Latest news with #PCIe
Yahoo
5 days ago
- Business
- Yahoo
This insane $1,000 Intel GPU could challenge Nvidia's $18,000 monster in workstation memory power
When you buy through links on our articles, Future and its syndication partners may earn a commission. Intel's Arc Pro B60 Dual offers pro-grade memory at a fraction of Nvidia's price This dual-GPU rig from Maxsun delivers workstation power Each GPU gets one DisplayPort and one HDMI, avoiding OS overload in multi-GPU workstations At Computex 2025, Maxsun unveiled a striking new entry in the AI hardware space: the Intel Arc Pro B60 Dual GPU, a graphics card pairing two 24GB B60 chips for a combined 48GB of memory. Servethehomeclaims Maxsun envisions these cards powering dense workstation builds with up to four per system, yielding as much as 192GB of GPU memory in a desktop-class machine. This development appears to have Intel's implicit approval, suggesting the company is looking to gain traction in the AI GPU market. The Arc Pro B60 Dual GPU is not designed for gaming. Instead, it focuses on AI, graphics, and virtualization tasks, offering a power-efficient profile. Each card draws between 240W and 300W, keeping power and thermal demands within reach for standard workstation setups. Unlike some alternatives, this card uses a blower-style cooler rather than a passive solution, helping it remain compatible with conventional workstation designs. That matters for users who want high-end performance without building custom cases or cooling systems. Still, the architecture has trade-offs. The card relies on x8 PCIe lanes per GPU, bifurcated from a x16 connector. This simplifies design and installation but limits bandwidth compared to full x16 cards. Each GPU also includes just one DisplayPort and one HDMI output. That design choice keeps multi-GPU setups manageable and avoids hitting OS-level limits, older Windows versions, for example, may have trouble handling more than 32 active display outputs in a single system. The card's most intriguing feature may be its pricing. With single-GPU B60 cards reportedly starting around $375 MSRP, the dual-GPU version could land near $1,000. If that estimate holds, Maxsun's card would represent a major shift in value. For comparison, Nvidia's RTX 6000 Ada, with the same 48GB of VRAM, sells for over $5,500. Two of those cards can push costs north of $18,000. Even so, Intel's performance in professional applications remains an open question. Many creative professionals still favor Nvidia for its mature drivers and better software optimization. These are the best AMD graphics cards you can buy now Take a look at the best mini PCs we've rounded up Rethinking power: how AI is reshaping energy demands in data centers
Yahoo
5 days ago
- Business
- Yahoo
This insane $1,000 Intel GPU could challenge Nvidia's $18,000 monster in workstation memory power
When you buy through links on our articles, Future and its syndication partners may earn a commission. Intel's Arc Pro B60 Dual offers pro-grade memory at a fraction of Nvidia's price This dual-GPU rig from Maxsun delivers workstation power Each GPU gets one DisplayPort and one HDMI, avoiding OS overload in multi-GPU workstations At Computex 2025, Maxsun unveiled a striking new entry in the AI hardware space: the Intel Arc Pro B60 Dual GPU, a graphics card pairing two 24GB B60 chips for a combined 48GB of memory. Servethehomeclaims Maxsun envisions these cards powering dense workstation builds with up to four per system, yielding as much as 192GB of GPU memory in a desktop-class machine. This development appears to have Intel's implicit approval, suggesting the company is looking to gain traction in the AI GPU market. The Arc Pro B60 Dual GPU is not designed for gaming. Instead, it focuses on AI, graphics, and virtualization tasks, offering a power-efficient profile. Each card draws between 240W and 300W, keeping power and thermal demands within reach for standard workstation setups. Unlike some alternatives, this card uses a blower-style cooler rather than a passive solution, helping it remain compatible with conventional workstation designs. That matters for users who want high-end performance without building custom cases or cooling systems. Still, the architecture has trade-offs. The card relies on x8 PCIe lanes per GPU, bifurcated from a x16 connector. This simplifies design and installation but limits bandwidth compared to full x16 cards. Each GPU also includes just one DisplayPort and one HDMI output. That design choice keeps multi-GPU setups manageable and avoids hitting OS-level limits, older Windows versions, for example, may have trouble handling more than 32 active display outputs in a single system. The card's most intriguing feature may be its pricing. With single-GPU B60 cards reportedly starting around $375 MSRP, the dual-GPU version could land near $1,000. If that estimate holds, Maxsun's card would represent a major shift in value. For comparison, Nvidia's RTX 6000 Ada, with the same 48GB of VRAM, sells for over $5,500. Two of those cards can push costs north of $18,000. Even so, Intel's performance in professional applications remains an open question. Many creative professionals still favor Nvidia for its mature drivers and better software optimization. These are the best AMD graphics cards you can buy now Take a look at the best mini PCs we've rounded up Rethinking power: how AI is reshaping energy demands in data centers
Tahawul Tech
5 days ago
- Tahawul Tech
Kingston accelerates storage innovation through PCIe Gen 5 SSDs
Khalil Yazbeck, Business Development Manager – UAE, Kuwait, Qatar and Oman at Kingston Technology, unpacks the evolution of PCIe SSDs and the practical value of Gen 3 to Gen 5 upgrades in conversation with TahawulTech Data-intensive applications, AI-driven workloads, and immersive gaming experiences are reshaping storage performance standards. In this exclusive interview for TahawulTech, Khalil Yazbeck of Kingston Technology explores the transition from PCIe Gen 3 to Gen 5, explains key differences in speed, form factor, and compatibility, and offers practical guidance for users considering their next upgrade. He also outlines Kingston's forward-looking plans in the high-performance storage landscape. Interview Excerpts: Can you walk us through the evolution of PCIe SSDs from Gen 3 to Gen 5 and explain why these generational upgrades are so important when it comes to storage performance? Over the years, PCIe SSDs have advanced considerably with each new generation doubling the available bandwidth and significantly improving read and write speeds. PCIe Gen 3 represented a major improvement over SATA SSDs, offering speeds of around 3,500 MB/s. Gen 4 doubled that to approximately 7,000 MB/s, and now Gen 5 exceeds 14,000 MB/s. These advancements lead to faster load times, quicker file transfers, and better performance in demanding tasks such as 8K video editing, large-scale simulations, high-end gaming, and real-time AI workloads. When it comes to speed and bandwidth, how do the data transfer speeds and bandwidth differ between Gen 3, Gen 4, and Gen 5 SSDs? PCIe Gen 3 SSDs deliver read speeds of up to 3,500 MB/s, with a maximum bandwidth of 16 GB/s. PCIe Gen 4 builds on that foundation, doubling performance to offer speeds of up to 7,000 MB/s and 32 GB/s of bandwidth. PCIe Gen 5 takes it even further, surpassing 14,000 MB/s in speed and reaching a bandwidth of 64 GB/s. At its core, each new generation approximately doubles the available bandwidth, resulting in significantly faster access to large files, reduced latency, and improved responsiveness in performance-intensive environments such as data-heavy applications, gaming, and professional content creation. Are there specific workloads or applications where these differences are most noticeable? The performance differences between PCIe generations become increasingly noticeable depending on the workload. Gen 3 SSDs are more than sufficient for everyday tasks such as web browsing, office applications, and media streaming. Gen 4 SSDs begin to show their advantage in more demanding scenarios like gaming, photo and video editing, and overall productivity, where faster data access can lead to smoother performance. Gen 5 SSDs, however, elevate performance to an entirely new level. They are particularly well-suited for intensive workloads such as professional video production, AI model training, large-scale software development, scientific simulations, and high-end gaming. In these environments, the dramatically increased read/write speeds and bandwidth can significantly reduce load times, accelerate data processing, and enhance overall system responsiveness. Are there differences in form factor between Gen 3, 4, and 5 SSDs that users should be aware of? Most consumers of PCIe SSDs, whether Gen 3, Gen 4, or Gen 5, use the M.2 2280 form factor, the industry-standard size compatible with the vast majority of modern laptops and desktops. While performance significantly improves with each generation, the physical dimensions remain largely unchanged. This consistency in form factor ensures a seamless upgrade path, allowing users to benefit from faster speeds without worrying about compatibility or installation issues. What compatibility factors should users consider before upgrading to Gen 5? Before upgrading to a Gen 5 SSD, users should ensure that their system is fully compatible with PCIe 5.0 technology. This includes verifying that the motherboard supports PCIe 5.0 lanes and features an M.2 slot specifically designed for Gen 5 SSDs. Equally important is the CPU, as only certain processors, such as Intel's 13th and 14th Gen chips or AMD's Ryzen 7000 and 9000 series, are capable of delivering the necessary PCIe 5.0 support. 'Without compatible hardware across both the motherboard and processor, users won't be able to harness the full performance potential of a Gen 5 SSD. In such cases, the drive may operate at reduced speeds, effectively limiting the benefits of the upgrade.' How important is thermal management, especially with higher-gen SSDs? Thermal management is absolutely essential, especially with higher-generation SSDs like Gen 5. These drives operate at significantly higher speeds, which naturally leads to increased heat output. Without effective cooling solutions, performance can throttle, resulting in reduced speed and efficiency during sustained workloads. Kingston addresses this challenge with its FURY Renegade Gen 5 SSD, engineered for optimal thermal performance. It features the Silicon Motion SM2508 controller, built on a 6nm process, which enhances energy efficiency and thermal regulation. A low-power DDR4 DRAM cache reduces overall power consumption, while an independent Buck IC ensures consistent and stable power delivery. Additionally, the 12-layer PCB design improves signal integrity and data reliability. These advanced design elements enable the Kingston FURY Renegade Gen 5 to maintain lower operating temperatures compared to many competing Gen 5 SSDs, ensuring sustained high performance even under demanding conditions. How should users weigh cost vs. performance benefits between Gen 3, Gen 4, or Gen 5 SSDs? When considering the balance between cost and performance, users should evaluate their specific needs and usage scenarios. For everyday computing, casual gaming, and general productivity, Gen 4 SSDs offer the ideal middle ground, delivering excellent performance at a more accessible price point. Gen 5 SSDs, on the other hand, are geared toward power users and professionals who demand top-tier speeds for tasks like 3D rendering, high-resolution video editing, scientific computing, or AI workloads. In these environments, the performance gains can translate directly into time savings and increased efficiency, making the higher investment worthwhile. Ultimately, users should align their choice with how intensively they rely on storage speed, opting for Gen 4 for balanced value, and Gen 5 when maximum performance is a priority. Can you speak to any misconceptions around needing the 'fastest' SSD available? One common misconception is that the fastest SSD is always the best choice. While higher speeds sound appealing, most users won't notice a significant real-world difference between Gen 4 and Gen 5 SSDs during everyday tasks. Unless your workflow involves heavy data transfers, high-resolution video editing, or other resource-intensive applications, a high-performance Gen 4 SSD such as the Kingston FURY Renegade, offers more than enough speed and responsiveness for the vast majority of users. In many cases, investing in the fastest SSD may not deliver proportional benefits relative to its cost. Which generation offers the best value for money today? Gen 4 currently offers the best balance of affordability, performance, and compatibility. It's an ideal choice for gaming, creative workloads, and even enthusiast builds, delivering high-speed performance without the premium cost associated with Gen 5. What practical tips would you give to someone looking to buy an SSD today? First, ensure that your motherboard and CPU are compatible with the generation of SSD you intend to purchase. Compatibility is crucial for unlocking the drive's full performance potential. Second, choose an SSD that aligns with your actual usage needs. Unless you're planning to future-proof your system for upcoming upgrades, there's no need to pay a premium for performance you won't utilize. For most users, a well-balanced Gen 4 SSD provides excellent value. Third, opt for a reputable brand known for reliability, consistent performance, and strong warranty support such as Kingston Technology. Finally, don't overlook thermal management, especially with Gen 4 and Gen 5 SSDs. These higher-performance drives can generate more heat, so proper cooling is essential for maintaining peak performance and longevity. Does Kingston offer any tools or resources to help users make informed purchasing decisions? Kingston provides several helpful resources. Our 'Ask an Expert' service offers personalized recommendations tailored to individual needs. The Kingston website features an easy-to-use configurator and compatibility tool to simplify the selection process. In addition, the company offers detailed datasheets and guides to help users choose the right drive. Where do you see SSD technology and PCIe performance heading? We anticipate continued advancements in sequential speeds, storage capacity, and energy efficiency, driven by ongoing innovation in SSD technology. With PCIe Gen 6 and future generations on the horizon, performance is set to scale even further, paving the way for faster data access, greater multitasking capabilities and improved support for next-gen workloads across both consumer and enterprise applications. Can you share what's next for Kingston in this space? Kingston has just launched our PCIe Gen 5 SSD, marking our biggest step forward in high-performance storage. While we don't have additional PCIe Gen announcements currently, we are looking to expand Gen 5 capacities to stay ahead of user needs across gaming, enterprise, and content creation.
Yahoo
21-05-2025
- Business
- Yahoo
CRDO vs. AVGO: Which AI-focused Semiconductor Stock is the Better Buy?
Credo Technology Group Holding Ltd. CRDO and Broadcom Inc. AVGO are semiconductor companies that specialize in high-speed connectivity solutions essential for AI-driven data centers. The data center market is undergoing rapid transformation as organizations accelerate their digital initiatives, including the adoption of hybrid and multi-cloud architectures, along with advanced technologies such as artificial intelligence (AI) and machine learning (ML). This surge in digital activity has resulted in explosive data growth, placing significant pressure on existing data center infrastructure. Consequently, there is a rising need for advanced data processing and storage solutions to support AI workloads. This trend presents strong growth opportunities for companies like Credo and Broadcom. So, now the question arises: Which stock makes a better investment pick at present? Let's dive into the pros and cons for each company. Credo is gaining a strong market presence in both Ethernet and Active Electrical Cables (AECs) solutions specialized for data centers. AEC products experienced triple-digit sequential growth in the third quarter of fiscal 2025, driven by their increasing adoption in the data center market. The demand for AECs is increasing as ZeroFlap AECs offer more than 100 times improved reliability than laser-based optical solutions. This made AECs an increasingly attractive option for data center applications, contributing to the new expansion of AEC usage and further solidifying Credo Technology's position in the market. Credo is also focused on expanding its product portfolio to include PCIe solutions, which will address the growing demand for AI scale-out and scale-up networks. CRDO's PCIe and Ethernet retimers saw strong customer interest, especially for scale-out networks in AI servers. This growing demand underscores the increasing importance of high-performance solutions in the rapidly expanding AI server market. PCIe retimer demand is expected to exceed $1 billion by 2027, positioning Credo for significant future revenue growth. Momentum in the optical business, particularly Optical Digital Signal Processors (DSPs), bodes well. It recently unveiled the innovative Lark Optical DSP family, engineered to transform 800G optical transceivers. The Lark portfolio has two distinct optical DSP products. The Lark 800 is a high-performance, low-power DSP optimized for fully retimed 800G transceivers, designed to meet the stringent power and cooling requirements of hyperscale AI data centers. The Lark 850 is an ultra-low-power 800G Linear Receive Optics DSP, consuming under 10W, making it an ideal solution for AI-driven data environments where power efficiency is exceptional. For the fourth quarter of fiscal 2025, CRDO expects revenues between $155 million and $165 million. The Zacks Consensus Estimate for fiscal fourth-quarter revenues is pegged at $160 million, suggesting growth of 163.2% from the year-ago quarter's reported figure. Broadcom is a diversified semiconductor company with a strong presence in AI chip development space. AVGO sees massive opportunities in the AI space as its three hyperscaler customers have started to develop their own XPUs. These hyperscalers are significantly ramping up investment in their next-generation frontier models, which do require high-performance accelerators and AI data centers with larger clusters. AVGO noted that by 2027, each of AVGO's three hyperscalers plans to deploy 1 million XPU clusters across a single fabric. Consequently, AVGO's Serviceable Addressable Market for XPUs and networks is expected to be between $60 billion and $90 billion in fiscal 2027 alone. Apart from these three customers, AVGO is also working with four other hyperscalers to aid them in building their own customized AI accelerators. XPUs are necessary to train Generative AI (GenAI) models, and they require complex integration of compute, memory, and I/O capabilities to achieve the necessary performance at lower power consumption and cost. Its next-generation XPUs are in 3 nanometers and will be the first of its kind to market in that process node. AVGO is now planning to tape out the industry's first 2-nanometer AI XPU packaging 3.5D and targets scaling clusters of 500,000 accelerators for hyperscale customers. In the last reported quarter, semiconductor solutions revenues (55% of net revenues) totaled $8.2 billion, rising 11% year over year, driven by strong AI semiconductor revenues. AI-related revenues grew 77% year over year to $4.1 billion, which was a major driver of overall revenue growth. This growth was driven by higher shipments of networking solutions to hyperscalers for AI applications. For second-quarter fiscal 2025, Broadcom expects revenues of $14.9 billion, while AI semiconductor revenues are projected to grow 44% year over year to $4.4 billion. Semiconductor revenues are anticipated to grow 17% year over year to $8.4 billion. Over the past month, CRDO and AVGO have registered gains of 67.8% and 36.3%, respectively. Image Source: Zacks Investment Research In terms of the forward 12-month Price/sales ratio, CRDO is trading at 15.50X, lower than AVGO's 16.05X. Image Source: Zacks Investment Research Analysts have kept their earnings estimates unchanged for CRDO's bottom line in the past 90 days. Image Source: Zacks Investment Research Meanwhile, for AVGO, there is a marginal upward estimate revision. Image Source: Zacks Investment Research Both CRDO and MRVL are well-positioned to gain from the rapidly growing AI driven data center market. CRDO at present flaunts a Zacks Rank #3 (Hold) while Broadcom carries a Zacks Rank #2 (Buy). Consequently, in terms of Zacks Rank, AVGO seems to be a better pick at the moment. You can see the complete list of today's Zacks #1 Rank (Strong Buy) stocks here. Want the latest recommendations from Zacks Investment Research? Today, you can download 7 Best Stocks for the Next 30 Days. Click to get this free report Broadcom Inc. (AVGO) : Free Stock Analysis Report Credo Technology Group Holding Ltd. (CRDO) : Free Stock Analysis Report This article originally published on Zacks Investment Research ( Zacks Investment Research Sign in to access your portfolio
Yahoo
15-05-2025
- Yahoo
Intel Core Ultra 200 CPUs May Bottleneck Gen 5 SSDs
It seems that Intel Core Ultra 200S CPUs have a design issue that can cause some latency when using Gen 5 SSDs. The SSD Review noticed that it wasn't getting anywhere near 14GB/s when plugging SSDs into the M.2 slots of Intel Z890-based motherboards. After significant testing and checking in with Intel, The SSD Review narrowed the suspect list to the CPU—when the SSD is plugged into the M.2 slot. The latency is dramatic: 12GB/s sequential read speeds have been the norm for the SSD reviewer with Core Ultra 200S CPUs in the test bench. Random write speeds are apparently lower than expected, too. Kudos go to Intel for responding to The SSD Review. It pointed out that the die-to-die path on its CPU is a little longer for Gen 5 memory than for other PCIe lanes. 'Intel can confirm that the PCIe Lanes 21 to 24 Gen5 root port on Intel Core Ultra 200S series processors may exhibit increased latencies compared to the PCIe Lanes 1 to 16 Gen5 root ports, owing to a longer die-to-die data path,' Intel noted to The SSD Review. 'However, any variations are contingent upon the specific workload and the capabilities of the PCIe endpoint device.' Credit: Intel As Tom's Hardware notes, Raptor Lake doesn't appear to suffer the same problem. The SSD Review tested the issue with Micron Pro and Samsung 9100 Pro SSDs on its test bench with Arrow Lake and Raptor Lake setups. The different CPU/motherboard combos highlighted the problem: The Raptor Lake setup logged 14.3GB/s, while the Arrow Lake setup landed at 12.3GB/s. That's not what you want to see from newer, more advanced hardware. That's an annoying problem, for sure, but you can circumvent it by picking up a PCIe add-in card instead of an M.2 card. The SSD Review had no problem getting full speeds from the PCIe 5.0 lane. And, although we're generally loathe to take the 'grin and bear it' approach, it may be the easiest path for you if you already have Gen 5 memory in an M.2 slot on an Intel Z890 board. After all, as The SSD Review points out, even at 12GB/s, the setup provides much faster speeds than a rig with Gen 4 speeds. If you have been using your PC without complaint until now, there's really no need to make a change.
