NVIDIA Partner Network

Manage NVIDIA Partner Network incentives, competencies and certifications in one dashboard.

NVIDIA reshaped the Partner Network: partner types regrouped into four families, competencies were updated, and each partner’s requirements are set in the Competency Exhibit they sign with NVIDIA. Rebates-On tracks it from your perspective, so you see what you’re owed, exactly what to do to earn more, and where NVIDIA rebate dollars are slipping away.

Trusted by IT partners managing 20+ vendor programs · Pre-built for the NVIDIA Partner Network

A laptop showing NVIDIA Partner Network tracking in Rebates-On, with rebate dollars trending upward
The program today

What is the NVIDIA Partner Network (NPN)?

The NVIDIA Partner Network (NPN) is NVIDIA’s enterprise channel program for everyone who sells, integrates, hosts, builds, or advises on NVIDIA products. New partners join at an entry level, and certain program elements, including the **Elite** membership level, are by invitation only and subject to approval. Partner types sit in four families, and the level structure varies by partner type.

The four partner-type families are AI Infrastructure, Cloud, Influencer, and Seller/Builder, reflecting NVIDIA’s shift toward the full AI-factory buildout, not just card resale. NPN is a curated program rather than a self-service one: you apply, and NVIDIA’s partner organization decides which partner type and competency fit you.

What you carry is then gated by competency, and the competency list depends on your partner type. The Solution Provider competencies are Compute, DGX AI Compute Systems, Embedded Compute, Networking, NVIDIA Enterprise Software, and Visualization. The AI Infrastructure programs have a single competency, NVIDIA Technologies.

What’s changing

NVIDIA reshaped NPN - here’s what moves your rebate dollars.

NVIDIA has reshaped NPN: partner types regrouped into AI Infrastructure, Cloud, Influencer and Seller/Builder families; new data-center buildout types arrived; and the competency schedules were updated. The families do not share one economic model, which is the part old assumptions miss. Resale and rebate economics sit with the Solution Provider track.

  • Four partner-type families. Types are now grouped as AI Infrastructure (Data Center Partner, Power & Cooling, Architecture/Engineering/Construction), Cloud (NVIDIA Cloud Partner / NCP), Influencer (Advanced Technology Partner, Education Services, ISV, GSI, Solution Advisor, Storage), and Seller/Builder (Solution Provider, OEM, System Partner, Distributor). The Solution Provider remains the core rebate-earning population.
  • The AI Infrastructure programs are not resale programs. NVIDIA states plainly that the AI Infrastructure programs do not support partners reselling NVIDIA products. They are measured on referrals, design qualification and competency attainment, and their sales benefit is demo and not-for-resale hardware plus a referral requirement. Reading them as a rebate motion is a category error.
  • Competency schedules differ by partner type. The Solution Provider schedule is set competency by competency, each with its own requirements per membership level, rather than as one company-wide bar. The AI Infrastructure programs carry a single competency, NVIDIA Technologies.
  • Supply and lead times are outside the program. Lead times on constrained systems can move a booking into a later period. That is a real planning problem, but it is not something NVIDIA’s partner program documents address, so treat it as market conditions rather than as a program rule.
60 days
to complete training after signing

NVIDIA's Competency Exhibit for Solution Providers requires the named number of sales, technical and marketing individuals to complete their training within sixty (60) days of signing the Exhibit. NVIDIA reviews membership levels annually and tracks compliance on an ongoing basis. What NVIDIA does not put in writing is what happens, or when, if a competency slips afterwards.

Where the money leaks

Where NVIDIA partners leave rebate money on the table.

NVIDIA partners lose rebate dollars in a few predictable places: competency-by-competency fragmentation, two-vendor incentive stacking on OEM AI servers, allocation slips that move deals into a later period, point-of-sale data that lands late or lands wrong, and slipping below the requirements for your level without noticing. Each is a deal that looks profitable on the quote while the money quietly never lands - something software can watch that a spreadsheet cannot.

Competency fragmentation

Your standing is earned per competency, not on one company-wide score, and a single deal can span several competencies at once. A line-item misclassification quietly credits the wrong competency, so whatever your agreement pays is calculated off the wrong base.

Two-vendor stacking on OEM systems

Most enterprise NVIDIA revenue moves as Dell, HPE, Lenovo, or Supermicro systems. The partner must claim the OEM’s rebate and any NVIDIA-side NPN benefit (plus distributor promos) on the same deal - and no single portal shows the combined picture.

Winning the deal is not getting the product

Winning a deal doesn’t secure Blackwell-class supply. Quotes expire, prices move, and lead-time slips push revenue and attainment into a later period, costing you both the timing you planned for and customer goodwill.

Point-of-sale reporting gaps

Membership level is measured against an Annual Revenue/POS Range Requirement, set per competency and per country schedule in the Competency Exhibit you sign. Point-of-sale data reaches NVIDIA through distribution, so when it lands late or lands wrong, the revenue credited toward your level is wrong too.

Slipping below your level

NVIDIA reviews membership levels annually and tracks program compliance on an ongoing basis. A partner that falls below the minimum requirements for its program, level or competencies may not be eligible for NPN benefits, and may be re-leveled or removed. NVIDIA does not publish when that takes effect, which is why you want to see the gap forming rather than hear about it at review.
What Rebates-On tracks for NVIDIA

The whole NPN program logic, maintained for you.

Rebates-On carries your NPN program logic so you don’t have to chase every NVIDIA update. One dashboard shows your incentive position by competency, your open deals and their allocation status, your level and competency progress, certification deadlines, and an order simulator that tells you what one more booking is worth toward the next requirement.

  • See your incentive position by competency - Compute, DGX AI Compute Systems, Networking, Visualization and the rest of your schedule - with Rebates-On tracking whichever NVIDIA incentive structures your agreement carries.
  • Track every open deal and its approval and allocation status, so a slipped lead time doesn’t quietly move revenue into a later period.
  • Watch your level and competency progress against the training and Annual Revenue/POS Range requirements in the Competency Exhibit you signed.
  • Catch the deadlines that put benefit eligibility at risk, including the training each signed Competency Exhibit requires, with alerts while you can still act.
  • Model the next move with the order simulator: what one more booking is worth toward the next requirement in your competency schedule.
Levels, requirements & certifications

Your standing is earned per competency - not company-wide.

Under NPN, you earn standing per competency, not on one company-wide score, and the level structure varies by partner type. Solution Providers qualify across Registered, Preferred and Elite, per competency, against the schedule in the Competency Exhibit they sign with NVIDIA. The AI Infrastructure programs (Data Center Partner, Power & Cooling, AEC) have two levels, Preferred and Elite, with new partners joining at Preferred. Elite is by invitation and subject to approval.

  1. Elite Invitation + requirements

    The top level: the highest training and revenue requirements per competency, and by invitation only, subject to NVIDIA’s approval. What Elite unlocks is set in your program agreement rather than published.

  2. Preferred Per competency

    A deeper investment: meet the training and revenue requirements set for that competency, at that level, on your country schedule. In the AI Infrastructure programs, Preferred is where new partners join.

  3. Registered Entry

    The entry level for many partner types: complete NVIDIA sales and technical training, and gain access to product, sales, and marketing tools.

Competency requirements are defined in the Competency Exhibit each partner signs with NVIDIA. They are set per competency, per membership level, and per country schedule, so your exact schedule depends on the country where your NPN Master Agreement is executed. Enablement runs on the NVIDIA Deep Learning Institute (DLI) and NVIDIA’s expanding certification portfolio, from Associate level (NCA-GENL, NCA-AIIO) through Professional level (NCP-AII, NCP-AIO, NCP-AIN) and beyond. EMEA partners also have the NVIDIA Partner Expert Program.

Because NVIDIA sets each level from the number of trained sales and technical staff you hold per competency, keeping that bench current is one of the most direct ways to protect your benefit eligibility. NVIDIA reviews membership levels annually and tracks compliance on an ongoing basis - exactly what the Rebates-On certifications module is built to watch.

Why partners use us for NVIDIA

How Rebates-On helps you avoid missed NVIDIA rebates.

NVIDIA incentive structures are set in the agreement you sign, and they differ by partner type, competency and country. Rebates-On tracks whichever structures apply to your agreement and handles the math, so nothing you have earned goes unclaimed.

  • No more tracking each competency in isolation. Your position across every competency you carry, held in one place against one calendar, so nothing gets forecast on its own.
  • No more manual period math. Rebates-On applies your agreement’s incentive structure to the revenue you have actually booked, so the running total is right at any point in the period, not just after the statement lands.
  • No more missed promotions. NVIDIA may make additional benefits and promotions available from time to time beyond the ones named in the program guide. Rebates-On flags the ones you qualify for and tracks them alongside everything else.
  • An NVIDIA number finance can rely on. Forecast your NVIDIA incentive income, then reconcile what NVIDIA actually paid against what you earned across each competency - instead of guessing at terms that were never published.
FAQ

NPN questions partners ask first.

The level structure depends on your partner type. Solution Providers qualify across Registered, Preferred and Elite, and they qualify per competency rather than company-wide. The AI Infrastructure programs (Data Center Partner, Power & Cooling, AEC) have two levels, Preferred and Elite, and new partners join at Preferred. Moving up means meeting the training and Annual Revenue/POS Range requirements defined in the Competency Exhibit you sign with NVIDIA, which are set per competency, per level, and per country schedule. Elite is by invitation and subject to NVIDIA’s approval on top of the requirements.
The specifics of what you earn are set in the agreement you sign, and NVIDIA does not publish them. What is clear from the documents partners are given: resale and rebate economics sit with the Solution Provider track, and standing is earned per competency, per level, and per country schedule. The AI Infrastructure programs are a different motion entirely, since NVIDIA states those programs do not support partners reselling NVIDIA products. Rebates-On tracks whichever incentive structures apply to your agreement.
NVIDIA regrouped partner types into four families - AI Infrastructure, Cloud, Influencer, and Seller/Builder - added data-center buildout types (Data Center Partner, Power & Cooling, AEC), and refreshed the competency schedules. The AI Infrastructure family is the part worth reading closely: those programs do not support partners reselling NVIDIA products, they carry a single competency (NVIDIA Technologies), and they run on two membership levels rather than three. If you assumed the whole network moved as one, that is the assumption to drop.
Approval criteria, validity windows and discount deltas are gated and set in your program agreement rather than published, so this page does not list them. Two things are worth knowing regardless: the published pricing benefit in the AI Infrastructure program guide is discounted pricing on specific demo and not-for-resale hardware, and on allocation-constrained Blackwell-class products, a won deal does not guarantee supply.
Almost all Solution Providers transact through NVIDIA-authorized distributors such as TD SYNNEX and Ingram Micro in a two-tier model. DGX systems require the DGX competency and flow through a restricted set of Elite partners, while OEM-branded HGX and Enterprise Reference Architecture servers are bought through Dell, HPE, Lenovo, and Supermicro programs with NVIDIA NPN benefits layered alongside - which is why the combined incentive picture is so hard to see.
Front-end margin on allocation-constrained GPUs and systems is compressed. Whatever your agreement carries on the back end, plus networking and storage attach, services, and NVIDIA AI Enterprise software, is where the rest of the economics sits. Partners who only track the front-end discount systematically understate - and under-collect - their true NVIDIA position.
Because nearly every lever is gated and moving: the terms sit in a signed agreement rather than a public rate card, and they vary per competency, per level and per country schedule; standing is measured per competency, not company-wide; OEM and distributor incentives stack on the same deal; point-of-sale data reaches NVIDIA through distribution; and allocation slips push attainment into a later period. That’s exactly the gap Rebates-On fills.

See exactly where your NVIDIA rebates stand - and what’s slipping.

Book a 30-minute demo, or start with an NVIDIA rebate audit that shows what you earned versus what NVIDIA actually paid.