Infrastructure 5 min read

GB10, DGX Spark, HP ZGX Nano, ASUS Ascent GX10: Eight Names for One 128 GB Box

ai.rs Sep 2, 2026
GB10, DGX Spark, HP ZGX Nano, ASUS Ascent GX10: Eight Names for One 128 GB Box illustration

If you have been researching a small AI box and come away confused, the confusion is not yours. The same machine is sold under at least eight names, and search results treat them as if they were eight different products.

They are not. There is one chip — the NVIDIA GB10 Grace Blackwell Superchip — and every box below is built on it, with the same 128 GB of memory, the same bandwidth and the same networking.

Eight vendor product names, all built on one NVIDIA GB10 Grace Blackwell Superchip

The eight names

you may have seen it is
NVIDIA DGX Spark NVIDIA's own box, the reference design
HP ZGX Nano AI Station HP's, model G1n
ASUS Ascent GX10 ASUS's
Dell Pro Max with NVIDIA AI Dev Dell's
Acer Veriton GN100 AI Mini Workstation Acer's
Gigabyte AI-TOP ATOM Gigabyte's
Lenovo Workstation AI Lenovo's
MSI Next-Level AI Power MSI's

And the terms that turn up alongside them, all pointing at the same thing:

  • GB10 — the superchip, not a product you buy on its own
  • Grace Blackwell — the architecture pairing an Arm "Grace" CPU with a Blackwell GPU on one package
  • Blackwell 128 GB — a description, not a product name
  • DGX — NVIDIA's product family; DGX Spark is the desk-sized one, not to be confused with rack DGX systems
  • Project DIGITS — the pre-launch codename you will still find in older coverage

What is identical, and what is not

What is identical across all eight boxes and what varies by vendor

Everything NVIDIA publishes as platform specification is the same in every one of them:

chip NVIDIA GB10 Grace Blackwell Superchip
memory 128 GB LPDDR5x, coherent unified CPU+GPU
memory bandwidth 273 GB/s
compute up to 1 PFLOP FP4
CPU 20-core Arm — 10× Cortex-X925 + 10× Cortex-A725
networking ConnectX-7 NIC at 200 Gb, plus 10 GbE
power 140 W chip, 240 W system

The one specification that genuinely varies is storage. ASUS ships the Ascent GX10 from 1 TB, with 2 TB and a 4 TB PCIe Gen 5 configuration above it. HP's ZGX Nano comes in 2 TB and 4 TB. NVIDIA's own DGX Spark goes up to 4 TB. If you are comparing prices and one box looks cheap, check the SSD before concluding anything.

Beyond that, what differs is what always differs between OEM builds of a reference design: chassis, cooling and therefore noise, whether wireless is included, bundled vendor software, and support, warranty and price. Real considerations — none of them change how fast a model runs.

So which one should you buy?

On performance grounds, it does not matter. The silicon is fixed, the memory is fixed, and the bandwidth that governs token generation is fixed at 273 GB/s regardless of whose logo is on the front.

Buy on the things that actually vary:

  • Storage, because it is the one spec that differs and the one you cannot change on some units
  • Noise, if it lives on your desk rather than in a rack
  • Support terms, if it is going into a business
  • Price, once the storage tiers are matched — comparing a 1 TB box against a 4 TB box tells you nothing

What one of these actually does

The naming is the easy part. The harder question is whether the machine suits your work, and that we can answer with measurements rather than specifications.

On one box. The 128 GB of unified memory is the entire point: it runs 70B models at 4-bit and 120B-class MoE that a 32 GB desktop card cannot load at all. The cost is bandwidth — 273 GB/s against roughly 1,792 GB/s on an RTX 5090 — so token generation runs at roughly a fifth the speed for models that fit both. We put numbers on that trade in RTX 5090 vs GB10.

On two boxes. Paired over the 200 Gb ConnectX link, we spent a month running three 180B-to-320B flagships across two of them. DeepSeek-V4-Flash measured 71.8 tok/s mean, and retrieval was verified at 835,348 tokens — twelve needles out of twelve. The interconnect turned out to be a non-event: tensor-parallel traffic used 1.3% of the link during decode, so the exotic networking is not what makes the pair work. That write-up is Best Model for a Dual DGX Spark, and it applies to any two of the eight boxes above.

One warning that surprises people. These are not faster than a desktop GPU at everything. Ask a pair to look at images rather than write text and a single RTX 5090 wins by about 2.4× — vision is almost entirely prefill, which is compute-bound, and one 5090 has more compute than two GB10s. Prompt processing and token generation are two different speeds, and this machine is strong at one of them.

The short answer

If you are trying to decide between a DGX Spark, a ZGX Nano, an Ascent GX10 and a Veriton GN100, you are not choosing a computer. You are choosing a case, a warranty and an SSD around a computer that has already been chosen for you.

Frequently Asked Questions

Is the NVIDIA DGX Spark the same as the HP ZGX Nano and ASUS Ascent GX10? +

Yes, in every way that affects performance. All three are built on the NVIDIA GB10 Grace Blackwell Superchip with 128 GB of coherent unified LPDDR5x at 273 GB/s, up to 1 PFLOP of FP4 compute, a 20-core Arm CPU and a ConnectX-7 NIC at 200 Gb. Acer's Veriton GN100, Gigabyte's AI-TOP ATOM, Dell's Pro Max with NVIDIA AI Dev, Lenovo's Workstation AI and MSI's box are the same platform again. What differs between them is storage, chassis and cooling, bundled software, support and price.

What is the NVIDIA GB10? +

GB10 is the superchip, not a product you buy on its own — the NVIDIA GB10 Grace Blackwell Superchip pairs a 20-core Arm CPU (10 Cortex-X925 and 10 Cortex-A725) with a Blackwell GPU on one package, sharing 128 GB of coherent unified LPDDR5x memory at 273 GB/s. You buy it inside a box: NVIDIA's own DGX Spark, or one of seven OEM versions of the same design.

What is the difference between DGX Spark and DGX? +

DGX is NVIDIA's product family name. DGX Spark is the desk-sized member of it, built on a single GB10 superchip with 128 GB of unified memory and a 240 W power supply. The rack-mounted DGX systems are an entirely different class of machine at a different scale and price. If you saw the name Project DIGITS in older coverage, that was this product's pre-launch codename.

Which GB10 box should I buy? +

On performance grounds it does not matter — the chip, the 128 GB, the 273 GB/s and the networking are fixed across all eight. Choose on the things that actually vary: storage (ASUS starts at 1 TB with 2 TB and a 4 TB Gen 5 option; HP offers 2 TB and 4 TB; NVIDIA's own goes up to 4 TB), noise if it sits on your desk, support terms if it is for a business, and price once the storage tiers are matched. Comparing a 1 TB box against a 4 TB box tells you nothing.

Do the OEM versions differ in speed? +

No. Token generation on this platform is bounded by the 273 GB/s of memory bandwidth, which is a property of the GB10 superchip and identical in every box. Prompt processing is bounded by the same 1 PFLOP of FP4 compute. Chassis and cooling differ, which shows up as noise rather than as throughput.

Is a GB10 box faster than an RTX 5090? +

Only for work the 5090 cannot do at all. For models that fit in 32 GB the RTX 5090 wins both phases by roughly 5-6x, because its memory bandwidth is about 1,792 GB/s against the GB10's 273. The GB10's advantage is capacity: 128 GB runs 70B models at 4-bit and 120B-class MoE that will not load on a 32 GB card. Vision is a further twist — a single 5090 processed video frames about 2.4x faster than a pair of GB10s, because vision work is almost entirely compute-bound prefill.

Can you connect two GB10 boxes together? +

Yes, over the ConnectX-7 link at 200 Gb, and it works well. Running three flagship models across a pair at tensor parallelism 2, we measured DeepSeek-V4-Flash at 71.8 tok/s mean and verified retrieval at 835,348 tokens, twelve needles out of twelve. The link itself is barely used — tensor-parallel traffic took 1.3% of it during decode — so the interconnect is not the reason the pair works, and a slower cable would give the same numbers.

Deploying AI for your business?

Inference, GPUs, and quantization choices look different in production. See where your business is on the readiness curve.

Take the AI Readiness Check
Share: Post Share

Read next