Microsoft might be working on a cheaper 12-inch Surface, and Apple's MacBook Neo looks like the reason. DigiTimes reports, citing supply chain sources, that Microsoft is reviewing the specifications of its entry-level 12-inch Surface to bring costs down, and is also rethinking which processor goes inside. The report's headline, spotted by harukaze5719 on X, goes as far as saying Apple is forcing Microsoft to downgrade the device, and hints at a comeback for AMD with an Arm chip. Reportedly, the processor, memory, camera, battery, and chassis are all being looked at, though nothing seems to be finalized yet.

The pricing gap between the two laptops is quite large. Microsoft only just refreshed the 12-inch Surface Pro with Snapdragon X2 Plus, and it now starts at $1,149, up from $799 last year, partly because the 8 GB base configuration was dropped. Apple launched the MacBook Neo at $599, and even after a $100 increase in June that Apple blamed on memory costs, it still sits at $699. As for the chip, AMD's "Sound Wave" Arm APU was previously linked to Microsoft's 2026 Surface lineup and has even shown up in shipping manifests, so it looks like a candidate. Leaks put it at 2 performance and 4 efficiency Arm cores, RDNA 3.5 graphics with 4 CUs, and a 5 W to 10 W power target. Then again, Microsoft could just as well stick with Qualcomm, and neither Microsoft nor AMD has confirmed anything.

DLSS Swapper finally supports bulk DLL swapping, with version 1.2.7.0 landing on GitHub on October 18. Brad Moore's open-source tool lets PC gamers upgrade or downgrade the DLSS, FSR, and XeSS files in their installed games without waiting for a developer patch, but until now, that meant going through every game one at a time. Videocardz points out that the new Bulk Swap button in the games page toolbar lets users tick multiple games, then pick a version for each DLL type, plus presets for DLSS Super Resolution, Frame Generation, and Ray Reconstruction. Every row defaults to "Don't change," so nothing gets touched unless it's chosen. Apparently, users had been asking for this on GitHub for quite a while, with some saying they had well over 50 DLSS games to update by hand.

Bulk restoration of the original DLLs made it in too. Moore had pointed out in that same thread that users would need an easy way back if a new DLSS build turned out to be a regression across, say, 160 games. Of course, a swapped DLL can still stop a game from launching, and titles with anti-cheat might flag modified files, so online games are probably best left alone. The update also adds DLSS Frame Generation Preset C, the faster performance mode NVIDIA pushed to its app beta last week, which modders have already brought to RTX 30 cards. A bug that could stop the game library from loading has also been fixed, and full library scans are faster too. Manual swappers can still grab individual files from TechPowerUp's DLSS DLL archive.

Apple's MacBook Neo, MacBook Air, and the latest MacBook Neo series of laptops are now susceptible to a significant performance drop if they remain powered on for 49.7 days without a reboot. Specifically, after exactly 49 days, 17 hours, 2 minutes, and 47.296 seconds of continuous operation, the entire TCP/IP networking stack shuts down and stops establishing new connections. This unusual bug stems from a kernel-level issue in Apple's XNU kernel involving tcp_now, a 32-bit unsigned integer that tracks elapsed milliseconds since boot for the TCP stack. When it reaches its maximum value of 4,294,967,295, the counter stalls instead of resetting, disrupting the arithmetic macOS uses to expire and discard closed connections.

These inactive connection sessions accumulate quickly, increasing CPU usage as the system struggles to handle an ever-growing backlog. Once the available port pool, typically around 16,384 ports, is fully used, no new connections can be opened. Applications that rely on network access begin to fail, although already-established sessions continue to function, and the machine still responds to pings, albeit much more slowly. This initially delayed the diagnosis of the bug. A reboot resets the counter and restores normal operation, but the issue will recur on the same schedule after another 49.7 days of continuous use. There are potential solutions to fix the bug by using larger 64-bit integers, but this would require more extensive code changes, performance testing, and additional work compared to a simple fix.

Last November, ASUS launched its GeForce RTX 5090 ROG MATRIX Platinum 30th anniversary 32 GB graphics card model. This strictly limited edition release—at the time, only 1000 units were produced—demanded a whopping $3999 (per unit) price tag. Unique selling points (USPs) included an unusual overall design, 3 oz copper PCB layers, a dual-power input involving the 16-pin 12V-2x6 connector and BTF-supplied power delivery, as well as the integration of a tiny Bosch Sensortec BMI323 inertial measurement unit (IMU). By early December, it was not clear whether initial quality issues had caused an official(?) recall of extremely overengineered MATRIX custom flagships. Months later, a Chinese market variant has hit the pre-order stage—over the weekend, ITHome reported on the pre-release status of this sanction-conforming model.

Despite featuring downgraded specifications, the customized GeForce RTX 5090 D V2 24 GB version still demands a hefty cost of entry—a main JD.com listing advertises a price tag of 29,999 yuan (~$4341 USD). Last month, Tony Yu—the manager of ASUS China—showcased a MATRIX GeForce RTX 5090 D V2 sample achieving a 3DMark Time Spy Extreme score of 28,638 points. This impressive tally was enabled via an 800 W operation mode, ITHome noted that this figure: "even surpasses NVIDIA's RTX PRO 6000 (21,760 cores vs. 24,064 cores), which is closer to (utilizing) the 'full-blooded' GB202 (GPU die's max. count of 24,576 units)." The older GeForce RTX 5090 D "V1" 32 GB card was no slouch in gaming and AI performance departments, but global tensions led to an eventual cut-off of supplies into China. Oddly, ITHome tracked the 25,999 yuan (~$3762 USD) listing of an MSI GeForce RTX 5090 D Gaming Trio 32 GB SKU on the JD.com e-tail platform, earlier on in 2026. In any case, local hardcore PC enthusiasts and AI companies can access top-tier "Blackwell" GPU hardware, whether in V2 or V1 guises. Black market options, including "truly banned" models are also (reportedly) available.

TrendForce's latest investigations indicate that DRAM suppliers in 1Q26 will continue to reallocate advanced process nodes and new capacity toward server and HBM products to support rising AI server demand. This shift has significantly limited supply in other markets, causing conventional DRAM contract prices to increase by about 55-60% QoQ. Meanwhile, disciplined capacity management by NAND Flash suppliers, along with robust server demand that is displacing other applications, is likely to increase contract prices across all NAND Flash product categories by 33-38%.

Despite weaker notebook shipments and slower growth in memory demand due to potential specification downgrades, PC DRAM prices are still set to rise sharply in the first quarter. DRAM suppliers have simultaneously tightened supply to PC OEMs and module makers, forcing some OEMs to procure memory at higher prices through modules makers. This is expected to raise DRAM suppliers' module pricing and significantly drive up PC DRAM prices.

On Monday, Fitch, a global credit rating agency, lowered Intel's long-term rating from BBB-plus to BBB and assigned a negative outlook. That change puts Intel just two notches above junk status. Fitch analysts cited growing challenges in maintaining strong demand for Intel's chips amid fierce competition from AMD, Broadcom, and NVIDIA. While Intel still holds a solid position in the PC and enterprise server markets, the agency warned that the company's financial structure is weaker than that of similarly rated peers and faces significant risk. To reclaim its prior rating, Intel must deliver successful product launches and reduce its net debt over the next 12 to 14 months. Despite the downgrade, Fitch described Intel's liquidity as robust: as of June 28, the company held $21.2 billion in cash and short-term investments, had an unused $7 billion credit revolver, and carried an undrawn $5 billion facility due in January 2026. Analysts say the move serves as a wake-up call for Intel's leadership as they navigate the company's various hurdles.

Credit ratings matter because they influence the cost of borrowing for a public company and can impact investor confidence. A downgrade typically leads to higher interest rates on new debt and may unsettle creditors and shareholders. For a capital-intensive business like Intel, increased financing costs can lead to cuts in research, development, and fab investment, thereby slowing progress on next-generation technologies. A negative outlook also highlights the risk of further downgrades if performance does not improve, potentially affecting the stock price. Intel has already seen similar actions from other agencies: S&P Global cut its rating to BBB in December, and Moody's lowered its senior unsecured debt grade last August. Investors will be watching how Intel's management balances cash flow, debt repayment, and product innovation. If the company can strike the right mix of investment and financial discipline, it may reverse the trend and secure a stronger rating in the months ahead. As one of the most important companies in the US semiconductor supply chain, Intel will need to overcome numerous issues to maintain a healthy business.

When we heard the news that NVIDIA will be halting shipments to Chinese customers for its GeForce RTX 5090D GPU in Q2, there were some speculations that the company is preparing to make a comeback. Leaving out the Chinese market is a tough decision from a business standpoint, but NVIDIA complies with export regulations imposed by the US administration under Donald Trump. However, the company now plans to keep its presence in China with the release of the export-abiding GeForce "RTX 5090DD" GPU. Carrying a GB202-240 die, it will be a cut-down version with reportedly only 14,080 CUDA cores present, paired with 24 GB of GDDR7 memory.

Chinese gamers will notice a significant difference, but so will anyone trying to acquire these GPUs for non-gaming purposes. As the US administration has banned shipments of high-end chips to China to stop them from training military-grade AI, gaming GPUs have been a target of AI firms. Now, a modified GeForce RTX 5090 DD, which not only cuts with raw compute but also firmware, should be sufficient for gaming. Carrying a PG145 SKU 40 board, the new RTX 5090DD GPU will also feature a modified 384-bit bus for its 24 GB of memory, down from the original RTX 5090D's 512-bit bus.

Update 09:45 UTC: According to one of the most accurate NVIDIA leakers, kopite7kimi, the core count for this GPU is supposed to be 21,760 CUDA cores, with 575 W TDP. The leaker did note that there is some surprise, so NVIDIA has definitely changed something big under the hood to comply with export regulations.

Around late April, Chinese industry insiders started whispering about a possible halting of NVIDIA GB202 "Blackwell" GPU shipments into the region. Team Green's local board partners released custom "market exclusive" GeForce RTX 5090D 32 GB cards; featuring slightly downgraded flagship silicon. Since launch, Chinese hardcore gaming enthusiasts and DIY AI firms have observed impressive performance credentials, despite the presence of factory limited specifications. On May 6, reports suggested a complete halt of GeForce RTX 5090D sales in China. Fresher rumors indicate NVIDIA's engineering team return to the drawing board; with their alleged preparation of an even weaker GeForce RTX 5090D design.

A group of Baidu tipsters have discussed a key area of (further) compromise: VRAM capacity. The original GeForce RTX 5090D configuration was armed with 32 GB of GDDR7 VRAM; mid-May speculation envisioned a 24 GB variant. Earlier today, harukaze5719 highlighted an intriguing Weibo post. A Manli webshop representative has "confirmed" that revised GeForce RTX 5090D cards will become available around July. Up until recent events, this Chinese brand was selling flagship GPU-based Gallardo (black) and Stellar (white) models. A web chat session was captured and shared on Weibo—in addition, the sales agent disclosed their belief that (NVIDIA's) supply of GeForce RTX 5090D GPUs has been "insufficient since launch" time. On a semi-positive note, customers will not be greeted by price hikes. Manli anticipates "unchanged" price tags; albeit associated with lesser specifications.