Selecting the Right NIC Interface Card for High-Performance Servers

Ready to start learning? Individual Plans →Team Plans →

Picking a wireless nic card for a server sounds simple until the machine starts dropping packets, spiking CPU, or falling behind under load. The same logic applies to any high-performance NIC interface card: the best card is not the one with the biggest speed label, but the one that fits the workload, PCIe slot, driver stack, cooling budget, and switch design.

Featured Product

Cisco CCNA v1.1 (200-301)

Learn essential networking skills and gain hands-on experience in configuring, verifying, and troubleshooting real networks to advance your IT career.

Get this course on Udemy at the lowest price →

Quick Answer

A wireless nic card or wired NIC interface card for high-performance servers should be chosen by workload, not headline speed. The right adapter balances throughput, latency, PCIe lane width, driver support, offloading features, and thermal fit. In real deployments, a well-matched 10GbE or 25GbE card can outperform a poorly integrated faster adapter.

Definition

A NIC interface card is a network adapter that connects a server to a network by sending and receiving frames, exposing link status, and handling packet processing between the operating system and the physical network medium. In a high-performance server, the NIC is a performance component, not just a connectivity part.

Primary Decision FactorsWorkload, PCIe compatibility, driver support, offloading, cooling, and redundancy
Typical Speed Tiers1GbE, 10GbE, 25GbE, 40GbE, 100GbE
Best Fit ExamplesVirtualization hosts, database servers, storage nodes, east-west traffic environments
Common BottleneckPCIe lane width or driver quality, not the port speed itself
Key Features to CheckChecksum offload, segmentation offload, receive-side scaling, queue depth, RDMA support
Validation MethodBenchmark with real traffic patterns before production rollout
CCNA RelevanceInterfaces, traffic flow, switching, duplex, and troubleshooting fundamentals from Cisco CCNA v1.1 (200-301)

What Does a NIC Interface Card Do in a High-Performance Server?

A NIC interface card moves network traffic in and out of a server, but that plain description hides the real job. It translates network frames into a form the operating system can process, exposes link state to the host, and helps determine how efficiently packets reach applications. That makes it directly relevant to performance, low latency, and reliability.

In a busy server, the NIC also affects CPU usage. A weak adapter may force the host to spend more time processing interrupts and packet bursts, while a stronger card with better offloading can reduce overhead. That is why NIC selection matters for database servers, virtualization hosts, and storage systems just as much as it matters for raw network access.

Onboard NICs vs PCIe Add-In Cards

Onboard NICs are built into the motherboard or server platform. They are convenient, usually adequate for general-purpose workloads, and often a good choice for management traffic, light application servers, or internal services with modest bandwidth needs.

PCIe add-in cards are separate adapters installed in expansion slots. They are preferred when you need higher port counts, more speed, better queue handling, or special capabilities like RDMA. In high-performance servers, PCIe cards are usually the serious choice because they give you more control over speed, lane width, and feature set.

Why Server NICs Affect More Than Connectivity

A server NIC influences packet handling efficiency, application responsiveness, and how hard the CPU has to work. If the card supports advanced queueing and offloading, the server can handle more traffic without choking on interrupts or spending excess time on repetitive packet work.

This is why a “faster” card can disappoint in practice. A 25GbE adapter with poor drivers or a bottlenecked PCIe slot may perform worse than a well-tuned 10GbE model in the same environment.

In server design, the NIC is not just a port. It is part of the data path, and the data path is where latency, throughput, and stability are won or lost.

For networking students working through Cisco CCNA v1.1 (200-301), this is where theory becomes practical. Understanding interfaces, switching, and traffic flow helps you see why the server side of the network is just as important as the switch port.

How Does a NIC Interface Card Work?

A NIC interface card works by moving frames between the host and the network while coordinating with the operating system, driver, and switch. The process is simple at a high level, but performance depends on how efficiently each step is handled.

  1. The host sends traffic to the driver. The operating system hands data to the NIC driver, which queues packets for transmission or receives packets from the wire.
  2. The NIC handles frame processing. The adapter builds or parses Ethernet frames, checks metadata, and uses hardware features to reduce host work when possible.
  3. The adapter communicates with the switch. Link negotiation, duplex mode, speed, and error handling determine whether traffic flows cleanly across the network.
  4. The NIC exposes status and counters. Administrators can review link state, drops, errors, and utilization to diagnose problems early.
  5. The CPU and memory subsystem share the load. With the right offloading and queue design, the NIC reduces bottlenecks that would otherwise hit the processor.

This is where link speed becomes only one part of the story. A card may advertise 25GbE, but if the workload is latency-sensitive or bursty, queue handling and interrupt moderation may matter more than the headline number.

Pro Tip

Check NIC statistics during testing, not after deployment. Errors, dropped packets, retransmissions, and queue overflows often reveal a mismatch between the adapter, the driver, and the workload before users notice a slowdown.

What Key Technical Terms Should You Know Before Choosing a NIC?

Before comparing hardware, you need the basic terms used in NIC specification sheets and server troubleshooting. These terms explain why one adapter behaves better than another even when the spec page looks similar.

  • Bandwidth is the maximum amount of data a link can move over time. It matters for peak transfer rate, but high bandwidth alone does not guarantee smooth application performance.
  • Latency is the delay between sending and receiving data. Low latency matters for databases, storage, real-time services, and interactive systems.
  • Duplex describes whether a port can send and receive traffic at the same time. Modern server links are typically full duplex, and mismatches can create severe problems.
  • Offloading is when the NIC performs tasks that would otherwise consume CPU cycles, such as checksum calculation or packet segmentation.
  • Queue depth is the number of packets the adapter can hold in line for processing. Higher queue capacity can improve burst handling.
  • Packet processing is the work involved in receiving, classifying, and forwarding frames through the host stack.

Bandwidth is easy to oversell because it is a single number. The problem is that real server traffic is rarely smooth. Virtual machines burst, databases issue many small requests, and backup jobs can flood a link at the worst possible time.

That is why advanced server teams look at queue behavior, interrupt handling, and counters instead of relying on the marketing label on the box. The right nic card should improve application behavior, not just move a speed test higher for a few seconds.

How Do You Match the NIC to the Workload?

The right NIC depends on what the server does most of the day. A virtualization host, a database node, and a storage server can all need different network characteristics, even if they sit in the same rack. Workload profiling should happen before purchase, not after the first performance complaint.

Virtualization Hosts

Virtualization environments often need strong packet-per-second performance, efficient queue handling, and good scalability. A host running many virtual machines may see lots of small flows rather than one huge transfer stream. In that situation, queueing and CPU efficiency matter more than pure top-end bandwidth.

Database and Application Servers

Database servers usually benefit from stable throughput and lower latency rather than simply chasing the fastest port available. A card with excellent driver support and low overhead can produce a better experience than a higher-speed model that increases jitter or CPU pressure. If the application is sensitive to response time, consistency matters more than a one-time benchmark win.

Storage and East-West Traffic

Storage-heavy servers and east-west rack traffic need predictable transfer behavior. Large file movements, replication jobs, and cluster communication can punish weak adapters. In those cases, efficient offloading and sustained throughput are the right priorities.

Virtualization Prioritize queue handling, packet-per-second performance, and driver maturity
Database Servers Prioritize latency consistency, stable throughput, and CPU efficiency
Storage Nodes Prioritize sustained bandwidth, offloading, and thermal stability

If you are studying networking through Cisco CCNA v1.1 (200-301), this workload-first approach is exactly the kind of thinking that builds better troubleshooting habits. It ties traffic patterns to interface behavior instead of treating the network as abstract cables and numbers.

How Do You Choose the Right Speed and Port Configuration?

The right speed depends on traffic volume, application behavior, and switch capacity. A wireless nic card is not the right comparison here for server use, but the same principle applies: speed is only useful when the rest of the system can keep up. For servers, common tiers include 1GbE, 10GbE, 25GbE, 40GbE, and 100GbE.

When One Fast Port Is Better

One faster port is often the better option when you want simple cabling, fewer failure points, and a cleaner traffic design. A single 25GbE port can be easier to manage than two 10GbE ports when the goal is straightforward high-throughput access to a storage or application network.

When Multiple Ports Make More Sense

Multiple ports help when you need redundancy, traffic separation, or aggregate throughput. A dual-port adapter can separate management and application traffic, or provide failover if one switch path fails. In environments with link aggregation, multiple slower ports may also offer design flexibility.

The tradeoff is real. Multiple ports increase switch port usage, cabling complexity, and troubleshooting overhead. One faster port reduces complexity, but only if the switch and server platform support it cleanly.

Here is the practical rule: choose the minimum port design that meets throughput, resilience, and segmentation requirements without wasting switch resources.

Note

Do not size a NIC in isolation. A 25GbE card on a 10GbE switch or a heavily oversubscribed uplink is a planning mistake, not an upgrade.

Why Do PCIe Compatibility and Bus Bottlenecks Matter?

A NIC can only perform as well as the bus that feeds it. If the PCIe slot is too narrow, wired incorrectly, or limited by platform design, the adapter may never reach its advertised capability. That is why PCIe generation and lane width are not small details; they are core performance constraints.

Server documentation should always be checked before purchase. A physically compatible card may still be electrically constrained, and a strong adapter installed in the wrong slot can create a bottleneck that looks like a network problem even though the issue is on the host side.

What to Verify Before You Buy

  • PCIe generation supported by the server motherboard or riser.
  • Lane width available in the intended slot.
  • Physical clearance for the card, bracket, and nearby components.
  • Thermal path so the adapter can move air correctly under load.
  • Firmware compatibility if the NIC requires a specific platform revision.

This is one of the most common mistakes in high-performance server builds. Teams buy for port speed and discover later that the platform cannot feed the adapter fast enough. The result is wasted budget and disappointing real-world throughput.

The lesson is simple: a fast NIC on a weak bus is still a weak design.

How Important Are Driver Quality and Operating System Support?

Driver quality is often the difference between an adapter that feels stable and one that causes mysterious performance swings. The same NIC can behave very differently depending on the operating system, kernel version, firmware, and vendor driver maturity.

Before deployment, review the official vendor documentation for supported operating systems and required driver versions. Microsoft documents network adapter support through Microsoft Learn, while Cisco publishes platform and interface guidance through its official documentation at Cisco. For Intel-based adapters, the vendor’s own network adapter docs are the right starting point, not third-party forum guesses.

What Bad Driver Support Looks Like

  • Reduced throughput under load
  • Higher CPU usage than expected
  • Link flapping or inconsistent negotiation
  • Feature loss, such as missing offload options
  • Unstable behavior after OS updates or kernel changes

Testing in a lab or pilot environment is essential. A NIC that looks perfect on a compatibility sheet can still fail in a specific hypervisor, storage stack, or patched OS environment. That is especially true in mixed fleets where one server platform is a generation older than the rest.

For readers building foundational networking skills, this is a practical extension of the Cisco CCNA v1.1 (200-301) mindset: know the interface, verify the link, and confirm behavior before assuming the hardware is fine.

Which Offloading and Queueing Features Actually Matter?

Offloading is the set of hardware tasks a NIC performs so the CPU does less work. That matters most on busy hosts where the processor is already handling virtualization, storage, security controls, and application logic. In the right environment, offloading improves efficiency without changing the application itself.

Common features include checksum offload, segmentation offload, and receive-side scaling. Each one solves a different bottleneck. The goal is not to collect features; the goal is to reduce host overhead where it hurts most.

Feature-by-Feature Breakdown

  • Checksum offload lets the NIC calculate packet checksums instead of the CPU doing it repeatedly.
  • Segmentation offload allows larger data blocks to be split into network-sized segments by hardware.
  • Receive-side scaling spreads incoming traffic across multiple CPU cores and queues.
  • Advanced queueing helps servers handle many simultaneous connections or many virtual machines more efficiently.
  • RDMA support can reduce latency and CPU overhead in specialized storage or clustering workloads.

The key is to match these features to the problem. If your bottleneck is not CPU saturation, a long list of acceleration features may not improve anything. On the other hand, a virtualization host or storage server can benefit a lot from the right queue model and offload combination.

The best NIC features are the ones that solve your actual bottleneck, not the ones that look strongest in a specification table.

How Do Reliability and Redundancy Change the NIC Decision?

A reliability-focused NIC choice is about uptime, not just throughput. If a server must stay online during switch maintenance, cable failure, or adapter replacement, then port count, failover design, and link monitoring become selection criteria, not extras.

Dual-port and multi-port adapters are especially useful when you want resilience without adding extra hardware. They can separate traffic paths, reduce downtime during maintenance, and provide fallback if one link fails. In a production environment, that flexibility often matters more than a small gain in peak bandwidth.

Operational Benefits of Redundancy

  • Maintenance can happen with less service interruption
  • Failover paths reduce the risk of a single cable failure causing outage
  • Traffic can be separated for storage, management, and production roles
  • Monitoring becomes easier when link state and counters are visible per port

Redundancy is not free. It adds cabling, switch ports, and design complexity. Still, for critical servers, the added resilience is usually worth the overhead. Monitoring link status and error counters should be part of normal operations, because small faults often show up before major outages.

For a broader infrastructure perspective, this is why server networking and switch design should be planned together. A NIC is only as resilient as the network path behind it.

Why Do Cooling, Power, and Physical Fit Matter So Much?

High-performance NICs generate heat, and heat affects stability. In dense racks, airflow can be uneven, and a card that performs well in a lab can throttle or become unstable in a hot production chassis. Power draw also matters because it influences thermal design and overall server headroom.

Before deployment, check the card’s physical size, bracket style, clearance needs, and any airflow assumptions. A tall heatsink, nearby GPU, or crowded riser can interfere with cooling. If the chassis is already near its thermal limit, adding a high-end adapter can create a new bottleneck that no benchmark from the lab will reveal.

What to Review in the Chassis

  • Airflow direction and whether the NIC sits in the main cooling path
  • Thermal headroom in a densely populated server
  • Power budget for the adapter and surrounding components
  • Bracket and slot fit to avoid physical installation issues
  • Rack density if multiple hot components are installed together

Overheating does not always show up as a clean failure. It can present as intermittent slowness, link instability, or degraded performance under sustained load. That makes cooling a networking issue as much as a hardware issue.

In practical terms, a cooler, less ambitious NIC can outperform a hotter, more powerful one if the chassis cannot support the thermal load.

What Should You Look for in Vendor Documentation and Real-World Validation?

Always start with official specifications. Vendor documentation tells you what the adapter is supposed to support, but it does not guarantee the behavior you will see in your environment. That is why real-world validation matters.

Use official product pages, installation guides, and compatibility matrices from the adapter vendor, and then test against your actual operating system and workload. For networking fundamentals and interface behavior, Cisco documentation is useful context, while Microsoft Learn is the correct source for Windows-based driver and platform guidance.

A Practical Validation Process

  1. Check official specs. Confirm speed, port count, PCIe requirements, and supported features.
  2. Verify OS support. Confirm the driver version and platform support for the exact server OS or hypervisor.
  3. Test with realistic traffic. Use representative application traffic, not only synthetic bandwidth tools.
  4. Measure CPU and error counters. Watch for signs of host overhead or link instability.
  5. Scale only after the pilot passes. A small validation failure is much cheaper than a fleet-wide mistake.

Community feedback can help identify recurring issues, but it should never replace vendor documentation. A deployment note from a similar server model is useful when it matches your OS, firmware, and workload profile. If those details differ, treat the report as a clue, not a guarantee.

How Do Cisco CCNA v1.1 Concepts Help You Make Better NIC Decisions?

Cisco CCNA v1.1 (200-301) reinforces the networking fundamentals that make NIC selection easier to reason about. If you understand interfaces, switching, duplex, and traffic flow, you are much less likely to buy hardware based only on the speed printed on the label.

That knowledge helps in three ways. First, it improves troubleshooting, because you know how to check link state, port behavior, and packet movement. Second, it improves procurement, because you can explain why a certain port speed or queue design fits the workload. Third, it improves collaboration between server, storage, and network teams, which is where many performance problems are actually solved.

Why the Fundamentals Matter

  • Interfaces tell you how the server connects to the switch.
  • Traffic flow helps you understand where bottlenecks form.
  • Switching behavior matters when ports are aggregated or segmented.
  • Duplex and link status help identify misconfiguration and instability.

A NIC selection conversation becomes much better when it is grounded in real networking terms instead of vague requests for “the fastest card.” That is the kind of practical literacy Cisco CCNA training helps build, and it applies directly to high-performance server design.

How Should You Compare NIC Options Side by Side?

Comparing NICs side by side helps you avoid choosing a card based on one impressive feature while missing a hidden limitation. A simple comparison table is often the fastest way to align procurement, engineering, and operations around the same requirements.

Onboard NIC Best for basic connectivity, lower cost, and simpler deployments; usually limited in performance tuning and expansion
Standard PCIe NIC Best balance of speed, driver support, and flexibility for most high-performance server workloads
Specialized High-Performance Adapter Best for demanding storage, virtualization, or latency-sensitive environments where offloading and queueing matter most

Evaluation Criteria That Matter Most

  • Speed — Does the port rate match real workload demand?
  • Port count — Do you need one fast link or multiple separated paths?
  • PCIe requirement — Can the server platform actually feed the card?
  • Driver support — Is the exact OS and kernel version supported?
  • Offloading — Will hardware acceleration reduce CPU pressure?
  • Thermal demand — Can the chassis cool the adapter under load?

Scoring NICs this way keeps the discussion practical. It also makes it easier to explain why a lower-speed adapter may be the better fit for a real server workload than a top-end model with hidden tradeoffs.

What Is the Best Step-by-Step Framework for Choosing a NIC?

The best NIC selection process starts with the workload and ends with validation. If you skip any of those steps, you risk buying the wrong card or discovering the limitation after deployment.

  1. Define the workload goal. Decide whether the server needs lower latency, higher throughput, more connections, or better redundancy.
  2. Check platform constraints. Confirm PCIe lane width, slot availability, chassis clearance, and thermal headroom.
  3. Verify OS and driver support. Use official documentation and confirm the exact version you plan to run.
  4. Match features to the need. Choose offloading, queueing, or RDMA only when they solve a real problem.
  5. Pilot the card. Test with realistic traffic and measure CPU usage, packet loss, latency, and error counters.
  6. Roll out only after success. Expand only if the adapter performs well in the target environment.

This framework works because it treats the NIC as part of the server system, not a standalone accessory. That mindset prevents the most expensive mistakes, especially in environments where one bad hardware choice can ripple across an entire fleet.

What Common Mistakes Do Buyers Make When Choosing a NIC?

The most common mistake is buying based only on advertised speed. A card that looks better on paper may be constrained by PCIe, driver problems, or thermal issues once it enters a real server.

Another mistake is ignoring the operating system until after installation. Driver support is not a footnote. It directly affects throughput, stability, and feature availability. A third mistake is selecting advanced features because they sound impressive, even when the workload does not benefit from them.

  • Buying for headline speed only instead of workload fit
  • Ignoring PCIe limits and slot wiring until performance disappoints
  • Skipping thermal planning in dense server builds
  • Overlooking driver support and OS compatibility
  • Failing to test real traffic before standardizing the hardware

These mistakes are avoidable. The fix is not more guesswork; it is a disciplined evaluation process grounded in the actual server environment. When teams test before rollout, they save time, money, and production stability.

Featured Product

Cisco CCNA v1.1 (200-301)

Learn essential networking skills and gain hands-on experience in configuring, verifying, and troubleshooting real networks to advance your IT career.

Get this course on Udemy at the lowest price →

What Is the Practical Bottom Line for High-Performance Server NIC Selection?

Choosing the right wireless nic card for a server, or more accurately the right wired NIC interface card for a high-performance server, is a system decision. You are not just buying speed. You are choosing how the server handles traffic, CPU load, redundancy, cooling, and future growth.

The best results come from matching the adapter to the workload, verifying PCIe and OS compatibility, and validating with real traffic before deployment. If you remember nothing else, remember this: the right NIC improves performance only when it fits the entire server ecosystem.

Key Takeaway

  • A NIC interface card affects throughput, latency, CPU usage, and reliability, not just connectivity.
  • Workload fit matters more than headline speed for servers running virtualization, databases, or storage traffic.
  • PCIe lane width, driver quality, and thermal headroom can bottleneck an otherwise powerful adapter.
  • Offloading and queueing features help only when they match the actual performance problem.
  • Validate every NIC choice with vendor documentation and real-world testing before broad rollout.

If you are building your networking foundation through Cisco CCNA v1.1 (200-301), this is exactly the kind of practical thinking that pays off in production. The more you understand interfaces, traffic flow, switching, and link behavior, the better your server and network decisions become.

CompTIA®, Cisco®, Microsoft®, AWS®, EC-Council®, ISC2®, ISACA®, and PMI® are trademarks of their respective owners.

[ FAQ ]

Frequently Asked Questions.

What are the key factors to consider when choosing a NIC interface card for high-performance servers?

When selecting a NIC for high-performance servers, it is essential to evaluate several critical factors beyond raw speed. First, ensure compatibility with your server’s PCIe slot version and bandwidth, as a mismatch can create bottlenecks despite high theoretical throughput.

Secondly, consider the driver quality and manufacturer support, as robust drivers can significantly impact stability, latency, and feature support. Proper cooling and physical design are also vital to prevent thermal throttling, which can degrade performance under heavy loads.

  • Throughput capacity aligned with your workload requirements
  • Compatibility with server hardware and PCIe specifications
  • Driver support and manufacturer reliability
  • Thermal management and physical design considerations

By carefully evaluating these factors, you can select a NIC that not only achieves high speeds but also maintains reliability and scalability for demanding server environments.

How does PCIe slot compatibility impact NIC performance in servers?

The compatibility between the NIC and the server’s PCIe slot is crucial because PCIe bandwidth limits can bottleneck network performance. Even if a NIC supports high data rates, installing it in a PCIe slot with insufficient bandwidth can reduce actual throughput.

For example, a 10GbE NIC installed in a PCIe 3.0 x8 slot will perform differently than in a PCIe 3.0 x16 slot, due to the available lane bandwidth. Ensuring that the NIC matches the PCIe version and lane count of your server’s motherboard is essential to maximize performance and avoid unexpected bottlenecks.

Consult your server’s hardware specifications to verify PCIe slot capabilities before choosing a NIC. Upgrading the server’s PCIe slots might be necessary if your workload demands higher throughput or lower latency.

Why is driver support important when selecting a network interface card for servers?

Driver support plays a vital role in the overall performance, stability, and feature set of a NIC. Well-supported drivers ensure that the hardware operates efficiently, reduces latency, and maintains reliability under high loads.

Subpar driver support can lead to issues such as packet loss, increased CPU utilization, or system crashes, which are detrimental in high-performance server environments. Manufacturers with a history of providing regular driver updates and good technical support can help mitigate these risks.

It’s also important to verify that the driver is compatible with your server’s operating system and that it supports advanced features like offloading, virtualization, and QoS, which are often critical in enterprise deployments.

How does thermal management influence NIC performance and longevity?

Thermal management is a critical aspect of NIC selection because excessive heat can cause thermal throttling, reducing the card’s performance during sustained high workloads. Proper cooling solutions, such as heatsinks or active fans, help maintain optimal operating temperatures.

Overheating not only impacts performance but also shortens the lifespan of the NIC hardware. Ensuring that the server chassis has adequate airflow and that the NIC design features efficient heat dissipation can prevent thermal issues.

In high-density or rack-mounted server environments, selecting NICs with advanced cooling features and planning for proper airflow is essential to maintain consistent performance and hardware reliability over time.

What are common misconceptions when choosing NIC interface cards for high-performance servers?

A common misconception is that the fastest NIC on paper will always deliver the best real-world performance. In reality, factors like driver support, PCIe compatibility, and thermal management greatly influence actual throughput and latency.

Another misconception is that more ports or higher speeds automatically translate to better scalability. While these features are important, they must align with the server’s architecture and workload demands to be effective.

Additionally, some believe that enterprise-grade NICs are unnecessary in smaller setups, overlooking that high-performance workloads benefit from robust features like offloading, virtualization support, and advanced error handling, which are more prevalent in higher-end models.

Related Articles

Ready to start learning? Individual Plans →Team Plans →
Discover More, Learn More
Choosing the Right Storage Solution for High-Performance Servers Discover how to select the best storage solutions for high-performance servers to… Python vs C++ for High-Performance AI Computing: Choosing the Right Tool for Scalable Intelligence Discover how to choose between Python and C++ for AI computing to… Selecting the Best Linux Distribution for Enterprise Servers Discover how choosing the right Linux distribution can prevent outages, ensure compliance,… CISM vs CISSP: Which Cybersecurity Certification is Right for You? Explore the differences between CISM and CISSP certifications to determine which aligns… Adobe InDesign vs Canva: Which is Right for Your Design Needs? Discover which design tool fits your workflow best and boost your productivity… CISA vs CISM: Choosing the Right Certification for Your Career Explore the differences between CISA and CISM certifications to choose the best…
FREE COURSE OFFERS