You had the box in your cart. A Ryzen AI Max+ 395 mini PC, 128GB of unified memory, the thing you have been reading about for a year as the way to run a 70-billion-parameter model on your desk without a rack of GPUs. Then a headline scrolled past: “AMD Ryzen AI Max 400.” A new number. A bigger number. And the cursor drifted away from the checkout button, because nobody wants to spend two thousand dollars on hardware that gets superseded a month later.
Here is the reassuring part. The 400 series, codenamed Gorgon Halo, is not the thing that supersedes your purchase. It is, by almost every measure that affects how a mini PC runs language models, the same chip with a new name. The question this article answers is the one the spec-news coverage skips: given what actually changed, should you wait, and if so, who exactly should wait?
What Gorgon Halo actually is
Strip away the model-number reshuffle and Gorgon Halo is a mid-cycle refresh of the Ryzen AI Max 300 (Strix Halo) parts that have been shipping since early 2025. Tom’s Hardware describes it as a “minor refresh”, and puts it bluntly: every chip in the new lineup uses the same Zen 5 CPU cores, the same RDNA 3.5 integrated graphics, and the same XDNA 2 NPU as Strix Halo. Their own summary is that you could scratch the “4” off the model number, write in a “3,” and the spec sheet would still be accurate.
The deltas are small enough to list in a sentence. Notebookcheck’s breakdown of the official lineup covers three Pro-tier chips: the flagship Ryzen AI Max+ 495 Pro, the Ryzen AI Max 490 Pro, and the Ryzen AI Max 485 Pro. The top part gets a 100 MHz clock bump (a 5.2 GHz boost versus 5.1 GHz on the 395), a Radeon 8065S iGPU that runs 100 MHz faster than the 8060S, and a 55 TOPS NPU instead of 50. Notice that AMD led with Pro SKUs this time, where Strix Halo debuted with consumer parts, and that there is still no firm calendar date: AMD says the Ryzen AI Max PRO 400 parts will reach HP, Lenovo, and other OEMs in Q3 2026, which is a window rather than a date you can buy against.
That is the whole list. If you were hoping the 400 series meant a new architecture, faster memory bandwidth, or a meaningfully quicker GPU for inference, it does not, at least not based on what AMD has shown so far.
The one real change, and its catch
The single capability Gorgon Halo adds that Strix Halo cannot match is memory capacity. Strix Halo tops out at 128GB of unified memory with up to 96GB addressable by the GPU. Gorgon Halo raises the ceiling to 192GB, of which Notebookcheck reports up to 160GB can be allocated to the integrated GPU. For a narrow set of workloads, that is a real, not cosmetic, difference.
The catch is bandwidth. As our practical Strix Halo guide lays out at length, token generation on these chips is almost entirely memory-bandwidth-bound, and the bottleneck sits at roughly 215 GB/s of real-world LPDDR5X throughput. Gorgon Halo runs on the same memory technology, so a larger model that now fits in 192GB still has to be read at the same speed. You can load something bigger; you cannot necessarily run it faster. More capacity lets you hold a model you could not hold before, but it does not move the tokens-per-second needle for models that already fit. That distinction matters, because “192GB” reads like a performance upgrade and is really a capacity upgrade.
Who should actually wait
This is where the spec-news coverage tends to go silent, and where a buyer needs a straight answer rather than a shrug. There is a real audience for whom the 192GB ceiling justifies waiting, and pretending otherwise would shortchange the reader.
If you run very large dense models at higher precision, the kind that spill past 128GB once you account for the key-value cache and a usable context window, the extra headroom is the difference between running them locally and not. The same is true if you want a big Mixture-of-Experts model like a 235B-class system resident at a usable quantization, or if your workflow keeps several large models loaded at once rather than swapping them in and out. People building multi-box Strix Halo clusters for the largest models already understand this ceiling intimately; for them, a single 192GB node can replace a more awkward two-machine setup.
There is a structural reason this is a buy-decision and not a wait-and-upgrade decision: the memory on these chips is soldered. As we covered in the soldered-RAM mini PC trap, you cannot start at 128GB and add more later. Whatever capacity you buy is the capacity you keep for the life of the machine. So if you genuinely know your models need more than roughly 120GB resident, that is a legitimate reason to hold out for a 192GB Gorgon Halo box rather than buy a 128GB one today and regret it.
For everyone else, and that is the large majority, 128GB is already more memory than the model needs. A 70B model at a sensible quant, a 30B coding assistant, the popular MoE models people actually run day to day: these live comfortably inside 128GB with room to spare. Buying a machine sized for a model you will never load is not future-proofing, it is overpaying. Our buyer’s guide to local-LLM mini PCs walks through matching the box to the model rather than to the headline number.
The timing math nobody is doing
Waiting only makes sense if the wait is short and the payoff is large. Here both work against it. Gorgon Halo has no announced launch date; coverage points to partner systems arriving later in 2026, and “announced” is a long way from “on a shelf at a price you would pay.” Then there is the next part after that. The genuine generational leap, Medusa Halo with Zen 6 cores and RDNA 5 graphics, is reportedly not due until CES 2028 according to HWCooling’s read of the leaks, which cites VideoCardz and the leaker Kepler_L2. Treat that date as a rumor rather than a promise, because AMD has not confirmed it, but the direction is clear: the real upgrade is years out, and Gorgon Halo is the stop-gap in between.
So the real framing of the wait is this. Hold off for Gorgon Halo and you wait an unknown number of months for a chip that runs your models at the same speed, to gain capacity most buyers will not use. Hold off for the part that genuinely changes the equation and, if the leaks hold, you are reportedly waiting until 2028. A Strix Halo box bought this week is not about to be leapfrogged; it sits comfortably as current hardware for a long time.
Timing cuts the other way too, and it cuts against the 192GB part specifically. The memory market is in the middle of a sustained price spike, a story we have tracked across the mini PC and single-board world. Tom’s Hardware made the same point about Gorgon Halo, calling a 192GB part arriving mid-shortage poorly timed and noting it would be “a minor miracle” if AMD ships that configuration consistently, pointing to Apple quietly pulling its highest-memory Mac Studio options as the cautionary tale. A part defined by more memory is the worst kind of product to launch into a memory crunch.
The boxes worth buying right now
If the verdict is buy, the next question is which one. The Strix Halo mini PC market has matured into several genuinely good options, and the right pick comes down to connectivity, form factor, and whether you need enterprise support. Each of these ships with the full Ryzen AI Max+ 395, the same silicon Gorgon Halo lightly refreshes. Prices below are list-price snapshots as of late May 2026 and move with the volatile memory market, so check the live figure before you buy.
GMKtec EVO-X2
- +Multiple RAM tiers (64/96/128GB)
- +quick-release SSD
- +full Ryzen AI Max+ 395
- +quad 8K display output
- -Dual 2.5GbE rather than 10GbE
- -soldered RAM
- -availability comes and goes
MINISFORUM MS-S1 MAX
- +USB4 V2 at 80Gbps
- +dual 10GbE
- +built-in 320W PSU
- +PCIe x16 slot
- -Premium price
- -soldered RAM
- -larger footprint than rivals
NIMO Mini PC AMD Ryzen AI Max+ 395
- +Compact 4L chassis
- +configurable power modes up to 120W
- +up to 8TB storage
- -2.5GbE only
- -single USB4 port
- -newer brand
HP Z2 Mini G1a
- +HP warranty and support
- +Thunderbolt 4
- +Flex IO customization
- +ISV-certified
- -Higher enterprise pricing
- -Mini DisplayPort only
- -2.5GbE instead of 10GbE
If you want the broader landscape, including how these boxes stack up against NVIDIA’s DGX Spark for local AI work, that comparison is worth reading before you commit. But the core decision is simpler than the headlines make it look. Gorgon Halo renamed the chip and raised a ceiling that most buyers will never reach. The hardware on shelves today is the hardware you want, and the one part that would genuinely change your mind is still a long way off.
