The Channel Company

Sponsored by:

Visit AMD Visit Supermicro

Capture the full potential of IT

AMD Instinct MI300 Series: Take a deeper dive in this advanced technology

Featured content

AMD Instinct MI300 Series: Take a deeper dive in this advanced technology

Take a look at the innovative technology behind the new AMD Instinct MI300 Series accelerators.

Learn More about this topic
  • Applications:
  • Featured Technologies:

Earlier this month, AMD took the wraps off its highly anticipated AMD Instinct MI300 Series of generative AI accelerators and data-center acceleration processing units (APUs). During the announcement event, AMD president Victor Peng said the new components had been “designed with our most advanced technologies.”

Advanced technologies indeed. With the AMD Instinct MI300 Series, AMD is writing a brand-new chapter in the story of AI-adjacent technology.

Early AI developments relied on the equivalent of a hastily thrown-together stock car constructed of whichever spare parts happened to be available at the time. But those days are over.

Now the future of computing has its very own Formula 1 race car. It’s extraordinarily powerful and fine-tuned to nanometer tolerances.

A new paradigm

At the heart of this new accelerator series is AMD’s CDNA 3 architecture. This third generation employs advanced packaging that tightly couples CPUs and GPUs to bring high-performance processing to AI workloads.

AMD’s new architecture also uses 3D packaging technologies that integrate up to 8 vertically stacked accelerator complex dies (XCDs) and four I/O dies (IODs) that contain system infrastructure. The various systems are linked via AMD Infinity Fabric technology and are connected to 8 stacks of high-bandwidth memory (HBM).

High-bandwidth memory can provide far more bandwidth and yet much lower power consumption compared with the GDDR memory found in standard GPUs. Like many of AMD’s notable innovations, its HBM employs a 3D design.

In this case, the memory modules are stacked vertically to shorten the distance the data needs to travel. This also allows for smaller form factors.

AMD has implemented the HMB using a unified memory architecture. This is an increasingly popular design in which a single array of main-memory modules supports both the CPU and GPU simultaneously, speeding tasks and applications.

Unified memory is more efficient than traditional memory architecture. It offers the advantage of faster speeds along with lower power consumption and ambient temperatures. Also, data need not be copied from one set of memory to another.

Greater than the sum of its parts

What really makes AMD CDNA 3 unique is its chiplet-based architecture. The design employs a single logical processor that contains a dozen chiplets.

Each chiplet, in turn, is fabricated for either compute or memory. To communicate, all the chiplets are connected via the AMD Infinity Fabric network-on-chip.

The primary 5nm XCDs contain the computational elements of the processor along with the lowest levels of the cache hierarchy. Each XCD includes a shared set of global resources, including the scheduler, hardware queues and 4 asynchronous compute engines (ACE).

The 6nm IODs are dedicated to the memory hierarchy. These chiplets carry a newly redesigned AMD Infinity Cache and an HBM3 interface to the on-package memory. The AMD Infinity Cache boosts generational performance and efficiency by increasing cache bandwidth and reducing the number of off-chip memory accesses.

Scaling ever upward

System architects are constantly in the process of designing and building the world’s largest exascale-class supercomputers and AI systems. As such, they are forever reaching for more powerful processors capable of astonishing feats.

The AMD CDNA 3 architecture is an obvious step in the right direction. The new platform takes communication and scaling to the next level.

In particular, the advent of AMD’s 4th Gen Infinity Architecture Fabric offers architects a new level of connectivity that could help produce a supercomputer far more powerful than anything we have access to today.

It’s reasonable to expect that AMD will continue to iterate its new line of accelerators as time passes. AI research is moving at a breakneck pace, and enterprises are hungry for more processing power to fuel their R&D.

What will researchers think of next? We won’t have to wait long to find out.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Supermicro debuts 3 GPU servers with AMD Instinct MI300 Series APUs

Featured content

Supermicro debuts 3 GPU servers with AMD Instinct MI300 Series APUs

The same day that AMD introduced its new AMD Instinct MI300 series accelerators, Supermicro debuted three GPU rackmount servers that use the new AMD accelerated processing units (APUs). One of the three new systems also offers energy-efficient liquid cooling.

Learn More about this topic
  • Applications:
  • Featured Technologies:

Supermicro didn’t waste any time.

The same day that AMD introduced its new AMD Instinct MI300 series accelerators, Supermicro debuted three GPU rackmount servers that use the new AMD accelerated processing units (APUs). One of the three new systems also offers energy-efficient liquid cooling.

Here’s a quick look, plus links for more technical details:

Supermicro 8-GPU server with AMD Instinct MI300X: AS -8125GS-TNMR2

This big 8U rackmount system is powered by a pair of AMD EPYC 9004 Series CPUs and 8 AMD Instinct MI300X accelerator GPUs. It’s designed for training and inference on massive AI models with a total of 1.5TB of HBM3 memory per server node.

The system also supports 8 high-speed 400G networking cards, which provide direct connectivity for each GPU; 128 PCIe 5.0 lanes; and up to 16 hot-swap NVMe drives.

It’s an air-cooled system with 5 fans up front and 5 more in the rear.

Quad-APU systems with AMD Instinct MI300A accelerators: AS -2145GH-TNMR and AS -4145GH-TNMR

These two rackmount systems are aimed at converged HPC-AI and scientific computing workloads.

They’re available in the user’s choice of liquid or air cooling. The liquid-cooled version comes in a 2U rack format, while the air-cooled version is packaged as a 4U.

Either way, these servers are powered by four AMD Instinct MI300A accelerators, which combine CPUs and GPUs in an APU. That gives each server a total of 96 AMD ‘Zen 4’ cores, 912 compute units, and 512GB of HBM3 memory. Also, PCIe 5.0 expansion slots allow for high-speed networking, including RDMA to APU memory.

Supermicro says the liquid-cooled 2U system provides a 50%+ cost savings on data-center energy. Another difference: The air-cooled 4U server provides more storage and an extra 8 to 16 PCIe acceleration cards.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

AMD drives AI with Instinct MI300X, Instinct MI300A, ROCm 6

Featured content

AMD drives AI with Instinct MI300X, Instinct MI300A, ROCm 6

Learn More about this topic
  • Applications:
  • Featured Technologies:

AMD this week formally introduced its AMD Instinct MI300X and AMD Instinct MI300A accelerators, two important elements of the company’s new push into AI.

During the company’s two-hour “Advancing AI” event, held live in Silicon Valley and live-streamed on YouTube, CEO Lisa Su asserted that “AI is absolutely the No. 1 priority at AMD.”

She also said that AI is both “the future of computing” and “the most transformative technology of the last 50 years.”

AMD is leading the AI charge with its Instinct MI300 Series accelerators, designed for both cloud and enterprise AI and HPC workloads. These systems offer GPUs, large and fast memory, and 3D packaging using the 4th gen AMD Infinity Architecture.

AMD is also relying heavily on cloud, OEM and software partners that include Meta, Microsoft and Oracle Cloud. Another partner, Supermicro, announced additions to its H13 generation of accelerated servers powered by 4th Gen AMD EPYC CPUs and AMD Instinct MI300 Series accelerators.

MI300X

The AMD Instinct MI300X is based on the company’s CDNA 3 architecture. It packs 304 GPU cores. It also includes up to 192MB of HBM3 memory with a peak memory bandwidth of 5.3TB/sec. It’s available as 8 GPUs on an OAM baseboard.

The accelerator runs on the latest bus, the PCIe Gen 5, at 128GB/sec.

AI performance has been rated at 20.9 PFLOPS of total theoretical peak FP8 performance, AMD says. And HPC performance has a peak double-precision matrix (FP64) performance of 1.3 PFLOPS.

Compared with competing products, the AMD Instinct MI300X delivers nearly 40% more compute units, 1.5x more memory capacity, and 1.7x more peak theoretical memory bandwidth, AMD says.

AMD is also offering a full system it calls the AMD Instinct Platform. This packs 8 MI300X accelerators to offer up to 1.5TB of HBM3 memory capacity. And because it’s built on the industry-standard OCP design, the AMD Instinct Platform can be easily dropped into an existing servers.

The AMD Instinct MI300X is shipping now. So is a new Supermicro 8-GPU server with this new AMD accelerator.

MI300A

AMD describes its new Instinct MI300A as the world’s first data-center accelerated processing unit (APU) for HPC and AI. It combines 228 cores of AMD CDNA 3 GPU, 224 cores of AMD ‘Zen 4’ CPUs, and 128GB of HBM3 memory with a memory bandwidth of up to 5.3TB/sec.

AMD says the Instinct MI300A APU gives customers an easily programmable GPU platform, high-performing compute, fast AI training, and impressive energy efficiency.

The energy savings are said to come from the APU’s efficiency. As HPC and AI workloads are both data- and resource-intensive, a more efficient system means users can do the same or more work with less hardware.

The AMD Instinct MI300A is also shipping now. So are two new Supermicro servers that feature the APU, one air-cooled, and the other liquid-cooled.

ROCm 6

As part of its push into AI, AMD intends to maintain an open software platform. During CEO Su’s presentation, she said that openness is one of AMD’s three main priorities for AI, along with offering a broad portfolio and working with partners.

Victor Peng, AMD’s president, said the company has set as a goal the creation of a unified AI software stack. As part of that, the company is continuing to enhance ROCm, the company’s software stack for GPU programming. The latest version, ROCm 6, will ship later this month, Peng said.

AMD says ROCm 6 can increase AI acceleration performance by approximately 8x when running on AMD MI300 Series accelerators in Llama 2 text generation compared with previous-generation hardware and software.

ROCm 6 also adds support for several new key features for generative AI. These include FlashAttention, HIPGraph and vLLM.

AMD is also leveraging open-source AI software models, algorithms and frameworks such as Hugging Face, PyTorch and TensorFlow. The goal: simplify the deployment of AMD AI solutions and help customers unlock the true potential of generative AI.

Shipments of ROCm are set to begin later this month.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

How AMD’s hardware-based security can keep your customers safer

Featured content

How AMD’s hardware-based security can keep your customers safer

AMD’s Infinity Guard hardware-level security suite is built into the company’s EPYC server processors. It guards against internal and external threats via a multilayered approach designed to prevent various types of attacks. 

Learn More about this topic
  • Applications:
  • Featured Technologies:

Helping your customers protect themselves against cyber attacks has never been more important.

In a recent survey, nearly 8 in 10 companies worldwide (77%) said they had experienced at least 1 cyber incident in the last 2 years. Virtually all said the attacks were serious.

In North America alone, it was even worse. There, 85% of the survey respondents said they’d been attacked in the last 2 years.

Kaspersky, which conducted the survey, estimates that nearly two-thirds of these attacks were due to human error. So the idea that antivirus software and employee training are enough is clearly wrong.

Why a new approach to security is needed

Fortunately, a relatively new and effective approach is available to you and your customers: hardware-based security.

To be sure, software-based solutions and dedicated firewalls are still effective weapons in the war against cybercrime. But as cybercriminals become increasingly sophisticated, IT managers have no choice but to harden security further by employing security features built in at the silicon level.

That’s because attacks can infect devices below the operating system level. When that happens, the malware gains control of a system before its OS has time to boot up and deploy the security software.

This threat is made even worse by today’s remote workforce. That’s because corporate firewalls can protect workers only when they’re connected to their organizations’ networks.

But remote workers often use networks that are insecure. They may visit a multitude of public websites, download apps, receive email attachments, and even let family and friends use their company-issued devices.

All that might be okay if not for the propensity of viruses and other malware to spread across networks like wildfire. A ransomware attack on a company laptop can, if not isolated, quickly spread to an entire network via a remote connection to a corporate data center.

From there, the ransomware can multiply and infect every other device attached to that same network. That’s how disasters happen.

Infinity Guard to the rescue

Put this all together, and you can see why hardware-level security tech like AMD Infinity Guard has become a must-have for modern data-center architecture.

AMD’s Infinity Guard hardware-level security suite is built into the company’s EPYC series server processors. There, it guards against internal and external threats via a multilayered approach designed to prevent various types of attacks. These include BIOS manipulation, in-memory return-oriented programming (ROP), and virtualized malicious hypervisor attacks.

Diving deep into the technology that underpins AMD Infinity Guard is like swimming to the bottom of the Mariana Trench—fascinating, but not for the faint of heart. A better option: consider Infinity Guard’s 4 primary safeguards:

  • AMD Secure Encrypted Virtualization (SEV): Provides encryption for every virtual machine on a server. SEV is bolstered by SEV-Secure Nested Paging (SEV-SNP), which includes memory integrity protection designed to prevent hypervisor-based attacks.
  • AMD Secure Memory Encryption: Guards against cold-boot attacks and other threats to the main memory. It’s a high-performance encryption engine integrated into the memory channel, which also helps accelerate performance.
  • AMD Secure Boot: Protects against bad actors by establishing a “root of trust.” This embedded security checkpoint validates a server’s initial BIOS software to ensure there’s no corruption. Secure Boot also ensures that only authorized firmware authenticated by the AMD Secure Processor can boot up.
  • AMD Shadow Stack: Maintains an ongoing record of return addresses so comparisons can be made to ensure integrity. Shadow Stack helps ward off ROP attacks in which an attacker directs control flow through existing code with malicious results.

‘Data-center security is easy,’ said no one ever

Maintaining a high level of data-center security is a full-time job. IT professionals can spend their entire careers playing digital defense against would-be cyberattackers.

Integrated, hardware-level security like AMD Infinity Guard gives those defenders a powerful tool to prevent ransomware and other attacks. That can help prevent incidents costing companies thousands, or even millions, of dollars.

Shifting your customers to servers with AMD Infinity Guard won’t stop the cyber arms race. But it will give them a hardware-based weapon for protecting themselves.

And Supermicro offers a wide range of servers with AMD EPYC CPUs. These help IT operators to keep their data secure and their systems protected.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Research Roundup: GenAI use, public-cloud spend, tech debt’s reach, employee cyber violations

Featured content

Research Roundup: GenAI use, public-cloud spend, tech debt’s reach, employee cyber violations

Catch up on the latest research from leading IT market watchers and analysts. 

Learn More about this topic
  • Applications:
  • Featured Technologies:

Generative AI is already used by two-thirds of organizations. Public-cloud spending levels are forecast to rise 20% next year. Technical debt is a challenge for nearly 75% of organizations. And info-security violations by staff are nearly as common as attacks by external hackers.

That’s some of the latest research from leading IT market watchers and analysts. And here’s your Performance Intensive Computing roundup.

GenAI already used by 2/3 of orgs

You already know that Generative AI is hot, but did you also realize that over two-thirds of organizations are already using it?

In a survey of over 2,800 tech professionals, publisher O’Reilly found that fully 67% of respondents say their organizations currently use GenAI. Of this group, about 1 in 3 also say their organizations have been working with AI for less than a year.

Respondents to the survey were users of O’Reilly products worldwide. About a third of respondents (34%) work in the software industry; 14% in financial services; 11% in hardware; and the rest in industries that include telecom, public sector/government, healthcare and education. By region, nearly three-quarters of respondents (74%) are based in either North America or Europe.

Other key findings from the O’Reilly survey (multiple replies were permitted):

  • GenAI’s top use cases: Programming (77%); data analysis (70%); customer-facing applications (65%)
  • GenAI’s top use constraints: Lack of appropriate use cases (53%); legal issues, risk and compliance (38%)
  • GenAI’s top risks: Unexpected outcomes (49%); security vulnerabilities (48%); safety and reliability (46%)

Public-cloud spending to rise 20% next year

Total worldwide spending by end users on the public cloud will rise 20% between this year and next, predicts Gartner. This year, the market watcher adds, user spending on the public cloud will total $563.6 billion. Next year, this spend will rise to $678.8 billion.

“Cloud has become essentially indispensable,” says Gartner analyst Sid Nag.

Gartner predicts that all segments of the public-cloud market will grow in 2024. But it also says 2 segments will grow especially fast next year: Infrastructure as a Service (IaaS), predicted to grow nearly 27%; and Platform as a Service (PaaS), forecast to grow nearly 22.

What’s driving all this growth? One factor: industry cloud platforms. These combine Software as a Service (SaaS), PaaS and IaaS into product offerings aimed at specific industries.

For example, enterprise software vendor SAP offers industry clouds for banking, manufacturing, HR and more. The company says its life-sciences cloud helped Boston Scientific, a manufacturer of medical devices, reduce inventory and order-management operational workloads by as much as 45%.

Gartner expects that by 2027, industry cloud platforms will be used by more than 70% of enterprises, up from just 15% of enterprises in 2022.

Technical debt: a big challenge

Technical debt—older hardware and software that no longer supports an organization’s strategies—is a bigger problem than you might think.

In a recent survey of 523 IT professionals, conducted for IT trade association CompTIA, nearly three-quarters of respondents (74%) said their organizations find tech debt to be a challenge.

An even higher percentage of respondents (78%) say their work is impeded by “cowboy IT,” shadow IT and other tech moves made without the IT department’s involvement. Not incidentally, these are among the main causes of technical debt, mainly because they are not acquired as part of the organization’s strategic goals.

Fortunately, IT pros are also fighting back. Over two-thirds of respondents (68%) said they’ve made erasing technical debt a moderate or high priority.

Cybersecurity: Staff violations nearly as widespread as hacks

Employee violations of organizations’ information-security policies are nearly as common as attacks by external hackers, finds a new survey by security vendor Kaspersky. 

The survey reached 1,260 IT and security professionals worldwide. It found that 26% of cyber incidents in business occurred due to employees intentionally violating their organizations’ security protocols. By contrast, hacker attacks accounted for 30%—not much higher.

Here’s the breakdown of those policy violations by employees, according to Kaspersky (multiple replies were permitted):

  • 25%: Using weak passwords or failing to change passwords regularly
  • 24%: Visiting unsecured websites
  • 24%: Using unauthorized systems for sharing data
  • 21%: Failing to update system software and applications
  • 21%: Accessing data with an unauthorized device
  • 20%: Sending data (such as email addresses) to personal systems
  • 20%: Intentionally engaging in malicious behavior for personal gain

The issue is far from theoretical. Among respondents to the Kaspersky survey, fully a third (33%) say they’ve suffered 2 or 3 cyber incidents in the last 2 years. And a quarter (25%) say that during the same time period, they’ve been the subject of at least 4 cyberattacks.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Choose the right AMD EPYC server processor for the job

Featured content

Choose the right AMD EPYC server processor for the job

AMD designed its 4th generation EPYC processors to address a variety of workloads, ranging from general-purpose computing to high-performance data center, AI and the edge. Here's how to pick the right one.

Learn More about this topic
  • Applications:
  • Featured Technologies:

Specifying AMD EPYC processors for your customers is a great first step. The vital second step is choosing the right AMD EPYC processor series for each customer’s unique requirements. It’s not a case of one-size-fits-all.

That’s because AMD designed its 4th generation EPYC processors to address a variety of workloads. These range from general-purpose computing to high-performance data center, AI and the edge.

The AMD server processor portfolio also focuses on energy efficiency, security and what the company calls leadership performance. The latter has allowed AMD to achieve more than 300 world records across industry benchmarks.

Four ways to go

Fortunately, selecting the right AMD 4th gen EPYC CPU isn’t that complicated. The company has organized its server CPUs into just 4 groups.

Each group offers a selection of speeds, core counts, cache sizes, and thermal design power (TDP) best suited to specific tasks including design simulation and cloud computing.

1) AMD EPYC 9004 Series processors: general purpose & enterprise computing

AMD counts the EPYC 9004 Series as its flagship CPU. These server processors are designed to accelerate workloads and speed results for business-critical applications in both enterprise data centers and the cloud.

AMD says the EPYC 9004 series speeds time-to-market by 2.1x and performance-per-system-watt by 2.7x, helping to lower server energy costs.

2) AMD EPYC 9004 Series processors with AMD 3D V-Cache technology: technical computing

Your customers can deploy AMD EPYC 9004 Series processors with AMD 3D V-Cache Technology to meet complex design and simulation challenges. Breakthrough performance comes courtesy of up to 96 ‘Zen 4’ cores and a ground-breaking 1152MB of L3 cache per socket.

These processors are arguably some of the most powerful in the world. They can help your customers accelerate product design even as they help reduce both capital and operational expenditures.

3) AMD EPYC 97x4 Series processors: cloud-native computing

System-level thread density, extensive x86 software compatibility, and a full ecosystem of services make the AMD EPYC 97x4 Series processors ideal for cloud-native environments. Customers focused on performance and energy efficiency can deploy this 4th gen AMD silicon to foster cloud-native work-load growth and infrastructure consolidation.

The 97x4 series includes 3 models with up to 128 cores operating at 2.25GHz. Each has a 256MB L3 cache and a TDP between 320W and 360W.

All 3 processors in this series support a suite of hardware-level security features called AMD Infinity Guard. They also support AMD Infinity Architecture, which lets system builders and cloud architects get maximum power while still ensuring security.

4) AMD EPYC 8004 Series Processors: cloud services, intelligent edge & telco

If your customers are looking for a highly energy-efficient CPU for single-socket platforms, AMD’s EPYC 8004 processors may be the right fit.

AMD engineered this series to power intelligent edge and telco servers like the Supermicro H13 WIO. As such, the processors are designed for space- and power-constrained deployments. Combined with their ability to provide performance in a power envelope as low as 70 watts, the 8004 series is ideally suited to the task.

Your customers can choose from models with 8, 16, 24, 48 or 64 ‘Zen 4c’ cores. The cores are supported by up to 1.152GB of DDR5 memory and up to 128MB of L3 cache.

Right tool for the job

Considering the total cost of ownership (TCO) when deploying a local or remote data center is essential. Choosing a processor well-suited to your customer’s workload can help them save time and money in the long run.

AMD’s 4th generation EPYC processor series includes a wide array of options. Your mission, should you choose to accept it, is to pick the CPU that best suits your customer’s needs.

Good luck!

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Supermicro expands rack capacity so you get servers faster & greener

Featured content

Supermicro expands rack capacity so you get servers faster & greener

Supermicro recently announced that it has expanded its capacity and can now provide 5,000 fully integrated, liquid-cooled racks per month. 

Learn More about this topic
  • Applications:
  • Featured Technologies:

Would you and your customers like to get faster delivery of Supermicro rackmount systems while also helping the environment?

Now you can.

Supermicro recently announced that it has expanded its capacity and can now provide 5,000 fully integrated, liquid-cooled racks per month. That’s because Supermicro now has integration facilities in four countries: the United States, Taiwan, Netherlands and Malaysia.

Supermicro also keeps in stock a certain number of commonly ordered rack configurations, what the company calls “golden SKUs.”

Between those systems and the company’s global locations, Supermicro can now deliver its rackmount systems both faster and over shorter distances. For example, Supermicro could ship a system to a customer in, say, Michigan from its Silicon Valley facility rather than from halfway around the world from Taiwan.

That shorter shipping distance also means less fuel needed and less polluting greenhouse gas produced. That’s an environmental win-win.

Get rolling with a rack

You can rely on Supermicro for data center IT solutions including on-site delivery, deployment, integration and benchmarking to achieve optimal operational efficiency.

Here’s how Supermicro’s rack delivery works in 3 steps:

Step 1: You start with proven reference designs for rapid installation while considering your clients' unique business objectives.

Step 2: You then work collaboratively with Supermicro-qualified experts to design optimized solutions for specific workloads. A prototype is designed and created for small-scale testing.

Step 3: Upon delivery, the racks need only be connected to power, networking and the liquid-cooling infrastructure. In other words, it’s a nearly seamless plug-and-play methodology.

Two areas of special interest for Supermicro are AI and liquid cooling. For AI, Supermicro plans to support AMD’s forthcoming MI300X GPU/CPU system, expected to be formally announced later this year. As for liquid cooling, it’s a technology Supermicro expects will soon be adopted by as many as 1 in 5 data centers worldwide as CPUs and GPUs continue to get hotter.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Tech Explainer: What’s the deal with AMD’s 3D V-Cache technology?

Featured content

Tech Explainer: What’s the deal with AMD’s 3D V-Cache technology?

AMD’s unique 3D V-Cache technology offers a marked performance boost. By stacking the cache vertically, AMD produces a 3x increase in the amount of a CPU’s L3 cache. This enables faster calculations and a noticeable increase in overall processor speed. 

Learn More about this topic
  • Applications:
  • Featured Technologies:

The more cache, the merrier.

The cache’s job is to store as much data as possible directly onto the CPU die. And the cache’s proximity to the processor makes it an ideal data-delivery system.

That might not be much of a sales pitch to folks who ask comparatively little of their devices. But when it comes to high-performance data centers, extra on-die storage can make a world of difference.

That’s because the faster and more efficiently a CPU can access data, the quicker it can complete the complex calculations required to return the requested result.

The trouble is, the usefulness of a cache has so far been limited by its small storage capacity. Unlike other computer storage devices, such as RAM and SSDs, your average cache is relatively tiny.

In a modern computer, it’s not uncommon to get 1 to 4 terabytes of storage space. But on that same machine, the processor cache would typically hold only 50 to 60 megabytes.

It stands to reason, then, that a cache able to hold 3 times as much data could dramatically increase the system’s efficiency. And that’s what AMD’s latest innovation, known as AMD 3D V-Cache technology, can do.

A whole new dimension

Modern caches are designed with a succession of stages labeled L1, L2 and L3. The “L” stands for Level.

L1 is closest to the processor and offers the fastest speed, yet it also provides the smallest capacity. L2 is a bit bigger, but also a bit slower. And L3 cache always provides the most data storage.

AMD calls its innovation 3D V-Cache because of the unique design. All 3 layers of the L3 cache are stacked vertically on the die.

This vertical stacking also means that all 3 layers are the same distance from the processor. As a result, all 3 also offer the same speeds.

What’s more, AMD 3D V-Cache’s extra capacity enables the processor to store and stream more instructions, yet without increasing the die’s size. As a result, the CPU does its job much faster and more efficiently than could a similarly powered processor with a traditional cache.

Out in the wild

AMD first introduced 3D V-Cache technology in 2022 as part of its gaming-focused Ryzen 7 5800X3D processor. That first iteration offered a 96MB L3 cache feeding 8 cores, each with a maximum clock speed of 4.5GHz.

This past February, AMD introduced its Ryzen 7000-series processors for content creators. Two of those CPUs—the AMD Ryzen 9 7950X3D and 7900X3D—featured L3 caches with 128MB of capacity.

But AMD knows the market for 3D V-Cache-enabled chips is far bigger than just gamers. Power-hungry data centers are cropping up everywhere, helping to power our private and public clouds, AI-based applications and other miraculous virtual innovations.

To power those data centers, AMD now offers 4th generation EPYC processors featuring AMD 3D V-Cache technology. These data-crunching monsters are truly on the cutting edge. The biggest of them all, the AMD EPYC 9654P, packs 96 cores and an astounding L3 cache of 348MB.

Taking the show on the road

With the introduction of the Ryzen 9 7945HX3D mobile processor, AMD announced to the world that speed and portability aren’t mutually exclusive. The first laptop gaming chip is a 16-core phenom featuring AMD 3D V-Cache Technology, which adds an impressive 128MB of L3 cache.

This past August, AMD launched its new silicon marvel inside the ROG STRIX SCAR 17 X3D, a titanically powerful gaming laptop with a 240Hz QHD display. AMD managed to get a mobile version to market barely a year after the launch of its first AMD 3D V-Cache Technology-enabled desktop chip. That’s an impressive cadence by any standard.

This is a laptop that was always going to be faster than most. But the addition of the Ryzen chip makes the STRIX SCAR one of the fastest mobile gaming rigs on this—or any other—planet.

Go ahead, try to find one that’s faster.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Meet the newest 3rd gen AMD EPYC processors

Featured content

Meet the newest 3rd gen AMD EPYC processors

AMD extends 3rd Gen EPYC CPU lineup to deliver greater price/performance for mainstream server applications.

Learn More about this topic
  • Applications:
  • Featured Technologies:

Are you or your customers looking for server processors that offer great price/performance, modern security, and energy efficiency for less technically demanding mainstream business workloads?

If so, AMD has some new server CPUs for you.

“We have seen a clear opportunity to give our customers more options,” says Dan McNamara, GM of AMD’s server business.

Expanded value options

The six new SKUs are actually part of AMD’s 3rd generation EPYC processor family. Originally introduced nearly three years ago, the 3rd gen AMD EPYCs have since been joined by the company’s 4th Gen.

So why might your customers be interested in new 3rd gen CPUs from AMD?

Several reasons. One, they might not need the latest AMD processor features, which include support for both DDR5 memory and PCIe Gen 5 connectivity. The new members of AMD’s 3rd gen EPYC family, which use the company’s Zen 3 cores, support up to 8 channels of the older DDR4 memory and up to 128 lanes of PCIe Gen 4.

By contrast, AMD’s 4th gen EPYC processors, with their Zen 4 cores, support both DDR5 and PCIe Gen 5. For some companies, AMD says, the upgrade to these newer technologies is “still high on the cost curve.”

Price/performance, too

Another reason is to get in on the new price/performance gains. Four of the six new SKUs are 8- and 16-core processors, and their retail prices range from just under $340 to just over $600. The other two CPUs offer 48 and 56 cores.

Yet another reason: modern security features that include AMD Infinity Guard. It’s a full suite of security features, built into silicon, that includes encrypted virtualization; nested paging; memory encryption; and AMD Shadow Stack, which offers hardware-enforced protection against malware.

The new AMD processors are fully compatible with existing AMD EPYC 7003 series-based server systems. And AMD’s major partners, including Supermicro, have said they’ll support the new 3rd Gen AMD EPYC processors in their enterprise server solutions.

Supermicro works closely with AMD to offer a wide range of application-optimized servers in its H13 product line. These systems, which support the new AMD EPYC processors, are designed for applications that need powerful processing performance, but may have thermal constraints.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Research Roundup: GenAI, 10 IT trends, cybersecurity, CEOs, and privacy

Featured content

Research Roundup: GenAI, 10 IT trends, cybersecurity, CEOs, and privacy

Catch up on the latest IT research and analysis from leading market watchers.

Learn More about this topic
  • Applications:
  • Featured Technologies:

Generative AI is booming. Ten trends will soon rock your customers’ world. While cybersecurity spending is up, CEOs lack cyber confidence. And Americans worry about their privacy.

That’s some of the latest from leading IT market watchers. And here’s your Performance Intensive Computing roundup.

GenAI market to hit $143B by 2027

Generative AI is quickly becoming a big business.

Market watcher IDC expects that spending on GenAI software, related hardware and services will this year reach nearly $16 billion worldwide.

Looking ahead, IDC predicts GenAI spending will reach $143 billion by 2027. That would represent a compound annual growth rate (CAGR) over the years 2023 to 2027 of 73%—more than twice the growth rate in overall AI spending.

“GenAI is more than a fleeting trend or mere hype,” says IDC group VP Ritu Jyoti.

Initially, IDC expects, the largest GenAI investments will go to infrastructure, including hardware, infrastructure as a service (IaaS), and system infrastructure software. Then, once the foundation has been laid, spending is expected to shift to AI services.

Top 10 IT trends

What will be top-of-mind for your customers next year and beyond? Researchers at Gartner recently made 10 predictions:

1. AI productivity will be a primary economic indicator of national power.

2. Generative AI tools will reduce modernization costs by 70%.

3. Enterprises will collectively spend over $30 billion fighting “malinformation.”

4. Nearly half of all CISOs will expand their responsibilities beyond cybersecurity, driven by regulatory pressure and expanding attack surfaces.

5. Unionization among knowledge workers will increase by 1,000%, motivated by fears of job loss due to the adoption of GenAI.

6. About one in three workers will leverage “digital charisma” to advance their careers.

7. One in four large corporations will actively recruit neurodivergent talent—including people with conditions such as autism and ADHD—to improve business performance.

8. Nearly a third of large companies will create dedicated business units or sales channels for machine customers.

9. Due to labor shortages, robots will soon outnumber human workers in three industries: manufacturing, retail and logistics.

10. Monthly electricity rationing will affect fully half the G20 nations. One result: Energy efficiency will become a serious competitive advantage.

Cybersecurity spending in Q2 rose nearly 12%

Heightened threat levels are leading to heightened cybersecurity spending.

In the second quarter of this year, global spending on cybersecurity products and services rose 11.6% year-on-year, reaching a total of $19 billion worldwide, according to Canalys.

A mere 12 vendors received nearly half that spending, Canalys says. They include Palo Alto Networks, Fortinet, Cisco and Microsoft.

One factor driving the spending is fear, the result of a 50% increase in the number of publicly reported ransomware attacks. Also, the number of breached data records more than doubled in the first 8 months of this year, Canalys says.

All this increased spending should be good for channel sellers. Canalys finds that nearly 92% of all cybersecurity spending worldwide goes through the IT channel.

CEOs lack cyber confidence

Here’s another reason why cybersecurity spending should be rising: Roughly three-quarters of CEOs (74%) say they’re concerned about their organizations’ ability to avert or minimize damage from a cyberattack.

That’s according to a new survey, conducted by Accenture, of 1,000 CEOs from large organizations worldwide.

Two findings from the Accenture survey really stand out:

  • Nearly two-thirds of CEOs (60%) say their organizations do not incorporate cybersecurity into their business strategies, products or services
  • Nearly half (44%) the CEOs believe cybersecurity can be handled with episodic interventions rather than with ongoing, continuous attention.

Despite those weaknesses, nearly all the surveyed CEOs (96%) say they believe cybersecurity is critical to their organizations’ growth and stability. Mind the gap!

How do Americans view data privacy?

Fully eight in 10 Americans (81%) are concerned about how companies use their personal data. And seven in 10 (71%) are concerned about how their personal data is used by the government.

So finds a new Pew Research Center survey of 5,100 U.S. adults. The study, conducted in May and published this month, sought to discover how Americans think about privacy and personal data.

Pew also found that Americans don’t understand how their personal data is used. In the survey, nearly eight in 10 respondents (77%) said they have little to no understanding of how the government uses their personal data. And two-thirds (67%) said the same thing about businesses, up from 59% a year ago.

Another key finding: Americans don’t trust social media CEOs. Over three-quarters of Pew’s respondents (77%) say they have very little or no trust that leaders of social-medica companies will publicly admit mistakes and take responsibility.

And about the same number (76%) believe social-media companies would sell their personal data without their consent.

Do more:

 

Featured videos

Events

Find AMD & Supermicro Elsewhere

Related Content

Pages