Update

Author: tolak

.gitbook/assets/automata-attestation.png

@@ -24,12 +24,14 @@ https://inference-api.phala.network/
 {
   "signing_address": "...",
   "nvidia_payload": "...",
-  "dcap_payload": "..."
+  "intel_quote": "..."
 }
 ```
 
 ### Verify the Attestation
 
+- Verify GPU Attestation
+
 You can copy the value of `nvidia_payload` as the whole payload as followed to verify:
 
 ```
@@ -39,6 +41,19 @@ curl -X POST https://nras.attestation.nvidia.com/v3/attest/gpu \
  -d '__COPY_FROM_ABOVE__'
 ```
 
+- Verify TDX Quote
+
+Theoretically, you can verify the Intel TDX quote with the value of `intel_quote` at anywhere that provide TDX quote verification service. The screenshot below is an example of how to verify the Intel TDX quote with the [Automata's on-chain attestation smart contract](https://explorer.ata.network/address/0xE26E11B257856B0bEBc4C759aaBDdea72B64351F/contract/65536_2/readContract#F6). For Automata example, just need to convert the returned base64 encoded quote to hex format (take Node for example).
+
+```sh
+console.log('Quote bytes:', '0x' + Buffer.from(intel_quote, 'base64').toString('hex'));
+
+// Use on-chain smart contract function `verifyAndAttestOnChain` https://explorer.ata.network/address/0xE26E11B257856B0bEBc4C759aaBDdea72B64351F/contract/65536_2/readContract#F6
+// to verify with the printed quote bytes above.
+```
+
+<figure><img src="../.gitbook/assets/automata-attestation.png" alt=""><figcaption></figcaption></figure>
+
 ## Chat API
 
 OpenAI-compatible API. See: https://platform.openai.com/docs/api-reference/chat

.gitbook/assets/hopper-arch-confidential-computing.jpeg

@@ -1,5 +1,11 @@
 # Host LLM in TEE
 
+Phala Network leverages NVIDIA's TEE GPU technology to provide confidential LLM inference services. By utilizing NVIDIA Blackwell's Confidential Computing capabilities, we ensure that AI model execution and data processing remain fully protected within secure enclaves. This solution enables organizations to run their LLM workloads with guaranteed privacy and security, preventing unauthorized access to both the model and user data during inference operations.
+
+<figure><img src="../.gitbook/assets/hopper-arch-confidential-computing.jpeg" alt=""><figcaption></figcaption></figure>
+
+*source: https://www.nvidia.com/en-us/data-center/solutions/confidential-computing/*
+
 The following diagram illustrates the architecture of the Phala Confidential AI Inference (private LLM Node) service. To host your own private LLM in TEE, you only need to wrap your LLM inference code into a docker image, then deploy your own container to the TEE network. To make the LLM fully verifiable, you need make sure the docker image is reproducible.
 
 <figure><img src="../.gitbook/assets/host-llm-in-TEE.png" alt=""><figcaption></figcaption></figure>