NVIDIA Partner Network

Manage NVIDIA Partner Network rebates, deal registration, competencies and certifications in one dashboard.

NVIDIA rebuilt the Partner Network - partner types regrouped into four families, competencies renamed, five invitation-only Specializations added - and locked nearly every rate inside the NPN Portal. Rebates-On tracks all of 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 revenue trending upward
The program today

What is the NVIDIA Partner Network (NPN)?

The NVIDIA Partner Network (NPN) is NVIDIA’s single, invitation-gated enterprise channel program for everyone who sells, integrates, hosts, builds, or advises on NVIDIA products. Since its latest refresh, partner types sit in four families and you qualify across three levels - Registered, Preferred, Elite. Your earnings live in back-end rebates tied to competencies, so your earnings follow what you carry.

Two ideas drive NPN. First, it is a curated, sell-with program rather than a self-service one: you apply, NVIDIA’s partner organization decides which partner type and competency fit you, and the top Elite tier is invitation-only on top of meeting thresholds. Second, what you can resell - and which rebates apply - is gated by competency: Compute, DGX AI Compute Systems, Embedded Compute, Networking, NVIDIA Enterprise Software, NVIDIA Technologies, and Visualization.

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.

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; competencies were renamed and updated; and five invitation-only Specializations went live. Your earnings still ride on back-end rebates and bonuses - but now across a portfolio that runs from edge to AI factory, so old assumptions miss income.

  • 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, Education Services), and Seller/Builder (Solution Provider, Distributor, OEM, System Partner, Solution Advisor, GSI, ISV, Storage). The Solution Provider remains the core rebate-earning population.
  • Five invitation-only Specializations. Agentic AI, AI Factory, Embedded AI, Physical AI, and Reference Platform NCP. These are badges layered on top of competencies, signaling NVIDIA’s move to reward practice-building over resale volume.
  • Competencies renamed and expanded. DGX AI Compute Systems, NVIDIA Enterprise Software (now including Run:ai), and Networking (NVLink, InfiniBand, Spectrum-X, absorbed from the legacy Mellanox PartnerFIRST program).
  • Allocation, not demand, is the constraint. Hyperscaler forward orders for Blackwell-class systems have consumed most supply well into the future, squeezing channel allocation - registering and winning a deal does not guarantee product.
Feb–Jan
NVIDIA fiscal year

NVIDIA’s fiscal year runs February–January, misaligned with most partners’ January–December books - so calendar-year rebate forecasts routinely mis-time Q4/Q1 attainment.

Where the money leaks

Where NVIDIA partners leave rebate money on the table.

NVIDIA partners bleed rebate revenue in five predictable places: competency-by-competency fragmentation, two-vendor incentive stacking on OEM AI servers, allocation slips that push deals across fiscal quarters, late point-of-sale reporting, and expired MDF. Each is a deal that looks profitable on the quote while the rebate quietly never lands - something software can watch that a spreadsheet cannot.

Competency fragmentation

Back-end rebates pay separately by competency - compute, DGX, networking and visualization each have their own rules. A single deal can span several of them, so a line-item misclassification silently forfeits rebate dollars.

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.

Deal registration vs. allocation

Registering and winning a deal doesn’t secure Blackwell-class supply. Quotes expire, prices move, and lead-time slips push revenue and rebate attainment across NVIDIA’s Feb–Jan quarter boundaries, costing the quarterly bonus and customer goodwill.

Point-of-sale reporting gaps

Rep-level NVIDIA Sales Rewards - and, per NVIDIA’s own terms, program standing - hinge on weekly, accurate POS reporting. Missing or late POS means forfeited credits, with only short dispute windows, and the same plumbing underlies partner-level rebate calculation through distribution.

Inception pass-throughs

Resellers fronting the NVIDIA Inception startup discount must reconcile NVIDIA’s credit; unclaimed or mismatched Inception credits are straight margin loss.

MDF expiry and tier slippage

Dedicated Elite MDF and proposal-based funds run on use-it-or-lose-it cycles. And because Elite is invitation plus threshold, an allocation-constrained year can drop a partner below revenue thresholds - losing the fixed rebate and dedicated MDF the following year.

Split shipments that pay out short

When a single order ships in parts across a quarter, the rebate is earned only on the portion that actually shipped - and those split amounts are easy to miscount, so deals that straddle a quarter-end quietly pay out below what was earned.
What Rebates-On tracks for NVIDIA

The whole NPN program logic, maintained for you.

Rebates-On carries the full NPN program logic so you don’t have to read every NVIDIA portal update. One dashboard shows your back-end rebate position by competency, every deal registration and its allocation status, your level and Specialization progress, MDF balances, certification deadlines, and an order simulator that tells you what one more booking is worth toward a threshold.

  • See your back-end rebate position by competency - compute, DGX, networking and visualization - with your estimated rebate this quarter against NVIDIA’s Feb–Jan fiscal calendar.
  • Track every deal registration and its approval and allocation status, so a slipped lead time doesn’t quietly move revenue across a quarter boundary.
  • Watch your level and Specialization progress - Registered, Preferred, Elite, plus eligibility signals for AI Factory, Agentic AI and the other invitation-only badges.
  • Manage NVIDIA MDF activities and claims before dedicated or proposal-based funds expire.
  • Catch the deadlines and dependencies that forfeit rebate dollars - POS reporting cadence, dispute windows, Inception credit reconciliation - with alerts while you can still act.
  • Model the next move with the order simulator: what one more booking is worth toward crossing a competency threshold or qualifying for a level.
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. You start Registered, reach Preferred by hitting training and revenue goals, and reach Elite at the highest thresholds plus an NVIDIA invitation. Elite is where the real rebate dollars sit - fixed back-end rebates, dedicated MDF, and, in practice, better allocation. Exact thresholds live only inside the NPN Portal.

  1. Elite Invitation + thresholds

    The top tier: highest training and revenue thresholds per competency plus an NVIDIA invitation. Elite unlocks fixed back-end rebates, dedicated MDF, named technical resources, and stronger big-deal participation.

  2. Preferred Per competency

    A deeper investment: meet training, revenue, and other goals per competency to unlock additional benefits, including proposal-based MDF, priority support, and better incentives.

  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.

Requirements vary by partner type, level, and competency, and the exact revenue targets and certified-individual counts are gated inside the NPN Portal - so treat any specific threshold numbers seen on third-party sites as unverified. 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 certifications are a heavy input to competency standing, keeping them current is one of the most direct ways to protect your rebate eligibility - 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’s mix of standard Fixed rebates and cumulative Quarterly Performance Bonuses (QPB) is notoriously complex. Rebates-On handles the complicated math, so nothing you have earned goes unclaimed.

  • No more juggling Fixed and QPB separately. Your standard Fixed rebates and your annual QPB targets tracked together in one place, so nothing is forecast in isolation.
  • No more QPB guesswork. Because the QPB is an annual target paid quarterly, Rebates-On calculates each quarter’s exact payout from what you’ve already earned - so the running total is always right.
  • No more missed Boost and Inception promos. Limited-time Boost and Inception promotions flagged and tracked automatically, so you capture every extra incentive without the manual guesswork.
  • An NVIDIA number finance can rely on. Forecast rebate and bonus income, then reconcile what NVIDIA actually paid against what you earned across each competency - instead of guessing at portal-gated terms.
FAQ

NPN questions partners ask first.

NPN has three levels - Registered, Preferred, and Elite. Registered requires completing NVIDIA sales and technical training. Preferred adds revenue and training goals per competency and unlocks additional benefits. Elite requires the highest thresholds plus an invitation from NVIDIA, and unlocks fixed back-end rebates, dedicated MDF, and named technical resources. Exact revenue and certification thresholds are published only inside the NPN Portal, so any numbers on third-party sites should be treated as unverified.
Yes. Elite-level Solution Provider and Solutions Integration partners earn a fixed back-end rebate by competency - compute, DGX, networking and visualization - typically alongside a quarterly performance bonus measured against an annualized goal. “Fixed” means a set rate on eligible revenue rather than a goal-attainment kicker. Rates are not public; rebate history is visible in the NPN Portal.
NVIDIA regrouped partner types into four families - AI Infrastructure, Cloud, Influencer, and Seller/Builder - added data-center buildout types (Data Center, Power & Cooling, AEC), refreshed competencies (DGX AI Compute Systems, NVIDIA Enterprise Software with Run:ai, Networking with Spectrum-X and NVLink), and introduced five invitation-only Specializations: Agentic AI, AI Factory, Embedded AI, Physical AI, and Reference Platform NCP.
Partners register opportunities in the NPN Portal (networking deals historically ran through the legacy Mellanox PartnerFIRST portal, now folded in). Approved registrations carry preferential pricing and a degree of deal protection. The caveats: approval criteria, validity windows, and discount deltas are portal-gated, and on allocation-constrained Blackwell-class products, registration 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. NVIDIA’s partner economics deliberately push profit to back-end rebates, the quarterly bonus, MDF, networking and storage attach, services, and NVIDIA AI Enterprise software. Partners who only track the front-end discount systematically understate - and under-collect - their true NVIDIA economics.
Because nearly every lever is gated and moving: rates and eligible SKUs live in the NPN Portal and shift with NVIDIA’s Feb–Jan fiscal calendar; rebates pay separately across several competencies with different rules; OEM and distributor incentives stack on the same deal; rep-level Rewards and rebate calculation both depend on weekly POS reporting; and allocation slips push attainment across quarters. 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.