Latest news with #CPUs
Yahoo
3 days ago
- Business
- Yahoo
Intel experimenting with direct liquid cooling for up to 1000W CPUs
When you buy through links on our articles, Future and its syndication partners may earn a commission. Intel is testing a new way to tackle the growing heat output of its power hungry chips. At its recent Foundry Direct Connect event, the company showcased an experimental package-level water cooling solution designed to more efficiently cool CPUs . Intel has working prototypes for both LGA (Land Grid Array) and BGA (Ball Grid Array) CPUs, with demos using Intel's Core Ultra as well as Xeon server processors. The cooling solution doesn't apply coolant directly to the silicon die. Instead, a specially designed compact cooling block sits atop the package, featuring microchannels made of copper that precisely guide the coolant flow. These channels can be optimized to target specific hotspots on the die, potentially improving heat removal where it matters the most. Intel claims the system can dissipate up to 1,000 watts of heat using standard liquid cooling fluid. That kind of thermal load isn't typical for consumer CPUs, but it could be relevant for high-end AI (Artificial Intelligence) workloads, HPC (High Performance Computing), and workstation applications. Image 1 of 9 Image 2 of 9 Image 3 of 9 Image 4 of 9 Image 5 of 9 Image 6 of 9 Image 7 of 9 Image 8 of 9 Image 9 of 9 The cooling assembly is also said to make use of solder or liquid metal TIM (Thermal Interface Material), which is said to offer better contact than polymer based TIM. Compared to a traditional liquid cooler mounted to a delidded bare die, Intel says this solution can deliver 15–20% better thermal performance. Notably, Intel's approach isn't just a lab experiment. The company has reportedly been working on this technology for years. With rising thermal demands from modern chip designs, Intel is now exploring how to produce this system for real-world deployment. While Intel refines its prototype, the enthusiast community is already experimenting with similar concepts. YouTuber octppus recently modified the heatspreader of an Intel Core i9-14900KS, machining it into a functioning miniature water block. With internal channels carved into the IHS (Integrated Heat Spreader) and sealed under acrylic, the mod somewhat mirrors Intel's concept in DIY fashion. Intel hasn't confirmed when or if this cooling approach will hit mainstream products, but the demonstration is critical for CPU thermal design. As power consumption and package density increase, direct cooling may become a necessity for both professional and enthusiast hardware in the coming future. Follow Tom's Hardware on Google News to get our up-to-date news, analysis, and reviews in your feeds. Make sure to click the Follow button.
The Verge
21-05-2025
- Business
- The Verge
AMD takes aim at Intel with new 96-core Threadripper 9000 series CPU
AMD has announced its latest Zen 5-based Ryzen Threadripper 9000 Series of CPUs at Computex today. The 9000 Series and 9000 WX-Series are built for the demanding workstation market, and the top Threadripper Pro 9995WX will ship with 96 cores and 192 threads. This flagship Threadripper chip is designed for professionals who are working on visual effects, simulations, and AI model development. The Threadripper Pro 9995WX also has up to 384MB of L3 cache and 128 lanes of PCIe Gen 5, making it ideal to pair with multiple GPUs. AMD claims that the Threadripper Pro 9995WX is 2.2x faster than Intel's 60-core Xeon W9-3595X processor in Cinebench 2024 multi-threaded rendering. If you don't need a 96-core CPU, AMD's Threadripper 9000 Series are also targeted at enthusiasts and creators who want workstation-like performance. The Ryzen Threadripper 9980X has 64 cores and 128 threads, a base frequency of 3.2GHz, and 320MB of L3 cache. All of these new Threadripper chips, pro or not, will run at a thermal design power (TDP) of 350 watts and will work (after a BIOS update) on existing motherboards that support the sTR5 socket. Both Threadripper 9000 Series and the Pro WX-Series processors will be available from retailers in July, but AMD isn't announcing pricing just yet. Given the its high-end 7980X Threadripper CPU retailed at $4,999 in 2023, it's fair to say these next-gen equivalents will be around that price.
Yahoo
19-05-2025
- Business
- Yahoo
Qualcomm to make data center processors that connect to Nvidia chips
By Stephen Nellis and Max A. Cherney SAN FRANCISCO/TAIPEI (Reuters) -Qualcomm on Monday said it will make custom data center central processing units, or CPUs, that use technology from Nvidia to connect to Nvidia's artificial intelligence chips. Nvidia's chips are dominant in the AI market but always paired with CPUs, a market traditionally dominated by Intel and Advanced Micro Devices. Nvidia has jumped into the CPU market itself, designing a chip using technology from Arm Holdings to develop its own "Grace" CPU. On Monday, Qualcomm said it would return to the data center CPU market. In the 2010s, Qualcomm began developing an Arm-based CPU that it tested with Meta Platforms, but curtailed those efforts amid cost cuts and legal challenges. But after acquiring a team of ex-Apple chip designers in 2021, Qualcomm has quietly revived those efforts, again holding discussions with Meta about a data center CPU. Qualcomm last week confirmed that it has a letter of understanding with Saudi Arabian AI firm Humain to develop a custom data center CPU. On Monday, Qualcomm said that its future chips would use technology from Nvidia that will help them communicate quickly with Nvidia's graphics processors (GPUs), which are the mainstay of its AI chip portfolio. "With the ability to connect our custom processors to Nvidia's rack-scale architecture, we're advancing a shared vision of high-performance energy-efficient computing to the data center," Cristiano Amon said on Monday. 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
CNA
19-05-2025
- Business
- CNA
Qualcomm to make data center processors that connect to Nvidia chips
SAN FRANCISCO/TAIPEI :Qualcomm on Monday said it will make custom data center central processing units, or CPUs, that use technology from Nvidia to connect to Nvidia's artificial intelligence chips. Nvidia's chips are dominant in the AI market but always paired with CPUs, a market traditionally dominated by Intel and Advanced Micro Devices. Nvidia has jumped into the CPU market itself, designing a chip using technology from Arm Holdings to develop its own "Grace" CPU. On Monday, Qualcomm said it would return to the data center CPU market. In the 2010s, Qualcomm began developing an Arm-based CPU that it tested with Meta Platforms, but curtailed those efforts amid cost cuts and legal challenges. But after acquiring a team of ex-Apple chip designers in 2021, Qualcomm has quietly revived those efforts, again holding discussions with Meta about a data center CPU. Qualcomm last week confirmed that it has a letter of understanding with Saudi Arabian AI firm Humain to develop a custom data center CPU. On Monday, Qualcomm said that its future chips would use technology from Nvidia that will help them communicate quickly with Nvidia's graphics processors (GPUs), which are the mainstay of its AI chip portfolio. "With the ability to connect our custom processors to Nvidia's rack-scale architecture, we're advancing a shared vision of high-performance energy-efficient computing to the data center," Cristiano Amon said on Monday.
Reuters
19-05-2025
- Business
- Reuters
Qualcomm to make data center processors that connect to Nvidia chips
SAN FRANCISCO/TAIPEI, May 19 (Reuters) - Qualcomm on Monday said it will make custom data center central processing units, or CPUs, that use technology from Nvidia to connect to Nvidia's artificial intelligence chips. Nvidia's chips are dominant in the AI market but always paired with CPUs, a market traditionally dominated by Intel and Advanced Micro Devices. Nvidia has jumped into the CPU market itself, designing a chip using technology from Arm Holdings to develop its own "Grace" CPU. On Monday, Qualcomm said it would return to the data center CPU market. In the 2010s, Qualcomm began developing an Arm-based CPU that it tested with Meta Platforms, but curtailed those efforts amid cost cuts and legal challenges. But after acquiring a team of ex-Apple chip designers in 2021, Qualcomm has quietly revived those efforts, again holding discussions with Meta about a data center CPU. Qualcomm last week confirmed that it has a letter of understanding with Saudi Arabian AI firm Humain to develop a custom data center CPU. On Monday, Qualcomm said that its future chips would use technology from Nvidia that will help them communicate quickly with Nvidia's graphics processors (GPUs), which are the mainstay of its AI chip portfolio. "With the ability to connect our custom processors to Nvidia's rack-scale architecture, we're advancing a shared vision of high-performance energy-efficient computing to the data center," Cristiano Amon said on Monday.
