a feint on the test processor Core i5-12400F, but, alas, it is not too voracious for us to see noticeable changes in energy consumption, which are more like an error.
    nine0003

    Conclusion

    As one of our good comrades, an enthusiast and a blogger once said: «In achieving the goal, all means are good!»

    Our task was to reduce the temperature of the Intel Core i5-12400F processor as efficiently as possible, so that later it could be used in an SFFPC gaming compact PC with a low-profile cooling system and even with an Intel Laminar RM1 box cooler at an average rotation speed and a comfortable noise level at an acceptable heating. We used available methods and decided to share the results with you, dear i2hard readers!
    nine0003

    The advertised clamping frame Thermalright LGA1700-BCF is not an empty phrase! The pressure became really better, and the average gain in temperature was 3-5 ° C, allowing, for example, to set the CO revolutions at the best values.

    The scalping method is the most effective method! Despite the use of solder on the 12th generation Intel Core, getting an additional 5-7°C reduction in heat even with a low-end processor is considered a good indicator. Therefore, the procedure will be useful to owners of Core i7 and Core i9 processors..

    Polishing the processor cover is a tedious process and takes personal time. But this will not stop a real enthusiast, but every overclocker will want to get a mirror surface and, as a result, an ideal clamp.

    But we recommend that beginner gamers start experimenting with the voltage reduction method, which does not lead to a loss of warranty or additional expenses, is formally safe and accessible to everyone!

    In conclusion, we want to say that our little investigation has come to an end! The Intel Core i5-12400F processor is 12-15°C cooler with the current cooling system, and if we had a board with an external clock generator, this would allow it to be slightly overclocked. In the meantime, we will quietly reduce the fan rotation curve at the peak from 2200 rpm to 1600 rpm, making the PC almost silent!
    nine0003

    Overview of the Intel Core i5 660 processor / Sudo Null IT News With the release of the first Intel processors for LGA1156, many advanced users «buried» this connector. Intel’s second blow didn’t keep us waiting long. Processors have entered the market that are even cheaper, cooler and even better overclocked. We will talk about one of these processors today. The Intel Core i5 660 (3.33GHz) is one of the older dual-core models of the 32nm Intel family codenamed Clarkdale. The Intel Core i5 660 will compete with the Inte Core 2 Duo E8600 (3.33GHz), which is considered the fastest dual core of the Core architecture. nine0003

    Representatives of Intel at their press conferences constantly tell the IT community about the Tick-Tock system. The name is chosen in the best possible way, and it reflects the real policy of Intel on the production of chips. As we know, Intel processors are usually divided into generations. The Tik-Tok system under consideration says that Intel devotes two technical processes (two stages) to the development of processors for each generation. The first is the running-in of a new architecture, the so-called “performance boom”. When the success of the new generation is consolidated, Intel transfers processors based on the same architecture to a new, thinner process technology, thereby winning on several indicators at once — heat dissipation and overclocking, as well as the greater availability and prevalence of the old process technology. nine0003

    Clarkdale — interesting in that Intel puts it on one of the top steps in its evolution of processors and chipsets. Most recently, we were left without the usual heatsink on motherboards that cools the north bridge. Now it is not needed there — the memory controller is integrated into the processor. Following the memory controller, the PCI-E bus controller followed in the brain center of the system. Now in the processor we also get a video card — an integrated graphics core. nine0003

    The image above shows the Intel Core i5 6xx and Intel H55 from left to right. It is the chipsets with the «H» designation that tell users that they support the integrated graphics built into the new Intel processors. Naturally, one can only dream of the crazy performance of the integrated graphics core. Probably, many remember the story that has spread all over the Internet. The graphics core built into the Intel processor showed very good performance in 3DMark Vantage, outperforming even discrete low-end solutions. But it was worth renaming the executable file to Vintage (instead of Vantage), as a significant part of the parrots was lost somewhere. nine0003

    In the Clarkdale line, the Intel Core i5 660 occupies one of the top positions. Its analogue can be considered the Intel Core i5 661 processor, which in terms of its processor characteristics does not differ from the i5 660 at all. But the i5 661 has a higher graphics core frequency — 900 MHz versus 733 MHz for the i5 660. The heat dissipation level, therefore, due to the lower graphics core frequency, does not exceed the threshold value of 73W.

    The new dual-core processors, produced according to the 32 nm process technology, are interesting to compare with the previous generation of dual-core Core processors, namely with the competitor of the same frequency — Intel Core 2 Duo E8600. A processor that has gained popularity among many users may now be beaten. This is exactly what I decided to find out by comparing the Intel Core i5 660 with the Inte Core 2 Duo E8600. nine0003

    The test bench was based on the Asus Maximus III Formula motherboard, which left a double impression of itself. Incorrect work with the TurboV utility left only negative emotions, but in general, the performance of this motherboard is at a good level.

    Test stand (LGA1156):
    Processor Intel Core i5 660
    Cooling Zalman CNPS10X Extreme
    Motherboard Asus Maximus III Formula
    RAM Kingston HyperX KHX2000C9D3T1K3/6GX 32048MB CL9 (two sticks of memory were used)
    Video card AMD Radeon HD4890
    Power Supply Enermax Revolution 1050W
    Hard drive 500 GB Seagate Barracuda 7200. 12 ST3500418AS, 7200 rpm.

    Test stand (Socket 775):
    Processor Intel Core 2 Duo E8600
    Cooling Zalman CNPS10X Extreme
    Motherboard Asus Rampage Extreme
    RAM Kingston HyperX KHX2000C9D3T1K3/6GX 32048MB CL9 (two sticks of memory were used)
    Video card AMD Radeon HD4890
    Power Supply Enermax Revolution 1050W
    Hard drive 500 GB Seagate Barracuda 7200.12 ST3500418AS, 7200 rpm.

    The operating system used for testing was Windows 7 Ultimate x86 (32 bit), version of the driver for the Catalyst 9.12 video card.

    Results.

    CineBench R10 is a benchmark based on image rendering. The entire computational load during the processing of the task goes to the processor. In this test, the winner is the Intel Core i5 660 due to Hyper-Threading, which this benchmark actively uses. The architecture itself also plays a small role, which is undeniably faster. nine0003

    As shabby as the Clarkdale memory controller is called, when compared to older models, it manages to beat the Core 2 Duo E8600 in the Everest app’s memory reading test as well.

    There were also no surprises with the recording. The Core i5 660 is slightly faster.

    CPU Test of the popular 3DMark 2006 graphics benchmark also gives laurels to the winner of the new product from Intel. It’s again about the notorious Hyper-Threading, support and optimization of which leaves no chance for the E8600. nine0003

    HAWK is a popular gaming benchmark that is well-optimized to work with a large number of cores. But although the Core i5 660 is slightly faster, its victory is not so convincing.

    FarCry2 changes the balance of power in the struggle for the laurels of the winner. A game that actively uses cache memory is more favorable to the E8600. Still, 6 MB of cache versus 4 MB for the i5 660 do their job.

    Crysis is a game that has captivated a large number of people, it is famous for its tough hardware requirements. Architecture still takes up, albeit with a very meager advantage. nine0003

    Extreme overclocking of Intel Core i5 660.

    Recently, the Internet is full of information about the successful overclocking of processors above 6GHz. I decided to check this information on my own by cooling the processor with liquid nitrogen.

    An important factor affecting overclocking is coldbug. Coldbug is the temperature of the processor, below which it stops working, and the system stops showing signs of life. So Clarkdale processors are famous for having Coldbug at very low temperatures, close to the boiling point of liquid nitrogen -196 degrees Celsius. But on one of the foreign forums dedicated to extreme overclocking, you can find information that such temperatures are subject only to Evga motherboards. One of the developers of the company shares a secret mod that allows you to drop to such low temperatures.

    In fact, all the above words turned out to be true. Motherboard Asus Maximus III Formula lost performance at -80 degrees Celsius. This temperature is not enough up to high frequencies. nine0003

    valid.canardpc.com/show_oc.php?id=907841

    Subdued frequency — 5528 MHz. The voltage set at the same time in bios was 1.7 V. Well, not the most successful overclocking attempt. Not only the instance of the processor should be blamed, but also the motherboard. Initially, I planned to overclock the processor with the Asus TurboV utility. But when changing any bios parameter, for example, the processor voltage, the system hung. Therefore, I had to limit myself to bios tools.

    We managed to pass the wPrime 32M multithread test at 5477 MHz. The result obtained is 8.736 sec.

    The wPrime 1024M had to drop another 100 MHz to pass the long test. Result 4 min 44.232 sec.

    Result 43767 3Dmarks — this is the result of 3DMark05. This test was passed at a processor frequency of 5477 MHz, the frequencies of the Radeon HD5970 video card were set to 970/1200 MHz, respectively for the core/memory.

    Conclusions. nine0014

    The impressions from the Intel Core i5 660 processor were very positive.