AMD has banned the use of Power Play Table (PPTable) in the Windows registry on its Radeon RX 7000 graphics cards.This limits the overclocking capabilities of the graphics cards and, in particular, prevents fine-tuning GPU power and frequency curves outside of their official specifications.This has been pointed out by well-known enthusiast Buildzoid.Image source: AMDThe Power Play tables are an alternative method of changing GPU clock frequency, power and voltage via the Windows registry instead of modifying or replacing the BIOS.At the same time, it is the safest method of running GPUs outside of their official specifications, as it prevents any risk that may arise from BIOS re-flashing.Setting up Power Play tables or changing the BIOS has several advantages over traditional AMD overclocking methods and features offered by Radeon Software Adrenalin.First, users can bypass GPU hardware limitations associated with the firmware and overclock the graphics card as much as possible.Secondly, Power Play tables allow for fine-tuning of the voltage/frequency curve to improve the power efficiency of the graphics card.In other words, this method allows you to fine tune the video card's GPU behavior in more detail, which is not possible with AMD's usual overclocking software.Last generation Radeon RX 6000 video cards do not have the ability to fine tune power via Power Play tables blocked.Why AMD decided to prohibit access to it on Radeon RX 7000 video cards is unknown.However, one AMD engineer on the Reddit forum in a thread discussing power consumption issues with Radeon RX 7000 graphics cards reported that \"architecturally RDNA3 has some changes in how to manipulate the power tables.\"
The Overclocked Gaming Systems (OGS) enthusiast team overclocked the Galax GeForce RTX 4090 HOF graphics card GPU to an impressive 3705 MHz clock rate.
Of course, we're talking about overclocking with extreme liquid nitrogen cooling.As a result, the graphics card set new performance records in the tests 3DMark Port Royal and GPUPI v3.3 (32B).
The fact that the company Galax prepares to release gpu GeForce RTX 4090 HOF, reported the other day, when the Network appeared on the pictures of its circuit board.This is the first known version of the GeForce RTX 4090, equipped with two 12 + 4-pin 12VHPWR power connectors, which together are able to transmit to the video card to 1200 watts of power.
Increase the maximum frequency of the GeForce RTX 4090 GPU to 3705 MHz, which is 65.77% higher than the reference value of 2520 MHz, overclocking team OGS managed during the test of computing performance GPUPI for one gas pedal.
The frequency of video memory was not changed by the enthusiasts.When overclocked, the video card managed to compute in the GPUPI test in just 49 seconds and 49 milliseconds.
The closest result in this test is 57 seconds and 724 milliseconds.GeForce RTX 4090 showed it too, but with GPU frequency only at 3210 MHz.
Galax GeForce RTX 4090 HOF in the GPUPI test.Source image: HWBOTThe OGS team also managed to break the previous graphics performance record in the 3DMark Port Royal synthetic test.
Here the GeForce RTX 4090 HOF graphics chip was running at 3570 MHz.The gpu scored 31,096 points in the test, more than 1,000 points above the previous record.
Galax GeForce RTX 4090 HOF in the 3DMark Port Royal test.
The Galax GeForce RTX 4090 HOF is obviously the most powerful non-standard version of this graphics card at the moment.
With the absence of Kingpin models from EVGA, which decided to leave the graphics card market, as well as the lack of solutions from competitors Galax version of the graphics card could be the choice for most overclocking enthusiasts.
Gigabyte has shown how its Intel Z790 motherboards overclock DDR5 memory to 8000, 8333 and even 9300 MHz
Gigabyte has announced that some of its motherboard models on the Intel Z790 chipset for Raptor Lake processors have gained support for Intel XMP 3.0 profiles to automatically overclock DDR5 memory to effective frequencies of 8000 and even 8333 MHz.In addition, the Taiwanese company has shown support for manually overclocking memory up to 9300MHz.Image source: GigabyteThe manufacturer notes that the Z790 Aorus Tachyon motherboard for Intel Raptor Lake processors has successfully tested the DDR5-8333 modules using Intel XMP 3.0 profiles.It also demonstrated the ability to manually overclock the RAM to an effective frequency of 9300 MHz.Z790 Aorus Xtreme and Z790 Aorus Master models were tested with Intel XMP 3.0 profiles for DDR5-8000 RAM.Gigabyte has confirmed its words with screenshots of DDR5 memory modules overclocking tests. A set of memory modules overclocked manually to an effective frequency of 9300 MHz demonstrated the ability to run at CL46-58-58-62-120 timings on Z790 Aorus Tachyon motherboard. Automatically overclocked with Intel XMP 3.0 profiles, DDR5-8000 RAM set was tested with CL36-45-45-75-120 timings on Z790 Aorus Master board. The DDR5-8333 overclocking profile (shown in the picture above) on the Z790 Aorus Xtreme motherboard passed the test with CL38-48-48-78-126 timings.
The new Intel Core i9-13900K flagship processor went on sale today and set a new all-time overclocking record of 8.8 GHz.Thus, the previous record, held for 8 years, was broken.The new achievement probably won't last that long.Image source: HWBOT / ElmorThe previous world record for the highest CPU overclocking frequency was set in 2014.That CPU turned out to be the AMD FX 8370, released the same year.At that time, the chip was overclocked on a single core to a frequency of 8722.78 MHz.In all the time past, none of the presented Intel processors could not even come close to the previous record overclocking representative of the ancient AMD FX series.That all changed with the release of the new Raptor Lake series processors, the flagship of which can automatically overclock up to 5,8 GHz on 1-2 cores out of the box.The leaks of recent months clearly hinted that the Core i9-13900K has a huge potential for overclocking.It was obvious that the processor was capable of conquering higher clock frequencies.But few expected that the new chip which just went on sale would immediately set a new world overclocking record.An enthusiast with the nickname Elmor from Sweden entered into the HWBOT database his result of an extreme overclocking experiment with the Core i9-13900K using liquid nitrogen.The overclocker managed to overclock the chip on a single core to a frequency of 8812.85 MHz, which is an absolute world record.The ASUS Z790 ROG Maximus APEX motherboard was also used in the test system. Source image: CPU-Z ValidatorNo less interesting detail is the fact that the last 12 years in setting new world records win systems based on ASUS motherboards, which will be a separate reason to be proud of this Taiwanese manufacturer.
Raptor Lake's flagship 24-core Core i9-13900KF processor was overclocked to 6.2GHz, but only one core, but using only a conventional unattended (AIO) liquid cooling system.Picture courtesy of Overclock.net.Source image: VideoCardzThe flagship 24-core Core i9-13900KF Raptor Lake processor is officially allowed to automatically overclock one of its cores to 5.8GHz with Thermal Velocity Boost technology.The suffix \"KF\" in the model name indicates that the chip lacks integrated graphics.When overclocking the CPU to 6.2GHz, a Corsair iCUE H150I RGB Elite liquid cooling system was used as well as a Thermal Grizzly contact frame to protect against socket and poor waterblock contact with the chip cover.The test system also contained an ASUS ROG Maximus Z690 APEX motherboard, a GeForce RTX 4090 graphics gas pedal and G.Skill DDR5-6000 memory modules with CL36 delays. Intel Core i9-13900KF overclocking.Source image: CodyeThe system after overclocking was tested in Cinebench and 3DMark synthetic tests.In the single-core benchmark Cinebench R23, the processor showed a score of 2352 points after overclocking.This allowed it to rank 13th in HWBOT's single-core performance ranking.The user also reported that the processor scored 41,940 points in the multi-core performance test. Intel Core i9-13900KF overclocking.Image source: CodyeThe owner of the Core i9-13900KF notes that despite being able to overclock the chip to 6.2GHz on a single core, the processor temperature did not rise above 75°C when using the aforementioned coolant.The Core i9-13900KF was overclocked at 1.474V. The enthusiast also added that he was limited in time, so he has not yet conducted thorough tests on the processor.At the same time, standard overclocking profiles of ASUS ROG Maximus Z690 APEX motherboard were used in the experiment.
Graphics processors of NVIDIA GeForce RTX 4090 and GeForce RTX 4080 16GB graphics cards can run at 3.0 GHz.This is indicated by leaks collected by VideoCardz.Recall that the GeForce RTX 4090 and GeForce RTX 4080 16GB GPUs have officially announced clock speeds of 2520 and 2505 MHz under load, respectively.Image source: BillibiliSam NVIDIA previously demonstrated that the GeForce RTX 4090 is able to run up to 2.85 GHz under load and without additional overclocking.The transition to the new architecture Ada Lovelace and more advanced GPU manufacturing process in next-generation graphics cards, apparently, also laid a huge potential for additional manual overclocking of new gas pedals. According to a leak appeared in the Chinese social network Bilibili, in the stress test graphics subsystem with MSI Kombustor application graphics processor GeForce RTX 4090 card worked without problems at 3.0 GHz, while the power consumption of the whole graphics card was almost 426 W.On this basis, we can assume that in less stressful conditions, the graphics card chip will be able to work at 3.1 GHz.In addition, in the stress test Furmark GPU GeForce RTX 4090 worked at 2.64 GHz, but the peak power consumption of the video card was almost 616 W.Recall that the officially announced power consumption figure GeForce RTX 4090 is 450 watts.Unfortunately, the leak does not disclose data on the temperature performance of the GPU and other elements on the graphics card. According to another leak, GeForce RTX 4080 16GB is also capable of running at 3.0 GHz.It is very likely that here we are also talking about additional manual overclocking of the gas pedal.The data was provided by the same source which previously reported on the performance of the graphics card in the synthetic tests 3DMark Fire Strike Ultra and 3DMark Time Spy. Source image: ChirhellThe new leak indicates that the maximum temperature of the graphics processor was 60.8 ° C, and the hottest point of the graphics card temperature reached 75.4 ° C.According to GPU-Z software, the peak power consumption of the gas pedal was 333W and 290.7W for the GPU.Recall that the official stated power consumption figure GeForce RTX 4080 16GB is 320 watts. Source image: CrystalThe overclocking performance of GeForce RTX 4080 16GB graphics card was tested in the 3DMark Time Spy synthetic test.Due to the higher frequency after manual overclocking the gas pedal got away from its competitors even more, as you can see in the table below. Source image: VideoCardzThe flagship GeForce RTX 4090 will go on sale on October 12.Its recommended price is $1599.The GeForce RTX 4080 16GB will go on sale in November.Its suggested price is $1199.
At the Innovation 2022 event, Intel confirmed that its 13th generation Core processors support RAM up to DDR5-5600 standard without overclocking.Now we know that the flagship Core i9-13900K processor also announced support for DDR5-7600 memory, but overclocked with Intel XMP 3.0 profiles.Image source: VideoCardzIntel XMP technology allows you to automatically select the best memory profile for your intended use, determining the frequency and latency values when booting up the system.Overclocking memory tweaks can usually take hours to complete and test results, and Intel has offered to save time with default memory tuning profiles from the RAM manufacturer.Intel has a list of XMP 3.0 profiles which specify the optimal settings for certain memory modules to be overclocked.Recently Intel has updated this list.The fastest memory modules listed in the XMP 3.0 profiles are G.Skill's ones that were tested on ASUS Z790 ROG Maximus Hero and Apex motherboards.The lists indicate that DDR5-7466 and DDR5-7600 standard RAM modules can run at 1.4V and CL36 timings, as part of configurations with a Core i9-13900K processor.Below is a list of memory modules that work at these settings:G-Skill DDR5-7466 CL36 - F5-7466J3648G16GX2-TZ5RKG-Skill DDR5-7466 CL36 - F5-7466J3648G16GX2-TZ5RKG-Skill DDR5-7600 CL36 - F5-7600J3648G16GX2-TZ5RK Image Source: IntelWe should clarify that currently the fastest DDR5 memory kit available from G.Skill is the Trident Z5 RGB DDR5-6600.As for the RAM kits listed above, they have not yet gone on sale.
AMD Ryzen 9 7950X overclocked under liquid nitrogen to 6.5GHz and updated several records in Cinebench
AMD Ryzen 7000 series processors have not yet gone on sale but are already actively setting new records in various benchmarks, as evidenced by HWBot ratings.Earlier this week it was reported about four records set by Ryzen 9 7950X with liquid cooling, and now Tom's Hardware writes that some of those records have been surpassed - with Ryzen 9 7950X and liquid nitrogen.For starters, let us remind you that Ryzen 9 7950X with Corsair H115i liquid cooling system with 280mm radiator enthusiasts managed to overclock all 16 cores to 5.35-5.5GHz.This allowed them to update 16-core processor records in multi-threaded benchmarks such as Cinebench R23, Cinebench R20 and Cinebench R15 as well as in the 7-Zip archiver test. Now AMD Ryzen 9 7950X flagship is overclocked even more, but in extreme conditions with liquid nitrogen.In multi-threaded Cinebench R23 test AMD processor was overclocked to 6.45GHz and gained 48,235 points, 16.2% over the maximum score of Core i9-12900KS, also overclocked with liquid nitrogen. On the other hand, Ryzen 9 7950X overclocked to 6.5GHz in Cinebench R20 test and achieved 18,605 points.This is 15.8 % more than the previous best result achieved by liquid nitrogen cooled processors (Ryzen 9 5950X to 6 GHz).We should note that obviously more record updates are waiting for us in the near future, as it usually happens with the release of new generation processors.Ryzen 7000 will be released on September 27.Also in the second half of October we expect the release of Intel Core 13th generation processors (Raptor Lake), which will also compete for new HWBot ratings records.
Intel continues to educate the public about its Arc A770 and Arc A750 gaming gas pedals, giving away information bit by bit.Earlier, the company announced the official specs of the graphics cards, and shared their performance data with ray tracing and XeSS scaling technology.Now, Ryan Shrout and Tom Petersen from the company have shared information about how to manually overclock these graphics cards.Image source: IntelAs usual, manufacturers rarely talk about manual overclocking, let alone talk about the potential of one or another video card from them.Intel has decided to go against the long-standing tradition.Ryan Schrout and Tom Petersen showed overclocking of the reference Arc A750 through the proprietary Arc Control app, increasing the GPU Power Limit from 185W to 228W and changing the voltage settings. The graphics card was overclocked using its regular cooling system.In the end, the graphics chip frequency was increased to 2719 MHz.To test it, the game Hitman 3 was run.The overclocking did not result in any significant increase in game performance.Before the manipulations with GPU frequency settings, the card was producing 90 frames per second, and after that - 95-96 frames per second.The temperature of the GPU itself rose to 81 degrees Celsius after overclocking, and the temperature of memory chips - to 88 degrees. Intel representatives also showed the flagship gas pedal Arc A770 Limited Edition apart.As it turned out, the reference version of the graphics card is equipped with a cooling system, which includes a solid aluminum radiator, copper evaporation chamber, as well as four flat heat pipes measuring 10 × 3 mm. The entire design is cooled by two fans.According to Intel, the noise level of the cooling system is less than 39 dBA.In the reference version, the Arc A770 has RGB backlighting. The reference Arc A770 motherboard has a VRM subsystem with six power phases, eight GDDR6 memory chips (total capacity of 8 or 16 GB), three DisplayPort 2.0 connectors, as well as one HDMI 2.1 port, which provides chip PCON.Non-reference variants of Intel's Arc graphics cards may also be equipped with an HDMI 2.1 connector, but it will be up to manufacturers themselves if they wish to use an additional PCON chip in their circuitry.Additionally on the reference Arc A770 one 8-pin and one 6-pin power connectors can be noted.
Intel held an Intel Technology Tour press event in Israel this month, where employees from the local office talked about their achievements in developing both past and future generation central processing units.The Raptor Lake family is the brainchild of Intel's Israeli engineers and they proudly announced that the processors are overclocked to 8GHz.Image: Intel, Tom's HardwareThe figure may seem impressive to the untutored layman, but keep in mind that such experiments usually involve using liquid nitrogen to cool the processors and you can't expect to get any serious payload at that frequency range.Moreover, not even all of the processor's processing cores are usually active when reaching a record frequency.The progress of the Raptor Lake family, based on achieving a frequency of 8 GHz with liquid nitrogen, also not so great - at least the statistics overclocking Intel Core i9-12900KS processor contains evidence of reaching 7.8 GHz in June this year with the activity of eight cores and eight threads.It turns out that the successor was able to surpass this result by 200 MHz.Not the fact that the frequency of 8 GHz is limited, of course, because practice shows the ability of individual instances of processors of the new family in the future to improve the results of extreme overclocking in comparison with the debut experiments, which usually involve the use of early engineering samples.In the Intel processor history the 8GHz milestone is not unique either, as Cedar Mill's Celeron D generation processors once hit the 8.5GHz milestone, even though they were manufactured using the \"mature\" 65nm technology.Along the way Tom's Hardware representatives point out that Raptor Lake processors can dynamically boost frequencies to 6.0GHz using standard cooling systems.That's 500 MHz higher than the default limiting frequency of the Core i9-12900KS processor, which is considered the flagship of the Alder Lake line.However, the Intel presentation slides from the day before mentioned only 5.8 GHz as the Raptor Lake limiting frequency.Perhaps the beautiful value of 6.0 GHz is achieved for one or two active cores in a special mode of operation.
During the nightly presentation of Ryzen 7000 processors and new Socket AM5 platform, AMD also officially introduced new technology that is designed to make overclocking of DDR5 RAM much easier.It is named AMD Extended Profiles for Overclocking or AMD EXPO for short.Image source: AMDEXPO is an alternative to Intel XMP 3.0 technology.It is specifically designed for Socket AM5 platform and Ryzen 7000 processors, and allows the user to easily select the desired speed of RAM operation within RAM profiles, announced by the manufacturer.According to AMD company, using EXPO technology allows to easily increase the gaming system performance at 1080p resolution by up to 11%, and reduce the memory lag to around 63 ns. AMD said it aims to distribute EXPO under a free and royalty-free license, which will accelerate its adoption by memory manufacturers.The company said that many manufacturers have already adopted AMD EXPO and are about to release more than 15 bundles of EXPO memory modules with effective frequencies up to 6,400 MHz.These companies include ADATA, Corsair, GeiL, G.Skill, Kingston and others, for example. The first DDR5 RAM modules with support for overclocking profiles have already been introduced.For example, such solutions were announced by GeIL, Corsair and G.Skill.
MSI has released new BIOS versions for its AMD X570 and B550 chipset-based motherboards.The fresh firmware adds more customization options for Ryzen 7 5800X3D processors.These chips officially do not support manual overclocking, however, some motherboard manufacturers go for various tricks, allowing to raise CPU frequency.Image source: MSIMSI has released beta BIOS for its AMD X570 and B550 chipsets with Kombo Strike feature.At the moment, it is hard to say what exactly the benefit of this feature is - the company itself does not talk about it in detail.The description to the new BIOS version only mentions \"improved Ryzen 7 5800X3D performance\".At the moment we know that Kombo Strike has three operating modes and an option to turn it off.Probably the third mode provides the biggest performance boost for the processor.It is quite possible that MSI has found a way to change the CPU power consumption level.Officially this setting is locked out for the CPU as well as the others which aim to overclock.Corresponding beta BIOS versions with Kombo Strike have been released for X570S Unify, Ace, X570 Godlike, B550 Unify X, B550 Unify, B550 Gaming Edge WiFi, B550I Gaming Edge MAX WiFi and B550M Mortar mainboards.After installing the new BIOS version, the Kombo Strike feature can be found under Overclocking/Advanced CPU Configuration.Keep in mind that this is a beta firmware version and therefore errors and general system instability are possible during its operation.Note that MSI is releasing the second BIOS version with additional settings for Ryzen 7 5800X3D within a week.The previous version adds CPU offset, Precision Boost Overdrive and Curve Optimizer.The new beta BIOS also adds CPU Offset Voltage setup.It is intended for enthusiasts who want to experiment with the settings of the operating voltage indicator of the chipRyzen 7 5800X3D.But it is recommended to be used only by experienced users who know what they are doing.MSI is not the only motherboard manufacturer who tries to expand the possibilities of changing Ryzen 7 5800X3D settings.For example, ASUS has also included new features in their latest firmware versions that work on all Ryzen 5000 series processors, including the Ryzen 7 5800X3D.Because of this, some owners of these chips were able to overclock them to a frequency above 5 GHz at the official maximum of 4.5 GHz.
At Computex 2022, it was confirmed that AMD Ryzen 7000 processors will receive support for EXPO and Smart Access Storage technologies.The first technology has been previously reported.EXtended Profiles for Overclocking or EXPO is an analogue of Intel XMP 3.0 automatic overclocking profiles to increase the frequency of DDR5 RAM.Image source: VideoCardz / MSIAMD said that Smart Access Storage technology will be based on the already well-known Microsoft DirectStorage, which is used in Xbox Series game consoles and accelerates the loading of games.As noted by the company, the traditional game download scheme involves the participation of the CPU in the unpacking of game data and its movement, which increases the delay and requires a lot of CPU resources.Smart Access Storage will redirect the unpacking of game data to the GPU, bypassing the CPU, which will accelerate the loading of game levels and textures.Note that SAS which will not only rely on Microsoft DirectStorage, but also technology Smart Access Memory (SAM), which provides processor access to the entire memory of the graphics gas pedal.Image source: AMDAmd did not touch the topic of overclocking memory profiles EXPO, so the exact details about it yet unknown.However, the company used overclocked DDR5-6000 memory modules during the presentation of the Ryzen 7000 processor series. Source image: MSIThe EXPO technology was mentioned in MSI promo slides, which the company prepared to announce its motherboard based on chipset AMD X670.Another interesting feature of MSI promo slide is an indication that Socket AM5 platform supports 28 lanes of PCIe 5.0 instead of 24 lanes as AMD itself said during its presentation.It is possible that one of the manufacturers could have made a mistake.However, it is also possible that these four additional lines will be available for the next generation of Ryzen processors within the Socket AM5 platform, for example, for Ryzen 8000 conditional chips.Another interesting topic of discussion is the claimed TDP value of 170 W for the older Ryzen 7000 models.AMD has confirmed that this refers not to the nominal power consumption of the chip, but to the Peak Package Tracing (PPT) power value of the entire processor package.In other words, we're talking about the CPU's maximum power consumption during heavy loads.The actual TDP rating of the older Ryzen 7000 processor models is likely to be 105-125W.However, AMD does not specify this point.Still, the aforementioned MSI promotional slide clearly indicates exactly the TDP figure.If we're talking about TDP value, it is higher than current Ryzen 5000 processor's value of 142 W.VideoCardz says that MSI has asked the media not to publish the above slide.The manufacturer did not specify the reason.Perhaps because it indicates the EXPO technology, or maybe because it contains incorrect data.
AMD will unveil Radeon RX 7000 graphics cards on the latest RDNA 3 architecture graphics processors tonight. The start of the presentation,...