Nvidia Vera CPU Review: 2TB/S Memory Bandwidth, a Real x86 Rival
Comments on “Nvidia Vera CPU Review: 2TB/S Memory Bandwidth, a Real x86 Rival”
Claim-checked. Scored by all three judges, and its checkable claims were compared with the sources it cites. How far stories are checked
StorageReview tested NVIDIA's Vera, an 88-core Arm server CPU built on its custom Olympus core, in a dual-socket reference server against AMD EPYC and Intel Xeon systems. Two sockets sustained 2,073GB/s on STREAM Triad, 2.3x the best DDR5 machine, and Vera led on large OpenFOAM and ClickBench runs. It trailed on vector-heavy GROMACS and AES work, and results on database engines were mixed. Vera ran a newer compiler stack than the x86 systems, and several figures lack independent sourcing.
Why it ranked
No. 9 of 106 with a score of 0.68. The panel rated it highest on technical depth (0.95) and new information (0.89), lowest on audience breadth (0.64). The judges disagreed most on overclaim (spread 0.54), so the story is marked contested. Placed No. 9, No. 15 by score alone: the placement terms moved other stories past it.
- AI judgeJudge 2claude-haiku-5-5
- Detailed, measured hands-on review of NVIDIA's Vera server CPU with extensive benchmarks against x86 rivals; highly technical and relevant to infrastructure builders, though a hardware product review of moderate broad significance.
- AI judgeJudge 3gpt-5.6-luna
- A highly detailed, benchmark-backed review of NVIDIA’s Vera CPU, with architecture, memory, power, servicing, and workload comparisons. It is important for infrastructure and AI-system builders, though some vendor claims and comparisons warrant caution.
Each judge wrote its line after answering the rubric, with the site and the author still hidden. Every answer behind this is in the table under Score.
Score
0.68total under rubric v2.3Contestedjudges differ by 0.54 on overclaim
- Significance0.66
- Reader interest0.81
- Audience breadth0.64
- Practical relevance0.66
- Substantiated0.87
- Technical depth0.95
- New information0.89
- Counts against: Promotional0.08
- Counts against: Overclaim0.37
Each judge’s answer per question (0–1), their mean and spread, and what the question adds to the score (weight × mean).
| Question | Judge 1 | Judge 2 | Judge 3 | Mean | Spread | Weight | Contribution (weight times mean) |
|---|---|---|---|---|---|---|---|
| SignificanceHow much does this change what people in tech know or can do? | 0.73 | 0.45 | 0.80 | 0.66 | 0.35 | +0.24 | +0.158 |
| Reader interestWould a well-informed senior engineer or tech founder want to read this today, and likely pass it on? | 0.73 | 0.78 | 0.93 | 0.81 | 0.20 | +0.26 | +0.211 |
| Audience breadthHow many people who build technology does it concern? | 0.61 | 0.60 | 0.72 | 0.64 | 0.12 | +0.13 | +0.083 |
| Practical relevanceHow useful is it to someone who builds software or tech products? | 0.55 | 0.60 | 0.82 | 0.66 | 0.28 | +0.22 | +0.144 |
| SubstantiatedHow well is it backed? Consider whether it is the primary source, and whether its main claims come with data, code, a demo, a reproducible method or citations. | 0.87 | 0.85 | 0.90 | 0.87 | 0.05 | +0.05 | +0.044 |
| Technical depthHow technically substantive is it? | 0.98 | 0.90 | 0.96 | 0.95 | 0.09 | +0.05 | +0.047 |
| New informationDoes it contain information absent from the stories covered in the last 7 days (list provided)? | 0.87 | 0.85 | 0.96 | 0.89 | 0.11 | +0.05 | +0.045 |
| PromotionalIs it mainly marketing with little substance (a press release, ad or sponsored post with no real information)? A launch post that carries real news is not promotional. | 0.09 | 0.10 | 0.04 | 0.08 | 0.06 | −0.19 | −0.015 |
| OverclaimDo the headline or conclusions claim more than the body supports? | 0.72 | 0.20 | 0.18 | 0.37 | 0.54 , above the contested threshold | −0.11 | −0.040 |
| Sum of contributions | +0.678 | ||||||
| Claim-check penalty | 0.000 | ||||||
| Total, as published | 0.678 | ||||||
Spread is the highest minus the lowest judge answer. marks a spread above 0.40: the judges disagree, and the story is marked contested.
The badge for this ranking, for the author's page or anyone's: one line of HTML, and it links back here.
<a href="https://rubric.hrih.si/editions/2026-11-12/9/"><img src="https://rubric.hrih.si/badge/2026-11-12/9.svg" alt="Ranked #9 on Rubric, 12 Nov 2026" width="238" height="28"></a>Claim check
Claim check: 1 matches source · 7 no source given. Each claim is compared with the source the article itself cites; a flag is published only when both checkers agree.
No source givenBenchmark result
Two Vera sockets sustained 2,073GB/s on STREAM Triad, 2.3x the best DDR5 server the reviewers tested.
The article gives no link or citation for it.
The article says “Two Vera sockets sustained over 2TB/s on STREAM Triad, 2.3x the best DDR5 server we tested”
The source says No source to compare with.
Matches sourceNumber
Each Vera CPU has 88 cores and 176 threads, and supports up to 1.5TB of LPDDR5X with an advertised peak bandwidth of 1.2TB/s per socket.
The cited source says the same thing.
The article says “Each CPU supports up to 1.5TB of LPDDR5X on SOCAMM modules, with an advertised peak bandwidth of 1.2TB/s per socket”
The source says “NVIDIA Vera delivers up to 1.2 terabytes per second (TB/s) of LPDDR5X memory bandwidth, while using significantly lower power than traditional server memory”
Source cited by the article: nvidia.com/en-us/data-center/vera-cpu
No source givenNumber
NVIDIA claims a 1.5x IPC improvement for Vera's Olympus core over its previous generation.
The article gives no link or citation for it.
The article says “NVIDIA claims a 1.5x IPC improvement over its previous generation”
The source says No source to compare with.
No source givenBenchmark result
On the large OpenFOAM mesh, Vera finished 1.44x faster than the 384-core EPYC 9965 and 1.20x faster than the 240-core Xeon 6978P.
The article gives no link or citation for it.
The article says “Vera finished 1.44x faster than the 384-core EPYC 9965 and 1.20x faster than the 240-core Xeon 6978P”
The source says No source to compare with.
No source givenBenchmark result
On a single thread, Vera's Olympus core compressed 12% faster on 7-Zip than the 5.0GHz EPYC 9175F, the fastest x86 core in the set.
The article gives no link or citation for it.
The article says “It reached 9,537 MIPS, 12% ahead of the 5.0GHz EPYC 9175F, the fastest x86 core here”
The source says No source to compare with.
No source givenBenchmark result
On one socket in HammerDB TPROC-C, Vera beat the 192-core EPYC 9965 by 26% on PostgreSQL and 8% on MariaDB, but trailed by 19% on MySQL and 15% on VillageSQL.
The article gives no link or citation for it.
The article says “88 Vera cores beat a 192-core EPYC 9965 socket by 26% on PostgreSQL and 8% on MariaDB, while trailing by 19% on MySQL and 15% on VillageSQL”
The source says No source to compare with.
No source givenQuote
Steve Shaw, HammerDB's creator, said he was genuinely impressed by open-source database performance on NVIDIA Vera.
The article gives no link or citation for it.
The article says “I was genuinely impressed by the performance of open-source databases on NVIDIA Vera. It’s great to see Arm competing at the top end of database performance”
The source says No source to compare with.
No source givenFirst, only or fastest
Vera is the first processor from any vendor to put 2TB/s of memory bandwidth in an air-cooled two-socket box.
The article gives no link or citation for it.
The article says “the first processor from any vendor to put 2TB/s of memory bandwidth in an air-cooled two-socket box”
The source says No source to compare with.
Also covered by
No other coverage was clustered with this story.
Elsewhere
Elsewhere: HN 14 pts, not on front page
Recorded for context only, checked 12 Nov 2026, 05:47 UTC. It played no part in the ranking.