In 2018, AMD was able to achieve very noticeable success in the processor market. According to the company itself, the volume of shipments of chips for PCs in the fourth quarter of last year exceeded the same indicator a year ago by one and a half times, which turned out to be due to both the general increase in the popularity of Ryzen-branded products and the increase in the price of the average processor. A significant increase in buyers’ interest in Ryzen can be seen in everything: in the systematic growth of AMD’s share in the processor market; and in the intensification of guerrilla marketing by Ryzen fans online; and in reports from individual stores pointing to the emerging turnaround in the mood of PC buyers, who are increasingly abandoning the familiar Intel platform.
And there is nothing surprising here. AMD has been very lucky with market conditions. Intel made serious miscalculations in production planning, which ultimately led to a prolonged shortage and a rise in prices for inexpensive Core and Pentium processors. As a result, Ryzen turned out to be a much better value proposition in terms of price-performance ratio in the middle and lower price segments purely automatically — only due to the fact that AMD did not adjust its pricing policy in response to the problems of its only competitor.
Democratic prices have always been one of AMD’s trump cards. The company has always tried to make sure that its processors, at the same cost as Intel’s offerings, surpass them in some important characteristics, for example, in the number of computing cores and threads. But now the company’s position has further strengthened due to the fact that Intel products have become simply impossible to buy at a «fair» price set by the manufacturer.
And here another factor played into the hands of AMD. An important component of the marketing policy of this company is to offer the widest range of models. AMD is simply flooding the market with a variety of processors with slight differences in performance and slightly different prices. Previously, this allowed buyers to easily find the right option for themselves without having to make any financial compromises. But now, when Intel processors are often out of stock due to a shortage, AMD simply occupied the price lists, not giving a potential buyer the slightest opportunity to even think about choosing an alternative. A simple example: in the store of our partner, the Regard company, in the range from $6 to $100 for 17 different Ryzen modifications, only 6 current Intel processors for the LGA1151v2 platform are offered.
But the most interesting thing is that AMD does not stop there, and the variety of Ryzen processors continues to multiply further. Today we will get acquainted with two chips that have appeared on sale recently — Ryzen 5 2500X and Ryzen 3 2300X. This pair of processors expands the Pinnacle Ridge family downwards, adding cheaper offerings with four processing cores to its existing eight-core and six-core representatives. This means that the most recent variants of Ryzen with the Zen + microarchitecture, released on the 12nm process technology, are now available for under $100.
Formally, the Ryzen 5 2500X and Ryzen 3 2300X were released by AMD as solutions for OEM builders and should not be supplied to retailers. However, this, as usual, does not apply to the Russian market. Our feature of the «OEM-ness» of the quad-core Pinnacle Ridge is only that you can buy them only in the delivery variant without a box and a cooler. Which, however, is unlikely to stop anyone, especially since the Ryzen 5 2500X and Ryzen 3 2300X look very attractive, since they are superior in performance to any other AMD quad-core.
In April last year, AMD updated the design of its desktop processors — it was then that the second generation of Ryzen appeared, transferred to the 12nm process technology and received an improved Zen + microarchitecture. However, the changes affected only the eight-core and six-core representatives of the Ryzen family, while the spring upgrade did not touch the range of quad-cores. As a result, the benefits of Zen + were available exclusively to buyers of quite expensive processors, while in the middle price segment, users were forced to continue to focus either on the old Ryzen of the 2017 model, or on the Raven Ridge family of processors released a year ago, equipped with an integrated graphics core.
This situation lasted long enough, but last fall, AMD nevertheless announced quad-core CPU models of the 2000 series with the Zen + microarchitecture. The range has been replenished with two new products, Ryzen 5 2500X and Ryzen 3 2300X, which were obviously intended as a more modern replacement for the Ryzen 5 1500X and Ryzen 3 1300X. True, AMD, for some reason, did not want to completely abandon the models of the past generation in favor of new ones. Old quad-cores continue to be supplied to this day, and a rather limited niche is still allocated for new products: they are positioned as solutions for computer builders and are not offered in retail, in boxes and with proprietary cooling systems.
As a result, the lineup of desktop quad-core Ryzen processors has grown to eight representatives with quite different characteristics. For clarity, we have brought their specifications together.
|Design||Cores/Threads||Base frequency, GHz||Turbo frequency, GHz||L3 cache, MB||Memory||Graph. nucleus||PCI lanes||TDP, W|
|Ryzen 5 2500X||Pinnacle Ridge||4/8||3.6||4.0||eight||DDR4-2933||Not||24||65|
|Ryzen 5 2400G||Raven Ridge||4/8||3.6||3.9||four||DDR4-2933||Vega 11||sixteen||65|
|Ryzen 5 1500X||Summit Ridge||4/8||3.5||3.7||sixteen||DDR4-2666||Not||24||65|
|Ryzen 5 1400||Summit Ridge||4/8||3.2||3.4||eight||DDR4-2666||Not||24||65|
|Ryzen 3 2300X||Pinnacle Ridge||4/4||3.5||4.0||eight||DDR4-2933||Not||24||65|
|Ryzen 3 2200G||Raven Ridge||4/4||3.5||3.7||four||DDR4-2933||Vega 8||sixteen||65|
|Ryzen 3 1300X||Summit Ridge||4/4||3.5||3.7||eight||DDR4-2666||Not||24||65|
|Ryzen 3 1200||Summit Ridge||4/4||3.1||3.4||eight||DDR4-2666||Not||24||65|
It is clearly seen that the Ryzen 5 2500X and Ryzen 3 2300X stand out against the background of their brothers with a number of noticeable advantages that make them very interesting offers.
First, the Ryzen 5 2500X and Ryzen 3 2300X are indeed carriers of the newer 12nm Zen+ microarchitecture. This means that they are based on the Pinnacle Ridge design, which has better frequency potential and offers a slightly higher specific performance by lowering L2 and L3 cache latency. In addition, such processors have a memory controller formally compatible with DDR4-2933 SDRAM.
Secondly, although the Ryzen 5 2500X and Ryzen 3 2300X are based on exactly the same eight-core semiconductor crystals as in the older 2000-series processors, only one of the two CCX modules is active in them. And this is a huge advantage of the new products over the quad-core predecessors with the Summit Ridge design, where both CCXs remained operational, and the cores in them were blocked symmetrically. Now, thanks to changes in the structure, all active cores are collected in a single domain, and when transferring data or accessing the L3 cache, they do without using the Infinity Fabric bus, which often becomes a bottleneck in existing processors with the Zen / Zen + microarchitecture. As a result, the performance of new products has improved for this reason.
Third, the Ryzen 5 2500X and Ryzen 3 2300X are AMD’s highest clocked quad-core processors today. Their operating frequencies are higher than those of the quad-core Ryzen thousandth series by up to 300 MHz, and the frequencies in turbo mode reach 4 GHz, which quad-cores for Socket AM4 have not yet taken. True, in fairness it should be noted that the six- and eight-core Pinnacle Ridge processors use even higher frequencies, and in this regard, the Ryzen 5 2500X and Ryzen 3 2300X cannot be called champions.
However, the quad-core Ryzen 5 2500X and Ryzen 3 2300X still look very good. When compared to the same class of processors that AMD offered earlier, they win in almost everything. The only thing that can be seen as a deterioration in performance is the size of the L3 cache. Consisting of two CCX modules, the Ryzen 5 1500X offered 16 MB L3 cache, while the new Ryzen 5 2500X has half the L3 cache. However, this difference can be critical only in tasks that intensively interact with large amounts of data. In addition, do not forget that the monolithic L3 cache in the Ryzen 5 2500X is clearly faster than the two-part cache of the Ryzen 5 1500X.
⇡#Ryzen 5 2500X and Ryzen 3 2300X in action
Due to a couple of new products under consideration, AMD’s quad-core lineup has become additionally swollen, and potential buyers of inexpensive Socket AM4 processors are faced with a rather ambiguous choice. The new Ryzen 5 2500X is not yet very common on sale and costs about $10.00, that is, it is approximately equal in price to the Ryzen 5 2400G, but more expensive than the Ryzen 5 1500X, somewhere around $1500. A slightly different situation with prices for the second novelty, Ryzen 3 2300X. This CPU is already widely sold in the retail channel and is priced at $6.00, while the alternative Ryzen 3 2200G and Ryzen 3 1300X are sold for about $30 more.
Under these conditions, the Ryzen 5 2500X and Ryzen 3 2300X seem like a viable choice in cases where the processor does not need an integrated GPU, which the Pinnacle Ridge family does not have. The arguments in favor of new products are quite obvious, just look at how they work in real conditions.
For example, under a single-threaded load, the Ryzen 5 2500X will run at 3.95 GHz, while the Ryzen 5 2400G will run at 3.85 GHz in a similar situation, and the Ryzen 5 1500X at 3.875 GHz.
The spread in frequencies at full load on all cores becomes more noticeable. In this case, the Ryzen 5 2500X offers a 50MHz lead over the Ryzen 5 2400G, but a 200MHz lead over the Ryzen 5 1500X.
Naturally, this also affects performance. Even according to the benchmark built into the CPU-Z diagnostic utility, the new Ryzen 5 2500X offers 3-6% faster performance than older brothers in both single-threaded and multi-threaded tests. At the same time, you need to understand that this benchmark is purely synthetic in nature and does not use the L3 cache and memory subsystem. That is, in fact, the difference in performance here is due only to the difference in the operating frequency.
In real tasks, this will be supplemented by a small gain in IPC provided by the Zen + architecture, and a decrease in memory subsystem latencies, which is caused by the single-module structure of the Ryzen 5 2500X, which offloads the internal Infinity Fabric interface. The fact that all this is not an empty phrase is easy to verify by measuring the latency of the cache and memory. The graph below (taken on a DDR4-3200 system with 16-16-16-36 timings) shows that the Ryzen 5 2500X has the same L3 cache latency as the Ryzen 5 2400G, but the cache itself is larger. And the memory latency compared to the Ryzen 5 2400G and Ryzen 5 1500X is about 5-11% lower.
But the Ryzen 5 2500X does not offer noticeable advantages in power consumption and temperature conditions. For example, in the CPU-Z stress test, the maximum temperature of the Ryzen 5 2500X does not exceed 60 degrees (with a Noctua NH-U14S cooler), and the processor heat dissipation is within 67 watts. But the Ryzen 5 2400G and Ryzen 5 1500X chips produced using 14nm technology, at the same load, warm up to 62 and 53 degrees, respectively, consuming 63 and 64 watts.
The situation is similar with the Ryzen 3 2300X. In particular, with a single-threaded load, the new quad-core processor without SMT can operate at 4.0 GHz, and its predecessors, Ryzen 3 2200G and Ryzen 3 1300X, at 3.7 and 3.9 GHz, respectively.
As the load increases, the Ryzen 3 2300X noticeably increases its frequency advantage. With simultaneous operation of all cores, the frequency reaches 3.925 GHz, while the Ryzen 3 2200G and Ryzen 3 1300X are content with 3.6 GHz.
This is reflected in the CPU-Z benchmark. The single-threaded result of the Ryzen 3 2300X exceeds the performance of previous Ryzen 3 generations by 8-9%, and in the multi-threaded test, the new processor is 9-11% faster than its 14nm predecessors.
In terms of power consumption and heat dissipation, under load, the temperature of the Ryzen 3 2300X reached 57 degrees, and the power consumption measured by the sensor built into the processor reached 52 watts. These are slightly higher values when compared with the behavior of the Ryzen 3 2200G and Ryzen 3 1300X under similar conditions: processors of previous generations under load warmed up to 50-54 degrees, and their power consumption was 45-50 watts.
In addition, it is necessary to mention an important fact that concerns the support of Ryzen 5 2500X and Ryzen 3 2300X by the existing fleet of Socket AM4 motherboards. The fact is that manufacturers do not consider it necessary to include these processors in their compatibility lists. However, in reality there are no problems. The Ryzen 5 2500X and Ryzen 3 2300X are typical Pinnacle Ridge design carriers, so those motherboards that have a BIOS that supports 2nd generation Ryzen octa-core and six-core will work absolutely fine with the new quad-cores. In other words, the vast majority of motherboards are suitable for the Ryzen 5 2500X and Ryzen 3 2300X, including low-cost motherboards on low-end chipsets.
⇡ # Overclocking
There is another advantage that the Ryzen 5 2500X and Ryzen 3 2300X processors should have “by definition”. Thanks to the 12nm process technology, they should have a higher frequency potential available during overclocking than past quad-cores. All AMD processors still have a free multiplier, and the new items reviewed today are no exception. They accelerate in exactly the same way as other Ryzen — by increasing the multiplier. Fortunately, this does not require any special equipment, overclocking in the Socket AM4 platform is available even to owners of inexpensive motherboards based on the B350 and B450 chipsets.
However, do not place high expectations on the quad-core Pinnacle Ridge. Yes, such processors are indeed built on 12nm crystals, which are no different from those used, for example, in the Ryzen 7 2700X. But not everything is so simple: for the Ryzen 5 2500X and Ryzen 3 2300X, the manufacturer uses production rejection, that is, those semiconductor blanks that, for one reason or another, were not suitable for the older processors of the family. And this means that the overclocking potential of the new quad-cores may be worse than that of the older Ryzen 2000-series processors.
Practice has shown that this is approximately the case. While Pinnacle Ridge processors with eight cores often hit frequencies of the order of 4.2 GHz during overclocking, neither the Ryzen 5 2500X nor the Ryzen 3 2300X were able to overclock like that. The achievable maximum for both CPUs turned out to be 4.1 GHz, which, however, is still a hundred or two megahertz higher than that available when overclocking quad-core Summit Ridge.
The performance of the Ryzen 5 2500X at 4.1 GHz was achieved by increasing the supply voltage to 1.375 V and simultaneously enabling the Load-Line Calibration function.
And the simpler Ryzen 3 2300X without SMT support was able to take a frequency of 4.1 V at a slightly lower voltage — 1.35 V.
It’s funny that during stress testing of both processors with the described overclocking, they showed approximately the same practical power consumption at the level of 75-80 W and similar temperatures in the range of 75-78 degrees. Moreover, a similar performance at the level of 114-116 Gflops was also given by the Linpack stress test.
Separately, it is worth mentioning that when setting the limits of overclocking, we used a fairly productive and expensive Noctua NH-U14S air cooler. In real systems, where quad-core Ryzen have a chance to get, coolers will most likely be used simpler. This means that ordinary users can count on slightly weaker overclocking of the Ryzen 5 2500X and Ryzen 3 2300X, for example, to frequencies of the order of 4.0 GHz. But such a frequency is close to the nominal mode of these processors, so there is not so much sense in forced overclocking of the Ryzen 5 2500X and Ryzen 3 2300X — the resulting increase in performance will not be noticeable much.
⇡#Description of test systems and testing methods
In an ideal world, the quad-core processors of the Ryzen 5 family should be priced to act as an intermediate option between the Core i5 and Core i3, and Ryzen 3 was conceived as competitors for the lower versions of the Core i3 and older Pentium Gold. However, objective reality has little overlap with these attitudes. Due to the shortage, Intel processors have risen in price, and today mutual positioning looks completely different. Six-core Core i5s have become significantly more expensive than quad-core Ryzen 5s, which can now be considered in a completely different capacity — as a full-fledged alternative to quad-core Core i3 processors. The Ryzen 3 family is being sold at prices comparable to those of dual-core Pentium Gold.
With this in mind, we included a full set of quad-core Ryzen 5 and Ryzen 3 processors in testing; two six-core Ryzen 5 processors, one of which is a Ryzen 5 1600, comparable in price to the Ryzen 5 2500X; as well as Core i5-8400, Core i3-8300, Core i3-8100 and Pentium Gold G5500 — Intel processors, which are sold at more or less adequate prices.
Ultimately, the test systems used the following set of equipment:
- AMD Ryzen 5 2600X (Pinnacle Ridge, 6 cores + SMT, 3.6-4.2 GHz, 16MB L3);
- AMD Ryzen 5 1600 (Summit Ridge, 6 cores + SMT, 3.2-3.7 GHz, 16MB L3);
- AMD Ryzen 5 2500X (Pinnacle Ridge, 4 cores + SMT, 3.6-4.0 GHz, 8MB L3);
- AMD Ryzen 5 2400G (Raven Ridge, 4 cores + SMT, 3.6-3.9 GHz, 4 MB L3);
- AMD Ryzen 5 1500X (Summit Ridge, 4 cores + SMT, 3.5-3.9 GHz, 16MB L3);
- AMD Ryzen 3 2300X (Pinnacle Ridge, 4 cores, 3.5-4.0 GHz, 8MB L3);
- AMD Ryzen 3 2200G (Raven Ridge, 4 cores, 3.5-3.7GHz, 4MB L3);
- AMD Ryzen 3 1300X (Summit Ridge, 4 cores, 3.5-3.9GHz, 8MB L3);
- Intel Core i5-8400 (Coffee Lake, 6 cores, 2.8-4.0 GHz, 9 MB L3);
- Intel Core i3-8300 (Coffee Lake, 4 cores, 3.7 GHz, 8 MB L3);
- Intel Core i3-8100 (Coffee Lake, 4 cores, 3.6 GHz, 6 MB L3);
- Intel Pentium Gold G5500 (Coffee Lake, 2 cores + HT, 3.8 GHz, 4 MB L3).
- CPU cooler: Noctua NH-U14S.
- ASRock X470 Taichi (Socket AM4, AMD X470);
- ASRock Z390 Taichi (LGA1151v2, Intel Z390).
- Memory: 2×8 GB DDR4-3466 SDRAM, 16-16-16-36 (G.Skill Trident Z RGB F4-3466C16D-16GTZR).
- Video Card: NVIDIA GeForce RTX 2080 Ti (TU102, 1350/14000 MHz, 11 GB GDDR6 352-bit).
- Disk subsystem: Samsung 960 PRO 1TB (MZ-V6P1T0BW).
- Power Supply: Thermaltake Toughpower DPS G RGB 1000W Titanium (80 Plus Titanium, 1000W).
A separate explanation should be made regarding the memory modes that were activated in various test platforms. The fact is that the most popular LGA1151v2 motherboards based on the Intel H370, B360 and H310 chipsets do not allow overclocking the RAM. Therefore, the Core i5-8400 processor was tested with memory in DDR4-2666 mode 14-14-14-34, and the Core i3-8100 and Pentium Gold G5500 were tested with memory in DDR4-2400 mode 14-14-14-34. At the same time, Socket AM4 motherboards do not create any such problems, so all Ryzen were tested with DDR4-3200 16-16-16-36.
The Core i5-8400 was tested with the Multi-Core Enhancements option enabled, that is, with the abolition of any power consumption restrictions. We are aware that this mode is somewhat at odds with the specifications of Intel, but most users willy-nilly use processors in it. The problem is that all motherboard manufacturers, without exception, activate Multi-Core Enhancements by default, and this trend is not expected to change.
Testing was performed on the Microsoft Windows 10 Enterprise (v1803) Build 17137.1 operating system using the following set of drivers:
- AMD Chipset Driver 18.10;
- Intel Chipset Driver 10.1.1.45;
- Intel Management Engine Interface Driver 220.127.116.117;
- NVIDIA GeForce 417.71 Driver.
Description of the tools used to measure computing performance:
- BAPCo SYSmark 2018 — testing in Productivity scenarios (office work: spreadsheet processing, archiving and unzipping files, working with PDF and text documents, email, installing and uninstalling programs, creating presentations, optical recognition of a scanned document), Creativity (working on multimedia content — merging panoramas from several images, creating HDR photos, preparing images for printing, importing and exporting photos, face recognition in photos using AI algorithms, video transcoding, preparing videos for publication on the web), Responsiveness (launching «heavy» software packages, working in a browser with a large number of open tabs, installing and removing programs, switching between browser tabs and open applications, writing a set of documents to a folder).
- Futuremark 3DMark Professional Edition 2.4.4264 — testing in the Time Spy Extreme 1.0 scene.
- 7-zip 18.05 — archiving speed testing. The time taken by the archiver to compress a directory with various files with a total volume of 3.1 GB is measured. The LZMA2 algorithm and the maximum compression ratio are used.
- Adobe Photoshop CC 2018 19.1.3 — Graphics performance testing. Measured is the average execution time of a test script, which is a creatively redesigned Retouch Artists Photoshop Speed Test, which includes a typical processing of four 24-megapixel images taken by a digital camera.
- Adobe Photoshop Lightroom Classic CC 7.3 — performance testing for batch processing of a series of images in RAW format. The test scenario includes post-processing and export to JPEG at 1920 × 1080 resolution and maximum quality of two hundred 16-megapixel RAW images taken with a Fujifilm X-T1 digital camera.
- Adobe Premiere Pro CC 2018 12.1.0 — performance testing for non-linear video editing. It measures the rendering time to YouTube 4K format of a project containing HDV 2160p30 footage with various effects applied.
- Blender 2.79b — testing the speed of the final rendering in one of the popular free packages for creating three-dimensional graphics. The duration of building the final model from Blender Cycles Benchmark rev4 is measured.
- Stockfish 9 — testing the speed of the popular chess engine. The speed of enumeration of variants in the position «1q6/1r2k1p1/4pp1p/1P1b1P2/3Q4/7P/4B1P1/2R3K1 w» is measured.
- x264 r2851 — testing the speed of video transcoding to H.264/AVC format. To evaluate performance, we use the original 2160p@24FPS AVC video file with a bitrate of about 42 Mbps.
- Assassin’s Creed Odyssey. Resolution 1920 × 1080: Graphics Quality = Ultra High.
- Civilization VI. Resolution 1920×1080: DirectX 12, MSAA=4x, Performance Impact=Ultra, Memory Impact=Ultra.
- Far Cry 5. Resolution 1920 × 1080: Graphics Quality = Ultra, Anti-Aliasing = TAA, Motion Blur = On.
- Kingdom Come: Deliverance. Resolution 1920 × 1080: Overall Image Quality = Ultra High.
- Shadow of the Tomb Raider. Resolution 1920×1080: DirectX12, Preset=Highest, Anti-Aliasing=TAA.
- Total War: Warhammer II. Resolution 1920 × 1080: DirectX 12, Quality = Ultra.
- Watch Dogs 2. Resolution 1920 × 1080: Field of View = 70°, Pixel Density = 1.00, Graphics Quality = Ultra, Extra Details = 100%.
- World of Tanks. Resolution 1920 × 1080: Preset = Ultra Quality. Antialiasing = TSSAA HQ.
In all gaming tests, the results are the average number of frames per second, as well as the 0.01-quantile (first percentile) for FPS values. The use of the 0.01-quantile instead of the minimum FPS is due to the desire to clean up the results from random bursts of performance that were provoked by reasons not directly related to the operation of the main components of the platform.
⇡#Performance in complex tests
Usually, in complex tests like SYSmark 2018, Ryzen family processors do not show very high results. Working in common user scenarios that are built on common applications requires good single-threaded performance, low latency of the memory subsystem and high speed of the disk subsystem. In all these three areas, the AMD platform is inferior to Intel’s offerings, which led to the fundamental lag behind Ryzen from the representatives of the Core family.
However, in today’s testing, the situation turned out to be unexpectedly different. The Ryzen 5 2500X and Ryzen 3 2300X give a decent SYSmark 2018 score, comparable to the Core i3. The novelties have improved the results compared to their predecessors by 10 to 15%, which is explained not only by increased frequencies, but also by their single-module design with one active CCX instead of two.
The quad-core Pinnacle Ridge demonstrates the greatest performance gain against the background of Summit Ridge and Raven Ridge in the Productivity scenario, where we are talking about user interaction with office applications. However, in other scenarios, we have no particular complaints about the Ryzen 5 2500X and Ryzen 3 2300X: SYSmark 2018 clearly says that AMD has chosen the right design for them.
Ryzen 5 2500X and Ryzen 3 2300X also look good in the traditional 3DMark Time Spy Extreme gaming benchmark. The first processor from this pair turns out to be the fastest quad-core CPU of our time, ahead of both all its predecessors and competing members of the Core i3 family. However, the Ryzen 3 2300X processor, deprived of SMT support, does not look so confident against the background of its older brother, it noticeably loses to the Core i3-8100.
You also need to understand that 3DMark Time Spy Extreme is a synthetic benchmark heavily tuned for multithreading. And therefore, one should not be surprised that six-core processors in it show a huge lead over quad-core ones. For example, even the Ryzen 5 1600 boasts a 32 percent performance advantage over the Ryzen 5 2500X, which, of course, will not be observed in real gaming applications.
⇡#Performance in applications
Despite the fact that the quad-core Pinnacle Ridge does not carry anything new, but is built on long-used and well-tested technologies, their performance in applications is significantly higher than that of past AMD quad-cores. For example, the Ryzen 5 2500X outperforms the Ryzen 5 1500X by up to 10 percent, outperforming its predecessor only in 7-zip and Lightroom, where the L3 cache size is important.
But the truly impressive offering is the Ryzen 3 2300X, with a 9-17% lead over the Ryzen 3 1300X and a 12-27% lead over the Ryzen 3 2200G. As a result, in terms of performance, the Ryzen 3 2300X becomes a full alternative to the Core i3-8100 — on average, the performance of these quad-core processors is the same.
Unfortunately, the Ryzen 5 2500X cannot boast of such a successful confrontation with Intel alternatives. This processor is certainly faster than the older quad-core Core i3-8300, but its performance is significantly lower than the six-core Core i5-8400. And this means that it is still impossible to consider the Ryzen 5 2500X as a competitor to any Core i5 options: this processor is simply the best quad-core, but nothing more.
In addition, among AMD’s offerings is the Ryzen 5 1600, a six-core processor similar in price to the Ryzen 5 2500X. And if we are talking about which of AMD’s low-cost offerings is best for working with resource-intensive applications for processing and content creation, then the Ryzen 5 2500X will certainly not be a favorite here.
Work on the Internet:
⇡ # Performance in games
Before proceeding directly to the diagrams, it is worth once again commenting on why the gaming tests were carried out on a system with the highest performance graphics card. The fact is that this approach allows us to talk specifically about processor gaming performance, which is not limited by any side factors, in this case, insufficient graphics performance. In other words, we are considering an ideal (in terms of CPU evaluation) situation, while in real systems with weaker video cards, the gap between assemblies with different CPUs will be smaller in absolute value, but still remain qualitatively exactly the same.
As for the actual test results, they can be called somewhat unexpected. We’ve long been accustomed to the Ryzen family of processors being inferior in gaming performance to similarly priced offerings from Intel, but the new Ryzen 5 2500X and Ryzen 3 2300X break that rule. When creating the quad-core Pinnacle Ridge, AMD abandoned the design of the processor from two CCX modules, completely turning off one of them in a semiconductor chip. And it gave a very tangible effect. The Ryzen 5 2500X was on average 9% faster in games than the Ryzen 5 1500X, and the Ryzen 3 2300X is up to 20% faster than the Ryzen 3 1300X.
Thanks to this transformation, quad-core Ryzen can now be considered completely full-fledged alternatives to Intel’s quad-core Core i3 family. The Ryzen 3 2300X can be seen as a no-nonsense gaming replacement for the Core i3-8100, while the Ryzen 5 2500X can be seen as capable of delivering performance on par with the Core i3-8300. Given the shortage of Intel products, this is a very valuable quality of the new Ryzen, which will undoubtedly be appreciated by buyers of inexpensive gaming configurations.
However, it should be mentioned that at around $100, the new Ryzen 5 2500X may still not be the best processor for gaming in its price category. The fact is that the Ryzen 5 1600, thanks to its one and a half times superiority in the number of cores, can in some cases provide a higher frame rate, especially when it comes to fresh open-world games.
We are very familiar with the Pinnacle Ridge family of processors, so we know that AMD has directed all the advantages of the new 12nm process to increasing clock speeds, and not to improving energy efficiency. Therefore, you should not expect high efficiency from the Ryzen 5 2500X and Ryzen 3 2300X, after all, these CPUs have a standard 65-watt thermal package for quad-core processors.
Let’s illustrate this with measurements. The digital power supply of the Thermaltake Toughpower DPS G series, which we use in the test system, allows you to control the consumed and output electrical power, which we use for measurements. The graphs below show the total consumption of systems (without a monitor), measured «after» the power supply, which is the sum of the power consumption of all components involved in the system. The efficiency of the power supply itself in this case is not taken into account.
The Ryzen 5 2500X and Ryzen 3 2300X were not only faster, but also more power hungry than previous generations of quad-core processors. However, there is nothing criminal here: the power consumption of systems based on them at full processor load remains within acceptable limits. The only thing worth keeping in mind is that for quad-core Pinnacle Ridge you shouldn’t take a very simple cooler, especially considering the possibility of overclocking such processors.
It’s very unfortunate that AMD not only decided to do without the loud announcement of the Ryzen 5 2500X and Ryzen 3 2300X processors, but also released them exclusively as OEM products. After all, these are processors that everyone should know about: not only are they the best AMD quad-core processors to date, but in general they can be considered the most attractive offers in their price segment.
In the quad-core processors based on the Pinnacle Ridge design, AMD did a very important and right thing: it refused to use a pair of CCX modules in the design at the same time and offered quad-core processors that do not need to access the Infinity Fabric bus either during inter-core data exchange or when working with L3 cache. Together with a slight increase in clock speeds and Zen + microarchitectural improvements, this, without any exaggeration, put the performance of the reviewed new products one step higher than the quad-core processors of the Summit Ridge and Raven Ridge series.
As a result, the Ryzen 5 2500X was faster than the Ryzen 5 1500X and Ryzen 5 2400G by up to 10% and up to 15%, respectively. And the advantage of the Ryzen 3 2300X over the Ryzen 3 1300X and Ryzen 3 2200G is even more impressive — it reaches 15 and 20%, respectively. Moreover, the quad-core Pinnacle Ridge demonstrate the most significant increase in performance in games — exactly where the main complaints arose against AMD’s proposals.
All this finally made AMD’s quad-core processors full-fledged competitors to Intel’s quad-core processors, which it sells under the Core i3 trademark. In fact, in terms of performance, the Ryzen 3 2300X could become an analogue of the Core i3-8100, while the Ryzen 5 2500X can be considered an even better alternative to the Core i3-8300, since it offers 20-30% faster performance in multi-threaded resource-intensive tasks due to support for SMT technology.
Of course, the Ryzen 5 2500X seriously falls short of the performance level set by the six-core Core i5-8400 processor, but it absolutely does not need it, because today’s prices are such that the Ryzen 5 2500X plays in the same price category with the Core i3-8100, and Ryzen 3 2300X can generally be opposed to processors of the Pentium Gold family. And with this positioning, there can simply be no doubt about the overwhelming superiority of the Ryzen 5 2500X and Ryzen 3 2300X over competitor products.
AMD has always been able to come up with interesting low-cost offerings, but today’s situation is very different. Quad-core Ryzen Pinnacle Ridge series provide at least as good performance as Intel processors, which cost 40-50% more. It is clear that all this is primarily a consequence of the production problems of the microprocessor giant, but, be that as it may, there is no reason to buy a Core i3 or Pentium today, when Ryzen 5 2500X and Ryzen 3 2300X are next to them in the price list. meaning.