NVIDIABest for AI inference
L4
LLM and RAG inference, computer vision and video analytics at very low power
- GPU memory
- 24 GB GDDR6
- Power
- 72 W
- Form factor
- Single-wide
- Per server
- Up to 5
Max per server120 GB
Get pricing with this GPU
16th Generation PowerEdge · AI-ready tower server
Dell PowerEdge T560The AI-ready tower server for offices and branches without a data center.
Up to two Intel Xeon Scalable processors, 1.5 TB of DDR5, up to 24 drives and up to five NVIDIA L4 GPUs in a quiet 4.5U tower. Run private AI, ERP and virtualization on-premises. Configured, quoted and delivered by Forte Tech across Egypt and the GCC.
- Genuine Dell hardware
- Egypt & GCC delivery
- Local pre-sales engineers
- Tender & PO ready
Intel® Xeon® Scalable4th & 5th Gen, up to 2 CPUs
AI-ready towerUp to 5 × NVIDIA L4 or 2 × double-wide GPUs
2× Xeon
4th Gen (up to 32 cores) or 5th Gen (up to 28 cores) Intel Xeon Scalable
1.5TB
DDR5 across 16 RDIMM slots, up to 5200 MT/s
5× L4
NVIDIA L4 GPUs, or 2 × 300 W double-wide cards
24drives
2.5" front bays, up to 360 TB, or 12 × 3.5"
Built for
One tower for AI and everyday IT
Configure the T560 for GPU inference, virtualization, ERP and databases, or video surveillance and backup, all in one office-friendly server.
On-premises AI inference
Run private LLMs, RAG assistants and computer vision on up to five NVIDIA L4 GPUs, inside your own office.
Virtualization for branches
Up to 64 cores and 1.5 TB of DDR5 host VMware, Hyper-V or Proxmox for a whole branch or SME.
ERP, databases & file services
Xeon Scalable CPUs, NVMe and up to 24 drives for SQL Server, SAP Business One, Odoo and file shares.
Video surveillance & backup
Up to 12 × 3.5" drives (180 TB) for NVR recording, backup targets and archives, plus GPU video analytics.
AI & GPU acceleration
Bring AI into your office, not the cloud
The T560 is one of the easiest ways for a company in Egypt or the GCC to run AI on its own hardware. Add NVIDIA GPUs to a quiet tower and keep sensitive documents, customer data and camera footage inside your building.
- NVIDIA GPUs: up to 5 × NVIDIA L4 or up to 2 × 300 W double-wide cards such as the L40 or A16, in two PCIe Gen5 x16 slots.
- Intel AMX built in: Xeon Scalable processors accelerate smaller models, embeddings and pre- and post-processing directly on the CPU.
- Room for models and data: up to 1.5 TB of DDR5, NVMe drives for models and vector databases, and up to 360 TB for datasets and video.
- No data center required: an office-friendly tower with redundant power supplies up to 2800 W, ready for branches and SMEs.
Private LLMs & RAG
Arabic and English assistants that answer from your own policies, contracts and documents, with the data never leaving your building.
CCTV & video analytics
People counting, license-plate recognition and safety alerts across dozens of camera streams with NVIDIA L4 GPUs.
Document AI & OCR
Extract data from invoices, IDs and contracts automatically for banks, insurers, hospitals and government offices.
VDI & graphics
GPU-accelerated virtual desktops, CAD and design workstations for small teams with NVIDIA A16 or L40.
GPU (VGA card) options for the T560
NVIDIA data-center GPUs that fit the T560. Choose one and we'll quote a complete, compatible build.
NVIDIAMost memory per card
L40
Larger language models, generative AI, rendering and 3D graphics
- GPU memory
- 48 GB GDDR6
- Power
- 300 W
- Form factor
- Double-wide
- Per server
- Up to 2
Max per server96 GB
Get pricing with this GPU
NVIDIAVDI
A16
GPU-accelerated virtual desktops for many concurrent users
- GPU memory
- 64 GB GDDR6
- Power
- 250 W
- Form factor
- Double-wide
- Per server
- Up to 2
Max per server128 GB
Get pricing with this GPU
NVIDIAEntry
A2
Entry-level inference, OCR and edge AI on a budget
- GPU memory
- 16 GB GDDR6
- Power
- 60 W
- Form factor
- Single-wide
- Per server
- Up to 6
Max per server96 GB
Get pricing with this GPU
Not sure which GPU?
Tell us your model, data and number of users. Our engineers will recommend the right GPU and a complete, compatible build.
Ask our engineersSizing rule of thumb: a model needs about 2 GB of GPU memory per billion parameters at FP16, or about 1 GB at FP8/INT8, plus headroom for context and users. For example, an 8B model fits on one L4 (24 GB), and a quantized 70B model fits on 5 × L4 (120 GB) or 2 × L40 (96 GB).
Validated builds: GPU configurations need specific risers, fans and power supplies, and supported cards and quantities depend on the configuration. Our engineers check every GPU build against Dell's rules before quoting.
What's inside
16th-generation PowerEdge in a tower
The T560 brings data-center features, PCIe Gen5 and DDR5 to a tower that fits in any office.
Compute
Up to two Xeon Scalable CPUs
4th Gen Intel Xeon Scalable with up to 32 cores, or 5th Gen with up to 28 cores, each with built-in Intel AMX for AI.
Memory
Up to 1.5 TB DDR5
16 RDIMM slots at up to 4800 MT/s (4th Gen) or 5200 MT/s (5th Gen) to hold models, embeddings and VMs.
Storage
Up to 24 drives in one tower
Up to 24 × 2.5" or 12 × 3.5" SAS/SATA, or 8 × 3.5" plus 8 × 2.5" NVMe, with BOSS-N1 mirrored M.2 boot.
GPU & expansion
6 PCIe slots, two at Gen5
Two full-length x16 Gen5 slots for double-wide GPUs, plus four Gen4 slots for single-wide GPUs and NICs.
Office-friendly
Quiet tower design
A 4.5U tower with low-noise air cooling that can sit in an office or branch, with no data center or rack required.
Security & management
Silicon root of trust + iDRAC9
Signed firmware, Secure Boot, TPM 2.0 and System Lockdown, managed remotely with iDRAC9, Redfish and OpenManage Enterprise.
Popular configurations
Start from a proven build
Reference configurations our engineers adjust to your workload, licensing and budget.
AI Inference Tower
Private LLMs, RAG and video analytics in an office, with no data center needed.
- 2 × Intel Xeon Gold CPUs with Intel AMX
- 256 GB DDR5 RDIMM
- 2 to 5 × NVIDIA L4 GPUs
- 2 × NVMe SSD for models + 3.5" HDD for data
- BOSS-N1 mirrored boot
- Dual redundant Titanium PSUs
Branch Office Server
Active Directory, ERP, file shares and backup for a branch or SME.
- 2 × Intel Xeon Silver 4510 CPUs
- 64 GB DDR5 RDIMM
- 4 × 3.5" SAS HDD + SSD boot
- PERC H755 RAID
- 2 × 1 GbE LOM
- Dual redundant PSUs
VDI & Design Tower
GPU-accelerated virtual desktops and CAD for small engineering and design teams.
- 2 × Intel Xeon Gold CPUs
- 512 GB DDR5 RDIMM
- 2 × NVIDIA A16 or L40 GPUs
- 8 × 2.5" SSD
- OCP 3.0 10/25 GbE
- High-output Titanium PSUs
Need something different? Send us your requirements or BOM and our engineers will propose the right build.
Technical specifications
Dell PowerEdge T560 at a glance
- Form factor
- 4.5U tower server
- Processors
- Up to two 4th Gen Intel Xeon Scalable (up to 32 cores each) or 5th Gen Intel Xeon Scalable (up to 28 cores each), with Intel AMX
- Memory
- 16 DDR5 RDIMM slots, up to 1.5 TB, up to 4800 MT/s (4th Gen) or 5200 MT/s (5th Gen)
- Front storage
- Up to 24 × 2.5" SAS/SATA (360 TB) · up to 16 × 2.5" · up to 12 × 3.5" SAS/SATA (180 TB) · 8 × 3.5" + 8 × 2.5" NVMe (240 TB)
- RAID & boot
- PERC H965i, H755, H755N, H355, HBA355i · external HBA355e · BOSS-N1 (2 × M.2 NVMe, hardware RAID)
- GPU
- Up to 2 × 300 W double-wide or 6 × 75 W single-wide (up to 5 × NVIDIA L4)
- Expansion
- 6 PCIe slots: 2 × x16 Gen5 full-length, 4 × Gen4
- Networking
- 2 × 1 GbE LOM · 1 × OCP 3.0
- Power
- 600 W to 2800 W hot-swap redundant PSUs, Platinum and Titanium, plus DC options
- Cooling
- Air cooling with up to 8 standard or high-performance fans
- Management
- iDRAC9, iDRAC Direct, Redfish RESTful API, iDRAC Service Module, Quick Sync 2, OpenManage Enterprise, CloudIQ
- Security
- Silicon Root of Trust, signed firmware, Secure Boot, Secure Erase, System Lockdown, data-at-rest encryption, TPM 2.0 (FIPS, CC-TCG)
- Operating systems
- Windows Server with Hyper-V, Red Hat Enterprise Linux, SUSE Linux Enterprise, Ubuntu Server LTS, VMware ESXi
- Dimensions
- H 464 mm (with feet) × W 200 mm × D 678.2 mm (with bezel)
Specifications per Dell Technologies. Available options vary by configuration and region.
Download Dell spec sheet (PDF)
Why Forte-Tech
Your local partner for Dell infrastructure
We help banks, hospitals, factories, government entities and growing companies specify, source and deploy Dell PowerEdge servers and AI infrastructure.
Genuine Dell hardware
Factory-configured PowerEdge servers with Dell manufacturer warranty.
AI sizing by engineers
We size GPUs, memory and storage to your model, number of users and camera streams.
Delivery across Egypt & GCC
Delivered to your office, branch or data center, with optional installation.
Tender & PO ready
Formal quotations, compliance documents and invoicing for procurement teams.
How ordering works
Tell us your workloadPick a build or send us your requirements or BOM.
Get a validated quoteOur engineers validate compatibility and send a detailed quotation with pricing and lead time.
Approve & orderPurchase order, bank transfer or tender documentation.
Delivery & deploymentDelivered to your site in Egypt or the GCC, with optional installation.
FAQ
Questions about the T560
How much does the Dell PowerEdge T560 cost in Egypt?
The price depends on the processors, memory, drives and GPUs you choose. A ready-to-ship Dell PowerEdge T560 with 2 × Xeon Silver 4510 is listed in our store. For GPU and custom builds, request a quote and we'll send a validated configuration and price.
Is the Dell PowerEdge T560 good for AI?
Yes, for AI inference. The T560 supports up to five NVIDIA L4 GPUs or two 300 W double-wide GPUs such as the L40 or A16, plus Intel AMX on the CPUs. It is well suited to private LLMs, RAG assistants, OCR and video analytics. For training large models, choose a GPU server such as the PowerEdge XE9680.
Which GPU should I choose for the T560?
Choose the NVIDIA L4 for most AI inference and video analytics, the L40 for larger models, generative AI and graphics, the A16 for virtual desktops, and the A2 for entry-level inference. Tell us your model size and number of users and we'll recommend the right card and quantity.
Can the T560 run a private LLM like Llama or Qwen?
Yes. As a rule of thumb, a model needs about 2 GB of GPU memory per billion parameters at FP16, or about 1 GB at FP8/INT8. An 8-billion-parameter model runs on a single L4, and 5 × L4 (120 GB) or 2 × L40 (96 GB) can serve a quantized 70-billion-parameter model for a small team.
Is the T560 quiet enough for an office?
Yes. It is designed as an office-friendly tower with low-noise cooling, so it can run in an office or branch without a server room. Noise increases under heavy GPU load, so we recommend a ventilated space for GPU builds.
What is the difference between the T560 and the R760?
Both use Intel Xeon Scalable processors and DDR5. The T560 is a tower for offices and branches, with up to 1.5 TB of memory and up to 5 × L4 GPUs. The PowerEdge R760 is a 2U rack server for data centers, with up to 8 TB of memory and up to two 350 W GPUs such as the H100 or L40S.
What warranty is included?
Every server is supplied with Dell's manufacturer warranty. Extended Dell ProSupport or ProSupport Plus coverage can be added, and the exact warranty terms are stated on your quotation.
Do you deliver outside Egypt?
Yes. We supply customers across Egypt and the GCC, including Saudi Arabia, the UAE, Qatar, Kuwait, Oman and Bahrain.
Ready to bring AI on-premises?
Send us your requirements and our engineers will size the right T560 for you.