We have AMD’s 4700S Desktop Kit, a motherboard that almost certainly houses a defective version of the chip typically found in the Sony PS5, in for testing. This kit definitely isn’t a typical AMD product: The eight-core 16-thread Zen 2 ‘AMD 4700S’ chip has a 3.6 GHz base and 4.0 GHz boost, but it comes directly mounted onto a mini-ITX motherboard along with either 8 or 16GB of memory (ours had 16GB) and a cooler, while the chip’s custom integrated GPU with 36 RDNA2-based compute units (CUs) is disabled. That means the chip’s ability to compete with the best CPUs for gaming, and its position on our CPU Benchmark hierarchy, rests on its ability to push a discrete GPU to high levels of performance. Today we’ll put the chip through its paces in a full spate of tests, including gaming and applications.
How did such an odd product come to market? At last count, Sony has sold more than 10 million PS5 consoles, each with its own special custom ‘Oberon‘ chip designed by AMD. These custom chips come with eight Zen 2 cores and a powerful custom RDNA graphics engine, but an untold number of chips suffer from defects during the manufacturing processes, meaning that they won’t function correctly (if at all), which typically results in a trip to the trash bin. Sometimes the chips simply can’t meet certain clock speed criteria. Regardless of the issue with these chips, they can’t be used in a console, but AMD appears to have found a way to sell the defective silicon by creating a system board with most of the key components you need to craft small systems.
|Arch.||Price||Cores/ Threads||Base/ Boost Freq.||TDP||GPU Cores||GPU Freq. (MHz)||L3 (MB)|
|AMD 4700S||Zen 2||$400 w/ board and memory (est.)||8 / 16||3.6 / 4.0||75W||N/A||N/A||8|
|Ryzen 7 5700G||Zen 3||$359||8 / 16||3.8 / 4.6||65W||RX Vega 8||2000||16|
|Ryzen 5 5600G||Zen 3||$259||6 / 12||3.9 / 4.4||65W||RX Vega 7||1900||16|
|Ryzen 7 4750G||Zen 2||~$310||8 / 16||3.6 / 4.4||65W||RX Vega 8||2100||8|
AMD certainly isn’t known for selling nearly-complete systems, but it’s necessary for the 4700S Desktop Kit because AMD fused the custom 75W AMD 4700S chip and eight GDDR6 memory packages to the motherboard with BGA mounting.
GDDR6 may seem to be an odd addition — we typically see this type of memory used for GPUs — but the 4700S uses GDDR6 as main memory instead of the standard DRAM we’re accustomed to. This type of memory provides higher bandwidth than standard DRAM, but that comes at the cost of higher latency. While GPUs crave this higher bandwidth and aren’t as sensitive to latency, GDDR6 does have performance implications in latency-sensitive workloads that run on CPUs, like gaming. In addition, the chip only exposes a single PCIe 2.0 x4 interface for the user to connect PCIe devices, ultimately restricting GPU performance. It also has extremely limited I/O connectivity with only two SATA ports for storage devices.
Almost all the details about AMD’s 4700S kit for desktop PCs had come via product listings and leaks before AMD even listed the chip on its site, and the company launched the AMD 4700S without any of its normal communication with the press via briefings or press releases. AMD also didn’t release the specifications for the chip, like clock speeds, cache, and TDPs, until after we requested them several months after the systems began shipping.
AMD has declined to comment on whether or not the AMD 4700S chip consists of defective PS5 silicon, but the identical chip packaging and completely unique system requirements (like GDDR6) make it clear. AMD has also conspicuously avoided using Ryzen branding.
AMD hasn’t engaged in the usual level of support for the chip with independent software vendors (ISVs), so most common monitoring utilities can’t give accurate monitoring information, including for some basic parameters like clock speeds. AMD’s own Ryzen Master doesn’t support the chip, either. All of this combines to make this system, in many respects, a black box.
|Processor||AMD 4700S 8-Core Processor “Zen 2” architecture|
|Memory||8GB or 16GB GDDR6|
|PCIe Slots||One PCIe 2.0 x4|
|FCH||FCH AMD A77E Fusion Controller Hub|
|LAN||Asix AX88179 Gigabit Ethernet controller|
|Audio codec||Realtek ALC897|
|Internal connections||2x SATA 6Gb/s connectors — 1x USB3.2 Gen1 5Gbps connector — 1x Front panel audio connector|
|External I/O||4x USB2.0 — 3x USB3.2 Gen2 — 1x LAN — 1x USB3.2 Gen1 3x audio jacks|
What do we know? We know that AMD positions the Ryzen 4700S Desktop Kit as a “high-performance productivity” solution for mainstream, home office, small business, and enterprise uses. That means it’s suited for tasks like productivity, multi-tasking, and light 3D workloads. AMD sells the kit either to system integrators (SI), who then create full systems with the unit (over 80 designs are coming to market), or as a boxed item. Systems based on this kit were first available in China, but both systems and standalone kits have now filtered out to Europe. However, AMD hasn’t said if systems will be available in the US.
You’ll notice that AMD specifically doesn’t position the 4700S Desktop Kit for gaming systems, and you’ll soon see why. First, let’s take a trip through the system we bought from a Chinese retail shop, and then run through the benchmarks. If you’re looking for the benchmarks, you can skip ahead to the gaming benchmarks section.
AMD Ryzen 4700S Desktop Kit
We purchased a full system built with the 4700S Desktop Kit through a Chinese retailer, so our MaiBenBen PC47 ITX system only represents a product from a single vendor. We paid roughly $700 USD for the full system, but you can find the standalone 4700S Desktop Kit for roughly $450 USD in Europe.
The PC47’s shell is made of a sturdily-thick aluminum and has a multitude of slots for airflow, though only a single 80mm fan exhausts air from the rear — there isn’t an intake fan. We didn’t spend much time testing the system in this enclosure, as it isn’t intended to be sold outside of China. Besides, we’re more interested in the 4700S.
Inside we found a tightly-packed chassis with a PowerColor Red Dragon Radeon 550 GPU, 256GB ‘Alhua’ C8000 Series SATA SSD, and a 200W HuntKey SFX12V power supply (HK300-41FP). Because the system isn’t intended for sale in the US (we used a courier service), we disassembled the system before the first power-on.
As you can see in the last few images in the album above, the inward-facing side of the power supply was damaged from an overly long screw that should hold it snug to the mounting bracket. Instead, the screw collided with the power supply’s interior casing, causing it to bend. This damage was concealed from view, so we’re lucky to have caught it. Needless to say, we didn’t use the power supply for testing.
AMD recommends a minimum 250W power supply for the 4700S Desktop Kit but suggests 300W or greater. As such, this 200W power supply isn’t fit for use. Instead, we pulled the AMD 4700S Desktop kit out of the chassis for testing and used our go-to EVGA 1600W T2 power supply. This is complete overkill, of course, but entirely removes the influence of power restrictions from our testing.
Here we can see the actual Mini-ITX motherboard that comprises the 4700S Desktop Kit. Connectivity is extremely limited, and that starts with the single PCIe 2.0 x4 connection that rides the side of the motherboard. Even though it accepts x16 cards, the x4 electrical connection provides a theoretical maximum of 2 GB/s of throughput, which will restrict performance with higher-end graphics cards. For example, the PCIe 3.0 x16 connection found on most motherboards tops out at 15.75 GB/s, while PCIe 4.0 x16 delivers 31.5 GB/s. Below, we’ll show you how that looks in a throughput test with an RTX 3090.
The motherboard also has only two SATA 6GBps ports for storage devices, and you can’t use the PCIe slot for an M.2 SSD carrier card — you’re forced to populate the PCIe slot with a discrete GPU because the integrated graphics are disabled. As a result, your storage options are extremely limited.
Of course, you could use an external storage device across one of the three USB-A 3.1 (10 Gb/s), one USB-A 3.0 (5 Gb/s), or four USB-A 2.0 (480 Mb/s) ports on the rear I/O panel, but you’ll have to compromise on throughput.
The rear of the board also has a Gigabit LAN port (Asix AX88179 controller) and three audio jacks (Realtek ALC897 codec). Interior connectivity options consist of one USB 3.2 Gen 1 connector, though you’ll also find two fan headers (one for the CPU, one for a system fan).
Here we can see the AMD 4700S eight-core chip with the cooler stripped off the top. Unfortunately, the chip doesn’t have any identifying markings that we’re accustomed to seeing (it almost looks as if they’ve been sanded off of the very center of the die). However, we do see a 100-000000466 product number on the silver casing around the chip.
The system BIOS is extremely limited — there are very few configurable parameters. That includes no controls for the axial fan on the cheap aluminum heatsink, it’s clearly not among the best CPU coolers, and the 80mm fan is an obnoxiously loud affair that jumps to full speed during even the lightest of workloads.
The stock cooler uses a custom mounting bracket and hole alignment, so it isn’t compatible with off-the-shelf cooling solutions like we see with the AM4 ecosystem, leaving no clear way to step up to a better cooler. Instead, your best option would be to swap out the fan for a quieter model. AMD uses a standard TIM with the cooler, but even swapping to higher-quality variants didn’t improve performance. Make no mistake, this fan and cooler are unsuitable — even judging by Intel stock cooler standards. We’ll take a look at how that impacts temperatures below.
The DRAM slots are conspicuously absent, but we find their GDDR6 replacements on the bottom of the motherboard underneath a large, thick metal shroud that dissipates heat from the modules via thermal pads. Eight 1.75 GHz 16Gb SK Hynix H56CBM24MIR-S2C chips, for a total of 16GB, operate at 1.35V. We have throughput and latency testing below.
The board’s ‘Bolton-E4’ A77E Fusion Controller Hub (FHC), which connects to the chip via a PCIe 2.0 x4 interface, resides under the small heatsink adjacent to the processor. You can download the chipset, LAN, audio driver, and BIOS from AMD’s website.
AMD 4700S Desktop Kit Test Setup, Boost Clocks, Thermals, Power
This round of testing comes with a lot of caveats: We aren’t entirely sure that the clock frequencies and temperatures we’ve collected are accurate. In fact, common utilities like AIDA and HWinfo didn’t expose much useful information when we first received the system. As seen in one of the screenshots below, AIDA initially identified the system as the AMD “K17.4 PS5/XBox Scarlett APU.”
We followed up with FinalWire (Aida) and REALiX (HWinfo) and submitted debug reports to help them correctly enumerate the various sensors and telemetry data, but with only limited success. Both vendors released updates that improved the situation, but some of the parameters still can’t be logged, and some logged variables could be subject to inaccuracies. We followed up with multiple other ISVs, and it doesn’t appear that AMD has shared the technical information that enables monitoring that it normally does, despite these chips being on the market for several months.
AMD has confirmed to techy’s points that third-party monitoring tools, and its own Ryzen Master, won’t work correctly with the system. We’re not sure when, or if, that will be resolved.
As a result, the clock frequencies reported by the chip are still inaccurate (we often see spikes to 12 GHz), so we’re forced to resort to plotting the effective clock rates instead of the values reported by the chip. Our numbers should be close to the correct clock speeds, but there is room for some inaccuracy.
AMD says that it is possible for the 4700S to exceed 100C for brief periods, but only as the chip throttles to return to sub-100C operation. However, we’ve recorded extended periods (several minutes) of 103C+, with peaks at 104.5C. AMD also says the sensors don’t use an offset. The extended 100C+ periods do raise some questions about either the accuracy of the temperature output or the efficacy of the chip’s internal throttling mechanisms. Considering the frequency throttling we’ll show in the tests below, it could be a bit of both.
As such, take the frequency and temperature measurements with some salt. The system also doesn’t report accurate power consumption telemetry, so we’ve turned to our in-line PSU tester to measure power consumption at the 8-pin EPS12V and the 24-pin motherboard connectors.
Typically the EPS12V connector supplies the lion’s share of power to a desktop PC processor, while a small amount (a few watts) will feed through the 24-pin connector. That means we can simply intercept power at the EPS12V input for measurements.
However, our 4700S measurements reveal that less power comes across the 8-pin than ‘normal,’ suggesting that the power is distributed in a non-standard way. That makes accurate physical power measurements for just the chip impossible (at least with our equipment). As such, we include the full board power (8-pin + 24-pin), including the chipset and GDDR6 power draw, in the measurements below.
Physical power measurements can involve a bit of fuzzy math, as VRM inefficiencies can create deltas between the power delivered to the VRMs and the power fed to the processor. These deltas vary based on the components in each motherboard’s power delivery subsystem (typically ~10% to ~15%). Unfortunately, that means that we can’t compare our 4700S power measurements to the other processors in our library of results, as the latter is for chip power only and isn’t influenced by VRM inefficiencies.
The 4700S BIOS is spartan with only one configurable parameter that can impact performance — you can only enable/disable Cool and Quiet (we disabled the feature for maximum performance). The BIOS also doesn’t allow clock rate adjustments, making it impossible to test IPC with a locked frequency. Combined with the fact that we can’t log accurate clock rates, it’s hard to generate precise IPC measurements at this time. Testing from @aschilling suggests that IPC is largely identical to other Zen 2 chips, with a notable distinction that, as a result of customizations, the chip has two FPU ports instead of four. That impacts performance in floating point and vectorized operations, some of which could be offset through the increased throughput from GDDR6. We’ll see the impact of that design decision in our application test suite.
Well, that’s a lot of caveats. Here’s what we came up with.
As per our normal routine, we put AMD’s boost clocks to the test in both single- and multi-threaded workloads (methodology here). The lightly-threaded test regimen is designed to extract the highest boost clock rates possible as we step through ten iterations of the LAME encoder, then single-threaded POV-Ray and Cinebench runs, PCMark 10, and GeekBench.
To keep the charts ‘clean,’ we only plot the maximum frequency recorded on any one core during the test to determine the peak clock rates. Here we can see that the chip boosts to 4.0 GHz during lightly-threaded work.
We also ran our standard thermal and power testing with intense multi-threaded applications, like AVX-powered Blender and HandBrake distributions, and plotted both minimum and maximum frequencies.
This series of tests runs the Corona ray-tracing benchmark, several HandBrake runs, POV-Ray, Cinebench R20, and four different Blender renders. All of these tests make an appearance in our standard test suite and represent real-world usages.
We recorded a peak of 104.5C during the tests, with extended periods of 103C that resulted in clock rates of ~3.4 GHz and lower. Notably, these regressions and periods of overheating occurred during AVX-heavy workloads, with the multi-threaded y-cruncher test resulting in sub-2.5 GHz clock rates.
The overheating and throttling persisted even after we applied our own thermal paste (MX-4) and remounted the cooler. Power consumption peaked at 98W, but we recorded an average of 64W during the extended tests. Remember, we measured the total board power and VRM inefficiencies also have an impact on these readings.
AMD 4700S Cache and Memory Latency Benchmarks
|AIDA 64 Latency Measurements||AMD 4700S – GDDR6||Ryzen 7 4750G – DDR4|
The 4700S Desktop Kit is unique because it uses GDDR6 for main memory, just like consoles, but it doesn’t have to share the memory throughput with a greedy graphics engine. The general idea with graphics memory (GDDR) is to gain higher bandwidth and lower power by sacrificing latency, and our test results indicate that holds true for the 4700S CPU.
Latency measurements vary based upon the type of access pattern and test depth used, and here we can see that AIDA, which measures to a 64MB depth, measures GDDR6 latency at 145ns. That’s roughly twice the Ryzen 7 4750G that measures at 74ns with stock DDR4-3200 memory.
Using a memory latency tool created by Lamchester (from chipsandcheese.com) that measures to a 1GB depth, we recorded peak latencies of 246ns for the 4700S with GDDR6 memory, and 123ns for the Ryzen 7 4750G with stock DDR4-3200.
This higher latency will penalize latency-sensitive applications, like games.
We’ve already seen the impact of trading higher latency for additional bandwidth, but the eight GDDR6 memory packages that feed the CPU across an aggregate 256-bit pipe do provide a tremendous amount of throughput.
The AMD 4700S with GDDR6 memory achieves 92,892 MB/s of copy bandwidth, which is roughly twice the throughput of the 4750G with DDR4. Gains are more muted with strictly read or write workloads, but the bandwidth is still impressive. This increased throughput will benefit bandwidth-hungry applications and vectorized workloads.
AMD 4700S Desktop Kit Gaming Benchmarks
Below you can see the geometric mean of our gaming tests with a discrete GPU at 1080p and 1440p, with each resolution split into its own chart to give us a decent overall view of the current landscape. As usual, we’re testing with an Nvidia GeForce RTX 3090 to reduce GPU-imposed bottlenecks as much as possible, and differences between test subjects will shrink with lesser cards or higher resolutions. These are cumulative metrics, so individual results vary on a per-title basis, which you’ll find further below.
It’s notable that the 4700S doesn’t have the RTX 3090 on its support matrix, but we used the card to allow us to use our library of results for comparative information. It also help us define the maximum level of performance you can wring from this kit. AMD doesn’t position the 4700S Desktop Kit as a gaming machine, though, and keep in mind that it has a restricted list of supported GPUs:
- AMD Radeon 550, RX 550, RX 560, RX 570, RX 580, RX 590
- Nvidia GeForce GT 1030, GTX 1050, GTX 1050 Ti, GTX 1060
To be clear, AMD doesn’t position the 47000S Desktop Kit for gaming systems, and the geometric mean of our test results shows why. Bear in mind that these deltas won’t be as pronounced with a GPU from the recommended list as you’ll be more likely to encounter GPU limitations, but they are telling.
Even though the Zen 2-powered 4700S has the same number of cores and threads as the first-gen Ryzen 7 1800X with the Zen microarchitecture, the 1800X is 50% faster at 1080p and 36.5% faster at 1440p.
The quad-core Ryzen 5 3400G with the Zen+ architecture is even slower than the 1800X, yet is still 27% faster than the 4700S at 1080p and 33% faster at QHD.
Perhaps the most telling result comes from the lowly quad-core, Intel Core i3-10100, which is 71% faster than the 4700S at 1080p and 53% faster at QHD.
|AMD 4700S + RTX 3090 (PCIe 2.0 x4)||1.33 GB/s|
|AMD 4700S + Radeon 550 (PCIe 2.0 x4)||0.96 GB/s|
|Ryzen 7 4750G + RTX 3090 (PCIe 3.0)||13.43 GB/s|
|Ryzen 7 4750G + Radeon 550 (PCIe 3.0)||3.87 GB/s|
|Ryzen 7 3700X + Radeon 500 (PCIe 4.0)||25.5 GB/s|
|Ryzen 7 3700X + Radeon 500 (PCIe 4.0)||4.12 GB/s|
The impact of the higher GDDR6 memory latency will result in reduced gaming performance, but the dismal throughput of the 4700S Desktop Kit’s PCIe 2.0 x4 connection probably hides the impact — it appears to be the more severe bottleneck.
Above, we can see a comparison between the Zen 2 Ryzen 7 4750G and the AMD 4700S chips, both with either the GTX 3090 or AMD Radeon RX 550 attached, in the 3DMark PCIe feature test. The benchmark measures the amount of throughput available to your GPU by uploading a tremendous amount of vertex and texture data to the GPU for each frame, thus making the PCIe connection the limiting factor for performance.
Restricted PCIe bandwidth really isn’t an issue for any GPU connected to a PCIe 3.0 or 4.0 interface (provided it has sufficient lanes), but stepping back to the PCIe 2.0 x4 interface clearly exposes a bottleneck. Here we can see that the 4700S’ maximum attainable bandwidth weighs in at 1.33 GB/s with the GTX 3090, but dropping the same GPU into an X570 motherboard results in ~10X more throughput (13.43 GB/s) for a PCIe 3.0 connection and 19X more with PCIe 4.0 (25.5 GB/s). Even the Radeon 550 benefits from the faster interface. You get roughly four times the throughput with PCIe 3.0 and 4.0, easily beating the 4700S’ paltry 0.96 GB/s.
We’re going to skip the game-by-game commentary for the titles below because the results are redundant — the 4700S lands in last place in all real-world gaming benchmarks. However, that comes with the caveat that we’re testing it with a much higher-end GPU than you’d pair with this class of system. However, even most lower-end GPUs will suffer some performance impact from the low PCIe bandwidth and high GDDR6 latency, and the faster GPUs on the recommended list will suffer even more. You also shouldn’t expect much headroom, if any, for future GPU upgrades.
You will see at least some level of competitive performance in the synthetic benchmarks in the first album of results below, but those results don’t translate to real-world gaming benchmarks.
3D Mark, VRMark, Stockfish Chess Engine on AMD 4700S Desktop Kit
Borderlands 3 on AMD 4700S Desktop Kit
Far Cry 5 on AMD 4700S Desktop Kit
Hitman 2 on AMD 4700S Desktop Kit
Project CARS 3 on AMD 4700S Desktop Kit
Red Dead Redemption 2 on AMD Ryzen 4700S Desktop Kit
Shadow of the techyb Raider on AMD Ryzen 4700S Desktop Kit
AMD Ryzen 4700S Desktop Kit Application Benchmarks, the TLDR
The charts above provide the geometric mean of several of our application tests (listed in the chart title), representing broader trends in lightly- and multi-threaded applications. Be sure to check our application tests below for performance in specific applications.
The AMD 4700S is more competitive in this round of tests, but only during multi-threaded applications where the insufficient cooler probably restricts performance in the heavier workloads (we’ve already recorded chip throttling during heavy work).
The AMD 4700S is 8% faster than the Ryzen 7 1800X in threaded work, but the eight-core Ryzen 7 2700X with the Zen+ architecture is 4% faster than the 4700S. We can likely chalk that up to the 2700X’s higher 3.7/4.3 GHz base/boost clocks and 16MB of L3 cache (twice the 4700S). The modern Ryzen eight-core chips, like the Ryzen 7 3800X and 5800X, are vastly superior to the 4700S, though.
Flipping over to the cumulative measure of single-threaded performance finds the 4700S near the bottom of the chart. Surprisingly, the Ryzen 7 1800X provides the same level of performance in single-threaded work as the AMD 4700S, while all the other Zen-based chips lead to varying amounts.
Overall the AMD 4700S is decent in threaded work, even though the lackluster cooler probably holds it back. Still, the disappointing tradeoff of vastly lower single-threaded performance, likely due to higher GDDR6 latency, ruins the value proposition.
Rendering Benchmarks on AMD Ryzen 4700S Desktop Kit
The AMD 4700S falls to the first-gen Ryzen 7 1800X in quite a few of these rendering tests, like Corona, C-Ray, V-ray, and the bmw27 Blender render, setting the tone for lackluster performance in these types of applications. Given that there are more full-featured, newer and cheaper chips available that don’t require massive tradeoffs, like the Ryzen 5 5600G, the AMD 4700S fails to impress.
We suspect that throttling could have an impact on more than a few of these benchmarks due to the wholly insufficient cooler, but that’s hard to rectify given its odd mounting spacing, highlighting the drawback of buying a one-off product that includes several unique integrated system components.
Encoding Benchmarks on AMD Ryzen 4700S Desktop Kit
The lightly-threaded LAME and FLAC encoders expose the 4700S Desktop Kit’s critical weakness — single-threaded performance. The AMD 4700S is thoroughly unimpressive here.
The 4700S is more competitive in the HandBrake tests where it can leverage the improved throughput of the GDDR6 memory in AVX workloads.
Web Browser, Office and Productivity on AMD Ryzen 4700S Desktop Kit
AMD promotes the 4700S for light office work and productivity, but its laggardly performance in single-threaded work makes that a tough sell for most types of latency-sensitive applications, like web browsers and Microsoft Office.
You’ll also notice the relatively poor showing in the PCMark 10 application start-up benchmarks, highlighting the sluggish experience that was noticeable by our own subjective measure when we used the system for simple tasks.
Compilation, Compression, AVX Performance on AMD Ryzen 4700S Desktop Kit
The GeekBench results find the 4700S again showing its somewhat bipolar nature — it’s passable in threaded work, but lags in single-threaded tasks.
The AMD 4700S Desktop Kit probably wouldn’t make much sense anywhere except in this silicon-starved world where every chip sells for a premium, but it’s still a dicey value proposition even during these dark days of chip shortages.
Below, we have the geometric mean of our gaming test suite at 1080p and 1440p and a cumulative measure of performance in single- and multi-threaded applications. Bear in mind that we conducted the gaming tests with an RTX 3090, so performance deltas will shrink with lesser cards and higher resolution and fidelity settings.
AMD doesn’t position the 4700S for gaming, and that’s likely because it doesn’t have either the PCIe bandwidth or low-latency memory needed for a gaming PC. The combination of the PCIe 2.0 x4 interface to the GPU and the high-latency GDDDR6 memory for the CPU certainly isn’t a recipe for gaming success.
The 4700S also only officially supports up to either an AMD Radeon RX 590 or Nvidia GeForce GTX 1060, so your upgrade options are limited. That doesn’t really matter, though: Our tests show that the 4700S doesn’t have the heft to push higher-end GPUs as hard as even the first-gen Ryzen 7 1800X. Bear in mind that the limitations we noticed won’t be as pronounced with GPUs from the recommended list as you’ll more likely encounter a GPU bottleneck first, but there isn’t a sufficient amount of CPU horsepower for upgrades.
AMD says the 4700S Desktop Kit is for mainstream, home office, small business, and enterprise use and that it works well for productivity, multi-tasking, and light 3D workloads, but there are better alternatives for those types of workloads. The 4700S is serviceable for some types of heavily-threaded applications, but the included cooler and fan combo resulted in chip throttling during the heaviest workloads. You could replace the fan with a better model, but we’re not sure that would offset the poor quality of the aluminum heatsink. We’re not aware of any aftermarket coolers with the same spacing and mounting system, so upgrading the cooler isn’t an option.
You’ll also have to overlook the sluggish performance in single-threaded workloads. The AMD 4700S provides roughly the same level of single-threaded performance as the 1800X in many workloads, which is a rather extreme sacrifice compared to what you could simply get by going with a newer processor.
The 4700S Desktop Kit comes with its own unique integrated components, like the CPU, memory and cooler, but that prevents you from upgrading those components. The 4700S Desktop Kit’s lackluster connectivity is also an issue: A single PCIe 2.0 x4 connection is insufficient in a modern machine, as it will restrict the performance of the graphics card that you absolutely must place in the slot.
The single PCIe slot also means you cannot install an M.2 SSD or other PCIe peripherals. Additionally, yhe system has only two SATA ports, so your storage options are extremely limited. We found the fan to be egregiously loud, but you could swap it out for a better model to reduce the noise level.
Overall the AMD 4700S Desktop Kit requires too many extreme tradeoffs, so it won’t find any success in the enthusiast realm. You have little to no upgrade path for the most important components, including the CPU, memory, GPU, cooler, and storage. Additionally, you’ll have to accept sluggish performance in lightly-threaded work to gain access to the one positive aspect of the machine, which is the serviceable performance in some threaded applications.
AMD’s system integrator partners plan to bring 80 systems based on the kit to market, and from the ones we’ve seen, they’re overpriced considering they are the most basic of low-end machines. That relatively high pricing largely stems from the bundled GPUs that are unnaturally expensive given the current graphics card shortage. There’s no doubt the 4700S would make a whole lot more sense if it came with even the most basic of integrated GPUs. If you’re looking for a pre-built system, there are better options that don’t require as many severe tradeoffs.
You could also buy the 4700S as a standalone kit, but that isn’t an attractive option, either. Enthusiasts looking for a more solid building block for a basic system should turn to modern chips like the Ryzen 5 5600G, or look to previous-gen chips that can often be found at reduced pricing. Given the 4700S Desktop Kit’s lacking connectivity options and bipolar performance trends, previous-gen (even second-hand) hardware will often provide better overall performance and a more rounded feature set.
MORE: Best CPUs for Gaming
MORE: CPU Benchmarks Hierarchy
MORE: All CPUs Content
|AMD Socket AM4||Ryzen 47000S|
|AMD Ryzen 47000S Desktop Kit|
|8 x2GB Sk Hynix GDDR6, 256-bit (16×16)|
|AMD Socket AM4||Ryzen 7 5700G, Ryzen 5 5600G, Ryzen 3 5300G, 4750G, 3400G|
|ASUS ROG Strix B550-E|
|2x 8GB Trident Z Royal DDR4-3600 @ 3200, Kingston DDR4-3200|
|Intel Socket 1200 (Z590)||Core i5-11400, Core i3-10100|
|ASUS Maximus XIII Hero|
|2x 8GB Trident Z Royal DDR4-3600 – 10th-Gen: Stock: DDR4-2933, OC: DDR4-4000, 11th-Gen varies, outlined above (Gear 1)|
|AMD Socket AM4 (X570)||AMD Ryzen 7 5800X, Ryzen 5 5600X, Ryzen 7 3800X|
|MSI MEG X570 Godlike|
|2x 8GB Trident Z Royal DDR4-3600 – Stock: DDR4-3200, OC: DDR4-4000, DDR4-3600|
|All Systems||Gigabyte GeForce RTX 3090 Eagle – Gaming and ProViz applications|
|Nvidia GeForce RTX 2080 Ti FE – Application tests|
|2TB Intel DC4510 SSD|
|EVGA Supernova 1600 T2, 1600W|
|Windows 10 Pro version 2004 (build 19041.450)|
|Cooling||Corsair H115i, Custom loop|