Ryzen 9000 can boost to higher clock than the official spec says

When AMD announced the Ryzen 9000 desktop CPUs with Zen 5 architecture at Computex 2024 back in early June, the company already revealed full specifications including the clock speeds. The boost clock speeds given for the highest model turned out to be the same as with Zen 4 (5.7 GHz for the Ryzen 7950X), but that wasn’t quite the final answer. As in previous generations, the maximum clock speed is actually higher than the official number.

Although it is already clear that officially, Zen 5 will have the same maximum clock speed as Zen 4, it was not yet clear if that will be true in reality, whether the unofficial extra clock speed will not increase or decrease. The answer may now have been found, though it doesn’t have to be definitive yet.

Images from the testing of a 16-core Ryzen 9000 processor have appeared on the AnandTech forum, provided by an unknown source and posted by another user (nicknamed igor_kavinky). The person with access to the processor provided testing results from Blender, where the 16-core shows around 20% performance increase compared to what can be measured on the Ryzen 9 7950X. However, this is not a production processor in the final configuration, but an ES sample after tuning using Curve Optimizer. So these numbers might not be representative of how the final processor will perform.

Real (unofficial) maximum clock speed of Ryzen 9 9950X

However, the HWiNFO images provided by the person in question show something else as well – the maximum clock speeds detected by HWiNFO (“Frequency Limit – Global”). And these apparently can indicate just what the real maximum clock speed of the processor is (on its preferred cores, not necessarily on all cores). It should be a value of 5850 MHz, which would mean that in high performance models like the Ryzen 9 9950X, Zen 5 will again unofficially have 150 MHz extra compared to the official maximum boost.

This extra clock speed may or may not be utilized, for example it might require the temperature of the silicon to be low. The Thermal Velocity Boost from Intel is supposed to be something similar (although boards often ignore its own temperature threshold). So you can’t count on the processor always running at 5.85 GHz in a single-threaded application.

This 150 MHz extra would mean that the additional unofficial boost clock range – perhaps it could be called XFR, which was how extra boost was called in the first Ryzen CPU generation – is the same as for the Ryzen 9 5950X (officially 4.9 GHz, unofficially 5050 MHz) and Ryzen 9 7950X (officially 5.7 GHz, unofficially up to 5850 MHz). So in the end, it looks like Zen 5 will have the same official maximum clock speeds as well as the unofficial ones. This will make benchmarks comparing the last-gen and en 5 top models interesting as it should thus show purely how the architecture has improved (although the ability of the CPU to cool itself could also be a factor).

Official Parameters of AMD Ryzen 9000 Processors – Computex 2024 Presentation (Author: AMD, via Anandtech)

But it’s not that simple and for now, this information is more of a curiosity, not necessarily the ultimate answer. As it was already mentioned, the user posting the images from the testing of the processor in question does not possess a regular shipping unit of Ryzen 9 9950X, but an engineering sample (which also reports itself as AMD Engineering Sample, but its identification code is obscured). And at the same time, Precision Boost Overdrive and Curve Optimizer are applied to the processor. Thus, the maximum clock speeds reported in the HWiNFO might not match the specification.

ES processors are generally not supposed to have the same clock speeds as the final chips, especially the first batch of samples tends to run much slower. However, later generations of ES processors can also simulate final clock speeds, although this is not as common. On the other hand, the user himself states that he uses Curve Optimizer and PBO. The former overclocking function should not affect the maximum boost and neither should PBO (Precision Boost Overdrive). However, the AMD platform offers another function called Precision Boost Override that can be used to manually increase the maximum clock speed. Unlike the basic PBO, the user does not declare he is using that, but this adds some uncertainty, as with the factor that an ES processor is used.

Thus, it is not yet definitively clear whether 5850 MHz is really the maximum clock speed at which Zen 5 will run straight from the factory. However, it probably indicates that the reports about there being some hope of a clock speed increase with Zen 5 turned out to not be true. Some time ago there were rumours that the real unofficial Zen 5 maximum could be somewhere slightly above 6000 MHz, but this seems to have been unsubstantiated, in the current light.

CPUs are already on their way to stores

Apart from this news, final preparations for the release of the processors also seem to be underway these days. According to the latest reports, samples are already being received by editors of hardware testing media and various YouTube “influencers” (although perhaps only the biggest and most important names are getting sampled at this stage). But the first distributors and vendors have also already started to confirm the arrival of the processors. So hopefully nothing should change about the fact that Ryzen 9000s will start selling at the end of July (or at worst a few days later).

Sources: AnandTech (forum), igor’sLAB, VideoCardz

English translation and edit by Jozef Dudáš


Contents

AMD has started mass production of Zen 6. Only for Epycs for now?

It’s been nearly two years since AMD’s Zen 5 architecture launched (in late July and early August 2024) and time is ripe for a successor—especially given how far competitors such as Qualcomm’s Snapdragon X2 Elite and Apple’s M5 have advanced in the meantime. AMD’s answer in the form of the Zen 6 architecture is now finally getting closer—the company has announced the ramp up of production, which could signal a launch later this year. Read more “AMD has started mass production of Zen 6. Only for Epycs for now?” »

Advanced Shader Delivery on AMD: No more waiting for compilation

Gamers definitely know the experience: You want to launch a game, but it starts to compile shaders, which can take several minutes. The reason is that shaders have to be built from code specifically for your hardware. And the compilation must be repeated after every GPU driver update, so you can end up seeing this waiting screen quite frequently. But Windows is now getting an improvement that should largely eliminate this waiting. Read more “Advanced Shader Delivery on AMD: No more waiting for compilation” »

FSR 4.1 AI upscaling finally coming to older Radeon GPU users

When AMD launched the FSR4 AI-based upscaling technology for games last year, it was exclusively available for the new Radeon RX 9000 generation GPUs using the RDNA 4 architecture. This was despite the fact that a version using INT8 compute compatible with older GPUs had leaked out, apparently by accident. But owners of older Radeon cards are finally in luck now: FSR 4.1 is coming to RDNA 2 and RDNA 3 as well. Read more “FSR 4.1 AI upscaling finally coming to older Radeon GPU users” »

Leave a Reply

Your email address will not be published. Required fields are marked *