AMD launched limited regional promotion, within the framework of which buyers of ready-made computers with a video card Radeon RX 9060 8GB will receive a digital copy Call of Duty: Black Ops 7However, the offer extends only for the South Korean market and acts through selected partners.
The Radeon RX 9060 used in the promotion is OEM version based on RDNA 4 architecture, not intended for retail sale separately. The card is installed exclusively in pre-assembled or custom systems. RX 9060 is equipped with 1792 stream processors, 28 RT blocks и 56 tensor cores, providing support for ray tracing and AI acceleration. Video memory is 8 GB GDDR6 with 128-bit bus and throughput 288 GB / s, and the maximum frequency reaches 2990 MHz. The map uses one 8-pin power connector, takes up two slots and is designed for a power supply unit with a capacity of 500 W.
Customers who place an order for a PC with RX 9060 as part of the promotion will receive unique AMD Rewards code, which gives the right to download Call of Duty: Black Ops 7 via Battle.net. The promotion is limited to one code per system and email address, with a maximum of three games per household.
The action is held in fourth quarter of 2025 and operates until codes run outOther regions, DIY PC builders, and owners of non-OEM RX 9060 cards are still out of coverage.
The tech community found itself in the midst of a curious situation when the defunct AMD Ryzen 7 9700X3D processor entered the PassMark database and sparked a wave of media coverage. It all started with a joke by a group of Reddit enthusiasts who they changed the name of the real one CPU Ryzen 7 9700X in a Linux system file, passing it off as a supposedly new chip with 3D V-Cache technology.
After this substitution, a benchmark was run in PassMark, which added the results to the open database without verification. Thus, A post about a non-existent Ryzen 7 9700X3D appeared, which sparked a furious reaction from industry publications. Many media outlets rushed to report on the "new" AMD chip without verifying the information. Some even began speculating about its specifications: frequency of 5,8 GHz from the fake test was reduced to a “more realistic” 4,8 GHz, and TDP of 120 W was invented.
The situation got out of control when Some online stores have opened pre-orders "new product" based on just one test. Realizing the scale of what was happening, the authors of the joke published a detailed explanation on Reddit, in which described the substitution process in detail and called for a critical approach to isolated leaks, especially from benchmark databases.
The case was a stark reminder of how easily misinformation can spread in the technological environment. PassMark has not yet commented., and many media outlets were forced to remove or correct publications based on the fictitious processor.
Less than a week after the previous achievement, a new world record was set in the world of memory overclocking: DDR5 speed reached 13211 MT/sThe author of the overclocking was an enthusiast AiMax, who used GIGABYTE Z890 AORUS Tachyon ICE motherboard, which has already proven itself as the best choice for extreme RAM overclocking.
AiMax applied one Patriot Viper Xtreme 5 24 GB module, working in single-channel mode, Intel Core processor Ultra 7 265KTo cool and stabilize operation at such high frequencies, a a liquid nitrogen - for both CPU and memory. CPU-Z fixed the frequency 6605.7 MHz, which is equivalent 13211 MT / s, and the result was officially uploaded to the HWBot database.
This achievement surpassed the previous record of Saltycroissant (13153 MT/s) by only 58 MT/s, however AiMax took first place Both in the overall standings (out of 2762 results) and in the DDR5 category (out of 400). Memory timings remained the same as in previous entries: CL68-127-127-127-2, which indicates stability at such extreme frequencies.
Interestingly, the Breaking the 13000 MT/s mark has become a new standard for single modules on high-end platforms with extreme cooling. While these frequencies don't provide a noticeable boost in everyday workloads, they reflect the potential of modern hardware and the engineering achievements of enthusiasts. Given the stability of current systems, The 13500 MT/s mark no longer seems out of reach.
A new gaming PC, the PowerSpec G527, has appeared on the MicroCenter network, which is equipped with the unannounced Ryzen 5 7500X3D processor. At the time of publication AMD has not officially announced this model, and it not listed on the manufacturer's website, however, the characteristics are confirmed through the OPN code (100-000001904) and are completely consistent with previously appeared leaks.
The Ryzen 5 7500X3D is a 6-core processor with a base frequency of 4.0 GHz., which uses 3D V-Cache technology, similar to other chips in the X3D series. The PowerSpec G527 system includes motherboard with B850 chipset, 32GB DDR5-6000, Radeon RX 9060 XT graphics card with 16 GB of memory, which provides a noticeable advantage over the 8-gigabyte version of the same model.
The most interesting thing is - The assembly price is $999, while a similar PowerSpec G525 system with Ryzen 5 7600X3D costs $100 more expensive. This way, buyers get a more affordable configuration with the same features (except CPU), but saving a significant amount of money.
The B850M chipset supports PCIe 5.0, making it the optimal choice for a system with a Ryzen 5 7500X3D. Like other X3D processors, this chip does not support overclocking, so using the more expensive B650E or X670E boards doesn't make much sense.
Thus, the PowerSpec G527 with the new processor looks like an attractive option for casual gamers, who value high gaming performance and a balanced configuration without overpaying.
Microsoft is developing a special version Windows 11 26H1, which is intended exclusively for new devices based on the Snapdragon X2 and NVIDIA N1X platformsThis build is based on new Bromine architecture and, unlike the general update 25H2 (Germanium), is a full upgrade of the operating system, focused on supporting specific silicon.
The final version of the 26H1 with RTM assembly in the range 28000 expected in November 2025 of the year, after which manufacturers will be able to optimize firmware and drivers for future releases. for current users Windows 11 no action required - 26H1 will not be distributed through Windows Update and intended for new devices only.
According to Windows Central, Snapdragon X2 Elite and X2 Elite Extreme were announced in September 2025 and will appear in commercial laptops and mini-PCs in the first half of 2026. NVIDIA N1X — a platform that has not yet been announced, but it is known that a similar architecture is already being used in DGX Spark with chip Blackwell GB10, which uses DGX OS, which supports games through emulation.
In this way, Windows 11 26H1 will be the basis of a new generation of Arm-oriented AI-PCs, including devices with platforms from Qualcomm and, possibly, compact systems from NVIDIA. Later, in the second half of 2026, support for all architectures is expected to be unified. version Windows 11 26H2.
Well-known tech insider @InstLatX64 published a new batch of data about AMD's future processors. The most notable discovery was first confirmed CPUID for the Zen 6 architecture — B80F00According to the attached table, this architecture will receive the code name "Venice" for server solutions (SP8 socket) and Medusa Point for future APUs (FP10 socket).
The leak also sheds light on the structure of future chips. It appears Zen 6 will use a complex hybrid architecture similar to Intel's approach. The data suggests three different core types: "Classic Zen6", "Dense Zen6" и "LPE Zen6" (probably Low Power Efficient)This could mean that AMD is preparing chips with different core types for different tasks, which would be a significant improvement over Zen 5.
While Zen 6 is in development, an insider shared fresh data on Zen 5. A 64-core processor dump was detected. Threadripper Pro 9985WX, which belongs to the "Shimada Peak" (Zen 5) generation.
Finally, a leak confirmed CPUID for Zen 5 servers, which were previously reported. Chips "Sorano" (Zen 5c) for EPYC 8005 (SP6) received CPUID B10F00, and chips "Grado" (Zen 5) for EPYC 4005 (AM5) — CPUID B40F40This fills the gaps in information about AMD's Zen 5 server lineup.
The previously reported GeForce RTX 50 Super series has been delayed., could have a domino effect on Nvidia's entire roadmap. According to recent rumors, the release of updated Super cards postponed to the third quarter of 2026 (Q3 2026). Now analysts believe that this is inevitable. will also delay the release of the next generation, known under the code name "Rubin".
It was previously expected that Rubin architecture (RTX 6000) will be announced in late 2026, with the graphics cards themselves expected to go on sale in the first half of 2027. However, the new release schedule makes this scenario unlikely. Releasing an entirely new generation (RTX 6000) just one or two quarters after the RTX 50 Super refresh would be illogical from a marketing and product lifecycle perspective.
Thus, according to new forecasts, The release of the all-new 6000 series is delayed.The full launch of the GeForce RTX 6090, RTX 6080, and other models based on the Rubin architecture is now expected no earlier than the end of 2027.
As a reminder, a significant performance boost is expected from the Rubin architecture., comparable to the transition from the 3000 to the 4000 series. Rumors indicated that it was based on GeForce RTX 6090 will be based on TSMC's 3nm process technology, the card will receive 32GB of video memory and will be 80% faster than RTX 5090However, it now appears gamers will have to wait much longer for this upgrade.
A new test update for Mesa graphics drivers (version 25.3-rc4) has been released for Linux gamers, fixing a host of gaming issues. The release of the final stable version 25.3.0 is slightly delayed as developers fix the final bugs, but this release candidate already contains key changes.
Main news for owners Intel Arc video cardsIntel engineers have fixed a serious issue that was causing The graphics in many games are broken, launched through Steam Play (Proton). The problem was that the Intel driver handled textures differently in Windows-games and Linux games. Now the driver has been taught to "think" like Windows, when a D3D10+ game is running (almost all modern ones). This fixes visual bugs and glitches in a bunch of games, including Assassin's Creed Valhalla.
There is good news for owners too. AMD Radeon graphics cards. Special patches-"workarounds" have been added to the driver, which solve problems in specific gamesThis time the fixes have been updated Doom Dark Ages и No Man's SkyIf you've been experiencing crashes or bugs in these games on Linux, they should disappear soon.
The remaining fixes are minor and concern old drivers or general errors.
Nvidia CEO Jensen Huang is reportedly arriving in Taiwan today, November 8, 2025. His private jet is rumored to be flying directly to Tainan. The purpose of the visit is to visit 3nm TSMC fabs (Fab 18B) at the South Science Park, where he is expected to be personally greeted by TSMC Vice President Wang Yinglan.
Tomorrow, November 9, Huang will leave for Hsinchu to attend TSMC's annual sports dayThis is a landmark event where he is expected to appear on the same stage with the founder of TSMC. Morris ChangThis visit comes amid news that Nvidia has chosen Taipei (Beishike District) for its new Taiwanese headquarters.
Huang's visit is directly related to a sharp increase in orders for AI accelerators. According to sources, TSMC is planning significantly expand production of 3-nm chips to Fab 18B to meet additional demand from Nvidia. Production capacity is reportedly expected to increase from the current 100-110 wafers per month to 160 wafers per month.
This capacity growth will be between 45% and 50%It is reported that only additional Nvidia order TSMC's monthly capacity is 35,000 wafers. This move reflects the ongoing boom in AI infrastructure construction worldwide, providing TSMC with a strong growth impetus next year. Nvidia's additional orders are likely for Blackwell chips. Ultra, as well as for the platform Vera Rubin, which will replace them next year. Both generations will be produced using the same process technology TSMC N3P.
Nvidia's graphics card component supply problems appear to be worsening. Following rumors of a delay for the RTX 50 Super lineup, authoritative insider MEGAsizeGPU (@Zed__Wang) acted with a new warning. He stated that the popular mid-range graphics card, GeForce RTX 5060 Ti with 16GB of memory, "will very soon be in short supply."
The announcement comes amid a broader discussion initiated by Hardware Canucks that rising prices for DRAM memory have already led to a medium-term shortage. It was previously reported that a shortage of chips (especially 3GB GDDR7) forced Nvidia to delay or even cancel the Super line.
Now, apparently, low memory problems These changes are starting to affect more mainstream products as well. The RTX 5060 Ti 16GB model uses twice as many memory chips as its 8GB counterpart, which, given rising DRAM prices, makes its production either less profitable or more complex from a C-level logistical standpoint.
If the insider's forecast turns out to be correct, popular mid-range card with a large amount of video memory may soon disappear from store shelves or increase in price significantly
