The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →iTechGuides is reader-supported.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
Yes, in one specific setup—but not because the full model fits in 3 GiB. Edge0 reports running a 35B mixture-of-experts (MoE) model with about 2.9 GiB of peak active memory by keeping roughly 23 GB of weights on an SSD and streaming the needed expert weights into memory. Its published result used a Mac mini M4 Pro with 24 GB of unified memory, so it is not evidence that an arbitrary 35B MoE will run on a computer with only 3 GiB of system memory.
What “3 GiB of RAM” means in this result
Edge0’s roughly 2.9 GiB figure is the peak reported by the MLX allocator during a short-context test. It is not the model’s full storage footprint, the Mac’s total memory, or a guarantee that the entire inference process uses no more than 2.9 GiB. The test machine had 24 GB of unified memory, which is shared by the system and Apple Silicon workloads.
These are separate quantities:
- Active memory: memory used for the immediate working set during inference. This is what the reported 2.9 GiB allocator peak describes.
- System or unified memory: the computer’s physical memory pool. Edge0’s reported test used a 24 GB Mac mini M4 Pro.
- Disk storage: space for the complete model weights. Edge0 lists approximately 23 GB for its 35B tier.
So the result supports a narrow claim: a particular 35B MoE inference pipeline can keep its reported active working set below 3 GiB under the stated test conditions. It does not show that the model is a 3 GiB download or that a machine with only 3 GiB of physical memory can run it. Edge0’s project page provides the detailed figures.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
How a 35B MoE can use a small active working set
An MoE model uses a router to select experts for each input. That sparsity can reduce how many parameters take part in a given computation, but it does not remove the need to store the model’s full pool of weights. In Edge0’s approach, the expert weights reside on SSD and the inference pipeline loads the required pieces on demand.
#1 Best Overall
- 🔥【Excellent Performance】 Beelink SER3 equipped with AMD Ryzen 3 3200U (up to 3.5GHz), which adopts an 2-core/4-thread. The base frequency is 2.6GHz / Max turbo frequency can reach 3.5GHz. Ensure seamless multitasking and no-delay switching at work, provide the next generation of multitasking experience, and bring processing speed, energy efficiency, productivity, and all-around performance to new heights.
- 🔥【Capacity Storage】Beelink Mini PC driven by the AMD 14nm Processor and 16GB DDR4 2400MHz Memory(can upgrade to 32GB, 2 x 16GB), 500GB M.2 PCIE3.0 X4(2280) SSD, this High-Performance Mini PC designed by our talented European designers delivers enough power and storage for you to play, create and enjoy all day!
- 🔥【HD Graphics & Dual Display】Beelink 3200U integrates Radeon Vega 3 Graphics 3core 1200 MHz GPU to deliver powerful graphics processing power to easily handle the demands of complex design software, 4K@60Hz UHD video editing, and playback, or light gaming. And it can connect 2 screens efficiently handle your tasks, and meet your specific needs.
- 🔥【Multiple Interfaces & Wireless】Mini Desktop PC equipped with a 1000M LAN (RJ-45, supporting Gigabit file transfer speeds), Dual-band 2.4G 5G WiFi (802.11ac, stronger capacity of resisting disturbance), and built-in Bluetooth, high-speed wireless connection makes you step ahead. And 4*USB3.2 ports, 2*HDMI ports, 1*Audio Jack (HP&MIC), and 1*DC Jack, thus offering the user even greater versatility in use.
- 🔥【Lifetime After-Sales Service】Beelink has been dedicated to R&D Mini PC for many years. All Beelink Mini-PC have passed strict inspections before shipping. If you have any questions, please don’t hesitate to contact US. We are 100% guaranteed to solve your problems. We offer lifetime technical support, a 3 year warranty, and 24/7 after-sales service. All of our products obtained FCC, RoHS, and CE Certifications.
Edge0 also describes a trained “Prerouter” that predicts upcoming expert choices. The prediction lets the system begin reading likely-needed weights early, with some SSD reads overlapping computation. The project’s pipeline includes a separate recovery LoRA trained for its quantized, predicted-routing path. These are implementation choices in this system—not automatic properties of MoE models.
What Edge0 reports for speed and quality
The project’s detailed 35B-tier table reports 14.9–17.7 tokens per second for decode on a Mac mini M4 Pro with 24 GB of unified memory. Edge0 describes a test with an approximately 3.3k-token prompt, ten warm-up sampling steps, a timed 200-token decode, and two runs per model tier. It distinguishes cold prefill—the first request after startup—from warm prefill when pages are in the file cache.
Rank #2
- VALUE & PERFORMANCE MINI PC - GMKtec Nucbox M6 Ultra Series is equipped with the powerful AMD Ryzen 5 7640HS processor. This CPU is an upper mid-range processor (APU) of the Phoenix product family. It has 6 SMT-enabled Zen 4 cores (12 threads) running at 4.3 GHz base speed to turbo boost 5.0 GHz.With a TDP Boost of 45W-60W, the Ryzen 7640HS CPU is more energy efficient and delivers a 30% Performance increase over previous AMD Ryzen 7 6800H, 6600U.
- 32GB DDR5 RAM & 1TB PCIe SSD - Installed with DDR5 32GB RAM SO-DIMM Dual Channel (2x16GB), the Nucbox M6 Ultra mini pc support expansion to 128GB RAM. Featured with 1TB M.2 2280 PCIe 3.0 SSD, support dual slot expansion to PCIe 4.0 8TB SSD. (Upgrades not included)
- GAMING PC - The Radeon 760M iGPU has 8 CUs (512 shaders) running at up to 2,600 MHz. This desktop computer can play moderate gaming at a steady FPS, it also HW-encodes and HW-decodes the most widely used video codecs such as AV1, HEVC and AVC.
- DUAL NIC LAN 2.5G RJ45 - Fast Network Speeds: Enjoy up to 2500Mbps data transmission speed without worrying about lagging. Ideal for working, gaming, and surfing the internet. Great for Untangle, Pfsense or as a server office PC.
- TRIPLE 4K DISPLAY - Unlock unparalleled productivity with support for three simultaneous displays, including a stunning 8K@60Hz via USB4, plus 4K@60Hz through both HDMI 2.0 and DisplayPort, transforming your workspace into a command center for multitasking and immersive entertainment.
The September 2026 preprint’s abstract summarizes a 20-token-per-second result, while Edge0’s detailed table gives the 14.9–17.7 range. These are different summaries; the detailed table is the more specific figure to use for the 35B tier. The project also reports an average score of 79.2 for the complete Edge0 35B pipeline and 83.2 for its fp16 base model across five public benchmark evaluations. Its comparison includes quantization, adapters, and predicted routing, and the effects vary by task.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Clear out junk files and repair common Windows errorsFree Scan →For Prerouter specifically, Edge0 reports up to 59% decoding speedup in alternating tests against native routing with the same model, adapter, and workload. “Up to” describes the project’s maximum reported gain, not an expected speedup for every prompt or machine. The project’s figures and test description are available at Edge0’s results page; the authors’ preprint is at arXiv.
Rank #3
- 🎈【POWERFUL PERFORMANCE】Beelink SER3 equipped with AMD Ryzen 3 3200U (up to 3.5GHz), which adopts an 2-core/4-thread. The base frequency is 2.6GHz / Max turbo frequency can reach 3.5GHz. Ensure seamless multitasking and no-delay switching at work, provide the next generation of multitasking experience, and bring processing speed, energy efficiency, productivity, and all-around performance to new heights.
- 🎈【UHD 4K@60HZ & DUAL DISPLAY】Beelink 3200U integrates Radeon Vega 3 Graphics 3core 1200 MHz GPU to deliver powerful graphics processing power to easily handle the demands of complex design software, 4K@60Hz UHD video editing, and playback, or light gaming. And it can connect 2 screens efficiently handle your tasks, and meet your specific needs.
- 🎈【CAPACITY STORAGE】Beelink Mini PC driven by the AMD 14nm Processor and 16GB DDR4 2400MHz Memory(can upgrade to 32GB, 2 x 16GB), 500GB M.2 PCIE3.0 X4(2280) SSD, this High-Performance Mini PC designed by our talented European designers delivers enough power and storage for you to play, create and enjoy all day!
- 🎈【MULTIPLE INTERFACES & WIRELESS】Mini Desktop PC equipped with a 1000M LAN (RJ-45, supporting Gigabit file transfer speeds), Dual-band 2.4G 5G WiFi (802.11ac, stronger capacity of resisting disturbance), and built-in Bluetooth, high-speed wireless connection makes you step ahead. And 4*USB3.2 ports, 2*HDMI ports, 1*Audio Jack (HP&MIC), and 1*DC Jack, thus offering the user even greater versatility in use.
- 🎈【LIFETIME AFTER-SALES SERVICE】Beelink has been dedicated to R&D Mini PC for many years. All Beelink Mini-PC have passed strict inspections before shipping. If you have any questions, please don’t hesitate to contact US. We are 100% guaranteed to solve your problems. We offer lifetime technical support, a 3 year warranty, and 24/7 after-sales service. All of our products obtained FCC, RoHS, and CE Certifications.
What you need to reproduce the setup
- Compatible hardware and backend: Edge0’s published page identifies MLX on Apple Silicon. Its results do not establish equivalent memory, speed, or quality on Windows, Linux, or discrete GPUs.
- Storage for the complete weights: the 35B tier lists about 23 GB of weights. This is model storage, not a recommended minimum drive size; the source does not say an external SSD is required if internal storage is sufficient.
- Allowance for the rest of the system: the 2.9 GiB number is an allocator peak for a short-context workload, not total process memory. The published test ran on a 24 GB system.
- Comparable workload conditions: context length, cache state, hardware, and SSD performance can affect generation. The project notes that available memory, cache state, and SSD speed all affect generation speed.
Do not treat the result as a universal hardware requirement or a promise of the same tokens-per-second rate on a different SSD, context length, or system.
Does quantization alone make the model fit?
No. Quantization can reduce weight size and memory requirements, but the runtime’s loading path matters. Hugging Face’s GGUF documentation describes GGUF as a single-file format for model metadata and tensors, including quantized data types. It also notes that packed-weight handling depends on supported kernels: for Qwen3.5 MoE, absent the relevant kernels, a loader may dequantize weights at load time. Other architectures use the legacy loader, which dequantizes.
Rank #4
- 【Great power in a small computer】Get fast performance from the AMD Ryzen 5 3501U CPU (2.1GHz-3.7GHz, 4 Cores 8 Threads) inside this mini pc, TDP 15W up to 25W. It's perfect for all your home office and business use, like daily computing, web browsing, and smooth media streaming. This small desktop computer handles everyday tasks easily and quietly.
- 【Work on many things at once with lots of storage】This mini PC comes with 16GB of fast DDR4 RAM (expandable up to 32GB), allowing you to smoothly run multiple programs, dozens of browser tabs, and large files all at once. It also features a spacious 512GB NVMe SSD that provides ample storage and delivers dramatically faster boot-ups, app launches, and file transfers compared to a traditional hard drive.
- 【See everything clearly on one or two 4K screens】Connect one or two monitors for more space to work or play. Dual HDMI ports on this mini pc support super sharp 4K Ultra HD video. It's great for doubling your work area for business or watching movies in high definition.
- 【Fast modern connections in a tiny box】Enjoy a better and more stable internet connection with the latest WiFi 6. Use Bluetooth 5.3 to connect wireless headphones, keyboards, and mice without wires. This small pc is very compact to save desk space and has extra USB ports (USB 2.0×2, USB 3.0×2, Type-c 2.0×1, Type-c 3.2 full featured×1, HDMI×2) for your printer, webcam, or other computer accessories.
- 【Reliable Warranty and Support】We provides 1 year warranty for each Mini computers. So you don't need to worry about any product problems. If you have any questions about the product, please contact our customer service, we will provide 24-hour professional technical support and serve you at any time.
That means downloading a quantized file is not, by itself, proof of a particular runtime RAM footprint. Check the model architecture, the inference framework’s support for its quantized format, and whether the implementation keeps weights packed during inference. Model-card guidance can be variant-specific; for example, the community conversion’s instructions for this Qwen3.6 35B A3B GGUF apply to that conversion, not every MoE model.
How to judge a “runs in 3 GB” claim
Before comparing memory claims, identify what each number measures. A useful comparison asks:
Best Value
- 【Processor】AMD Ryzen 5 2400GE delivers fast, reliable performance for office work, web browsing, and everyday multitasking.
- 【Storage & Memory】16GB DDR4 RAM for smooth multitasking; 256GB SSD for quick boot times and plenty of room for files and applications.
- 【WiFi Included】A USB WiFi adapter is included in the box, so you can join a wireless network as soon as you power the machine on — no separate purchase needed. DisplayPort video output, multiple USB 3.0/3.1 ports, RJ-45 Gigabit Ethernet, and audio jacks cover everyday home and office needs.
- 【Ready to Use】Ships with Windows 11 Pro pre-installed and activated, plus a wired keyboard and mouse. Plug in and get to work.
- 【BUY WITH CONFIDENCE】Professionally refurbished, tested, and certified to look and work like new; 90-day warranty and technical support.
- Is the figure active allocator memory, total process memory, system RAM, or GPU memory?
- Where are the full weights stored, and how large are they?
- What hardware, operating system, inference backend, context length, cache state, and SSD were used?
- Is throughput a measured decode rate under a stated workload, and is the quality comparison against a clearly named base model and benchmark set?
Without those details, two “3 GB” claims may describe different things. Edge0’s result is notable specifically as SSD-streamed inference with a reported sub-3-GiB active allocator peak—not as proof that a complete 35B model fits in 3 GiB of total memory or storage.
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.
