logo
Mullen Group Ltd. Acquisitions Drive Growth in the Second Quarter of 2025

Mullen Group Ltd. Acquisitions Drive Growth in the Second Quarter of 2025

OKOTOKS, Alberta, July 24, 2025 (GLOBE NEWSWIRE) — (TSX: MTL) Mullen Group Ltd. ('Mullen Group', 'We', 'Our' and/or the 'Corporation'), one of Canada's largest logistics providers today reported its financial and operating results for the period ended June 30, 2025, with comparisons to the same period last year. Full details of the results may be found within our Second Quarter Interim Report, which is available on the Corporation's issuer profile on SEDAR+ at
www.sedarplus.ca
or at
www.mullen-group.com
.
'
It was a very active and productive quarter from a corporate office perspective. We completed a very successful, oversubscribed, long term bond issue ensuring the balance sheet is prudently structured well into the next decade. In addition to raising the funds required to repay the 2026 expiring notes and our current bank indebtedness, we raised additional funds to support the next phase of our acquisition strategy. We also finalized the Cole Group acquisition on June 1, 2025, a quality organization that I believe will be a tremendous addition to the Mullen Group portfolio of companies. Both initiatives were completed despite the uncertainties permeating throughout the economy and the logistics industry. Being bold when others cannot and planning for the future has always been one of the hallmarks of our success. I want to thank all of those that believed in our story, and to those that helped get these initiatives completed
,' commented Mr. Mullen K. Mullen, Chair and Senior Executive Officer.
'
Second quarter consolidated revenues were positively impacted by acquisitions which, as we have suggested for some time now, is the only plausible means of growing during uncertain times and given the current supply/demand imbalance situation in nearly all segments of the logistics industry. Overall freight demand remains solid but thus far there has been no sustained economic growth to generate additional demand, a necessity in our view to alter the current excess industry supply situation. As it currently stands, shippers hold the balance of power when it comes to pricing, making it difficult to grow profitably or to improve margins. Our forty-one Business Units faced difficult discussions with many customers last quarter, that were also addressing pricing pressures from their customers. This is why I am actually pleased with the results of our core Business Units last quarter, holding revenues and margins close to last year despite the challenging market
.'
Mr. Mullen continued, '
There is no doubt that the current market conditions remain uncertain. The tariff/trade debate has not been resolved and this negatively impacts capital investment, a necessary ingredient in our view to increasing demand and stabilizing industry pricing. However, there are early signs that provide a glimmer of optimism. There appears to be momentum building to approve 'nation building' projects in Canada. These projects would be capital intensive and would create high paying jobs, both ingredients we believe to be positive for the economy. In addition, it is only a matter of time before the freight markets rebalance, providing service providers and carriers with more leverage to negotiate fair pricing with shippers. Our current strategy is to add acquisitions that strengthen the market position of our Business Units and to continue to protect margins. Once the markets tighten, we will pivot from protecting margins to improving margins
.'
Second Quarter Highlights
1
Refer to the section entitled 'Non-IFRS Financial Measures'.
Second Quarter Commentary
Revenue: Increased by $45.3 million or 9.1 percent to $540.9 million, led by higher revenue in the L&W, US 3PL and LTL segments being somewhat offset by lower revenue in the S&I segment.
OIBDA - adjusted1: Generated $83.8 million of OIBDA - adjusted1, a slight decrease of $1.8 million, or 2.1 percent. OIBDA was $76.6 million, down 10.6 percent led by declines in the S&I and LTL segments along with higher Corporate costs, which were somewhat offset by higher OIBDA in the L&W and US 3PL segments.
Net income: Net income decreased by $7.3 million, or 22.2 percent to $25.6 million, or $0.29 per Common Share due to:
1
Refer to
the
sections entitled 'Non-IFRS Financial Measures' and 'Other Financial Measures'.
Financial Position
The following summarizes our financial position as at June 30, 2025, along with some key changes that occurred subsequent to the end of the second quarter:
1
Refer to
the
section entitled 'Other Financial Measures'.
Non-IFRS Financial Measures
Mullen Group reports its financial results in accordance with International Financial Reporting Standards ('IFRS'). Mullen Group reports on certain non-IFRS financial measures and ratios, which do not have a standard meaning under IFRS and, therefore, may not be comparable to similar measures presented by other issuers. Management uses these non-IFRS financial measures and ratios in its evaluation of performance and believes these are useful supplementary measures. We provide shareholders and potential investors with certain non-IFRS financial measures and ratios to evaluate our ability to fund our operations and provide information regarding liquidity. Specifically, net income - adjusted, earnings per share - adjusted, and OIBDA - adjusted are not measures recognized by IFRS and do not have standardized meanings prescribed by IFRS. For the reader's reference, the definition, calculation and reconciliation of non-IFRS financial measures are provided in this section. These non-IFRS financial measures should not be considered in isolation or as a substitute for measures prepared in accordance with IFRS. Investors are cautioned that these indicators should not replace the forgoing IFRS terms: net income, earnings per share, and revenue.
Net Income – Adjusted and Earnings per Share – Adjusted
The following table illustrates net income and basic earnings per share before considering the impact of the net foreign exchange gains or losses, and the change in fair value of investments. Management adjusts net income and earnings per share by excluding these specific factors to more clearly reflect earnings from an operating perspective.
OIBDA - Adjusted
OIBDA - adjusted is calculated by subtracting foreign exchange gains and losses recognized on U.S. denominated cash held with the Corporate Office from OIBDA. Management relies on OIBDA - adjusted as a measurement since it provides an indication of Mullen Group's ability to generate cash from its principal business activities prior to depreciation and amortization, financing, taxation in various jurisdictions and gains and losses recognized on U.S. cash held within the Corporate Office. Net income is also an indicator of financial performance, however, net income includes expenses that are not a direct result of Mullen Group's operating activities.
Other Financial Measures
Other financial measures consist of supplementary financial measures and capital management measures.
Supplementary Financial Measures
Supplementary financial measures are financial measures disclosed by a company that (a) are, or are intended to be, disclosed on a periodic basis to depict the historical or expected future financial performance, financial position or cash flow of a company, (b) are not disclosed in the financial statements of a company, (c) are not non-IFRS financial measures, and (d) are not non-IFRS ratios. The following are supplementary financial measures disclosed by the Corporation.
Operating Margin
Operating margin is a supplementary financial measure and is defined as OIBDA divided by revenue. Management relies on operating margin as a measurement since it provides an indication of our ability to generate an appropriate return as compared to the associated risk and the amount of assets employed within our principal business activities.
OIBDA -Adjusted
1
as a Percentage of Consolidated Revenue
OIBDA - adjusted1 as a percentage of consolidated revenue is a supplementary financial measure and is defined as OIBDA -adjusted1 divided by revenue. Management relies on this adjusted operating margin as a measurement since it provides an indication of our ability to generate an appropriate return from our principal business activities prior to depreciation and amortization, financing, taxation in various jurisdictions and gains and losses recognized on U.S. cash held within Corporate Office as compared to the associated risk of our principal business activities.
Capital Management Measures
Capital management measures are financial measures disclosed by a company that (a) are intended to enable users to evaluate a company's objectives, policies and processes for managing the entity's capital, (b) are not a component of a line item disclosed in the primary financial statements of the company, (c) are disclosed in the notes of the financial statements of the company, and (d) are not disclosed in the primary financial statements of the company. The Corporation has disclosed the following capital management measure.
1
Refer to the sections entitled 'Non-IFRS Financial Measures'
.
Total Net Debt – 2024 Notes Calculation
The term '2024
total net debt
' is defined in the 2024 Notes agreement as all debt including the Debentures, the Private Placement Debt, lease liabilities associated with operating equipment, the Bank Credit Facilities and letters of credit less any unrealized gain on Cross-Currency Swaps plus any unrealized loss on Cross-Currency Swaps, as disclosed within Derivatives on the condensed consolidated statement of financial position. 2024 total net debt specifically excludes any real property lease liabilities. 2024 total net debt is defined within our 2024 Notes agreement and is used to calculate our 2024 total net debt to 2024 operating cash flow covenant. Management calculates and discloses 2024 total net debt to provide users with an understanding of how our debt covenant is calculated.
About Mullen Group Ltd.
Mullen Group is one of Canada's largest logistics providers. Our network of independently operated businesses provide a wide range of service offerings including less-than-truckload, truckload, warehousing, customs and logistics, transload, oversized, third-party logistics and specialized hauling transportation. In addition, we provide a diverse set of specialized services related to the energy, mining, forestry and construction industries in western Canada, including water management, fluid hauling and environmental reclamation. The corporate office provides the capital and financial expertise, legal support, technology and systems support, shared services and strategic planning to its independent businesses.
Mullen Group is a publicly traded corporation listed on the Toronto Stock Exchange under the symbol '
MTL
'. Additional information is available on our website at
www.mullen-group.com
or on the Corporation's issuer profile on SEDAR+ at
www.sedarplus.ca
.
Contact Information
Mr. Murray K. Mullen - Chair, Senior Executive Officer and President
Mr. Richard J. Maloney - Senior Operating Officer
Mr. Carson P. Urlacher - Senior Financial Officer
Ms. Joanna K. Scott - Senior Corporate Officer
121A - 31 Southridge Drive
Okotoks, Alberta, Canada T1S 2N3
Telephone: 403-995-5200
Fax: 403-995-5296
Disclaimer
Mullen Group may make statements in this news release that reflect its current beliefs and assumptions and are based on information currently available to it and contains forward-looking statements and forward-looking information (collectively, 'forward-looking statements') within the meaning of applicable securities laws. This news release may contain forward-looking statements that are subject to risk factors associated with the overall economy, Mullen Group's strategy, and the natural resources industry. These forward-looking statements relate to future events and Mullen Group's future performance. All forward looking statements and information contained herein that are not clearly historical in nature constitute forward-looking statements, and the words 'may', 'will', 'should', 'could', 'expect', 'plan', 'intend', 'anticipate', 'believe', 'estimate', 'propose', 'predict', 'potential', 'continue', 'aim', or the negative of these terms or other comparable terminology are generally intended to identify forward-looking statements. Such forward-looking statements represent Mullen Group's internal projections, estimates, expectations, beliefs, plans, objectives, assumptions, intentions or statements about future events or performance. These forward-looking statements involve known or unknown risks, uncertainties and other factors that may cause actual results or events to differ materially from those anticipated in such forward-looking statements. Mullen Group believes that the expectations reflected in these forward-looking statements are reasonable; however, undue reliance should not be placed on these forward-looking statements, as there can be no assurance that the plans, intentions or expectations upon which they are based will occur. In particular, forward-looking statements include but are not limited to the following: (i) Mullen Group's view that the current market conditions remain uncertain; (ii) Mullen Group's belief that there are early signs that provide a glimmer of optimism; (iii) Mullen Group's view that it is only a matter of time before the freight markets rebalance, providing service providers and carriers with more leverage to negotiate fair pricing with shippers; (iv) Mullen Group's current strategy is to add acquisitions that strengthen the market position of our Business Units and to continue to protect margins
;
and (v) that once the markets tighten, Mullen Group will pivot from protecting margins to improving margins.
These forward-looking statements are based on certain assumptions and analyses made by Mullen Group in light of our experience and our perception of historical trends, current conditions, expected future developments and other factors we believe are appropriate under the circumstances. These assumptions include but are not limited to the following: (i) the tariff/trade debate has not been resolved and this negatively impacts capital investment, a necessary ingredient in our view to increasing demand and stabilizing industry pricing; (ii) there appears to be momentum building to approve 'nation building' projects in Canada. These projects would be capital intensive and would create high paying jobs, both ingredients we believe to be positive for the economy; (iii) that if freight markets rebalance, there will be more leverage for providers and carriers to negotiate fair pricing with shippers; (iv) that acquisition opportunities will present themselves to Mullen Group; (v) that we have the balance sheet to execute acquisitions; (vi) that Mullen Group will continue to work to protect margins; and (vii) that the market will tighten in the future and that when that occurs, Mullen Group will pivot from protecting margins to improving margins. For further information on any strategic, financial, operational and other outlook on Mullen Group's business please refer to Mullen Group's Management's Discussion and Analysis dated July 23, 2025, available for viewing on Mullen Group's issuer profile on SEDAR+ at
www.sedarplus.ca
('
Interim MD&A
'). Additional information on risks that could affect the operations or financial results of Mullen Group may be found under the heading 'Principal Risks and Uncertainties' starting on page 48 of the 2024 Annual Financial Review as well as in reports on file with applicable securities regulatory authorities and may be accessed through Mullen Group's issuer profile on the SEDAR+ website at
www.sedarplus.ca
. All capitalized terms used in this news release and not defined herein have the meaning ascribed to them in the Interim MD&A. The forward-looking statements contained in this news release are expressly qualified by this cautionary statement. The forward-looking statements contained herein is made as of the date of this news release and Mullen Group disclaims any intent or obligation to update publicly any such forward-looking statements, whether as a result of new information, future events or results or otherwise, other than as required by applicable Canadian securities laws. Mullen Group relies on litigation protection for forward-looking statements.
PDF available:
http://ml.globenewswire.com/Resource/Download/fb5f76a1-35da-484a-ad5d-10816d7dcc56
Orange background

Try Our AI Features

Explore what Daily8 AI can do for you:

Comments

No comments yet...

Related Articles

M37Labs Unveils EBIC.AI: Revolutionizing Brand Intelligence with Adaptive AI
M37Labs Unveils EBIC.AI: Revolutionizing Brand Intelligence with Adaptive AI

Business Upturn

time4 hours ago

  • Business Upturn

M37Labs Unveils EBIC.AI: Revolutionizing Brand Intelligence with Adaptive AI

New York, NY , Aug. 04, 2025 (GLOBE NEWSWIRE) — M37Labs, a pioneering deep tech AI company, today announced the commercial launch of (Enterprise Brand Intelligence Console)3. This platform is described as the world's first truly adaptive and self-evolving AI brand intelligence platform, marking a significant advancement in enterprise brand and reputation management. aims to address a critical $100 billion market opportunity by enabling Enterprise PR Comms and Brand teams to leverage AI for exponential operational scaling. The platform provides a unified console for all brand, PR, and reputation activities, designed to alert C-Suite executives on potential crises, evolving narratives, and stakeholder engagement. ' is designed to be an intelligent co-pilot to every enterprise comms and PR/Brand Consulting Founder,' stated Prashant Shivram Iyer, Co-Founder & CEO of M37Labs AI. He further added that the platform's agile, adaptive, and self-evolving nature, powered by state-of-the-art AI frameworks and Reinforcement Machine Learning (RML), will enable communications professionals to significantly enhance their potential, driving growth in enterprise comms teams through augmented human-AI creativity, collaboration, and productivity. Revolutionary Technology Delivers Unprecedented Capabilities proprietary technology stack integrates four key components: Real Time omnichannel brand monitoring: Enterprises and Agencies can customise their portfolio of brands into one single dashboard with a contextual AI co-pilot that summarizes and analyses all developments to help teams strategize faster on next action steps Enterprises and Agencies can customise their portfolio of brands into one single dashboard with a contextual AI co-pilot that summarizes and analyses all developments to help teams strategize faster on next action steps Real Time Crisis Detector with Predictive Analytics: Unlike traditional tools, advanced AI algorithms analyse millions of news sources and digital data points to predict potential reputation threats, providing crucial time for preventative strategies. Unlike traditional tools, advanced AI algorithms analyse millions of news sources and digital data points to predict potential reputation threats, providing crucial time for preventative strategies. Strategic Narrative kid with AI-Human Collaboration: Utilizing a unique reinforcement learning framework, combines AI with human strategic thinking to generate compelling brand narratives that resonate with target audiences while preserving authentic brand voice and values. Utilizing a unique reinforcement learning framework, combines AI with human strategic thinking to generate compelling brand narratives that resonate with target audiences while preserving authentic brand voice and values. Intelligence Hub with Competitive Advantage Mapping: This feature offers real-time competitive intelligence, extending beyond basic mention tracking to deliver strategic insights into competitor positioning, market share shifts, and emerging opportunities, thereby helping enterprises maintain a competitive edge. Zorawar Purohit, CAIO & Co-founder of M37Labs, commented, ' doesn't just monitor your brand – it becomes an intelligent extension of your strategic team , learning from every interaction and continuously improving its ability to protect and enhance your brand value. This is the future of enterprise brand management, and we're proud to be leading this transformation.' Strategic Market Positioning for Global Expansion M37Labs is positioned at the forefront of the rapidly expanding Enterprise AI marketing vertical sector, which is projected to grow from USD $10.2 billion in 2024 to USD $69.6 billion by 2034, reflecting a CAGR of 21.6%. The platform's unique combination of predictive analytics, competitive intelligence, and strategic narrative generation addresses current market gaps where most solutions focus on single-point services rather than comprehensive brand intelligence. The company's strategic approach includes global expansion through partnerships with leading consulting firms and technology integrators for rapid deployment across international markets. cloud-native architecture and multi-language capabilities facilitate immediate adoption by multinational enterprises seeking unified AI brand intelligence across diverse geographic markets. Industry Recognition and Strategic Partnerships launch has garnered significant attention from industry analysts and marketing leaders, with several major PR/consulting firms already integrating the platform into their enterprise workflows. M37Labs' innovative approach to AI-human collaboration is recognized as a breakthrough in enterprise AI adoption, effectively addressing the challenge of maintaining human strategic oversight while leveraging AI capabilities. The company has also established strategic partnerships with leading cloud infrastructure providers to ensure enterprise-grade security, compliance, and scalability, meeting the stringent requirements of global corporations in regulated industries such as financial services, healthcare, and telecommunications. About M37Labs: Pioneering Enterprise AI Transformation M37Labs is a deep tech AI firm founded on the principle that artificial intelligence should augment human strategic thinking rather than replace it. The company focuses on delivering transformative business outcomes for enterprise clients through a comprehensive approach that includes advanced AI design thinking, rapid prototyping, and go-to-market implementation for end-to-end digital transformation across key industries. M37Labs' AIgnition workshops are a gold standard for C-suite AI strategy development, providing insights for practical, goal-aligned AI adoption roadmaps. With operations in the United States and India, M37Labs combines strategic foresight with deep technical expertise to help organizations maintain a competitive advantage in an AI-first business environment. The company's team of seasoned experts brings decades of experience in artificial intelligence, enterprise software, and strategic consulting, enabling solutions that drive measurable improvements in operational efficiency and innovation capacity. Availability and Implementation is immediately available for enterprise deployment via M37Labs' direct sales channel and authorized partner network. The platform offers flexible implementation options, including cloud-based deployment, on-premises installation, and hybrid configurations to meet diverse enterprise requirements. For more information about and M37Labs' comprehensive AI transformation services, please visit or About Forward-Looking Statements This press release contains forward-looking statements regarding market projections, technology capabilities, and business performance. Actual results may vary based on market conditions, competitive factors, and other business considerations. Media ContactSimran Arora [email protected] CONCEPT PUBLIC RELATIONS

Maximize Your Gains with 100x Leverage, Double Deposit Bonus, and No KYC at BexBack
Maximize Your Gains with 100x Leverage, Double Deposit Bonus, and No KYC at BexBack

Business Upturn

time4 hours ago

  • Business Upturn

Maximize Your Gains with 100x Leverage, Double Deposit Bonus, and No KYC at BexBack

By GlobeNewswire Published on August 4, 2025, 22:56 IST SINGAPORE, Aug. 04, 2025 (GLOBE NEWSWIRE) — As Bitcoin continues to hit new all-time highs and the bull market returns, BexBack is helping traders build wealth quickly with a range of limited-time offers: 100x leverage, a double deposit bonus(Deposit 1 BTC and receive 2 BTC. The bonus can only be used as margin and cannot cover losses), and no KYC required, enabling traders to easily maximize their gains in this volatile market. Advantages of 100x Leverage Crypto Futures Amplified Profits: Control large positions with a small amount of capital, capturing more profits from market fluctuations. Low Capital Requirement: Participate in high-value trades with minimal investment, lowering the entry barrier. Increased Market Opportunities: Profit quickly from price fluctuations, especially in volatile markets. High Capital Efficiency: Leverage enables better use of your capital, expanding your investment potential. Profit from Both Up and Down Markets: Adapt to any market conditions, with opportunities to profit whether the market goes up or down. What Is 100x Leverage and How Does It Work? Simply put, 100x leverage allows you to open larger trading positions with less capital. For example: Suppose the Bitcoin price is $100,000 that day, and you open a long contract with 1 BTC. After using 100x leverage, the transaction amount is equivalent to 100 BTC. One day later, if the price rises to $105,000, your profit will be (105,000 – 100,000) * 100 BTC / 100,000 = 5 BTC, a yield of up to 500%. With BexBack's deposit bonus BexBack offers a 100% deposit bonus. If the initial investment is 2 BTC, the profit will increase to 10 BTC, and the return on investment will double to 1000%. Note: Although leveraged trading can magnify profits, you also need to be wary of liquidation risks. How Does the 100% Deposit Bonus Work? The deposit bonus from BexBack cannot be directly withdrawn but can be used to open larger positions and increase potential profits. Additionally, during significant market fluctuations, the bonus can serve as extra margin, effectively reducing the risk of liquidation. About BexBack? BexBack is a leading cryptocurrency derivatives platform offering up to 100x leverage on futures contracts for BTC, ETH, ADA, SOL, XRP, and over 50 other digital assets. Headquartered in Singapore, the platform also operates offices in Hong Kong, Japan, the United States, the United Kingdom, and Argentina. Like many top-tier exchanges, BexBack holds a U.S. MSB (Money Services Business) license and is trusted by more than 500,000 traders worldwide. The platform accepts users from the United States, Canada, and Europe, with zero deposit fees and 24/7 multilingual customer support, delivering a secure, efficient, and user-friendly trading experience. Website: Contact: [email protected] Contact:Amanda [email protected] Disclaimer: This content is provided by Bexback. The statements, views, and opinions expressed in this content are solely those of the content provider and do not necessarily reflect the views of this media platform or its publisher. We do not endorse, verify, or guarantee the accuracy, completeness, or reliability of any information presented. We do not guarantee any claims, statements, or promises made in this article. This content is for informational purposes only and should not be considered financial, investment, or trading in crypto and mining-related opportunities involves significant risks, including the potential loss of capital. It is possible to lose all your capital. These products may not be suitable for everyone, and you should ensure that you understand the risks involved. Seek independent advice if necessary. Speculate only with funds that you can afford to lose. Readers are strongly encouraged to conduct their own research and consult with a qualified financial advisor before making any investment decisions. However, due to the inherently speculative nature of the blockchain sector—including cryptocurrency, NFTs, and mining—complete accuracy cannot always be the media platform nor the publisher shall be held responsible for any fraudulent activities, misrepresentations, or financial losses arising from the content of this press release. In the event of any legal claims or charges against this article, we accept no liability or does not endorse any content on this page. Legal Disclaimer: This media platform provides the content of this article on an 'as-is' basis, without any warranties or representations of any kind, express or implied. We assume no responsibility for any inaccuracies, errors, or omissions. We do not assume any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information presented herein. Any concerns, complaints, or copyright issues related to this article should be directed to the content provider mentioned above. Photos accompanying this announcement are available at Disclaimer: The above press release comes to you under an arrangement with GlobeNewswire. Business Upturn takes no editorial responsibility for the same. Ahmedabad Plane Crash GlobeNewswire provides press release distribution services globally, with substantial operations in North America and Europe.

UPDATE – New MLPerf Storage v2.0 Benchmark Results Demonstrate the Critical Role of Storage Performance in AI Training Systems
UPDATE – New MLPerf Storage v2.0 Benchmark Results Demonstrate the Critical Role of Storage Performance in AI Training Systems

Business Upturn

time4 hours ago

  • Business Upturn

UPDATE – New MLPerf Storage v2.0 Benchmark Results Demonstrate the Critical Role of Storage Performance in AI Training Systems

SAN FRANCISCO, Aug. 04, 2025 (GLOBE NEWSWIRE) — Today, MLCommons® announced results for its industry-standard MLPerf® Storage v2.0 benchmark suite, which is designed to measure the performance of storage systems for machine learning (ML) workloads in an architecture-neutral, representative, and reproducible manner. This round of the benchmark saw dramatically increased participation, more geographic representation from submitting organizations, and greater diversity of the systems submitted for testing. The benchmark results show that storage systems performance continues to improve rapidly, with tested systems serving roughly twice the number of accelerators than in the v1.0 benchmark round. Additionally, the v2.0 benchmark adds new tests that replicate real-world checkpointing for AI training systems. The benchmark results provide essential information for stakeholders who need to configure the frequency of checkpoints to optimize for high performance – particularly at scale. Version 2.0 adds checkpointing tasks, delivers essential insights As AI training systems have continued to scale up to billions and even trillions of parameters, and the largest clusters of processors have reached one hundred thousand accelerators or more, system failures have become a prominent technical challenge. Because data centers tend to run accelerators at near-maximum utilization for their entire lifecycle, both the accelerators themselves and the supporting hardware (power supplies, memory, cooling systems, etc.) are heavily burdened, minimizing their expected lifetime. This is a chronic issue, especially in large clusters: if the mean time to failure for an accelerator is 50,000 hours, then a 100,000-accelerator cluster running for extended periods at full utilization will likely experience a failure every half-hour. A cluster with one million accelerators would expect to see a failure every three minutes. Worse, because AI training usually involves massively parallel computation where all the accelerators are moving in lockstep on the same iteration of training, a failure of one processor can grind an entire cluster to a halt. It is now broadly accepted that saving checkpoints of intermediate training results at regular intervals is essential to keep AI training systems running at high performance. The AI training community has developed mathematical models that can optimize cluster performance and utilization by trading off the overhead of regular checkpoints against the expected frequency and cost of failure recovery (rolling back the computation, restoring the most recent checkpoint, restarting the training from that point, and duplicating the lost work). Those models, however, require accurate data on the scale and performance of the storage systems that are used to implement the checkpointing system. The MLPerf Storage v2.0 checkpoint benchmark tests provide precisely that data, and the results from this round suggest that stakeholders procuring AI training systems need to carefully consider the performance of the storage systems they buy, to ensure that they can store and retrieve a cluster's checkpoints without slowing the system down to an unacceptable level. For a deeper understanding of the issues around storage systems and checkpointing, as well as of the design of the checkpointing benchmarks, we encourage you to read this post from Wes Vaske, a member of the MLPerf Storage working group. 'At the scale of computation being implemented for training large AI models, regular component failures are simply a fact of life,' said Curtis Anderson, MLPerf Storage working group co-chair. 'Checkpointing is now a standard practice in these systems to mitigate failures, and we are proud to be providing critical benchmark data on storage systems to allow stakeholders to optimize their training performance. This initial round of checkpoint benchmark results shows us that current storage systems offer a wide range of performance specifications, and not all systems are well-matched to every checkpointing scenario. It also highlights the critical role of software frameworks such as PyTorch and TensorFlow in coordinating training, checkpointing, and failure recovery, as well as some opportunities for enhancing those frameworks to further improve overall system performance.' Workload benchmarks show rapid innovation in support of larger-scale training systems Continuing from the v1.0 benchmark suite, the v2.0 suite measures storage performance in a diverse set of ML training scenarios. It emulates the storage demands across several scenarios and system configurations covering a range of accelerators, models, and workloads. By simulating the accelerators' 'think time' the benchmark can generate accurate storage patterns without the need to run the actual training, making it more accessible to all. The benchmark focuses the test on a given storage system's ability to keep pace, as it requires the simulated accelerators to maintain a required level of utilization. The v2.0 results show that submitted storage systems have substantially increased the number of accelerators they can simultaneously support, roughly twice the number compared to the systems in the v1.0 benchmark. 'Everything is scaling up: models, parameters, training datasets, clusters, and accelerators. It's no surprise to see that storage system providers are innovating to support ever larger scale systems,' said Oana Balmau, MLPerf Storage working group co-chair. The v2.0 submissions also included a much more diverse set of technical approaches to delivering high-performance storage for AI training, including: 6 local storage solutions; 2 solutions using in-storage accelerators; 13 software-defined solutions; 12 block systems; 16 on-prem shared storage solutions; 2 object stores. 'Necessity continues to be the mother of invention: faced with the need to deliver storage solutions that are both high-performance and at unprecedented scale, the technical community has stepped up once again and is innovating at a furious pace,' said Balmau. MLPerf Storage v2.0: skyrocketing participation and diversity of submitters The MLPerf Storage benchmark was created through a collaborative engineering process by 35 leading storage solution providers and academic research groups across 3 years. The open-source and peer-reviewed benchmark suite offers a level playing field for competition that drives innovation, performance, and energy efficiency for the entire industry. It also provides critical technical information for customers who are procuring and tuning AI training systems. The v2.0 benchmark results, from a broad set of technology providers, reflect the industry's recognition of the importance of high-performance storage solutions. MLPerf Storage v2.0 includes >200 performance results from 26 submitting organizations: Alluxio, Argonne National Lab, DDN, ExponTech, FarmGPU, H3C, Hammerspace, HPE, JNIST/Huawei, Juicedata, Kingston, KIOXIA, Lightbits Labs, MangoBoost, Micron, Nutanix, Oracle, Quanta Computer, Samsung, Sandisk, Simplyblock, TTA, UBIX, IBM, WDC, and YanRong. The submitters represent seven different countries, demonstrating the value of the MLPerf Storage benchmark to the global community of stakeholders. 'The MLPerf Storage benchmark has set new records for an MLPerf benchmark, both for the number of organizations participating and the total number of submissions,' said David Kanter, Head of MLPerf at MLCommons. The AI community clearly sees the importance of our work in publishing accurate, reliable, unbiased performance data on storage systems, and it has stepped up globally to be a part of it. I would especially like to welcome first-time submitters Alluxio, ExponTech, FarmGPU, H3C, Kingston, KIOXIA, Oracle, Quanta Cloud Technology, Samsung, Sandisk, TTA, UBIX, IBM, and WDC.' 'This level of participation is a game-changer for benchmarking: it enables us to openly publish more accurate and more representative data on real-world systems,' Kanter continued. 'That, in turn, gives the stakeholders on the front lines the information and tools they need to succeed at their jobs. The checkpoint benchmark results are an excellent case in point: now that we can measure checkpoint performance, we can think about optimizing it.' We invite stakeholders to join the MLPerf Storage working group and help us continue to evolve the benchmark suite. View the Results To view the results for MLPerf Storage v2.0, please visit the Storage benchmark results . About MLCommons MLCommons is the world's leader in AI benchmarking. An open engineering consortium supported by over 125 members and affiliates, MLCommons has a proven record of bringing together academia, industry, and civil society to measure and improve AI. The foundation for MLCommons began with the MLPerf benchmarks in 2018, which rapidly scaled as a set of industry metrics to measure machine learning performance and promote transparency of machine learning techniques. Since then, MLCommons has continued using collective engineering to build the benchmarks and metrics required for better AI – ultimately helping to evaluate and improve AI technologies' accuracy, safety, speed, and efficiency. For additional information on MLCommons and details on becoming a member, please visit or email [email protected] . Disclaimer: The above press release comes to you under an arrangement with GlobeNewswire. Business Upturn takes no editorial responsibility for the same. Ahmedabad Plane Crash

DOWNLOAD THE APP

Get Started Now: Download the App

Ready to dive into a world of global content with local flavor? Download Daily8 app today from your preferred app store and start exploring.
app-storeplay-store