All news briefs
Developer platform

Nvidia positions Vera Rubin for continuous post-training of AI agents

The story in five sentences

  1. Nvidia describes continuous post-training as a central computing workload for agentic AI.
  2. The company proposes intelligence per dollar as a metric alongside cost per token.
  3. NeMo Gym and NeMo RL are intended to standardize distributed training environments and reinforcement learning.
  4. Nvidia claims that Vera Rubin can train the largest models with one quarter of the Blackwell GPUs.
  5. Most performance and efficiency figures come from Nvidia or the partner companies mentioned.

Full explanation

Nvidia presents continuous post-training as a recurring production cycle for AI agents. Tools and simulated or production-related environments generate rollouts, those attempts are evaluated, and the results are used for further weight updates. NeMo Gym supplies environments for that process, while NeMo RL coordinates distributed reinforcement learning. The stack is intended to make repeated improvement a standard workload.

Alongside cost per token, Nvidia promotes intelligence per dollar to assess both the initial creation and continued improvement of a model. The company cites Nemotron 3 Ultra, a 550-billion-parameter mixture-of-experts model, and reports 71.7 percent on SWE-bench. This is an Nvidia figure. Nvidia also claims that Vera Rubin can train the largest models with one quarter of the GPUs required by Blackwell.

Other figures come from named partners. Prime Intellect reports a 30 percent increase in CPU throughput, while Perplexity describes synchronization for trillion-parameter models in under two seconds. These remain vendor or partner statements. Plans involving Together and Prime Intellect describe intended future use rather than established production. The evidence supports Nvidia’s product positioning, not a neutral comparison proving its claimed advantage.

Privacy setting

May we measure pseudonymous page views and session replays?

With your consent, PostHog EU measures which pages are opened and additionally records your session: mouse movement, clicks, scrolling and the rendered page content are stored as a replayable reconstruction. Typed input is masked before sending. The contents of sign-in, sign-up, password recovery, the founder chat, the newsletter field, your email display, your quiz answers and the checklists are not recorded at all; on the sign-in and sign-up pages recording is paused. Addresses are always stored without query parameters. A random device identifier and technical connection data such as the IP address are processed as well. In addition, strictly defined usage events are measured: how far a post was read, which article element was used, which call to action was clicked, how a newsletter sign-up ended, how far you get in a course or in the feed (recording only whether a task was correct or incorrect, never your answer itself), which video you start, whether you switch the language, and how a sign-in or sign-up attempt ended — without the email address, without the password, and without the error message. Only values from a fixed list and whole numbers from fixed ranges are transmitted — no input and no free text. We also label production, internal, and automated test visits separately and derive whether a visit came from a known interface such as ChatGPT, Claude, or Perplexity from a known referring domain or a strictly allowed campaign value. The full referring address and query parameters are not sent, and this cannot identify a specific AI model. If PostHog's IP-based geo enrichment is enabled in the project, the service can derive an approximate country, continent, and region; we do not request browser or GPS location. Every visit is also assigned to a page area from a fixed list (for example home, blog, tools, account) — the area is sent, not the address. Page loading and stability metrics are measured as well (Core Web Vitals: LCP, CLS, INP, FCP), without network payloads. A click on a link leading away from the site is recorded only as the target domain from a fixed list, without path, query parameters, or link text. For the Idea Wheel, AI Labelling and Limit Reset tools only the kind of action is measured — spin, option changed, copy or export, for instance — never your input and never a result. Learn more about privacy

Without your consent no analytics code is loaded and nothing is recorded. You can withdraw at any time: withdrawal ends collection immediately; data already collected may have been transmitted by then and is deleted after the storage period.