SendTech Times
العربية
Chips & SemiconductorsNews|July 26, 2026, 06:47 PM
CAPACITY TEST:

AMD Helios AI Racks Bring Epyc CPUs To Nvidia Compute Challenge

BySendTech Times Chips & Compute DeskNewsroom-edited, source-reviewed coverage
Newsroom brief

Data Center Knowledge reported that AMD moved Helios into production with MI455X GPUs, Epyc processors, Pensando networking and ROCm software, while vendor performance claims still lack third-party benchmark validation.

Verified against source materialEdited by SendTech Times Chips & Compute Desk
AMD Helios AI Racks Bring Epyc CPUs To Nvidia Compute Challenge
Image source: datacenterknowledge.com

AMD has moved its Helios rack-scale AI system into full production, giving the chipmaker an integrated platform to compete with Nvidia beyond individual accelerators.

Data Center Knowledge reported that hardware partners are scheduled to begin shipments by the end of the third quarter of 2026.

The system combines Instinct MI455X accelerators, Epyc server processors, Pensando networking and ROCm software.

AMD said Microsoft and Anthropic are the first named deployment routes for the integrated architecture.

A Rack-Scale Platform Built Around Venice

Helios follows the integrated design now shaping high-end AI infrastructure: processors, accelerators, networking and software are engineered as one rack rather than assembled as separate components.

The approach is intended for increasingly complex agentic workloads that require CPUs to coordinate data retrieval, tool calls and other tasks around GPU reasoning.

At the July 23 Advancing AI event in San Francisco, Chief Executive Lisa Su said AI agents could expand from millions to billions.

Her presentation positioned server processors as an important part of that growth, not simply as hosts for accelerators.

AMD said the Venice family includes configurations with up to 256 cores: SP8 variants span 8 to 128 cores, the SP7 range reaches 256 cores and the LP host-node processor reaches 72 cores.

The lineup covers enterprise servers, cloud systems and specialised AI host nodes.

Performance Claims Centre On The Full System

Su said Helios delivers 15% more AI compute, 50% more memory, 50% more scale-out bandwidth and up to 30% more tokens per dollar than Nvidia's competing platform.

AMD's launch material also compared a 256-core Venice configuration with Nvidia's 88-core Vera CPU, claiming 2.2 times the throughput, and cited a 20% per-core advantage for a 96-core Venice chip.

According to AMD's launch benchmarks, MI455X provides 34 times the token throughput and up to 18 times the token-cost performance of the previous MI355X generation.

Taken together, the vendor's figures present a performance and capacity case for evaluating the rack as an alternative to Nvidia's integrated designs.

Microsoft And Anthropic Establish Deployment Routes

AMD identified Microsoft as a customer for frontier-model inference and Azure AI services.

Su also said Anthropic intends to deploy as much as 2 GW of accelerators in the rack-scale systems.

Those commitments move the launch beyond a component roadmap and support a broader effort to sell CPUs, accelerators, networking and software as one infrastructure stack.

In the same presentation, Su projected a $1.4 trillion AI infrastructure market by 2030, replacing her earlier $500 billion estimate for 2028.

She put the server CPU opportunity at $200 billion by 2030, up from $25 billion today.

The presentation also attributed 46% of latest-quarter server-market revenue to the processor family and listed deployments at more than 60% of Fortune 100 companies.

The Processor Roadmap Runs Through 2030

AMD said its roadmap schedules the SP7 version for the fourth quarter of 2026, SP8 for the first half of 2027 and the LP host-node processor for the second half of that year.

It places seventh-generation Florence CPUs in 2028 and an eighth generation in 2030.

On the accelerator side, the roadmap calls for a new GPU and Helios generation each year, including Instinct MI500 in 2027 and MI600 in 2028.

ROCm.AI is intended to support that cadence with command-line tools, reusable skills and AI-assisted software optimisation across the hardware portfolio.

The remaining evidence gap is operational.

Public material establishes production status, shipment timing, product specifications and customer commitments, but it does not provide order sizes, deployed-cluster economics or independently reproduced benchmarks.

Share this article
inXf

Related articles

More
US Export-Control Shift Opens UAE AI Chip Access, DCD Reports
Chips & Semiconductors

US Export-Control Shift Opens UAE AI Chip Access, DCD Reports

Data Center Dynamics reported that the US government has moved the UAE into a lower-restriction export-control category, giving UAE companies broader access to advanced AI chips from Nvidia and AMD. The account names G42 and hyperscaler data-centre projects, but it does not list chip volumes, approved licences or specific UAE projects.

Compute Exchange Opens Used Nvidia H100 And A100 GPU Marketplace
Chips & Semiconductors

Compute Exchange Opens Used Nvidia H100 And A100 GPU Marketplace

SiliconANGLE reported that Compute Exchange has launched a marketplace for used and refurbished Nvidia H100 and A100 GPUs. Compute Exchange said requests range from hundreds to tens of thousands of GPUs, but the public launch did not name suppliers, customers or pricing benchmarks.

Oxmiq Raises $35 Million For GPU IP And AI Factory Design
Chips & Semiconductors

Oxmiq Raises $35 Million For GPU IP And AI Factory Design

Oxmiq Labs raised $35 million in Series A funding, taking total funding to $60 million, as the startup expands from GPU IP toward data-centre-scale hardware, orchestration software and AI factory design, EE Times reported. Other large customers, production contracts and independent benchmarks remain outside the public record.

Microsoft Adds AMD Helios AI Racks To Azure Without Order Size
Chips & Semiconductors

Microsoft Adds AMD Helios AI Racks To Azure Without Order Size

Microsoft will deploy AMD Helios rack-scale AI accelerators for Azure AI workloads, with watts, dollars and rack counts still absent from the public terms of the commitment.

Maybank Sees AI Demand Driving Singapore Semiconductor Growth
Chips & Semiconductors

Maybank Sees AI Demand Driving Singapore Semiconductor Growth

Maybank Investment Bank said AI-related demand should remain the main growth driver for Singapore semiconductor companies, while warning that large AI infrastructure spending still needs stronger revenue proof.

AMD Ramps Venice EPYC CPUs On TSMC 2nm Process
Chips & Semiconductors

AMD Ramps Venice EPYC CPUs On TSMC 2nm Process

AMD says its 6th Gen EPYC processor, codenamed Venice, has entered production ramp on TSMC 2nm technology, with future plans for TSMC Arizona production.

Gulf AI Plans Still Depend On Nvidia Chips Despite Supplier Push
Chips & Semiconductors

Gulf AI Plans Still Depend On Nvidia Chips Despite Supplier Push

Rest of World reported that Saudi Arabia and the UAE are trying to diversify AI chip supply while major projects still rely on Nvidia hardware. The report cited Humain data-centre plans, G42’s Stargate project and analyst warnings that U.S. approvals, TSMC capacity and high-bandwidth memory remain constraints.

Cerebras Plans 8x To 10x Manufacturing Scale-Up For AI Inference
Chips & Semiconductors

Cerebras Plans 8x To 10x Manufacturing Scale-Up For AI Inference

Cerebras chief executive Andrew Feldman said the company plans to scale manufacturing capacity by 8x to 10x this year and claimed its systems can run inference 10, 15, 20 or 30 times faster than GPUs. The interview-led source named customers including AlphaSense, Cognition AI, OpenAI, Block and GlaxoSmithKline, The public record still lacks third-party benchmark methodology.

Keep Reading

More Stories

Latest
Dubai Airports Adds Smart Gate Pre-Check Before Summer PeakEconomyJul 26, 2026Dubai Airports Adds Smart Gate Pre-Check Before Summer PeakEconomy Middle East reported that Dubai Airports launched a Smart Gates Eligibility Pre-Check at DXB, letting passengers confirm eligibility through Pocket Flights or terminal QR codes before passport control.Hong Kong Indoor 5G Gap Shows Building-Level Network DivideTelco & ConnectivityJul 26, 2026Hong Kong Indoor 5G Gap Shows Building-Level Network DivideA Speedtest Intelligence study says Hong Kong’s indoor 5G experience varies sharply by building, putting the focus on shared in-building systems rather than outdoor coverage alone.CIRCIA Rule Faces September Deadline As Industry Seeks Narrower FilingsCapital & PolicyJul 26, 2026CIRCIA Rule Faces September Deadline As Industry Seeks Narrower FilingsCISA is working toward a September target for the delayed CIRCIA rule as industry groups seek fewer covered entities, narrower incident triggers and leaner reporting requirements. The law sets 72-hour incident and 24-hour ransomware-payment reporting deadlines, while the proposed rule could cover more than 300,000 entities.Ropedia Raises US$22 Million For Physical AI Data LayerAIJul 26, 2026Ropedia Raises US$22 Million For Physical AI Data LayerSingapore-based Ropedia has raised US$22 million in pre-Series A funding, e27 reported, as robotics developers look for data that records motion, depth, audio and human interaction rather than only online text.Azure West US Outage Exposes Fiber-Maintenance Routing RiskCloud & Data CentersJul 26, 2026Azure West US Outage Exposes Fiber-Maintenance Routing RiskThe Register reported that Microsoft’s West US Azure region was disrupted for almost five hours after routine fiber-related maintenance removed more routes than intended. Microsoft’s preliminary review said 27 services were affected before full recovery at 19:41 UTC on July 23.Coinbase Adds x402 Payments For AI Agent TransactionsFintech & Digital PaymentsJul 26, 2026Coinbase Adds x402 Payments For AI Agent TransactionsCoinbase Business users will be able to accept payments from AI agents through x402, while exchange users get live order views and developers get an SDK.Microsoft Azure AI Deals Add Cobalt Chips And French CapacityCloud & Data CentersJul 26, 2026Microsoft Azure AI Deals Add Cobalt Chips And French CapacityMicrosoft expanded Azure AI agreements with Databricks and Mistral, combining Cobalt processor adoption, Microsoft product integrations and Mistral-operated French data-centre capacity for customers with control and residency requirements.AI Kill Switch Bill Would Give Homeland Security Emergency Model ControlCapital & PolicyJul 26, 2026AI Kill Switch Bill Would Give Homeland Security Emergency Model ControlThe proposed AI Kill Switch Act would give Homeland Security emergency authority to slow, limit or shut down covered frontier AI systems, while exempting red-team incidents such as the OpenAI sandbox escape that triggered the debate.Japan Funds Noetra’s 140MW Physical AI Data CentreCloud & Data CentersJul 26, 2026Japan Funds Noetra’s 140MW Physical AI Data CentreTech Wire Asia reported that Japan is funding Noetra to build domestic physical AI models around a 140MW FRONTia facility using Nvidia Rubin GPUs, while model access terms remain open.Etched Raises $300 Million As $1 Billion Pre-Orders Test Inference Rack PlanChips & SemiconductorsJul 26, 2026Etched Raises $300 Million As $1 Billion Pre-Orders Test Inference Rack PlanEE Times reported that Etched raised $300 million at a $10 billion pre-money valuation and has $1 billion in pre-orders, but the AI chip startup has not made public performance figures for racks due to ship this summer.Nvidia-SK AI Deal Carries $500 Billion Value With Few Build DetailsChips & SemiconductorsJul 26, 2026Nvidia-SK AI Deal Carries $500 Billion Value With Few Build DetailsTom's Hardware reported that Nvidia and SK Group signed letters of intent for a strategic relationship valued at more than $500 billion, anchored by SK Telecom's planned 2-gigawatt South Korea AI data centre and a long-term SK hynix memory supply agreement.Upstart Bank Charter Approval Leaves Fed And FDIC Steps PendingFintech & Digital PaymentsJul 25, 2026Upstart Bank Charter Approval Leaves Fed And FDIC Steps PendingUpstart received conditional OCC approval to form a national bank, but the AI lending marketplace still needs deposit-insurance and bank-holding-company approvals before the charter can operate.