Senior Solutions Architect - CSP System
About the job
- A Bachelor's, Master's, or PhD degree in Computer Science, Computer Engineering, Electrical Engineering, Networking, or a related field; equivalent industry experience is highly valued.
- At least 7 years of hands-on experience in datacenter network architecture, cloud-native networking, DPU/SmartNIC development, or large-scale L4-L7 traffic management systems, with experience in North-South dataplane design, API Gateway systems, or service mesh infrastructure.
- Deep understanding of datacenter North-South network architecture, including load balancing, reverse proxies, API Gateways such as Envoy, Nginx, and HAProxy, service meshes such as Istio and Cilium, and cloud-native networking including CNI, the Kubernetes networking model, and Gateway API.
- Strong programming proficiency in C/C++ and Python; familiarity with dataplane programming such as DPDK, eBPF/XDP, P4, or DOCA, network protocol stacks including TCP/IP, RDMA, HTTP/2, and gRPC, and high-performance packet processing.
- Hands-on experience working with major Chinese CSPs or global hyperscalers on network infrastructure projects, including in-depth knowledge of their cloud networking architecture, traffic management mechanisms, and North-South dataplane evolution.
- Understanding of AI inference serving architecture, including vLLM, TensorRT-LLM, and Triton, and how network infrastructure supports prefill/decode disaggregation, KV-Cache transfer, token streaming, and multi-model Agent orchestration.
- Technical communication and presentation skills to explain complex network systems and DPU acceleration technologies to technical engineers, architecture teams, and business stakeholders.
- Ability to collaborate cross-functionally in a global matrix team and prioritize multiple high-value technical projects under fast-paced business demands.
- Hands-on engineering capability and ability to independently own end-to-end technical project delivery; must be result-oriented, self-driven, committed, proactive, and capable of sustained high-quality technical output for long-term strategic CSP projects.
- Partner with Sales, BD, and CPM teams to introduce NVIDIA BlueField DPU and DOCA Agent networking technologies into top-tier Chinese CSP accounts, drive technical adoption, and support sustainable business growth in the Agent infrastructure market.
- Serve as the primary technical authority for NVIDIA Agent North-South network infrastructure solutions for Chinese CSPs, providing end-to-end consultation on DPU-accelerated API Gateway architecture, Agent traffic routing and orchestration, North-South telemetry, and inline AI inspection pipeline design.
- Design and deliver BF4/DOCA 3.0 Agent-specific North-South solutions, including hardware-accelerated Agent API Gateways to replace Envoy/Nginx software dataplanes, inline token-level Agent traffic metering, multi-agent communication mesh acceleration, and North-South KV-Cache access optimization for disaggregated inference architectures.
- Lead open-source network architecture contributions for NVIDIA Agent infrastructure stacks, upstream optimized integrations with cloud-native ecosystems including Kubernetes Gateway API, Envoy xDS, Cilium, and eBPF, build China-localized best practices, and shape industry technical standards for Agent networking.
- Analyze North-South traffic patterns for Agentic AI workloads, including tool-calling, RAG retrieval, multi-agent orchestration, and human-agent interaction; implement DPU-level and DOCA-level dataplane optimizations; and deliver production-ready reference designs and tuning guidelines for CSP deployment.
- Act as the technical liaison between Chinese CSP customers and NVIDIA Networking BU global engineering, product, and R&D teams; collect local Agent networking requirements, drive BlueField/DOCA product roadmap iteration, and ensure alignment with NVIDIA global technical policies.
- Lead technical workshops, hands-on training, PoC and production pilot projects for key CSP accounts; quantify and demonstrate the business value of DPU-accelerated Agent networking, including latency reduction, throughput improvement, and TCO savings versus software-only solutions; and accelerate technology adoption and large-scale replication.
- Monitor industry trends in Agentic AI infrastructure, LLM serving network topology, MCP/A2A protocol standardization, disaggregated inference architecture, and cloud-native service mesh; produce technical insights to support team and product strategy.
- Mentor junior SA team members in DPU and networking domains, standardize CSP Agent networking technical engagement and solution delivery processes, and develop high-value technical best practices.
- Hands-on BlueField DPU/DOCA dataplane development, including DPA programming, hardware steering rules, and crypto/compression offload for North-South traffic.
- Experience building or operating large-scale API Gateway/L7 proxy systems at CSP traffic volumes of 1M+ RPS, with demonstrated optimization for latency-sensitive AI inference traffic.
- Proven contributions to cloud-native networking open-source projects such as Envoy, Cilium, Istio, Kubernetes SIG-Network, or Gateway API, with public upstream records.
- Understanding of Agentic AI infrastructure patterns, including MCP, A2A, LangChain/CrewAI/AutoGen runtime architecture, Agent tool-calling traffic patterns, and multi-agent communication topology.
- Experience designing disaggregated inference serving architectures where networking supports prefill-decode separation, remote KV-Cache access over RDMA/RoCE, and North-South token routing for mixture-of-experts models.
- Familiarity with DPU-based security and observability, including inline AI traffic inspection, token-level metering and billing, Agent identity/authentication offload, and zero-trust micro-segmentation for multi-tenant Agent platforms.
- Track record of independently leading CSP network infrastructure PoC, pilot verification, and large-scale production deployment projects with measurable latency, throughput, and TCO outcomes.
- Familiarity with NVIDIA networking products, including BlueField DPU, ConnectX SmartNIC, DOCA SDK, Spectrum switches, and NetQ.
- Familiarity with semiconductor and datacenter technology export compliance requirements in the China market.
About the company
NVIDIA designs the chips and software that power artificial intelligence and sells complete data center systems to cloud providers, AI labs, and other businesses, along with GeForce graphics cards for gamers. Its GPUs, chips that perform thousands of calculations at once, became the main hardware for AI after researchers used them to train the AlexNet image-recognition model in 2012. NVIDIA connects many GPUs with its own processors and networking gear into servers and rack-sized computers, and developers program them through its CUDA software. These systems bring in most of its sales, and it also makes chips for cars and robots. Jensen Huang co-founded NVIDIA in 1993, and it is based in Santa Clara, California.
Company Size
10,001+
Company Stage
IPO
Headquarters
Santa Clara, California
Founded
1993
Get referred to NVIDIA
See people who can refer or advise you
Simplify's Take
What believers are saying
- AI demand is still surging: data center sales rose 117% to $89 billion in the quarter ending July 2026.
- Stock grants are worth more: shares hit a record in October 2026, up 28% for the year.
- Engineering dominates hiring: 1,722 of NVIDIA's 2,686 job listings in October 2026 were in engineering.
What critics are saying
- Big buyers are hedging: Meta's February 2026 agreement covers 6 gigawatts of AMD chips.
- Financing its buyers is getting harder: lenders pushed back on NVIDIA's $500 billion chip-backed loan plan in October 2026.
- Staff described long hours: Bloomberg's August 2024 interviews reported seven-day weeks, late nights and shouting matches.
What makes NVIDIA unique
- Rivals must match its software too: over 7.5 million developers use CUDA and other NVIDIA tools to program its chips.
- The CEO has a broad direct staff: Jensen Huang says over 60 people report to him, without one-on-ones.
- People rarely leave: NVIDIA's yearly staff turnover is 3.7%, versus a 21.9% average across the tech industry.
Help us improve and share your feedback! Did you find this helpful?
Benefits
Company Equity
401(k) Company Match
Growth & Insights and Company News
6 month growth
0%1 year growth
-1%2 year growth
-2%Nvidia said Synopsys and Siemens are adopting its CUDA-X libraries for chipmaking. The announcement places Synopsys' adoption within a broader partnership with Nvidia on engineering and design. Nvidia said Siemens uses cuLitho for chip-pattern refinement and mask preparation. Siemens also works with Nvidia on industrial AI and manufacturing digital twins. Nvidia said cuPotential is available in early access. NanoVDB is open source within OpenVDB, while NPP is available through the CUDA Toolkit.
Nvidia CEO Jensen Huang announced investments in five UK startups at a London event, backing Boltz Bio, Foresight Works, Hivergé AI, PrimaMente, and quantum computing firm Riverlane — extending Nvidia's AI-ecosystem strategy across Europe's deep-tech capital.
Two weeks before ChatGPT launched, Nvidia's earnings call for the third quarter of fiscal 2023 held early signs of the generative AI boom. The November 16, 2022 report looked weak: revenue fell 17% year over year, gaming sales halved to $1.6 billion, and the company took a $702 million inventory charge tied to new US export rules on China. Yet data center revenue rose 31% to a record $3.8 billion, nearly two-thirds of the total. Management cited large language models and generative AI as growth drivers, and chief executive officer Jensen Huang called it "the beginning of the era of generative AI." Nvidia also began shipping its H100 data center chip that quarter. The stock was worth roughly $390 billion then, and $5.5 trillion in October 2026.
Nvidia is opening more of its Vera Rubin infrastructure designs to industry partners, though its shares fell 2.8% to $222.22 at 12:12 p.m. ET on October 14. The company's MGX 1.2 rack designs support up to 250 kilowatts. They include a liquid-cooled 5,000-amp busbar, standardised power and cooling connections, and support for 19-inch equipment. Nvidia's Vera CPU rack also handles orchestration and tool execution for AI agents. Separate grid-resilience specifications call for systems to recover to near-full load in under two seconds after disturbances. The figures describe design limits, not installed capacity or booked revenue. Nvidia's October 12 announcement gave no customer order value. Investors will watch for operating facilities and repeat purchases.
NVIDIA announced on October 14, 2026 that semiconductor leaders are adopting its CUDA-X libraries to build AI-defined fabs, factories where AI helps engineers optimise processes and fix production issues. The news came at SEMICON West in San Francisco. The company introduced cuPotential, a library for building AI agents that simulate how atoms and molecules interact. In early-access testing, it sped up already GPU-optimised workflows by an additional 2-3x. Synopsys is among the early users. Lam Research expects NanoVDB, an open source 3D data library, to speed up some wafer-etch modelling by up to 30x. Applied Materials, Cadence, KLA, and Siemens are also using NVIDIA's tools across chip design, manufacturing, and testing.