Servers

Add R740xd – GPU Ready (no GPU included) Build, Refurbished to the cart
Add R740xd – Dual V100 Build, refurbished to the cart
Add R740xd – A40 Build, Refurbished to the cart
Add AI Inference Rack – A40 Edition (70B Model Ready, Turnkey System) to the cart
Add Dell R7525 AI Ready with 2 x Nvidia GPU A100 to the cart
Add NVDGXSPARK-PB, NVIDIA DGX Spark to the cart
Add Dell PowerEdge R640 8SFF 1, Refurbished to the cart
Add Dell PowerEdge R640 8SFF 2, Refurbished to the cart
Add Dell PowerEdge R640 8SFF 3, Refurbished to the cart
Add Dell PowerEdge R740XD, Refurbished to the cart
Add HP ProLiant DL360 Gen10 8SFF Rack Server 1, Refurbished to the cart
Add HP ProLiant DL360 Gen10 8SFF Rack Server 2, Refurbished to the cart
Add HP ProLiant DL360 Gen10 8SFF Rack Server 3, Refurbished to the cart
The Dirty Connector Problem: MPO Inspection and Cleaning

Contamination is the leading cause of fiber link failure, and one dirty position on an MPO ferrule presents as a...

My Optic Won't Link: Reading DOM and Working the Optical Loss Budget

Before replacing a transceiver, read its DOM. Transmit and receive power split a dead link into three buckets in under...

What Can You Actually Run on a 48GB NVIDIA A40? Measured Model Sizes and Tokens Per Second (2026)

One A40 48GB runs 27–34B models at 23–29 tokens/sec, fits a 70B at Q4 at 12–13.5 tokens/sec, and serves 1,705...

The Dirty Connector Problem: MPO Inspection and Cleaning

Contamination is the leading cause of fiber link failure, and one dirty position on an MPO ferrule presents as a failing module. The inspect-clean-inspect discipline, the right tools, what never to do, and the symptom patterns that should send you to the scope instead of the spares shelf.

My Optic Won't Link: Reading DOM and Working the Optical Loss Budget

Before replacing a transceiver, read its DOM. Transmit and receive power split a dead link into three buckets in under a minute. Per-lane readings, the loss budget arithmetic, and the failures DOM is blind to.

What Can You Actually Run on a 48GB NVIDIA A40? Measured Model Sizes and Tokens Per Second (2026)

One A40 48GB runs 27–34B models at 23–29 tokens/sec, fits a 70B at Q4 at 12–13.5 tokens/sec, and serves 1,705 tokens/sec of aggregate output on an 8B model at 100 concurrent requests. The measured numbers, the KV cache math behind the 72B cliff, and the three things an Ampere card cannot do.

Keith L.