iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
If a local AI model is slow or runs out of memory, check where it is running before buying hardware. In Ollama, use ollama ps to see whether the model is on the GPU, CPU, or split across both; then check logs, GPU access, context length, and concurrent requests. These checks can reveal a configuration or detection problem rather than a hardware limit.
First, identify what is slow or failing
“Slow” can mean a long wait while the model loads, slow prompt processing, or slow generation after the response starts. Record which stage is affected and compare repeated runs with the same model, prompt, and settings. There is no universal speed target in Ollama’s documentation, so a consistent comparison is more useful than an unsupported tokens-per-second threshold.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
MINISFORUM MS-02 Ultra Workstation Mini PC, Intel Core Ultra 9 285HX (24C/24T, up to 5.5GHz), PCIe... | $1,659.00 | Buy on Amazon |
| 2 |
|
GMKtec EVO-X2 AI Mini PC Ryzen Al Max+ 395 Superchip 128GB LPDDR5X 2TB SSD | $3,649.99 | Buy on Amazon |
Note the model and quantization, runtime, operating system, GPU and available VRAM (or Mac unified memory), context setting, and number of requests running at once. For an out-of-memory error, record when it occurs: loading, processing a long prompt, or serving concurrent requests.
Free tools Windows power users keep installed
One-click scans. No signup required.
Check whether Ollama is using the GPU
Run ollama ps while the model is loaded. Ollama says this command shows the models in memory; its output includes processor placement and context information. A model may be fully on GPU, fully on CPU, or partly on each. CPU execution or partial offloading can help explain slow responses, but it does not by itself prove that the GPU has too little memory.
#1 Best Overall
- High-Performance AI Processor：The MS-02 Ultra features an Intel Core Ultra 9 285HX (24C/24T, up to 5.5 GHz, 13 TOPS NPU), delivering fast and efficient performance for AI inference, algorithm development, and media workloads. A PCIe x16 expansion slot supports desktop-class GPU upgrades for advanced model training and accelerated computing tasks. It's ideal for creators, engineers, and teams handling intensive parallel workloads.
- 4 × M.2 PCIe 4.0 + 4 × DDR5 SODIMM slots:Four DDR5 SODIMM slots support up to 256 GB of memory, while ECC helps maintain data integrity in mission-critical environments. Four PCIe 4.0 M.2 slots support up to 24 TB of storage, supporting RAID 0/1/5/10, combining high-speed performance with data protection. It allows for the creation of independent scratch disks, media libraries, and project drives, providing high-throughput for production workflows.
- PCIe & USB 4.0 v2: Up to three PCIe slots can be equipped, including a dual-slot x16 GPU. The main slot supports PCIe 5.0, meeting the needs of high-bandwidth creative and computing workloads. USB 4.0 v2 (80Gbps) supports high-bandwidth external storage and displays.
- Ultra-fast Networking: Wi-Fi 7 further enhances wireless performance with next-generation speeds and low-latency stability. Intelligent bandwidth switching optimizes throughput in different network environments, ensuring optimal performance for enterprise or local networks. Dual 25GbE ports (providing up to approximately 3.125 GB/s bandwidth, about 25 times faster than traditional 1GbE), enabling seamless large-scale file transfers and parallel computing. 10GbE and 2.5GbE ports, with support for Intel vPro technology, ensure enterprise-grade remote management and deployment flexibility.
- Server-grade thermal architecture: Utilizing a dedicated CPU/GPU airflow design, equipped with a 6-pipe dual-fan cooler, it maintains stable performance even under sustained loads, delivering up to 140W Turbo power while maintaining a 100W TDP, and operating with noise levels as low as 36 dB. An integrated 350W power supply ensures stable and reliable output for demanding computing tasks and fully loaded extended configurations.
Compare the reported placement with available GPU and system memory. Then inspect the logs for initialization, driver, or backend errors before changing hardware. Ollama’s FAQ explains the command and placement indicators.
Check logs and GPU access
Ollama’s troubleshooting guide gives platform-specific log locations and instructions for enabling debug logging. Look for GPU initialization failures, driver incompatibilities, and backend errors. If the model falls back to CPU, logs can help distinguish a detection issue from insufficient memory.
Linux with an NVIDIA GPU in Docker
To test whether Docker can access the GPU, Ollama suggests running:
docker run --gpus all ubuntu nvidia-smi
If the container cannot access the GPU, check the host driver and container GPU configuration. Ollama’s guide also lists NVIDIA UVM and driver checks; use the instructions that match your system.
AMD and other configurations
For AMD, check device access and Ollama’s logs, and use the AMD-specific diagnostic commands and environment variables in its troubleshooting guide. One version-specific issue to watch for on Linux: Ollama says its ROCm 7 libraries require a compatible ROCm 7 kernel driver. An older ROCm 6.x-or-earlier driver can cause GPU discovery to time out and Ollama to fall back to CPU. Do not apply that diagnosis to other operating systems or driver versions without checking the applicable guidance.
See Ollama’s troubleshooting documentation for the relevant platform-specific steps.
Reduce context and unnecessary parallel requests
Context length and simultaneous requests both affect memory use. Ollama states that increasing context length increases required memory. Its current documentation lists these defaults by available VRAM:
Rank #2
- EVOLUTION RYZEN AI MAX+ 395 MINI PC - GMKtec EVO-X2 is the next evolution in AI mini PC Ryzen Strix Halo series. Thanks to AMD Simultaneous Multithreading (SMT) the core-count is effectively doubled, to 32 threads. Ryzen AI Max+ 395 has 64 MB of L3 cache and can boost up to 5.1 GHz, depending on the workload. The Ryzen AI Max+ 395 is currently rated as the "most powerful x86 APU" on the market for AI computing.
- AI NPU with XDNA 2 ARCHITECTURE - Powered by 16 “Zen 5” CPU cores, 50+ peak AI TOPS XDNA 2 NPU and a truly massive integrated GPU driven by 40 AMD RDNA 3.5 CUs, the Ryzen AI MAX+ 395 is a transformative upgrade and delivers a significant performance boost over the competition. The Ryzen AI Max+ 395 excels in consumer AI workloads like the llama.cpp-powered application: LM Studio. Shaping up to be the must-have app for client LLM workloads, LM Studio allows users to locally run the latest language model without any technical knowledge required and unleash their creativity and productivity.
- AMD RADEON 8090S iGPU GAMING PC - The AMD Radeon RX 8060S offers all 40 CUs with up to 2.9 GHz graphics clock and uses the new RDNA 3.5 architecture. The powerful iGPU is positioned between an RTX 4060 and 4070 laptop GPU and therefore enables gaming in FHD at maximum details in most demanding games. The 8060S can also utilize the full 128GB pool, which is perfect for running LLMs such as Deepseek 70B Q8, which runs comfortably on this machine.
- EIGHT CHANNEL LPDDR5X - LPDDR5X is a new ground breaking memory small form factor installed on-board. With blazing speeds up to to 8000MT/s, it runs 1.5x faster than the DDR5 SODIMMs; 90% better performance over DDR5 SODIMMs in video conferencing and photo editing; 30% better performance in productivity apps; 12% better performance in digital content workloads.
- QUAD SCREEN 8K DISPLAY SUPPORT - EVO-X2 AI Mini PC support 4-screen 4K/8K output via HDMI 2.1 (8K@60Hz), DisplayPort 1.4 (4K@60Hz), and dual USB 4 40Gbps Transfer speed (supporting PD3.0/DP1.4/DATA). Ideal for gaming, video editing, and multitasking, it provides expansive and crisp multi-display support.
| Available VRAM | Ollama context default |
|---|---|
| Below 24 GiB | 4k |
| 24–48 GiB | 32k |
| 48 GiB or more | 256k |
These are Ollama defaults in its rolling documentation accessed in 2026, not universal hardware-sizing rules. A larger context may be useful for a task that needs it, but it also raises memory requirements. Set the context to what your workload actually needs; the context-length documentation describes Ollama’s behavior.
Parallel requests add to the memory burden. Ollama’s FAQ gives the example of a 2K context with four parallel requests producing an 8K effective context allocation, and says required RAM scales with OLLAMA_NUM_PARALLEL multiplied by OLLAMA_CONTEXT_LENGTH. If memory errors occur under load, reduce concurrency and retest.
Try memory-saving options, then retest quality
Ollama documents Flash Attention and quantized KV-cache options as ways to reduce memory use. Cache quantization can trade quality for lower memory consumption: Ollama describes q8_0 as using approximately half the f16 cache memory with very small quality loss, and q4_0 as using approximately one quarter with small-to-medium quality loss. The effect depends on the model and task and may be more noticeable at higher context sizes.
Change one setting at a time, then compare memory use, response speed, and output quality on the same workload. The Ollama FAQ describes the cache options and their trade-offs.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Decide whether memory capacity is actually the limit
Consider a hardware change only after placement, logs, GPU access, context, and concurrency have been checked. If the workload genuinely exceeds available memory, compare the model and its quantization, required context, GPU or unified memory, CPU/GPU placement, driver and runtime support, and complete system compatibility. The available documentation does not establish a particular GPU as the right choice for every model.
Ollama’s quickstart recommends 8 GB of available VRAM, or unified memory on a Mac, for its specific Gemma 4 E2B example. It lists that example’s download at about 7.2 GB and notes that using system RAM with less VRAM may make responses slower. These figures apply to that example, not to all local models; larger contexts also need more memory. See the Ollama quickstart.
A scheduling result is not a general speed guarantee
Ollama reported a scheduling comparison on September 23, 2025, in which token generation rose from 52.02 to 85.54 tokens per second for gemma3:12b at 128k context on one NVIDIA GeForce RTX 4090. The same example reported VRAM use of 19.9 GiB to 21.4 GiB and GPU placement of 48/49 to 49/49 layers. This is a vendor-reported result for that configuration, not an independent benchmark or a speed increase to expect on other systems. The details are in Ollama’s September 23, 2025 scheduling post.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Recommended Free Tools
