Intel's 192-Core Diamond Rapids CPU Takes Aim at AMD's Server Dominance
Diamond Rapids represents Intel's most aggressive server CPU push yet, splitting 192 performance cores across four 48-core tiles using multi-die architecture. You're looking at 16-channel DDR5 memory support delivering up to 1.6 TB/s bandwidth—double what current systems handle. That's the kind of memory throughput your AI training workloads have been desperately craving.
To get more in-depth news and expert analysis on global affairs from WPR, sign up for our free Daily Review newsletter.
The platform upgrades tell the real story. A new 9324-pin LGA socket, PCIe Gen 6.0 support, and configurations scaling to 768 cores per system suggest Intel is betting big on density. But here's where reality bites: each socket consumes 500W, meaning a fully-loaded quad-socket server hits 2000W. Your facility's power and cooling infrastructure might need serious upgrades.
Intel's timing couldn't be more critical. AMD's Epyc Venice, also targeting 2026, promises 256 Zen 6 cores on TSMC's 2nm process—33% more cores than Diamond Rapids. That's like bringing a smartphone to a smartwatch fight; the size difference matters more than you'd expect.
Both processors target the same sweet spot: environments blending traditional server tasks with AI workloads. Diamond Rapids includes Advanced Performance Extensions (APX) and enhanced Advanced Matrix Extensions (AMX), plus native support for TF32 and FP8 floating-point formats. These aren't just spec sheet bragging rights—they're essential for modern AI inference that doesn't choke your general computing performance.
Your infrastructure decisions in 2026 will hinge on more than core counts. Power consumption, platform compatibility, and total cost of ownership matter more than raw specifications. Diamond Rapids' 500W requirement might seem excessive, but it's targeting workloads that currently require multiple lower-core processors.
The server CPU wars are entering their most intense phase since the original Opteron vs. Xeon battles. Your next hardware refresh cycle will determine whether you're riding the performance wave or drowning in outdated infrastructure, especially as Intel's comeback plan takes shape.

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles


Digital Trends
11 hours ago
- Digital Trends
You can now unlock AMD's FSR 4 upscaling on even more games
AMD's FidelityFX Super Resolution 4 (FSR 4) upscaling is no longer limited to a short list of supported game titles. Thanks to a recent update to OptiScaler, Radeon RX 9000 series GPU users can now enable FSR 4 in virtually any game that uses DirectX 11 or 12. OptiScaler is an open-source utility that allows you to swap a game's existing upscaler with a more advanced version. For instance, if a game supports an older upscaling method like FSR 2, OptiScaler can intercept that and substitute it with FSR 4 instead. This gives players access to the latest upscaling technology, even in games that haven't officially been updated to support it. Recommended Videos Simply put, PC gamers can now bring FSR 4, AMD's latest and most advanced image upscaling technology, to previously unsupported games. These include ones that originally shipped with FSR 2, DLSS 2, or Intel XeSS, with no input required from game developers. While this isn't an official method endorsed by AMD, OptiScaler only modifies local game installation files. Hence, there's little risk involved as long as you stick to single-player titles and back up your game folders. Requirements and limitations Before we explain how to enable FSR 4 using OptiScaler, there are certain requirements you need to meet: GPU compatibility : FSR 4 is exclusively supported on AMD's Radeon RX 9000 series (RDNA 4) GPUs, including the RX 9060 and RX 9070 series. These cards feature dedicated AI acceleration hardware required to run FSR 4. Older generations of AMD GPUs currently do not support FSR 4. : FSR 4 is exclusively supported on AMD's Radeon RX 9000 series (RDNA 4) GPUs, including the RX 9060 and RX 9070 series. These cards feature dedicated AI acceleration hardware required to run FSR 4. Older generations of AMD GPUs currently do not support FSR 4. Game engine support : The game must run on DirectX 11 or DirectX 12. OptiScaler does not fully support Vulkan or other graphics APIs at this time. : The game must run on DirectX 11 or DirectX 12. OptiScaler does not fully support Vulkan or other graphics APIs at this time. Game type : OptiScaler may not work with multiplayer or online games that use strict anti-cheat protections like Easy Anti-Cheat (EAC) or BattlEye. This is because the injected files could trigger security flags and potentially lead to the suspension of your game account. : OptiScaler may not work with multiplayer or online games that use strict anti-cheat protections like Easy Anti-Cheat (EAC) or BattlEye. This is because the injected files could trigger security flags and potentially lead to the suspension of your game account. Per-game setup: OptiScaler needs to be configured manually for each game as there is no universal toggle. How to enable FSR 4 in your games Force enabling FSR 4 via OptiScaler requires a few manual steps for each game. All you need to do is copy special FSR 4 library files into a game's installation folder and configure OptiScaler to use it as the upscaler. While that may sound easy, it can be a little tricky so here's a step-by-step guide: Download the latest release of OptiScaler from its official GitHub page. No special installation is required, simply extract all the files into a folder. Make sure you are updated to the latest AMD Radeon graphics drivers. Head over to your game's main install directory and find the folder that includes the game's executable or .EXE file. Copy all the OptiScaler files that you previously extracted into this location and accept any prompts for replacing existing files. Once the OptiScaler files are copied into your game's directory (which includes the game executable), double click on the file named OptiScaler Setup or alternatively Windows Setup. A Windows Security prompt should appear, which you can bypass by clicking on Run Anyway. A command prompt-based setup for OptiScaler should fire up immediately where you need to select a filename for the new OptiScaler DLL file, followed by which GPU you are using since the tool works with Nvidia, AMD and Intel GPUs. There will also be an option to choose DLSS-based inputs, which is not necessary if the game already supports FSR 3 or above. Once all the selection is done, OptiScaler will be installed for your game. To set up OptiScaler, launch your game and bring up the OptiScaler in-game overlay by pressing the Insert key on your keyboard. Within OptiScaler you will have the option to select the upscaler you wish to use. To enable FSR 4, first select FSR 3x from the drop down menu and hit the Change Upscaler button. You should now have FSR 4 enabled with a dedicated settings tab within the tool. Unfortunately there is no universal installer meaning that one has to perform these steps for each game individually. In practice, copy the essential files into the new game's folder and configure OptiScaler the same way. Each title gets its own OptiScaler setup. In case you are not able to run FSR 4, head over to the OptiScaler GitHub Wiki page for more information. Once you've followed the steps and replaced the necessary files, the game will render using FSR 4, which should deliver improved image quality and higher frame rates compared to older upscaling versions. However, don't expect performance on par with NVIDIA's DLSS, which still holds a clear advantage in both visual fidelity and AI-driven upscaling. While OptiScaler's approach is somewhat of a workaround, it effectively extends FSR 4's availability from a limited number of officially supported titles to potentially hundreds of DirectX 11 and 12 games. That said, performance and visual improvements may vary from game to game depending on engine compatibility, asset quality, and how the original upscaling implementation was handled, so your mileage may vary.


Business Insider
14 hours ago
- Business Insider
‘No Magic Formula,' Says Top Investor About Intel Stock
Investors in Intel (NASDAQ:INTC) watched their fortunes plummet following the company's Q2 earnings call on Thursday after the close of the markets. INTC shares fell close to 9% on Friday, not exactly a resounding endorsement. Elevate Your Investing Strategy: Take advantage of TipRanks Premium at 50% off! Unlock powerful investing tools, advanced data, and expert analyst insights to help you invest with confidence. To be fair, there were some successes, as Intel's Q2 revenues of $12.9 billion surprised on the upside. EPS, however, came in at -$0.10, much lower than the projections of $0.01 for the quarter — which the company blamed on expenses related to 'onetime items and impairments.' CEO Lip-Bu Tan also shared that the company has been reevaluating its capex spending, and will be suspending various projects in Germany, Poland, and Costa Rica. There is no doubt that major changes are underfoot at Intel, which has been implementing large rounds of layoffs towards its goal of having 75,000 employees by the end of the year. One top investor known by the pseudonym JR Research applauds the CEO's 'pragmatic approach.' Nevertheless, the 5-star investor points out that there is no magic formula for Intel. 'CEO Lip-Bu Tan is no magician. He has to undo years of poor execution, strategic missteps, and a bloated internal bureaucracy, while facing extremely competitive and capable rivals in the semiconductor industry,' states the 5-star investor, who is in the very top 1% of TipRanks' stock pros. JR explains that the company is on an 'arduous journey' to unseat Taiwan Semiconductor's primacy in the foundry business. The investor points out that TSM has a massive leg up on Intel, so much so that companies like AMD believe that the 5% to 20% premium they could be paying for TSM's Arizona-manufactured chips over its Taiwan-produced hardware is 'worth it.' (JR reminds investors that Intel is based in the U.S. – 'what a pity!') 'Intel isn't anywhere close to winning enough business for the 14nm process to justify expanding aggressively now to compete with TSM,' adds JR. Intel can trumpet some positives, such as its 55% share of the server market, a signal that the company retains some of its historic cache. JR also points out that Intel's demise could be averted if it succeeds in executing well on its 18A process. For now, though, JR remains cautious. 'With structural constraints hampering its execution, Intel's hard times aren't expected to be over anytime soon,' sums up JR Research, who is maintaining a Hold (i.e. Neutral) rating. (To watch JR Research's track record, click here) That's exactly where most of Wall Street finds itself as well. With 26 Hold ratings – and only 1 Buy and 4 Sell ratings – INTC has a consensus Hold rating. Its 12-month average price target of $22.16 has an upside of ~7%. (See INTC stock forecast) To find good ideas for stocks trading at attractive valuations, visit TipRanks' Best Stocks to Buy, a tool that unites all of TipRanks' equity insights.
Yahoo
14 hours ago
- Yahoo
Intel CEO outlines 'hard but necessary decisions'; Trader Guilfoyle: CEO Tan 'Has a Chance'
Intel CEO outlines 'hard but necessary decisions'; Trader Guilfoyle: CEO Tan 'Has a Chance' originally appeared on TheStreet. Lip-Bu Tan did not hold back. "I know the past few months have not been easy," the Intel CEO said in a July 24 message to employees. "We are making hard but necessary decisions to streamline the organization, drive greater efficiency and increase accountability at every level of the company." 💵💰Don't miss the move: Subscribe to TheStreet's free daily newsletter 💰 Tan, who was named chief executive in March, said the chipmaker () plans to reduce headcount about 15% via dismissals and attrition, finishing the year with a global workforce of about 75,000. "We will become a faster, more agile and more vibrant company," he said. "We will eliminate bureaucracy and empower engineers to innovate with greater speed and focus." Intel's second-quarter results beat Wall Street's revenue expectations but came up short on earnings. The Santa Clara, Calif., company's shares were off 9.3% at last check; they're down 34% from a year ago. Intel CEO: Approaching AI Much Differently "Our goal is to reduce inefficiencies and redundancies," Tan said during the earnings call. More Tech Stocks: Analyst who correctly predicted Rocket Lab stock surge resets forecast Verizon Q2 earnings report surprises with remarks on tax reform Fund manager who forecast Nvidia stock rally reboots outlook He told analysts that Intel wouldn't move forward with planned projects in Germany and Poland. The company also intends to consolidate its assembly and test operations in Costa Rica into its larger sites in Vietnam and Malaysia. And it will further slow the pace of construction in Ohio to ensure spending is aligned with market demand. Turning to artificial intelligence, Tan said that Intel previously had approached AI 'with a traditional silicon and training centric mindset, without a cohesive silicon systems software stack and strategy,' he said."We recognize the need to move up the abstraction stack into system and software. This is the area where Intel has traditionally been weak or entirely absent.' 'But we intend to incubate and grow these important skillsets and capabilities under my leadership,' Tan continued. 'This will take time, but it will be vital for Intel to stay relevant in the next wave of computing.' TheStreet Pro's Guilfoyle: Intel CEO Tan has a chance TheStreet Pro's Stephen Guilfoyle offers a terse assessment of Intel's results. "Tough quarter," he said. "Tough few years." Guilfoyle, a veteran trader whose career dates back to the floor of the New York Stock Exchange in the 1980s, opines that "Intel has not had a competent CEO since Brian Krzanich resigned in June 2018 after violating the firm's non-fraternization policy with an employee," he said, "This new CEO has a chance.""At least he understands that Intel is no longer considered an elite American tech firm and that to be competitive a lot of work has to be done," Guilfoyle added. "He has no delusions." Tan has to improve cash flows and the balance sheet and compete against the likes of Nvidia () and Advanced Micro Devices () , "two chip designers at the top of the industry with an incredible level of corporate leadership stemming from their respective C-suites," he said. Intel's CEO "also seems to understand that the foundry business may grow but is not competitive against Taiwan Semiconductor () and is not close to turning a profit," he noted. Guilfoyle, known on Wall Street as Sarge, said the opportunity to make money on the short side has passed. "I don't see any catalyst that would prompt me to invest in Intel any time soon," he said. "The other semis are still better plays. ... I don't even see viable options-related plays here. There's just not enough premium to be paid for taking on the risk over time." "Kind of reminds me of an old Motley Crue song... 'Don't Go Away Mad (Just Go Away).'" Judge Dismisses Suit Against Intel; JPM Lifts Target Separately, a federal judge dismissed a lawsuit claiming Intel defrauded shareholders by concealing its business problems, leading to a $32 billion one-day tumble in its market value, Reuters reported. U.S. District Judge Trina Thompson in San Francisco said that while she "understands plaintiffs' frustrations," the company did not wait too long to reveal a $7 billion operating loss in its foundry business, Intel's stock price sank 26% on Aug. 2, 2024, a day after the chipmaker unveiled more than 15,000 job cuts and suspended its dividend, hoping to save $10 billion in the Q2 earnings report, JP Morgan raised its price target on Intel to $21 from $20 and affirmed an underweight rating on the shares, according to The Fly. The investment firm sees both continued demand and competitive headwinds for Intel. While the company decides whether to go fabless, its turnaround remains slow and customers buying before tariffs kick in creates risk to second-half 2025 estimates, it said.. Fabless refers to companies that design and market chips but outsource manufacturing to specialized CEO outlines 'hard but necessary decisions'; Trader Guilfoyle: CEO Tan 'Has a Chance' first appeared on TheStreet on Jul 27, 2025 This story was originally reported by TheStreet on Jul 27, 2025, where it first appeared. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data