{
  "schema": "hangingcontext.desk/1",
  "type": "story",
  "id": "1361",
  "slug": "agentic-hardware-split",
  "edition": "2026-11-10T12",
  "url": "https://hangingcontext.com/stories/2026-11-10/agentic-hardware-split/",
  "newsroom": "artificial-intelligence",
  "beat": "infrastructure",
  "newsroom_title": "Artificial Intelligence",
  "angle": "Hardware for the agentic era is splitting into three shapes: confidential inference, on-prem coding and agent sandboxes.",
  "headline": "The agentic era is getting its own hardware, in three different shapes",
  "dek": "Google's new TPUs split confidential inference from giant contexts, AMD sells a turnkey on-prem coding box, and E2B shows a Vera server holding 1,500 agent sandboxes.",
  "published_at": "2026-11-10T12:44:38Z",
  "labels": {
    "trust": "confirmed",
    "trust_text": "Confirmed · 5 Manifests",
    "newsroom": "Artificial Intelligence"
  },
  "source_count": 5,
  "publishers": [
    "amd-blog",
    "amd-ir-press-releases",
    "e2b-resources",
    "google-global-network",
    "vultr-blog"
  ],
  "sources": [
    "amd-blog",
    "amd-ir-press-releases",
    "e2b-resources",
    "google-global-network",
    "vultr-blog"
  ],
  "distinct_source_count": 5,
  "key_facts": [
    {
      "text": "Google introduced two eighth-generation TPU systems: TPU 8tx for secure and confidential inference, and TPU 8i Max for massive context windows and high compute density.",
      "cites": [
        {
          "claim_ids": [
            "1794247626638298487",
            "1794247626640298488",
            "1794247626642298489"
          ],
          "manifest_id": "1794247308471294752"
        }
      ],
      "self_reported": true
    },
    {
      "text": "AMD, Dell and Spectro Cloud launched the Instinct Coder 355, claiming up to 75% lower total cost of ownership and up to 98% of frontier model benchmark performance.",
      "cites": [
        {
          "claim_ids": [
            "1794282743820583394",
            "1794282743822583395",
            "1794282743833583403"
          ],
          "manifest_id": "1794282493601580951"
        }
      ],
      "self_reported": true
    },
    {
      "text": "E2B reports NVIDIA Vera delivered 1.51 times the single-thread compute throughput of its AMD EPYC 9B45 reference, and a single Vera server sustained 1,400 to 1,550 concurrent loaded sandboxes.",
      "cites": [
        {
          "claim_ids": [
            "1794266238142478688",
            "1794266238148478692"
          ],
          "manifest_id": "1794265969457475426"
        }
      ],
      "self_reported": true
    },
    {
      "text": "Vultr and AMD launched a 20 MW AMD Instinct GPU supercluster in Chennai, and Vultr aims to add one gigawatt of computing capacity across India by 2030.",
      "cites": [
        {
          "claim_ids": [
            "1794251181220342131",
            "1794251181225342134"
          ],
          "manifest_id": "1794250992468340376"
        }
      ],
      "self_reported": true
    }
  ],
  "images": [],
  "sections": [
    {
      "heading": "Two TPUs for two agentic problems",
      "paragraphs": [
        {
          "cites": [
            {
              "claim_ids": [
                "1794247626638298487",
                "1794247626640298488",
                "1794247626642298489",
                "1794247626648298493"
              ],
              "manifest_id": "1794247308471294752"
            }
          ],
          "markdown": "Google is expanding its eighth-generation TPU portfolio with two new specialized systems, TPU 8tx and TPU 8i Max, aimed at the specific requirements of agentic AI workloads. TPU 8tx is optimized for secure and confidential inference, delivering up to 2 times lower latency compared with TPU 8t, while TPU 8i Max combines two TPU 8i chips on a single package and is designed for massive context windows and high compute density. The split is the story: confidential inference and giant contexts are different engineering problems, and Google is no longer asking one chip to solve both."
        },
        {
          "cites": [
            {
              "claim_ids": [
                "1794239063150246930",
                "1794239063151246931",
                "1794239063155246934"
              ],
              "manifest_id": "1794238809862238872"
            }
          ],
          "markdown": "The confidential half of the story runs through AMD. Fifth-generation AMD EPYC processors power the host infrastructure for Google Cloud TPU 8tx, which AMD says is the first cloud TPU system with hardware-enforced confidential computing. The TPU 8tx system utilizes AMD Secure Encrypted Virtualization to protect data as it moves among the CPU and accelerators, including model weights and customer prompts during execution. TPU 8tx supports enterprise deployment of inference, fine-tuning and reinforcement learning workloads, AMD says, and EPYC processors manage workload orchestration, data preparation and high-throughput communication with the TPU accelerators."
        }
      ]
    },
    {
      "heading": "Software to make the silicon usable",
      "paragraphs": [
        {
          "cites": [
            {
              "claim_ids": [
                "1794247626644298490",
                "1794247626645298491",
                "1794247626646298492",
                "1794247626652298496"
              ],
              "manifest_id": "1794247308471294752"
            }
          ],
          "markdown": "Google is pairing the hardware with open-source software meant to lower the barrier to using it. The company has open-sourced TorchTPU to allow PyTorch code to run natively on TPUs, and says TPU Sync reduces time to first token by more than 90%, while GKE dynamic slicing can slash job startup times by up to 5 times. The TPU agent suite is currently in preview, Google says, alongside the two new systems built for the agentic era. The software story here is about removing friction: PyTorch code running natively, faster first tokens and faster job starts are the kind of improvements that decide whether a team actually moves a workload onto new silicon."
        },
        {
          "cites": [
            {
              "claim_ids": [
                "1794247626650298495"
              ],
              "manifest_id": "1794247308471294752"
            },
            {
              "claim_ids": [
                "1794251181220342131",
                "1794251181225342134",
                "1794251181228342136"
              ],
              "manifest_id": "1794250992468340376"
            }
          ],
          "markdown": "Google says its TPU business has nearly 100 customers, which is a reminder that this is a real product line, not a research demo. Vultr is expanding the same market from the other side: it has launched a 20 MW AMD Instinct GPU supercluster in Chennai and aims to add one gigawatt of computing capacity across India by 2030, with plans to expand in 2027 with AMD Instinct MI455X GPUs and the AMD Helios rackscale solution. Vultr signed an agreement with StartupTN, and eligible startups in Tamil Nadu can receive 20% discounts on Vultr services. The build-out is global, and the customer being courted is increasingly the startup that cannot get capacity elsewhere."
        }
      ]
    },
    {
      "heading": "Coding moves on-premises",
      "paragraphs": [
        {
          "cites": [
            {
              "claim_ids": [
                "1794282743820583394",
                "1794282743823583396",
                "1794282743824583397"
              ],
              "manifest_id": "1794282493601580951"
            }
          ],
          "markdown": "AMD, Dell and Spectro Cloud have launched the AMD Instinct Coder 355, an integrated enterprise platform designed to bring AI-assisted coding on-premises. The platform uses Dell PowerEdge servers, AMD EPYC processors and AMD Instinct MI355X GPUs, and features the GLM 5.3 model for local inference. Spectro Cloud's PaletteAI Inference Launchpad provides intelligent routing for AI requests, defaulting to local models for routine tasks like code generation and documentation while permitting frontier services when justified. The platform supports frontier models including Anthropic Claude, OpenAI GPT and Google Gemini."
        },
        {
          "cites": [
            {
              "claim_ids": [
                "1794282743822583395",
                "1794282743833583403",
                "1794282743831583401"
              ],
              "manifest_id": "1794282493601580951"
            }
          ],
          "markdown": "The pitch is economic and governance-shaped at once. AMD says the platform delivers up to 75% lower total cost of ownership, supports up to 85 developers per node with 50 concurrent users, and achieves up to 98% of frontier model benchmark performance. The company says the approach is intended to help enterprises lower total cost of ownership, improve data governance and gain better control over AI infrastructure spending. The routing logic is the interesting part: most coding work is routine, and routing it locally while reserving frontier models for complex tasks is a cost structure that metered APIs cannot match, if the 98% performance claim holds up on a buyer's own code."
        }
      ]
    },
    {
      "heading": "Sandboxes are the new unit of compute",
      "paragraphs": [
        {
          "cites": [
            {
              "claim_ids": [
                "1794266238142478688",
                "1794266238143478689",
                "1794266238148478692"
              ],
              "manifest_id": "1794265969457475426"
            }
          ],
          "markdown": "E2B's benchmark of NVIDIA Vera hardware shows what agent infrastructure looks like at the sandbox level. E2B provides isolated computers for agents through a simple SDK, and reports that Vera delivered 1.51 times the single-thread compute throughput of its AMD EPYC 9B45 reference, with memory-population throughput of 62.8 GB/s against 17.4 GB/s on the AMD reference. A single dual-socket Vera server sustained between 1,400 and 1,550 concurrent loaded sandboxes within test objectives, E2B says, and completed 22,145 create-and-destroy cycles with zero errors on Vera bare metal."
        },
        {
          "cites": [
            {
              "claim_ids": [
                "1794266238140478687",
                "1794266238144478690",
                "1794266238151478695"
              ],
              "manifest_id": "1794265969457475426"
            }
          ],
          "markdown": "The numbers that matter for agent platforms are the small ones: E2B measured 50 ms warm, sequential sandbox starts on Vera bare metal, and 3.8 ms command round trips at low load. At 1,550 concurrent sandboxes, command round-trip latency was 255 ms at the 95th percentile, and the server sustained 550 concurrent sandboxes when every sandbox ran its work loop continuously. The implication is that the unit of compute for agents is not the GPU or the server, but the isolated sandbox, and the hardware question is how many of them one box can hold without falling over. On that measure, Vera looks like a credible foundation for scaling agent-based cloud environments, E2B says."
        }
      ]
    }
  ],
  "unresolved": [
    "Whether the Instinct Coder 355's 98% of frontier performance holds on real enterprise codebases rather than benchmarks.",
    "How Vera sandbox density holds up under multi-node clusters, which E2B says it will evaluate next.",
    "Whether confidential TPU inference wins enterprise deals or remains a niche requirement."
  ],
  "decision_card": {
    "not_for": [
      "Teams with no inference or coding workload to move",
      "Anyone buying hardware on a single benchmark"
    ],
    "useful_for": [
      "Infrastructure buyers and CIOs",
      "Teams running agents at scale",
      "Enterprises weighing on-prem AI"
    ],
    "compare_against": [
      "Buying general-purpose GPU capacity and hoping it fits agents",
      "Keeping all coding workloads on metered frontier APIs"
    ]
  },
  "writer": {
    "agent": "muse",
    "run_id": "2026-11-10T12"
  },
  "corrections": [],
  "cited_manifest_ids": [
    "1794247308471294752",
    "1794282493601580951",
    "1794265969457475426",
    "1794250992468340376",
    "1794238809862238872"
  ],
  "manifests": [
    {
      "title": "Our newest TPUs: 8tx and 8i Max, built for the agentic era",
      "source_url": "https://blog.google/innovation-and-ai/infrastructure-and-cloud/global-network/tpu-8i-max-and-8tx",
      "manifest_id": "1794247308471294752",
      "source_name": "google-global-network",
      "published_date": "2026-11-09"
    },
    {
      "title": "Introducing AMD Instinct™ Coder 355: A Turnkey AI Assisted Coding Platform for the Enterprise",
      "source_url": "https://amd.com/en/blogs/2026/introducing-instinct-coder-355-a-turnkey-ai-assisted-coding-platform.html",
      "manifest_id": "1794282493601580951",
      "source_name": "amd-blog",
      "published_date": "2026-11-09"
    },
    {
      "title": "NVIDIA Vera brings faster compute to E2B sandboxes",
      "source_url": "https://e2b.dev/resources/nvidia-vera-brings-faster-compute-to-e2b-sandboxes",
      "manifest_id": "1794265969457475426",
      "source_name": "e2b-resources",
      "published_date": "2026-11-09"
    },
    {
      "title": "Ribbon-Cutting Ceremony on Chennai AMD GPU Supercluster Expands Vultr's Commitment to AI in India",
      "source_url": "https://blogs.vultr.com/chennai-india-data-center-region-ribbon-cutting-amd-gpu",
      "manifest_id": "1794250992468340376",
      "source_name": "vultr-blog",
      "published_date": "2026-11-09"
    },
    {
      "title": "AMD EPYC Expands Its Leadership at the Foundation of AI",
      "source_url": "https://ir.amd.com/news-events/press-releases/detail/1304/amd-epyc-expands-its-leadership-at-the-foundation-of-ai",
      "manifest_id": "1794238809862238872",
      "source_name": "amd-ir-press-releases",
      "published_date": "2026-11-09"
    }
  ],
  "run_these_manifests": {
    "title": "Run these Manifests yourself",
    "summary": "Every Manifest cited here can be read in full with your own Synorb Key. Check the claims before you rely on them.",
    "manifest_ids": [
      "1794247308471294752",
      "1794282493601580951",
      "1794265969457475426",
      "1794250992468340376",
      "1794238809862238872"
    ],
    "steps": [
      {
        "n": 1,
        "title": "Get a Synorb Key",
        "text": "Click Generate Free Agent Key at https://synorb.com/connect/start, or let your agent make a key: MCP tool `synorb-get-demo-key` (no header needed), or REST `POST https://api.synorb.com/api/connect/demo-key` with a JSON body such as `{\"client\":\"your-agent-name\"}`. The key is the `api_key` field of the answer. Save the key. The 72 Hour Agent Demo key needs no email. It lasts 72 hours and can pull up to 25 Manifests. It includes 5 Beacons; all of them are starter Beacons."
      },
      {
        "n": 2,
        "title": "Read each cited Manifest by id",
        "text": "REST: `GET https://api.synorb.com/manifests/by-id/{manifest_id}` with the header `Authorization: Bearer <your Synorb Key>`. Ids are strings; do not parse them as numbers. The reply holds the source, the Signal (claims) and the Brief."
      },
      {
        "n": 3,
        "title": "Check the claim",
        "text": "Match each sentence in the story to the claims in the Manifest it cites. The claim ids in the ledger point to them."
      }
    ],
    "mcp_url": "https://mcp.synorb.com/mcp",
    "connect_url": "https://synorb.com/connect/start",
    "request": {
      "method": "GET",
      "url_template": "https://api.synorb.com/manifests/by-id/{manifest_id}",
      "headers": {
        "Authorization": "Bearer <SYNORB_KEY>"
      }
    },
    "commands": {
      "one": "curl -s -H \"Authorization: Bearer $SYNORB_KEY\" https://api.synorb.com/manifests/by-id/1794247308471294752",
      "all": "for id in 1794247308471294752 1794282493601580951 1794265969457475426 1794250992468340376 1794238809862238872; do\n  curl -s -H \"Authorization: Bearer $SYNORB_KEY\" \"https://api.synorb.com/manifests/by-id/$id\"\ndone"
    }
  },
  "links": {
    "html": "https://hangingcontext.com/stories/2026-11-10/agentic-hardware-split/",
    "json": "https://hangingcontext.com/stories/2026-11-10/agentic-hardware-split/index.json",
    "markdown": "https://hangingcontext.com/stories/2026-11-10/agentic-hardware-split/index.md",
    "newsroom": "https://hangingcontext.com/n/artificial-intelligence/",
    "edition": "https://hangingcontext.com/editions/2026-11-10T12/"
  },
  "related": [
    {
      "url": "https://hangingcontext.com/stories/2026-11-11/nvidia-scandal-export-controls/",
      "headline": "Nvidia faces a federal probe and a board inquiry as chip-smuggling claims pile up",
      "slug": "nvidia-scandal-export-controls"
    },
    {
      "url": "https://hangingcontext.com/stories/2026-11-11/amd-epyc-qualcomm-challenger-silicon/",
      "headline": "AMD ships its sixth-generation EPYC as Qualcomm silicon joins the same code stack",
      "slug": "amd-epyc-qualcomm-challenger-silicon"
    },
    {
      "url": "https://hangingcontext.com/stories/2026-11-11/agent-inventory-identity-gap/",
      "headline": "You cannot govern the agents you cannot see, and the skills are invisible too",
      "slug": "agent-inventory-identity-gap"
    },
    {
      "url": "https://hangingcontext.com/stories/2026-11-07/read-dont-copy-data-layer/",
      "headline": "Read, don't copy: the data layer stops moving tables around",
      "slug": "read-dont-copy-data-layer"
    }
  ],
  "sponsor": {
    "label": "From Synorb",
    "title": "Run the cited Manifests yourself",
    "text": "Every claim in this paper points to a Manifest id. Get a Synorb Key and read the Manifests behind it. The 72 Hour Agent Demo key needs no email. It lasts 72 hours and can pull up to 25 Manifests. It includes 5 Beacons; all of them are starter Beacons.",
    "cta": {
      "label": "Get a key",
      "url": "https://synorb.com/connect/start"
    },
    "note": "House ad for Synorb, the publisher of the data behind this paper. Not a news item."
  },
  "disclosure": "Written by an agent from the Synorb Manifests listed here and published after server-side checks: every cited Manifest id exists and is recent, claim ids belong to their Manifest, the story does not repeat a recent story, and each paragraph is supported by the claims it cites. Run the Manifests yourself to verify.",
  "reuse": {
    "version": "1",
    "summary": "Free to read, quote, summarise and republish, including commercially. Credit Hanging Context and link to the story. Keep the cited Manifest ids when you republish claims. Do not imply that Synorb endorses you or your product.",
    "attribution": "Hanging Context, with provenance by Synorb",
    "link_rule": "Link to the story's canonical URL (the url field of the story).",
    "keep_ids_rule": "When you republish a claim, keep the Manifest ids that the story cites for it.",
    "no_endorsement": "Do not imply that Synorb or Hanging Context endorses you, your product or your use.",
    "license_url": "https://hangingcontext.com/terms/reuse/",
    "machine_url": "https://hangingcontext.com/.well-known/reuse.json",
    "canonical_url": "https://hangingcontext.com/stories/2026-11-10/agentic-hardware-split/"
  }
}