ASUS Ascent GX10 Review: 680 tok/s at Batch 64, but a Slow 1TB SSD
StorageReview's vLLM tests found the ASUS Ascent GX10 tracking other GB10 boxes closely, while ServeTheHome praised its cooling; its 1 TB SSD is slower. Listed at $5,999 on Oct. 30, it showed no price on Nov. 2.
Local AI DeskEdited by Ege Erten
9 min readAI-assisted
Follows our story: GB10 Boxes Compared: Similar gpt-oss-120b Speed, Prices $4,759 to $7,967

The ASUS Ascent GX10, ASUS's GB10 desktop AI computer (ASUS product page).
Image: ASUS
Source · Press image from the ASUS product page, used for news reporting.
The ASUS Ascent GX10 runs gpt-oss-120b at 70 tok/s for one request and 680 tok/s across 64, StorageReview measured, but ASUS's US store listed it out of stock, with no price, on Nov. 2.
The GX10 is ASUS's version of NVIDIA's DGX Spark design: the same GB10 Grace Blackwell Superchip and 128 GB of unified memory in a 150 x 150 x 51 mm case. Two independent reviews have measured it. ServeTheHome published its review on Jan. 4, 2026, and StorageReview published a full vLLM, thermal and storage review on Mar. 3, 2026. Our Nov. 2 comparison of six GB10 boxes placed the GX10 in the field. This review adds what that piece did not cover: idle power, the storage limits, the cooling dispute and the price history.
The price is the part that has moved most. ServeTheHome reported the 1 TB GX10 at $3,000 and the 2 TB model at $3,500 in January, when NVIDIA's 4 TB DGX Spark cost $4,000. The ASUS eShop in the US listed the 128 GB, 1 TB model (GX10-GG0010BN) at $5,999 on Oct. 30, about double the January figure. When the page was read again on Nov. 2, it showed the unit as out of stock, with a notify-me button and no price displayed.
What Is New in the ASUS Ascent GX10
The chip is not new. ASUS's spec page lists the same 20-core Arm CPU NVIDIA uses, with 10 Cortex-X925 and 10 Cortex-A725 cores, beside an integrated Blackwell GPU. Memory is 64 GB or 128 GB of LPDDR5X-8533 shared by both. ASUS rates the system at up to 1 petaFLOP of FP4 AI performance. NVIDIA, for the same chip, says that figure assumes sparsity. The box ships with NVIDIA's DGX OS, an Ubuntu variant.
What ASUS changed is the case and the board. ServeTheHome found that ASUS builds its own motherboard rather than buying NVIDIA's complete board. It also found the chassis changed from the plastic version shown at GTC 2025 to improve cooling, with extra intake vents on the bottom and two 6 W Delta fans. ASUS moved the power button to the front, a detail ServeTheHome said it wished every GB10 box had once it ran five of them in a cluster. StorageReview noted the button also carries a power light, which many GB10 systems lack.
The other change is the entry configuration. ASUS sells 1 TB and 2 TB PCIe 4.0 models and a 4 TB PCIe 5.0 model, while ServeTheHome reported in January that NVIDIA sold its own Spark only with 4 TB; NVIDIA's page now says up to 4 TB. The ASUS spec page also lists a 64 GB memory option. The ASUS eShop showed no 64 GB GX10 on Nov. 2, although NVIDIA opened 64 GB GB10 sales through ASUS and five other partners on Oct. 23.
How Fast the GX10 Runs gpt-oss-120b
The GX10 puts a 120-billion-parameter model on a desk without a GPU server. ASUS says the 128 GB unit runs inference on models of up to 200 billion parameters, and the 64 GB unit up to 100 billion. By those rated limits, gpt-oss-120b fits one 128 GB box but sits above the limit ASUS gives the 64 GB unit; no lab has tested a 64 GB GX10.
Speed depends on memory more than compute. When a model writes text, every new token requires reading the model's active weights from memory, so the bandwidth of that memory sets the ceiling. StorageReview lists the GX10 at 273 GB/s, the same figure NVIDIA gives for the DGX Spark. Every GB10 box shares that ceiling, so their speeds land together.
StorageReview's vLLM online serving test shows what that means in practice. On gpt-oss-120b, with equal input and output lengths, the GX10 ran 70 tok/s at batch 1 and 680 tok/s in aggregate at batch 64. In a decode-heavy mix, which leans hardest on memory bandwidth, it ran 45 tok/s at batch 1 and 260 tok/s at batch 64. Smaller models go much faster: gpt-oss-20b reached 1,600 tok/s at batch 64, and Llama 3.1 8B in FP4 reached 2,750 tok/s. StorageReview did not state the vLLM version or the quantization it used.
The ConnectX-7 network card adds the second answer: models too big for one box. ASUS says up to four 128 GB units cluster to 512 GB over its two 200 Gbps ports. ServeTheHome noted the card is fed by two PCIe 5.0 x4 links rather than one x8, which shows up as four network interfaces and limits how clusters can be wired.
Pros and Cons
The speed case rests on StorageReview's numbers. StorageReview ran the GX10, Acer, Dell and Gigabyte boxes in separate reviews on Mar. 3 with the same method: the GX10 reached 680 tok/s at batch 64, against 713.18 for the Acer, 670.83 for the Dell Pro Max with GB10 (listed on Dell.com as Dell Pro Precision with GB10) and 649.62 for the Gigabyte. StorageReview's own conclusion was that every Spark system shares the same board and chip, so the differences are incremental.
Storage is where the GX10 fell behind. The review unit's 1 TB Phison Gen4 drive was the only 1 TB GB10 platform StorageReview had tested, and the one with the slowest writes. In NVIDIA's GPU Direct Storage test, 16K writes peaked at 0.72 GiB/s at 32 threads, then fell back as threads rose, while 1M reads topped out at 5.12 GiB/s. StorageReview said a better SSD would help considerably. ServeTheHome found the slot itself is wired for PCIe 5.0 x4, but it takes only 42 mm drives, and ASUS ships a Gen5 drive only on the 4 TB model.
Cooling is the one point where the two reviews disagree. ServeTheHome re-ran AI samples after an hour of load and called the result a testament to ASUS's cooling, but it published no temperatures and said its heat-soak load was not realistic, with load falling off before the GPU picked up. StorageReview logged peaks of 87.3°C on the CPU and 82°C on the GPU in the prefill-heavy phase of its one-hour ramp, which it placed in the upper-middle of the comparison for the CPU and slightly warmer than some competitors for the GPU, with temperatures settling under sustained decoding. ASUS itself claims 1.6 times more efficient thermal coverage than comparable compact systems; no independent test has checked that figure. The likely reason for the split is method: one outlet measured peak temperatures, the other measured whether speed held.
Idle power is on the high side. ServeTheHome measured 42 to 47 W at idle, which includes the ConnectX-7 card even when it is not linked, and called it one of the higher-power GB10 boxes. Under load it said it could easily push the GX10 into the 120 to 130 W range before adding optics, and that it ran under 200 W on the 240 W adapter; StorageReview says the box draws up to 180 W. StorageReview put the GPU's peak at 69.77 W. For comparison, ServeTheHome measured Gigabyte's GB10 box at 36 to 37 W idle with the network card linked, but that was a 1 TB PCIe 4.0 sample in a different test, so the gap is indicative only.
Who the ASUS Ascent GX10 Fits
The GX10 fits a developer who wants NVIDIA's DGX software stack and a 120B model at batch 1 speeds near 70 tok/s, and who does not need 4 TB on board. On Oct. 30 it was listed at $5,999 for 1 TB, $951 below the $6,950 shown for the 4 TB DGX Spark in an NVIDIA marketplace copy cached Oct. 31; the storage and read dates differ, so it is not a same-configuration saving.
It also fits a small team serving several users from one box. StorageReview's 680 tok/s at batch 64 is shared across 64 requests, so each user sees far less than that, but the box keeps scaling as requests stack up. A lab building a cluster is the third fit: ASUS says up to four 128 GB units cluster to 512 GB through NVIDIA Sync and ConnectX-7 200GbE, with an external switch required for four, and ServeTheHome expects two-unit single-cable setups to be the common case.
It does not fit a buyer who streams large datasets from local disk. With the 1 TB Gen4 drive's writes topping out near 0.72 GiB/s at 16K in StorageReview's test, that buyer should look at the 4 TB Gen5 GX10 or a box sold with 4 TB Gen5 storage. It also does not fit a buyer who needs a machine this month, since ASUS's US store had none to sell on Nov. 2.
How it compares
Against the NVIDIA DGX Spark, the GX10 has the same 128 GB, the same 273 GB/s and, by ServeTheHome's Geekbench 5 runs, the same CPU speed. StorageReview measured the Spark at 701 tok/s at batch 64 on gpt-oss-120b with 256/256 input and output lengths, but in its July 6 Ryzen AI Halo review, separate from the GX10's March one, so the gap to the GX10's 680 is not a ranked result. NVIDIA's marketplace showed the Spark at $6,950 with 4 TB in a copy cached Oct. 31, with stock unverified; StorageReview's July review gave the Founders Edition at $4,699, so the Spark price is itself unsettled.

StorageReview measured the GX10 at 680 tok/s at batch 64 on March 3; the DGX Spark (701) and Ryzen AI Halo (222) came from separate reviews on July 6.
Illustration (AI-generated)
Against the Acer Veriton GN100, the GX10 lost the thermal comparison. In the same StorageReview series, the Acer peaked at 74.7°C on the CPU and 69°C on the GPU, the coolest of the group, and reached 713.18 tok/s at batch 64. Acer announced it from $3,999 in North America in September 2025, and we found no current US listing that confirms that price.
The HP ZGX Nano posted 722.9 tok/s at batch 64 in a later StorageReview run, published Apr. 24. In a review updated Oct. 5, Notebookcheck reported it at $4,759 on Amazon (date and storage not given) and $7,399 at the HP Store for 4 TB. The Dell Pro Precision with GB10, tested by StorageReview as the Dell Pro Max with GB10, measured 670.83 tok/s and listed at $7,967 on Dell.com on Nov. 2. The Gigabyte AI TOP ATOM measured 649.62 tok/s, and CDW advertised it at $7,312 with 4 TB. At the Oct. 30 listing of $5,999 the GX10 was the second-lowest of these six listed US prices ($4,759 to $7,967, across different configurations and dates), with a 1 TB SSD against 2 TB for the Dell and 4 TB for the Gigabyte and the Spark.
Outside the GB10 family, AMD's Ryzen AI Halo was listed at $3,999 with 2 TB in StorageReview's July 6 review, with 128 GB at 256 GB/s. In that vLLM test it reached 222 tok/s at batch 64 against the Spark's 701, so for serving many requests it is not a like-for-like substitute.
The table below this story sets these machines side by side. What it shows is that the GB10 boxes are one machine on speed: the decision is price, stock, storage and cooling. At its Oct. 30 listing of $5,999, the GX10 was the second-lowest of the six listed US prices ($4,759 to $7,967, across different configurations and dates), with the smallest SSD.
What to watch
The first milestone is a GX10 restock on the ASUS eShop in the US, and the price it returns at. The second is a 64 GB GX10 listing: NVIDIA's 64 GB GB10 systems have been on sale through partners, ASUS among them, since Oct. 23 from $4,999, but no independent lab has tested one yet. The third is a newer vLLM run on the GX10 itself, since StorageReview's March figures predate its July Spark run; its reviews list no software versions, so what changed in between is unknown.
How it compares
Full table| Product | Memory | Bandwidth | AI compute | Measured speed | Price |
|---|---|---|---|---|---|
| HP ZGX Fury AI StationHP | 252 GB HBM3e GPU + 496 GB LPDDR5x CPU (4x 128 GB SOCAMM); 748 GB unified[1] | 7.1 TB/s (GPU HBM3e); 396 GB/s (CPU LPDDR5x)[1] | — | Two-unit cluster with an ASUS ET900N G3 over two 400G DACs: GLM-5.3 NVFP4 5,018 output tok/s at 256 streams, 512/512, vLLM 0.27.1 TP2 with MTP (StorageReview); no single-unit HP result[2] | Japan list price from 28,490,000 yen incl. tax (sales from Oct 27, 2026); no US list price published[3] |
| NVIDIA DGX StationNVIDIA (systems built by ASUS, Dell, GIGABYTE, HP, MSI, Supermicro and Exxact) | 748 GB coherent: 252 GB HBM3e (GPU) + 496 GB LPDDR5X (CPU)[4] | GPU HBM3e 7.1 TB/s; CPU LPDDR5X 396 GB/s; NVLink-C2C 900 GB/s between them[4] | 15 PFLOPS FP4 (without sparsity)[4] | Over 26,000 output tokens/s aggregate on Qwen3.6-35B-A3B NVFP4 at 512 concurrent users (24-30-token input, 1,024-token output), Gigabyte W775-V10-L01; idle never below 250 W.[5] | No NVIDIA list price; first public US OEM list price is Dell Pro Precision with GB300 at $160,195 default config (Dell.com, read Nov 1, 2026); ServeTheHome had heard about $100K[6] |
| NVIDIA RTX PRO 6000 Blackwell Workstation EditionNVIDIA (sold through PNY, TD SYNNEX and workstation makers) | 96 GB[7] | 1,792 GB/s[7] | 4,000 AI TOPS (FP4, with sparsity) (sparse)[7] | LM Studio: Mistral Small (26 GB) at 42.4 tokens/s vs 17 on an RTX 5090; Gemma 3 27B at 29 vs 5; tied with the RTX 5090 at about 81 tokens/s on DeepSeek Llama 8B Distil, where VRAM does not limit.[8] | $16,000 on NVIDIA's US marketplace (Workstation Edition), listed out of stock[9] |
| Apple Mac Studio (M5 Ultra)Apple | up to 512 GB unified (96 GB on 30-core M5 Ultra; 256 GB or 512 GB on 36-core/80-core M5 Ultra)[10] | 1.2 TB/s[10] | — | Qwen 3.8 27B 4-bit MLX in LM Studio: about 55 tok/s on M5 Ultra (36-core, 256 GB) vs about 40 on M3 Ultra (BGR)[11] | $22,299 minimum with 512 GB ($6,799 36-core CPU/80-core GPU M5 Ultra config plus $15,500 memory upgrade; Oct 28, 2026)[12] |
| NVIDIA RTX Spark (N1X) PCsNVIDIA (MediaTek collaborated on the custom CPU design) | Up to 128 GB LPDDR5X (20-core CPU / 6,144-core GPU version); up to 64 GB (18-core / 5,120-core laptop version)[13] | Up to 300 GB/s: LPDDR5X-9400 on a 256-bit bus (NVIDIA's product page lists no bandwidth)[14] | — | gpt-oss:20b via Ollama in a Codex CLI coding agent, 64K context: 58.5 tok/s generation, 3 of 4 tasks passed (Dell XPS 16, 20-core N1X, 64 GB; StorageReview)[15] | Laptops from $2,599; pricing and configurations vary by region[16] |
| ASUS Ascent GX10ASUSThis story | 64 GB or 128 GB LPDDR5X-8533 unified[17] | 273 GB/s (8533 MT/s), as listed by StorageReview; ASUS's spec page states no bandwidth[18] | Up to 1 PFLOP FP4 (GB10; NVIDIA's footnote: theoretical, using the sparsity feature). ASUS's pages say 1 petaFLOP / FP4 with no sparsity label (sparse)[19] | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 70 tok/s at batch 1 rising to 680 tok/s aggregate at batch 64 (StorageReview)[18] | $5,999 (128 GB, 1 TB, GX10-GG0010BN, ASUS eShop US, Oct 30, 2026)[20] |
| Dell Pro Precision with GB10 (model FCM1253; launched as Dell Pro Max with GB10)Dell | 128 GB LPDDR5x up to 8533 MT/s onboard (64 GB option also listed)[21] | 273 GB/s (128 GB LPDDR5x, 256-bit)[22] | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 68.57 tok/s at batch 1 rising to 670.83 tok/s aggregate at batch 64 (StorageReview)[23] | $7,967 (128 GB, 2 TB Gen4 QLC SSD default config, Dell.com US, read Nov 2, 2026 16:40 UTC)[21] |
| Framework Desktop (Ryzen AI Max+ PRO 495, 192 GB)Framework | 192 GB LPDDR5X-8533 soldered[24] | 273 GB/s (256-bit)[24] | — | — | from $6,799 (DIY Edition pre-order, Oct 30, 2026)[25] |
| GIGABYTE AI TOP ATOMGIGABYTE | 128 GB unified system memory[26] | 273 GB/s (LPDDR5x-8533, 256-bit)[27] | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 60.71 tok/s at batch 1 rising to 649.62 tok/s aggregate at batch 64 (StorageReview)[27] | $7,312.00 advertised at CDW US against a $7,999.99 list (128 GB, 4 TB, ATAGB10-9000; read Nov 2, 2026)[28] |
| GMKtec EVO-X5 Pro AI Mini PCGMKtec | 192 GB LPDDR5X-8533 onboard (8 x 24 GB, non-upgradeable); up to 160 GB allocatable as VRAM[29] | Up to 273 GB/s[30] | — | GMKtec Laboratory table: GLM-5.3-Flash UD-IQ4_XS (146.05 GB) decode 17.79 tok/s, prefill 70.18, TTFT 14.71 s; Qwen3.8-Flash-Next ROCmFP4-FAST (93.48 GB) decode 45.08 tok/s[29] | $6,799 (192 GB/2 TB) and $7,099 (192 GB/4 TB), GMKtec US store, Nov 2, 2026[29] |
| HP ZGX Nano AI Station (G1n)HP | 128 GB coherent unified system memory[31] | 273 GB/s (128 GB LPDDR5x)[32] | GB10: up to 1 PFLOP FP4, theoretical, using the sparsity feature; HP and reviewers list 1,000 TOPS FP4 (sparse)[19] | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 69.6 tok/s at batch 1 rising to 722.9 tok/s aggregate at batch 64 (StorageReview)[33] | $7,399.00 at the HP Store for the 4 TB model; $4,759 on Amazon (Notebookcheck, review updated Oct 5, 2026)[34] |
| Lenovo ThinkStation PGXLenovo | 128 GB LPDDR5X unified (256-bit)[35] | 273 GB/s[35] | — | — | about $4,100 for the 1 TB configuration (Mar 2026)[36] |
| MSI EdgeXpert (MS-C931)MSI | 64 GB or 128 GB LPDDR5x unified (256-bit)[37] | 273 GB/s[37] | 1 PFLOPS FP4 (sparse)[37] | — | No list price; MSI sells via quote (Get a quote)[37] |
| NVIDIA DGX SparkNVIDIA (partner GB10 systems from Acer, ASUS, Dell, GIGABYTE, HP, Lenovo and MSI) | 128 GB; a 64 GB configuration is sold only through OEM partners[19] | 273 GB/s[19] | Up to 1 PFLOP FP4 (theoretical, using the sparsity feature) (sparse)[19] | Qwen 3.8 27B (dense) with multi-token prediction: decode of about 20 tokens/s, with time-to-first-token staying interactive out to the full 262K context.[38] | $6,950.00 (128 GB, 4 TB) on NVIDIA's enterprise marketplace with an Add To Cart button in a copy cached 07:59 UTC on 2026-10-31; stock unverified (listed out of stock on 2026-10-21); 64 GB partner systems from $4,999[39] |
| NVIDIA DGX Spark 64 GBNVIDIA (sold by Acer, ASUS, Dell, Gigabyte, HP and MSI) | 64 GB LPDDR5X unified (two units pool 128 GB over ConnectX-7)[40] | 273 GB/s (256-bit; NVIDIA spec table covers both 64 GB and 128 GB)[19] | 1 PFLOPS FP4 (sparse; NVIDIA spec table covers both memory configurations)[19] | — | from $4,999 (partner systems, Oct 2026; no Founders Edition)[40] |
| AMD Ryzen AI Halo Developer PlatformAMD | 128 GB LPDDR5x-8000 unified[41] | 256 GB/s[41] | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL 256/256: 222 tok/s aggregate at batch 64 vs 701 on DGX Spark (StorageReview)[42] | $3,999 with 2 TB SSD (Jul 2026)[42] |
| Framework Desktop (Ryzen AI Max+ 395)Framework | 128 GB LPDDR5x-8000 soldered (32 GB and 64 GB options)[43] | 256 GB/s[44] | — | gpt-oss-120b MXFP4 in LM Studio (out-of-the-box setup): roughly 45 tok/s (ServeTheHome; Ryzen AI Max+ 395, 128 GB, Windows 11 Pro)[45] | $3,449 (DIY Edition, Max+ 395 with 128 GB, Oct 30, 2026)[25] |
| Acer Veriton GN100 AI Mini WorkstationAcer | 128 GB LPDDR5x coherent unified[46] | — | — | gpt-oss-120b, vLLM online serving, equal ISL/OSL: 69.65 tok/s at batch 1 rising to 713.18 tok/s aggregate at batch 64 (StorageReview)[47] | from $3,999 in North America (announced Sep 3, 2025; prices vary by region)[46] |
| ASRock Industrial AI BOX-A395ASRock Industrial | 128 GB LPDDR5X-8000 soldered[48] | — | — | AgentSTH V7, 32 threads: 15% ahead of the next-closest Ryzen AI Max+ 395 box (single-core within 1%); estimated SPEC CPU2026 int rate: top of a 44% spread across four 395 systems; measured by ServeTheHome, Ubuntu 26.04[49] | — |
| Dell Pro Precision with GB300 (model FCT6263; announced as Dell Pro Max with GB300)Dell | 252 GB HBM3e GPU + 496 GB LPDDR5X CPU[6] | — | — | — | $160,195 (Dell.com US default config with RTX PRO 2000 display GPU and 16 TB SSD, read Nov 1, 2026)[6] |
| GMKtec EVO-X2 AI Mini PCGMKtec | 64 GB or 128 GB onboard LPDDR5X-8000 (non-upgradeable)[50] | — | — | — | $3,499.99 (128 GB/1 TB, GMKtec US store, Oct 30, 2026; listed against $3,999.99)[50] |
| HP Z2 Mini G1a WorkstationHP | up to 128 GB unified, up to 96 GB assignable to the GPU, up to 8000 MT/s[51] | — | — | gpt-oss-120b GGUF (63.39 GB, full GPU offload) in LM Studio: close to 40 tok/s (StorageReview)[52] | — |
| Microsoft Surface RTX Spark Dev BoxMicrosoft | 128 GB unified memory[53] | — | up to 1 PFLOP of AI compute (Microsoft; precision and sparsity not stated on the page) (sparse)[53] | — | From $5,999.99 (Microsoft Store US, pre-order, read Nov 1, 2026)[53] |
| MSI XpertStation WS300MSI (NVIDIA DGX Station platform) | 252 GB HBM3e GPU + 496 GB LPDDR5X CPU (748 GB coherent)[54] | — | — | GPT-OSS-120B MXFP4, vLLM, 512/512 ISL/OSL, concurrency sweep to 128: peak 9,294 output tok/s vs 3,384 on an RTX PRO 6000 and 500 on DGX Spark (StorageReview, MSI-hosted unit, no add-in GPU)[54] | ServeTheHome calls it a $100K+ system; no MSI list price given[55] |
| Tenstorrent TT-QuietBox 2 (Blackhole)Tenstorrent | 128 GB GDDR6 on the accelerators + 256 GB DDR5-5600 system memory[56] | — | — | — | $9,999 (Oct 30, 2026)[56] |
| WiCi One (RTX 5060 Ti edition)WiCi | 16 GB GDDR7 VRAM; 1 TB NVMe 4 SSD for TurboStream weight paging (WiCi's claim: NVMe used as working VRAM). System RAM not stated[57] | — | — | — | $2,599 list (WiCi shop, Oct 2026); a $600-off code (WICIWELCOME) is applied at checkout. Early-access price was $1,999 (Notebookcheck, Engadget, Sep 2026)[57] |
| WiCi One (RTX 5090 edition)WiCi | 32 GB GDDR7 VRAM; 4 TB NVMe 5 SSD for TurboStream weight paging (WiCi's claim: NVMe used as working VRAM). System RAM not stated[57] | — | — | — | $9,999 list (WiCi shop, Oct 2026); a $600-off code (WICIWELCOME) is applied at checkout[57] |
Italic values are reported by an outlet or measured by an independent lab; the rest come from the maker. Measured speeds use different models, quantizations and tools, so read each with its source rather than against the others. Voltgrid News does not test hardware itself.
Sources (57)
- [1] HP
- [2] StorageReview
- [3] HP Japan
- [4] NVIDIA
- [5] ServeTheHome
- [6] Dell
- [7] NVIDIA
- [8] Gamers Nexus
- [9] NVIDIA
- [10] Apple
- [11] BGR
- [12] MacRumors
- [13] NVIDIA
- [14] ServeTheHome
- [15] StorageReview
- [16] NVIDIA
- [17] ASUS
- [18] StorageReview
- [19] NVIDIA
- [20] ASUS eShop
- [21] Dell
- [22] HotHardware
- [23] StorageReview
- [24] Framework Blog
- [25] Framework
- [26] GIGABYTE (PR Newswire)
- [27] StorageReview
- [28] CDW
- [29] GMKtec
- [30] GMKtec
- [31] HP
- [32] HotHardware
- [33] StorageReview
- [34] Notebookcheck
- [35] Lenovo PSREF
- [36] ServeTheHome
- [37] MSI
- [38] Tom's Hardware
- [39] NVIDIA Marketplace
- [40] NVIDIA Blog
- [41] AMD
- [42] StorageReview
- [43] Framework
- [44] Framework Blog
- [45] ServeTheHome
- [46] Acer
- [47] StorageReview
- [48] ServeTheHome
- [49] ServeTheHome
- [50] GMKtec
- [51] HP
- [52] StorageReview
- [53] Microsoft
- [54] StorageReview
- [55] ServeTheHome
- [56] Tenstorrent
- [57] WiCi







