We overclock and undervolt the AMD Ryzen 7 8700G Hawk Point APU up to 5350 MHz with the ASUS ROG Strix X670E-I Gaming WiFi motherboard and EK custom loop water cooling.

https://www.youtube.com/watch?v=hYuLogj8ZBs

This APU is the successor to the 5700G which I overclocked in SkatterBencher #24, almost 3 years ago. APUs are always interesting because they’ve got reasonably performing integrated graphics. However, in this guideI’m only focused on overclocking the Ryzen 7 8700G CPU cores. I will check out the IGP overclocking in a next Skatterbencher guide.

Alright, let’s get going. I hope you enjoy the post!

AMD Ryzen 7 8700G: Introduction

The Ryzen 7 8700G is the flagship processor of AMD’s Zen 4-based Ryzen 8000 desktop APU product line codenamed “Hawk Point.” The Hawk Point processors were announced on January 8, 2024. Hawk Point is the successor to the Ryzen 5000 Cezanne APUs launched in 2021. There are more than a few differences between these chips. Here’s the shortlist:

• The CPU architecture moves from Zen 3 to Zen 4 and Zen 4c cores, depending on the SKU
• Hawk Point comes with an integrated XDNA AI accelerator
• The chips are produced using TSMC N4P process
• Hawk Point supports DDR5 (but not PCIe Gen 5)
• The integrated graphics is based on the RDNA 3 architecture.

In today’s guide I’m focusing specifically on the CPU side of this 8700G APU and I’ll cover overclocking the integrated graphics for another SkatterBencher guide.

A lot has been said about the difference between Zen 4 and Zen 4c cores. I won’t regurgitate everything here, but the point is that the core architecture is identical while Zen 4c cores have a smaller footprint. The benefit is that you get more compute per area, but the tradeoff is that you also have higher hotspot temperatures and lower frequencies.

The Ryzen 8000 series APUs come to market in with two distinct configurations. The Ryzen 7 8700G and 8600G feature the Phoenix1 die with Zen 4 cores and the Ryzen 5 8500G and 8300G feature the Phoenix2 die with a mix of Zen 4 and Zen 4c cores.

The Ryzen 7 8700G has eight Zen 4 cores with 16 threads. The base frequency is 4.2 GHz, and the listed maximum boost frequency is 5.1 GHz. The TDP is 65W, and the TjMax is 95 degrees Celsius.

In this guide, we will cover four different overclocking strategies:

• First, we enable Precision Boost Overdrive 2 and enable DOCP.
• Second, we tune the frequency using the Precision Boost Overdrive 2 toolkit.
• Third, we try out the ASUS AI Overclocking technology.
• Lastly, we try a manual overclock.

However, before we jump into overclocking, let us quickly review the hardware and benchmarks used in this guide.

AMD Ryzen 7 8700G: Platform Overview

The system we’re overclocking today consists of the following hardware.

Item	SKU	Price (USD)
CPU	AMD Ryzen 7 8700G	329
Motherboard	ASUS ROG Strix X670E-I Gaming WiFi	425
CPU Cooling	EK-Pro QDC Kit P360	750
Fan Controller	ElmorLabs EFC	20
Memory	G.SKILL Trident Z5 DDR5-6400	120
Power Supply	Antec HCP 1000W Platinum	250
Graphics Card	ASUS ROG Strix RTX 2080 TI	490
Storage	Kingston SSDNow V300 120GB SSD	50
Chassis	Open Benchtable V2	200
Other	ElmorLabs PMD-USB	60

ElmorLabs EFC

The ElmorLabs EFC is original Easy Fan Controller which formed the basis for its fancy brother, the (EFC-SB). I’ve used these fan controllers in many previous SkatterBencher guides.

I explained how I use the EFC devices in many previous SkatterBencher guides. I monitor the ambient temperature (EFC), water temperature (EFC), and fan duty cycle (EFC). I include the measurements in my stability test results.

I also use the device to map the radiator fan curve to the water temperature. Without going into too many details, I have attached an external temperature sensor from the water in the loop to the EFC. Then, I use the low/high setting to map the fan curve from 25 to 40 degrees water temperature. I use this configuration for all overclocking strategies. The main takeaway from this configuration is that it gives us a good indicator of whether the cooling solution is saturated.

AMD Ryzen 7 8700G: Benchmark Software

We use Windows 11 and the following benchmark applications to measure performance and ensure system stability.

BENCHMARK	LINK
SuperPI 4M	https://www.techpowerup.com/download/super-pi/
Geekbench 6	https://www.geekbench.com/
Cinebench 2024	https://www.maxon.net/en/cinebench/
CPU-Z	https://www.cpuid.com/softwares/cpu-z.html
V-Ray 5	https://www.chaosgroup.com/vray/benchmark
Corona Benchmark	https://corona-renderer.com/benchmark
AI-Benchmark	https://ai-benchmark.com/
3DMark Night Raid	https://www.3dmark.com/
3DMark CPU Profile	https://www.3dmark.com/
Returnal	https://store.steampowered.com/app/1649240/Returnal/
Shadow of the Tomb Raider	https://store.steampowered.com/app/750920/Shadow_of_the_Tomb_Raider_Definitive_Edition/
Final Fantasy XV	http://benchmark.finalfantasyxv.com/na/
OCCT	https://www.ocbase.com/

Returnal

I’m introducing a new game benchmark in this SkatterBencher guide: Returnal. I don’t have too much to share about the game itself, other than that it’s powered by the Unreal Engine and, most importantly, offers a very convenient in-game benchmark. I use the in-game benchmark for this guide with the visual settings set to Epic and DLSS enabled.

AI Benchmark

A major focal point of the Hawk Point processors is the integration of the AI engine. While I’m intrigued by the addition of new IPU of the Ryzen 7 8700G and would love to test it out, at the moment of writing, I didn’t find a suitable benchmark yet that could test the capabilities of the integrated inference processing unit.

It seems that support for the IPU will come in an updated version of AI Benchmark in the near future, but for this guide the AI Benchmark performance measurement is based on the CPU performance

I’ll cover IPU performance and overclocking in a future blog post.

AMD Ryzen 7 8700G: Stock Performance

Before starting overclocking, we must check the system performance at default settings. The default Precision Boost 2 parameters for the Ryzen 7 8700G are as follows:

Experienced Ryzen overclockers are familiar with most of these limiters, however two are not that common: STAPM and HTFMax. I cover the STAPM limiter in more detail in my first overclocking strategy.

The HTFMax is a frequency limiter that kicks in when the CPU is running at a high temperature. It is not active on all Ryzen processors. For example, we didn’t see this limiter on the Ryzen 9 7950X but it was present on the Ryzen 7 7700X.

On this 8700G, the HTFMax lowers the maximum frequency when the temperature exceeds 85 degrees Celsius from the default 5150 MHz. The frequency decreases gradually to 5050 MHz at 95 degrees Celsius.

Here is the benchmark performance at stock:

• SuperPI 4M: 43,613 seconds
• Geekbench 6 (single): 2,710 points
• Geekbench 6 (multi): 13,794 points
• Cinebench 2024 Single: 107 points
• Cinebench 2024 Multi: 978 points
• CPU-Z V17.01.64 Single: 700.1 points
• CPU-Z V17.01.64 Multi: 7,357.1 points
• V-Ray 5: 13,023 vsamples
• Corona 10: 5.49 MRays/s
• AI Benchmark: 3,746 points
• 3DMark Night Raid: 70,172 marks
• Returnal: 125 fps
• Tomb Raider: 171 fps
• Final Fantasy XV: 173.91 fps

Here are the 3DMark CPU Profile scores at stock

• CPU Profile 1 Thread: 1,038
• CPU Profile 2 Threads: 2,064
• CPU Profile 4 Threads: 4,045
• CPU Profile 8 Threads: 6,949
• CPU Profile 16 Threads: 8,212
• CPU Profile Max Threads: 8,203

When running the OCCT CPU AVX2 Stability Test, the average CPU core effective clock is 4063 MHz with 1.042 volts. The average CPU temperature is 57.7 degrees Celsius. The ambient and water temperatures are 26.3 and 30.0 degrees Celsius. The average CPU package power is 64.9 watts.

When running the OCCT CPU SSE Stability Test, the average CPU core effective clock is 4400 MHz with 1.136 volts. The average CPU temperature is 63.5 degrees Celsius. The ambient and water temperatures are 26.2 and 30.0 degrees Celsius. The average CPU package power is 69.5 watts.

Of course, we can increase the maximum power consumption limit using Precision Boost Overdrive. That’s what we’ll do in our first overclocking strategy.

However, before we get going, make sure to locate the CMOS Clear header. Shorting the Clear CMOS header will reset all your BIOS settings to default, which is helpful if you want to start your BIOS configuration from scratch. The Clear CMOS header is located on the ROG FPS-II card.

OC Strategy #1: PBO + D.O.C.P.

In our first overclocking strategy, we simply take advantage of enabling Precision Boost Overdrive 2 and AMD EXPO.

Precision Boost Overdrive 2

With the launch of Zen 3, AMD introduced an improved version of the Precision Boost Overdrive toolkit, allowing for manual tuning of the parameters affecting the Precision Boost frequency boost algorithm. Precision Boost Overdrive 2 builds on the PBO implementation of Zen 2. In addition to the overclocking knobs from Zen+ (PPT, TDC, EDC) and Zen 2 (Boost Override and Scalar), Precision Boost Overdrive 2 also introduced Curve Optimizer.

There are essentially 3 levels of Precision Boost Overdrive

1. AMD’s stock values, which can be set by disabling PBO
2. The motherboard vendor values, which are programmed into the BIOS to match the motherboard VRM specification and can be set by enabling PBO
3. Custom values, which can be programmed by the end-user

I most recently took a deep dive into the Precision Boost Overdrive 2 toolkit in my Ryzen 7000 launch content. If you want to learn more about the impact of each of these settings, I suggest you check out that guide.

In this overclocking strategy, we’re just enabling Precision Boost Overdrive, whereas, in the following strategies, we’ll explore tuning the parameters. By enabling Precision Boost Overdrive, we rely on the motherboard pre-programmed PBO parameters. We find that the following values have changed:

Increasing the PPT and, to a lesser extent, the TDC and EDC limit will help unleash the frequency in multi-threaded workloads previously limited by the EDC.

Precision Boost: STAPM Limit

A configuration option that’s not part of Precision Boost Overdrive but still affects the Precision Boost frequency boosting algoritm is STAPM. That’s an acronym for Skin Temperature Aware Power Management. I talked about it briefly in my Raphael Launch article when discussing how Precision Boost 2 works.

The STAPM limiter is primarily relevant for mobile devices. As the name implies, STAPM takes into account the device temperature for assessing the boost behavior. It represents the maximum power the processor can use within the thermal budget of the whole system. It is similar to the package power but also considers the system thermal capacitance headroom.

Since the Ryzen 7 8700G APU is a derivative of a notebook chip, it appears this STAPM limiter is also present. Usually we can disable STAPM in the BIOS, or the STAPM limit would change when we adjust any of the package power limits such as PPT or SPL. But in this case it was a little bit more complicated as none of the usual settings did the trick.

For this particular BIOS, the only way to adjust the STAPM limit was to adjust the Sustained Power Limit (SPL) in the SmartShift Control menu.

SmartShift is a set of AMD technologies that aim to optimize the power usage for mobile devices. Some of that technology enables dynamic power shifting between the CPU and GPU depending on the workload needs. But it also includes shifting to power-saving when the notebook is in battery mode.

I’m pretty confident that future BIOSes will adjust this value through the use of auto-rules when the user enables Precision Boost Overdrive. For this guide, however, I’ll just leave it in the BIOS configuration overview, just in case someone runs into this particular issue when overclocking.

EXPO – Extended Profiles for Overclocking

EXPO stands for AMD Extended Profiles for Overclocking. It is an AMD technology that enables ubiquitous memory overclocking profiles for AMD platforms supporting DDR5 memory. EXPO allows memory vendors such as G.SKILL to program higher performance settings onto the memory sticks. If the motherboard supports EXPO, you can enable higher performance with a single BIOS setting. So, it saves you lots of manual configuration.

BIOS Settings & Benchmark Results

Upon entering the BIOS

• Switch to the Advanced Mode view and stay in the AI Tweaker menu
• Set Ai Overclock Tuner to EXPO I
• Enter the Precision Boost Overdrive submenu
  • Set Precision Boost Overdrive to Enabled
• Switch to the Advanced menu
  • Enter the AMD CBS submenu
    • Enter the SMU Common Options submenu
      • Enter the SmartShift Control submenu
        • Set SmartShift Control to Manual
          • Set Sustained Power Limit to 1000000

Then save and exit the BIOS.

We re-ran the benchmarks and checked the performance increase compared to the default operation.

• SuperPI 4M: +7.56%
• Geekbench 6 (single): +3.47%
• Geekbench 6 (multi): +9.78%
• Cinebench R23 Single: +4.67%
• Cinebench R23 Multi: +11.35%
• CPU-Z V17.01.64 Single: +0.49%
• CPU-Z V17.01.64 Multi: +1.55%
• V-Ray 5: +6.27%
• Corona 10: +11.49%
• AI Benchmark: +21.30%
• 3DMark Night Raid: +4.71%
• Returnal: +0.80%
• Tomb Raider: +8.19%
• Final Fantasy XV: +5.80%

Here are the 3DMark CPU Profile scores:

• CPU Profile 1 Thread: +0.10%
• CPU Profile 2 Threads: +0.63%
• CPU Profile 4 Threads: +0.72%
• CPU Profile 8 Threads: +3.58%
• CPU Profile 16 Threads: +4.62%
• CPU Profile Max Threads: +4.72%

We get significant performance gains in all multi-threaded benchmark applications by enabling Precision Boost Overdrive. In addition, also lighter workloads see a small improvement from increasing the memory performance with EXPO. The highest performance increase is in AI Benchmark, with +21.30% over stock performance.

When running the OCCT CPU AVX2 Stability Test, the average CPU core effective clock is 4661 MHz with 1.263 volts. The average CPU temperature is 95.2 degrees Celsius. The ambient and water temperatures are 26.8 and 32.7 degrees Celsius. The average CPU package power is 137.8 watts.

When running the OCCT CPU SSE Stability Test, the average CPU core effective clock is 4773 MHz with 1.312 volts. The average CPU temperature is 95.2 degrees Celsius. The ambient and water temperatures are 26.6 and 32.0 degrees Celsius. The average CPU package power is 125.4 watts.

We use the Shamino Boost Curve and NOPBench to check the CPU’s boost behavior and per core maximum effective clock frequency.

The boost frequency at 1 active thread is about 5127 MHZ and trails off to 4829 MHz when all cores are active. That is remarkably close to the chip’s Fmax of 5150 MHz. Unlike with previously overclocked AMD CPUs, we don’t see a C-State Boost Limiter clipping the frequency when too many cores are active.

We also find each of the eight cores boosting to the Fmax of 5150 MHz. That’s remarkable, as typically AMD CPUs would have only a few cores that boost to the maximum frequency. It’s also promising as this may indicate lots of overclocking potential.

OC Strategy #2: PBO Tuned

In our second overclocking strategy, we tune the 8700G’s Precision Boost dynamic frequency technology using the Precision Boost Overdrive 2 toolkit.

PBO 2: Fmax Boost Override

Fused maximum frequency, or Fmax, is one of the Precision Boost infrastructure limiters constraining the CPU performance. The limiter determines the maximum allowed processor frequency across all CPU cores inside your CPU.

Boost Clock Override or Fmax Override is one of the overclocker tools available in the PBO 2 toolkit. It allows the user to override the arbitrary clock frequency limit between -1000 MHz and +200 MHz in steps of 25 MHz.

It’s important to note that the Fmax override only adjusts the upper ceiling of the frequency and doesn’t act as a frequency offset. Ultimately, the Precision Boost 2 algorithm still determines the actual operating frequency.

When we increase the Fmax boost limit by 200 MHz, the new Fmax is 5350 MHz.

PBO 2: Scalar

Scalar is a tool that allows the user to override the warranted silicon stress level, or FIT, to achieve higher frequency. You can adjust the maximum allowed FIT level to 10 times higher than the factory-fused limit. While the tool offers precise granularity, typically, you’ll find the available options to range between 1X and 10X in steps of 1 multiple.

The effect of increasing Scalar is that the Precision Boost algorithm should aggressively pursue higher voltages as it is less concerned with CPU lifespan. Usually the FIT is not the limiting factor for achieving maximum performance, however for this particular CPU it helped me get slightly higher frequencies.

PBO 2: Curve Optimizer

Curve Optimizer is an important new feature of Precision Boost Overdrive 2.

Curve Optimizer allows end-users to adjust the factory-fused VFT or voltage-frequency-temperature curve for each CPU core separately. The VFT curve is a unique curve for each core inside your CPU that defines the required voltage for a given frequency at a given temperature. Higher frequencies or higher operating temperatures require higher voltage.

Curve Optimizer adjusts the VFT curve by offsetting the voltages of the factory-fused VFT curve. By setting a positive offset, you increase the voltage point. Conversely, you decrease the voltage point by setting a negative offset.

You can offset the entire curve by up to 50 steps in a positive or negative direction. Each step represents approximately 5mV.

The traditional overclocking approach for AMD Ryzen CPUs is to set a negative curve optimizer. Two things happen when you adjust the VFT curve with a negative point offset.

1. First, we effectively undervolt the CPU and tell it needs less voltage for a given frequency. And, as a consequence, at a given voltage, it can apply a higher frequency.
2. Second, the CPU temperature will be lower because you use less voltage at a given frequency. That extra thermal headroom will also encourage the Precision Boost algorithm to target higher voltages and frequencies.

In my AMD Raphael overclocking launch guide, I explored the ins and outs of Curve Optimizing with Precision Boost Overdrive 2. If you want to learn more about the impact of Curve Optimizer, I suggest you check out that guide.

Like in the past, per-core tuning offers a real benefit as it provides some cores with a lot more frequency headroom.

AMD Ryzen 7 8700G Curve Optimizer Tuning Process

The manual tuning process for Curve Optimizer can become quite convoluted since it affects the CPU core voltage in all scenarios ranging from very light single-threaded workloads to heavy all-core workloads.

Usually, I spend a lot of time on per-core curve optimization. For this guide, I wanted to rely on the OCCT toolkit to find the right Curve Optimizer settings. Here’s my broad approach.

1. I started with a negative curve optimizer of -15 for each core.
2. Then, I use OCCT CPU Stability test with small data set, extreme mode, and steady load type. I cycle through each core every 10 seconds with 2 operating threads, starting first with AVX-512, then AVX2, then SSE. I do this for no more three times for each core
   1. I increased the negative curve optimizer magnitude by 5 steps for the cores that didn’t crash.
   1. I reduced the negative curve optimizer magnitude by 5 steps for the cores that did crash.
3. Once all cores were dialed in, I ran another OCCT AVX2 and SSE stability test with the goal to pass for 30 minutes without any core failing or clock stretching. Again, when a core fails, I reduce the negative curve optimizer magnitude by 5 steps.

Once all that’s done, I run through the usual benchmark suite and stability tests used in this guide. If there’s no instabilities I consider my Curve Optimizer settings as stable.

BIOS Settings & Benchmark Results

Upon entering the BIOS

• Switch to the Advanced Mode view and stay in the AI Tweaker menu
• Set Ai Overclock Tuner to EXPO I
• Enter the Precision Boost Overdrive submenu
  • Set Precision Boost Overdrive to Enabled
  • Set Precision Boost Overdrive Scalar to Enabled
    • Set Customized Precision Boost Overdrive Scalar to 10X
  • Set CPU Boost Clock Override to Enabled (Positive)
    • Set Max CPU Boost Clock Override(+) to 200
  • Enter the Curve Optimizer submenu
    • Set Curve Optimizer to Per Core
    • Set Core 0 to Core 7 Curve Optimizer Sign to Negative
    • I set the Curve Optimizer for each core according to my test result. The curve Optimizer results range from -50 to -25
• Switch to the Advanced menu
  • Enter the AMD CBS submenu
    • Enter the SMU Common Options submenu
      • Enter the SmartShift Control submenu
        • Set SmartShift Control to Manual
          • Set Sustained Power Limit to 1000000

Then save and exit the BIOS.

We re-ran the benchmarks and checked the performance increase compared to the default operation.

• SuperPI 4M: +12.77%
• Geekbench 6 (single): +7.01%
• Geekbench 6 (multi): +12.20%
• Cinebench R23 Single: +7.48%
• Cinebench R23 Multi: +16.67%
• CPU-Z V17.01.64 Single: +3.41%
• CPU-Z V17.01.64 Multi: +7.35%
• V-Ray 5: +11.29%
• Corona 10: +16.33%
• AI Benchmark: +24.93%
• 3DMark Night Raid: +6.50%
• Returnal: +1.60%
• Tomb Raider: +9.36%
• Final Fantasy XV: +6.24%

Here are the 3DMark CPU Profile scores:

• CPU Profile 1 Thread: +2.79%
• CPU Profile 2 Threads: +2.42%
• CPU Profile 4 Threads: +3.11%
• CPU Profile 8 Threads: +8.49%
• CPU Profile 16 Threads: +9.16%
• CPU Profile Max Threads: +9.35%

After the impressive Curve Optimizer undervolt, I expected to see great improvements in performance. While the improvements aren’t hitting the heights I was hoping for, we do see a substantial performance uplift across the board. The highest performance improvement is in AI Benchmark with an uplift of +24.93%

When running the OCCT CPU AVX2 Stability Test, the average CPU core effective clock is 4887 MHz with 1.237 volts. The average CPU temperature is 95.2 degrees Celsius. The ambient and water temperatures are 27.6 and 33.5 degrees Celsius. The average CPU package power is 137.2 watts.

When running the OCCT CPU SSE Stability Test, the average CPU core effective clock is 5053 MHz with 1.288 volts. The average CPU temperature is 94.4 degrees Celsius. The ambient and water temperatures are 27.5 and 32.9 degrees Celsius. The average CPU package power is 125.8 watts.

We use the Shamino Boost Curve and NOPBench to check the 8700G CPU’s boost behavior and per core maximum effective clock frequency.

The boost frequency at 1 active thread is about 5272 MHZ and trails off to 5128 MHz when all cores are active. That’s pretty impressive! Remember, the out of the box advertised boost frequency is 5.1 GHz and we’re now getting slightly higher than then when all cores are active!

Unfortunately, we also find that despite the aggressive undervolt with Curve Optimizer, we don’t get any core to hit the new Fmax of 5350 MHz. That’s a little surprising given all cores could hit the standard Fmax of 5.15 GHz.

OC Strategy #3: ASUS AI Overclock

We use the Asus AI Overclocking feature integrated into the ASUS ROG BIOS in our third overclocking strategy.

ASUS AI Overclocking

For many years, board vendors have tried to implement automatic overclocking features in their BIOS for more straightforward performance enhancement. This has always been a mixed bag, as most preset OC profiles are overly optimistic in frequency target or overly generous with the voltage selection. So often, you would end up with a slightly unstable or overheating system.

ASUS AI overclocking uses a unique strategy for automatic overclocking. Instead of working with preset frequency and voltage profiles, the system will monitor the CPU and cooling system throughout an initial testing phase. Based on its findings, it will then predict the optimal settings. The system automatically guides the overclocking process and adjusts voltages and frequency to match your cooling system.

The better your cooling, the higher your AI overclock.

There are three steps to enabling AI overclocking. First, reset the BIOS to default settings. Then, reboot and enter the operating system. Run heavy workloads, such as Prime95 or Cinebench, for 10 to 30 minutes. Then, return to the BIOS and enable AI Overclock when ready.

In addition to automatic overclocking, AI Overclocking provides advanced information and suggestions in the AI Features menu. The information includes:

• P0 VID and SP values for each CPU core
• Precision Boost Overdrive 2 suggested overclocking parameters
• Dynamic OC Switcher suggested overclocking parameters

The SP value is based on the maximum boost frequency, temperature, and P0 VID combination. Generally, it indicates the quality of a particular core. A higher SP value would indicate a better-quality core with superior overclocking capabilities, though it’s not an exact science. The overclocking suggestions are based on a continued evaluation of your CPU thermal solution.

After enabling AI Overclock, the following settings have changed:

• CPU Core Ratio: AI Optimized
• Precision Boost Overdrive: Enabled
  • Max CPU Boost Clock Override: +100MHz
  • Scalar: 8X
  • Curve Optimizer: 0
• Dynamic OC Switcher: Enabled
  • CCX Ratio: 51.00X
  • Core VID: 1.241V
  • Current Threshold: 55A

As you can see, AI Overclock provides a conservative +100 MHz increase of the Precision Boost maximum frequency. It also boosts the 8700G CPU cores to 5100 MHz in all-core workloads by enabling Dynamic OC Switcher.

BIOS Settings & Benchmark Results

Upon entering the BIOS

• Switch to the Advanced Mode view and stay in the AI Tweaker menu
• Set Ai Overclock Tuner to EXPO I
• Set CPU Core Ratio to AI Optimized

Then save and exit the BIOS.

We re-ran the benchmarks and checked the performance increase compared to the default operation.

• SuperPI 4M: +13.05%
• Geekbench 6 (single): +3.32%
• Geekbench 6 (multi): +7.42%
• Cinebench R23 Single: +5.61%
• Cinebench R23 Multi: +16.78%
• CPU-Z V17.01.64 Single: +0.49%
• CPU-Z V17.01.64 Multi: +0.65%
• V-Ray 5: +10.64%
• Corona 10: +15.66%
• AI Benchmark: +17.65%
• 3DMark Night Raid: +5.85%
• Returnal: +0.80%
• Tomb Raider: +8.77%
• Final Fantasy XV: +5.50%

Here are the 3DMark CPU Profile scores:

• CPU Profile 1 Thread: +1.54%
• CPU Profile 2 Threads: +0.48%
• CPU Profile 4 Threads: +0.96%
• CPU Profile 8 Threads: +1.09%
• CPU Profile 16 Threads: +3.58%
• CPU Profile Max Threads: +3.67%

We get a pretty decent performance lift over stock by enabling AI Overclock but it’s not as much as in our previous overclocking strategy. That’s likely due to AI Overclock not using Curve Optimizer. Still, we get a performance uplift of +17.65% in AI Benchmark.

When running the OCCT CPU AVX2 Stability Test, unfortunately, the CPU is not stable. AI Overclock targets the typical use-case and stability requirements, not a worst-case scenario like an extreme all-core AVX2 workload.

When running the OCCT CPU SSE Stability Test, the average CPU core effective clock is 5100 MHz with 1.216 volts. The average CPU temperature is 82.4 degrees Celsius. The ambient and water temperatures are 26.7 and 31.6 degrees Celsius. The average CPU package power is 114.6 watts.

OC Strategy #4: Manual Overclocking

In our fourth overclocking strategy, I will try manual overclocking. Since the AI Overclock didn’t pass our worst-case workload of OCCT with AVX2 enabled, I decided it would be interesting to find out at what frequency the 8700G would pass this worst-case stress test.

But first, let’s have a look at the Hawk Point frequency and voltage topology.

CPU Core Clocking Topology

The clocking of Hawk Point is similar to other AMD Zen 4 CPUs.

The standard Hawk Point platform has a 48 MHz crystal input to the integrated CGPLL clock generator. The CGPLL then generates a 48 MHz clock for the USB PLL and a 100 MHz reference clock for the FCH, which contains the CCLK PLL for the CPU cores and several SOC PLLs.

The CCLK PLL 100MHz reference clock drives the 200 MHz VCO, which is then multiplied by an FID and divided by a DID. As a whole, this provides a CPU clock frequency granularity of 25 MHz.

Each CCX has its own PLL, with the cores within that CCX running at the same frequency. In a typical operation, all cores within a CCX will run at the same frequency. Since each CCX contains all eight Zen 4 cores, all cores run the same clock. The effective clock will differ if the core is in a different P-state.

The SOC PLLs include a wide range of PLLs. The ones most relevant for overclocking are:

• FCLK for the data fabric
• UCLK for the memory controller
• MCLK for the system memory
• GFXCLK for the integrated graphics
• IPUCLK for the inference processing unit

The SOC PLLs are not particularly relevant for manual CPU Core overclocking.

Just like Raphael, Hawk Point supports up to two external clock modes in addition to the internal clock mode. In eCLK0 Mode, an external 100MHz reference clock is used for both the CPU and SOC PLLs. In eCLK1 Mode, there are two distinct external 100MHz reference clocks. One clock provides the 100MHz input for the CPU PLL, and another provides the 100MHz reference clock for the SOC PLLs.

CPU Core Voltage Topology

From the voltage topology perspective, there are a few minor changes. Like Raphael, the processor still relies on an internal and external power supply to generate the processor voltages.

There are four primary power supplies from the motherboard VRM to the processor: VDDCR, VDDCR_SOC, VDDCR_MISC, and VDDIO_MEM_S3.

The VDDCR voltage rail provides the external power for two internal voltage regulators: VDDCR_CPU and VDDCR_VDDM.

VDDCR_CPU provides the voltage for the CPU cores in CCX. The voltage rails can work in either regular or bypass mode, but it is always in bypass mode on Raphael. That means the voltage is always equal to the VDDCR external voltage.

The VDDCR voltage rail can be directly controlled via the SVI3 interface via the CPU configuration. The end user can also directly control the VDDCR voltage rail as the motherboard BIOS provides the means to program the voltage regulator directly.

Choosing the proper manual voltage is always a matter of finding a suitable trade-off between three factors: increased overclocking potential, the thermal challenges caused by increased voltage, and of course, consideration of CPU lifespan.

AMD Ryzen 7 8700G CCX Frequency Tuning Process

On many-core processors like the Ryzen Threadripper, per-CCX tuning can be a very tedious task as there are many CCXs to tune independently. However, on this Ryzen 7 8700G we only have one. So the tuning process can go much faster.

Our limiting factor will ultimately be the cooling solution as power consumption increases exponentially with operating voltage and temperature scales (somewhat) linearly with power consumption.

The maximum voltage will be determined by the application we’re tuning for. So, the first step in our tuning process would be to set a fixed CPU ratio and check the maximum temperature when running our workload. If there’s thermal headroom left, increase the operating voltage.

Once we know the maximum voltage, we can tune the CCX. Simply increase the ratio until the application shows signs of instability, then back off. No more, no less. In our case, we can push the 8700G to nearly 5 GHz.

Now that we know the ins and outs of Ryzen 7 8700G manual overclocking, let’s jump into the BIOS.

BIOS Settings & Benchmark Results

Upon entering the BIOS

• Switch to the Advanced Mode view and stay in the AI Tweaker menu
• Set Ai Overclock Tuner to EXPO I
• Enter the CPU Core Ratio (Per CCX) submenu
  • Set Core VID 0 to 1.250
  • Set CCD0 CCX0 Ratio to 49.50

Then save and exit the BIOS.

We re-ran the benchmarks and checked the performance increase compared to the default operation.

• SuperPI 4M: -3.19%
• Geekbench 6 (single): +1.96%
• Geekbench 6 (multi): +11.56%
• Cinebench R23 Single: +1.87%
• Cinebench R23 Multi: +14.31%
• CPU-Z V17.01.64 Single: +0.01%
• CPU-Z V17.01.64 Multi: +4.30%
• V-Ray 5: +8.68%
• Corona 10: +14.71%
• AI Benchmark: +21.30%
• 3DMark Night Raid: +5.78%
• Returnal: +0.80%
• Tomb Raider: +8.77%
• Final Fantasy XV: +4.68%

Here are the 3DMark CPU Profile scores:

• CPU Profile 1 Thread: -2.70%
• CPU Profile 2 Threads: -1.89%
• CPU Profile 4 Threads: -1.16%
• CPU Profile 8 Threads: +3.30%
• CPU Profile 16 Threads: +7.00%
• CPU Profile Max Threads: +6.88%

As expected, we see a performance deficit against stock performance in single-threaded applications as we lose benefit of the Precision Boost 1T frequency when we manually overclock. In multi-threaded applications, however, we see a nice performance uplift up to +21.30% in AI Benchmark.

When running the OCCT CPU AVX2 Stability Test, the average CPU core effective clock is 4950 MHz with 1.219 volts. The average CPU temperature is 91.4 degrees Celsius. The ambient and water temperatures are 26.1 and 31.5 degrees Celsius. The average CPU package power is 140.0 watts.

When running the OCCT CPU SSE Stability Test, the average CPU core effective clock is 4950 MHz with 1.224 volts. The average CPU temperature is 82.5 degrees Celsius. The ambient and water temperatures are 26.7 and 31.9 degrees Celsius. The average CPU package power is 114.5 watts.

AMD Ryzen 7 8700G: Conclusion

Alright, let us wrap this up.

I always find APUs interesting to overclock. Not just because they have reasonably powerful integrated graphics, but also because the chip design is different from mainstream desktop CPUs. The core architecture is monolithic instead of chiplet based.

The Ryzen 7 8700G has regular Zen 4 cores and the overclocking experience is very similar to Ryzen 7000 CPUs. I was surprised that every core could hit the Fmax out of the box and that made me hopeful for the overclocking potential. However, even after Curve Optimizing very aggressively, we couldn’t get any of the eight Ryzen 7 8700G core to hit the elevated Fmax of 5350 MHz. That was surprising.

The main tools for unlocking more performance on the Ryzen 7 8700G are undoubtedly unleashing the power limiters and enabling higher speed memory with EXPO.

Anyway, that’s all for today! I also plan to check out the Phoenix2 based APU in the future, so stay tuned if you’re interested. I want to thank my Patreon supporters for supporting my work. If you have any questions or comments, please drop them in the comment section below. 

See you next time!