- \n
- Purpose\u2011built silicon platform addressing data\u2011movement bottlenecks as AI inference shifts to real\u2011time, revenue\u2011generating workloads<\/i><\/li>\n<\/ul>\n
CARLSBAD, Calif.–(BUSINESS WIRE<\/a>)–
\nMaxLinear, Inc. (NASDAQ: MXL), a leading provider of high\u2011performance connectivity and data movement solutions, today announced that it is showcasing Panther V, the latest generation of its storage accelerator platform, at Dell Technologies World 2026, May 18\u201321, at The Venetian, Las Vegas (Booth 204).<\/p>\nThis press release features multimedia. View the full release here: https:\/\/www.businesswire.com\/news\/home\/20260505801716\/en\/<\/a><\/p>\n
![\"MaxLinear](\"https:\/\/mms.businesswire.com\/media\/20260505801716\/en\/2794942\/4\/Panther-Scaling.jpg\")
<\/p>\nMaxLinear Panther \u202fV accelerates AI inference efficiency by reducing data\u2011movement bottlenecks in data centers<\/p>\n<\/div>\n
\nPanther V addresses one of the most critical constraints emerging in large-scale AI inference data centers: the cost, latency, and inefficiency of data movement across storage, memory, and compute. As AI workloads transition from experimental pilots to persistent, production-scale inference, system performance is increasingly constrained by how efficiently data is staged, prepared, and activated for inference.<\/p>\n
Optimized for AI Inference and Time\u2011to\u2011First\u2011Token (TTFT)<\/b><\/p>\n
\nPanther V reduces end\u2011to\u2011end latency and improves responsiveness and throughput for modern AI inference by tightly coupling CPU, accelerator, and GPU resources to keep data moving efficiently through the system. Inline execution of data transformation, compression, encryption, and integrity operations eliminates unnecessary CPU involvement and memory round\u2011trips, reducing GPU idle time and accelerating time\u2011to\u2011first\u2011token while freeing up host CPUs to focus on model execution and coordination.<\/p>\n
\nAs agentic inference grows and workloads become increasingly latency\u2011sensitive, Panther\u2011based accelerators enable the same CPUs and GPUs to support many more simultaneous inference agents. This improves utilization, scalability, and overall system efficiency for interactive, real\u2011time AI services.<\/p>\n
Built for Today\u2019s Inference\u2011Dominated Workloads<\/b><\/p>\n
\nAs AI inference becomes the primary driver of production AI deployments, Panther V is purpose\u2011built to support the most demanding inference scenarios, including:<\/p>\n
- \n
- Low\u2011latency inference<\/b>, where fast TTFT is essential for conversational AI and real\u2011time applications\n<\/li>\n
- Retrieval\u2011Augmented Generation (RAG)<\/b>, accelerating document retrieval and preparation from enterprise data stores\n<\/li>\n
- KV\u2011cache\u2011intensive inference<\/b>, enabling reuse of pre-fill\u2011stage key\u2011value data across users and agents without impacting GPU hot\u2011path performance\n<\/li>\n<\/ul>\n
\nBy accelerating compression, decompression, encryption, and integrity validation in silicon, Panther V enables smaller, verified data to move faster through storage, memory, and network fabrics, which improves inference economics without increasing power or infrastructure cost.<\/p>\n
Key Panther V Capabilities<\/b><\/p>\n
\nPanther V combines scalable performance, deep CPU offload, and advanced security and integrity acceleration to enable efficient, high\u2011concurrency AI inference at scale.<\/p>\n
- \n
- Scalable Performance:<\/b> Supports system architectures exceeding 6Tbps, delivering up to 450Gbps per accelerator\n<\/li>\n
- CPU Offload:<\/b> Dedicated hardware engines perform single\u2011pass compression, encryption, and checksum processing entirely in silicon, avoiding multiple PCIe pass\u2011throughs\n<\/li>\n
- Advanced Accelerations:<\/b> GZIP, Zlib, Deflate, XP10, AES encryption (ECB, CBC, CTR, XTS, GCM), and SHA\u20111\/2 hashing and checksums\n<\/li>\n
- Data Integrity:<\/b> Real\u2011time, end\u2011to\u2011end verification with CRC validation and NVMe T10 DIF\/DIX support\n<\/li>\n
- Software Flexibility:<\/b> SDK supporting synchronous and asynchronous APIs, kernel and user space, NUMA\u2011aware queues, and peer\u2011to\u2011peer DMA\n<\/li>\n
- ZFlush\u2122 for OpenZFS:<\/b> A hardware\u2011accelerated OpenZFS implementation that integrates seamlessly with Panther V to improve file\u2011system performance\n<\/li>\n
- Industry\u2011Standard Form Factors:<\/b> Available in PCIe and OCP NIC 3.0 configurations\n<\/li>\n<\/ul>\n
Powering the Economics of AI Data Centers<\/b><\/p>\n
\nThe AI inference market is expanding rapidly, with sustained double\u2011digit growth projected into the early 2030s. As inference becomes persistent and monetized, infrastructure buyers are prioritizing system efficiency, power optimization, and time\u2011to\u2011value over peak compute metrics alone. Panther V enables data center operators to scale inference concurrency, support longer context windows, and deliver faster user experiences without linear increases in cost or power consumption.<\/p>\n
\n\u201cAI inference is rapidly becoming a real\u2011time, revenue\u2011generating workload, and data movement, not compute, is emerging as the primary system bottleneck,\u201d said Vikas Choudhary, SVP & GM of the Connectivity and Storage Business at MaxLinear. \u201cBy accelerating faster node bring\u2011up, growing context sizes, and KV\u2011cache compression, Panther V enables more efficient and low latency inference pipelines along with scalable AI inference economics. We believe that the size of the serviceable market for purpose-built silicon accelerator solutions, such as Panther V, is approximately $5 billion.\u201d<\/p>\n
\nMaxLinear representatives will be on site at Dell Technologies World 2026, May 18\u201321, at Booth 204, to discuss how Panther V supports scalable AI inference, RAG architectures, and next\u2011generation data center platforms<\/p>\n
About MaxLinear, Inc.<\/b><\/p>\n
\nMaxLinear, Inc. (Nasdaq: MXL) is a leading provider of radio frequency (RF), analog, digital, and mixed-signal integrated circuits for access and connectivity, wired and wireless infrastructure, and industrial and multimarket applications. MaxLinear is headquartered in Carlsbad, California. For more information, please visit https:\/\/www.maxlinear.com\/<\/a>.<\/p>\n
\nMaxLinear, the MaxLinear logo, any other MaxLinear trademarks are all property of MaxLinear, Inc. or one of MaxLinear’s subsidiaries in the U.S.A. and other countries. All rights reserved.<\/p>\n
\nAll third-party marks and logos are trademarks or registered trademarks of their respective holders\/owners.<\/p>\n
Cautionary Note About Forward-Looking Statements<\/b><\/p>\n
\nThis press release contains forward-looking statements within the meaning of Section 27A of the Securities Act of 1933, as amended, and Section 21E of the Securities Exchange Act of 1934, as amended. Forward-looking statements include, among others, statements relating to MaxLinear\u2019s Panther V product and the functionality, performance and benefits of such product, statements about the potential market opportunity for Panther V; the potential growth of the AI inference market; the serviceable market for purpose-built silicon accelerator solutions; and statements by MaxLinear\u2019s SVP & GM, Connectivity and Storage Business Unit. These forward-looking statements involve known and unknown risks, uncertainties, and other factors that may cause actual results to be materially different from any future results expressed or implied by the forward-looking statements and our future financial performance and operating results forecasts generally. Forward-looking statements are based on management\u2019s current, preliminary expectations and are subject to various risks and uncertainties. In particular, our future operating results are substantially dependent on our assumptions about market trends and conditions. Additional risks and uncertainties affecting our business, future operating results and financial condition include, without limitation; risks relating to: the development, testing, and commercial introduction of new products and product functionalities; risks related to the market opportunity for Panther V not developing or MaxLinear not being able to capture share of the market; the capabilities of our technology; our terminated merger with Silicon Motion and related arbitration and class action complaint and the risks related to potential payment of damages; the effect of intense and increasing competition; increased tariffs, export controls or imposition of additional trade barriers; impacts of global economic conditions; the cyclical nature of the semiconductor industry; a significant variance in our operating results and impact on volatility in our stock price, and our ability to sustain our current level of revenue, which has previously declined, and\/or manage future growth effectively, and the impact of excess inventory in the channel on our customers\u2019 expected demand for certain of our products and on our revenue; escalating trade wars, military conflicts and other geopolitical and economic tensions among the countries in which we conduct business; international geopolitical and military conflicts; our ability to obtain or retain government authorization to export certain of our products or technology; the loss of, or a significant reduction in orders from major customers; legal proceedings or potential violations of regulations; information technology failures; a decrease in the average selling prices of our products; failure to penetrate new applications and markets; development delays and consolidation trends in our industry; inability to make substantial and productive research and development investments; delays or expenses caused by undetected defects or bugs in our products; substantial quarterly and annual fluctuations in our revenue and operating results; failure to timely develop and introduce new or enhanced products; order and shipment uncertainties and differences between our estimates of customer demand and product mix and our actual results; failure to accurately predict our future revenue and appropriately budget expenses; lengthy and expensive customer qualification processes; customer product plan cancellations; failure to maintain compliance with government regulations; failure to attract and retain qualified personnel; any adverse impact of rising interest rates on us, our customers, and our distributors and related demand; risks related to compliance with privacy, data protection and cybersecurity laws and regulations; risks related to conforming our products to industry standards; risks related to business acquisitions and investments; claims of intellectual property infringement; our ability to protect our intellectual property; security vulnerabilities of our products; use of open source software in our products; failure to manage our relationships with, or negative impacts from, third parties; and future decisions relating to our stock repurchase program.<\/p>\n
\nIn addition to these risks and uncertainties, investors should review the risks and uncertainties contained in our filings with the Securities and Exchange Commission, including our Current Reports on Form 8-K, as well as the information to be set forth under the caption “Risk Factors” in MaxLinear’s Quarterly Report on Form 10-Q for the quarter ended March 31, 2026. All forward-looking statements are based on the estimates, projections and assumptions of management as of the date of this press release, and MaxLinear is under no obligation (and expressly disclaims any such obligation) to update or revise any forward-looking statements whether as a result of new information, future events, or otherwise.<\/p>\n
<\/p>\n
View source version on businesswire.com: <\/span>https:\/\/www.businesswire.com\/news\/home\/20260505801716\/en\/<\/a><\/span><\/p>\n
MaxLinear Press Contact:
\n<\/b>
Debbie Brandenburg
\n
Sr. Marketing Communications Manager
\n
Tel: +1 669.265.6083
\n
dbrandenburg@maxlinear.com<\/a><\/p>\nKEYWORDS:<\/b> California Nevada United States North America<\/p>\n
INDUSTRY KEYWORDS:<\/b> Data Management Technology Other Technology Networks Artificial Intelligence Internet Hardware<\/p>\n
- CPU Offload:<\/b> Dedicated hardware engines perform single\u2011pass compression, encryption, and checksum processing entirely in silicon, avoiding multiple PCIe pass\u2011throughs\n<\/li>\n
- Retrieval\u2011Augmented Generation (RAG)<\/b>, accelerating document retrieval and preparation from enterprise data stores\n<\/li>\n
- Low\u2011latency inference<\/b>, where fast TTFT is essential for conversational AI and real\u2011time applications\n<\/li>\n
![\"Photo\"](\"https:\/\/mms.businesswire.com\/media\/20260505801716\/en\/2794942\/3\/Panther-Scaling.jpg\")
<\/td>\n<\/tr>\n![\"Logo\"](\"https:\/\/mms.businesswire.com\/media\/20260505801716\/en\/2734491\/3\/MaxLinearR_Black-Red_Logo.jpg\"){kind=logo}
<\/td>\n<\/tr>\n