Predictika · Server processor orders

An AI sales agent's Xeon 6 processor order, checked before it goes out

An AI sales agent prepares an order for Intel Xeon 6 server processors with the CPU model, sockets per server, platform socket, memory type, speed and channels, and a CPU subtotal. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the processor specifications Intel publishes and returns a verdict, every reason and the figures it computed.

Scope: Intel Xeon 6 processor orders: CPU model, sockets per server, platform socket, memory type, speed and channels, and the CPU subtotal at Intel's listed prices. See everything this validator covers

This validator models orders for four Intel Xeon 6 processors (6980P, 6760P, 6780E and 6515P), not just the scenarios shown below.

  • Sockets per server — Intel lists all four models as 2S, so an order with four processors in a server is blocked.
  • Platform socket — the 6980P uses socket FCLGA7529; the 6760P, 6780E and 6515P use FCLGA4710. A processor paired with the other platform is blocked.
  • Memory — DDR5 at up to 6400 MT/s on every model; MRDIMM at up to 8800 MT/s is listed on the 6980P page only. Memory channels are 12 on the 6980P and 8 on the others.
  • CPU subtotal — CPUs (servers times sockets) multiplied by the Recommended Customer Price Intel lists for the model, compared with the subtotal the order states.

Any Xeon 6 order that touches these rules is in scope — not only the scenarios pictured here. Not covered: OEM bid, volume or contract pricing and discount approval, other Xeon 6 models and families, motherboards and memory modules. The prices are the Recommended Customer Prices Intel lists, not a quotation.

The AI sales agentBuilds the processor order from the conversation and says it is ready.
Predictika validatorChecks every rule at once and returns the verdict, each finding, and the computed figures.

7 scenarios in 3 groups · each group runs from the agent’s first attempt to the outcome · click one to switch

What the agent proposes

The AI sales agent says

Illustrative agent statement. The validator receives the structured request below.

Proposing agent: AI sales agent (scripted)

Before sending, the agent must call Recorded run

Predictika validator


      

        
        
        
      

      
Waiting for the call.

      
    

  

  
  

    
Where rules-based CPQ gets it wrong
Click to expand
    
Our assessment of what a typical rules-based CPQ (product rules, price rules and validation rules, as in Salesforce CPQ) does with the same cases. It is a judgment about effort and reliability, not a claim that nothing can be built.

    
What the validator caughtTypical rules-based CPQPredictika
A processor paired with the wrong platform socketOnly with care
Each model has its own socket, so every processor and platform pairing needs its own compatibility rule.		The CPU and the platform socket are in one request and compared on every call.
Memory type, speed and channels that this processor does not listOnly with care
Type, speed ceiling and channel count differ by model, so each combination needs its own rule and limits.		The CPU, memory type, speed and channels are compared on every call, in any order.
A subtotal that disagrees with servers, sockets and priceOnly with workarounds
The subtotal is a derived figure; checking a stated total against it takes extra price rules and formulas that interact with the rest.		The stated subtotal is a field in the request and is compared with the computed one.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the order is saved or a step is completed, so a rep often fixes one error and meets the next.		All findings come back together, whatever order the choices arrived in.

  

  

    
What this validator catches: findings from its test cases
Click to expand
    
Every rule is checked on every call, in any order the choices arrive; one problem never hides another. Blocking findings stop the quote; approval findings need a person to sign off; everything else is computed.

    
1The order asks for memory at 8,801 MT/s, but the listed maximum for the selected memory type is 8,800 MT/s.
2The order asks for memory at 6,401 MT/s, but the listed maximum for the selected memory type is 6,400 MT/s.
3The order specifies MRDIMM memory, but the Intel specifications page for the selected processor lists DDR5 at up to 6,400 MT/s only.
4The order pairs the selected processor (socket FCLGA7529) with a platform of the other socket type, FCLGA4710.
5The order pairs the selected processor (socket FCLGA4710) with a platform of the other socket type, FCLGA7529.
6The order populates 9 memory channels per socket, but the selected processor has 8 memory channels.
7The order populates 13 memory channels per socket, but the selected processor has 12 memory channels.
8The order puts four processors in each server, but Intel lists the selected Xeon 6 processor as 2S (up to two sockets).
9The order states a CPU subtotal of 4,369,500 USD, but 1,000 CPUs at the Intel-listed recommended customer price of 8,739 USD each come to 8,739,000 USD.
10The order states a CPU subtotal of 0 USD, but 1,000 CPUs at the Intel-listed recommended customer price of 8,739 USD each come to 8,739,000 USD.
11The order asks for memory at 9,000 MT/s, but the listed maximum for the selected memory type is 8,800 MT/s.
12The order populates 12 memory channels per socket, but the selected processor has 8 memory channels.

  

  

    
Scenarios and results
Click to expand
    
The recorded result of the validator for each scenario on this page: the request the AI agent sent, the verdict, and every finding, in the order they appear above.

    
1. Memory and total: A Xeon 6760P order with its own memory settings and total

    
The agent's request. Please quote 250 servers, each with two Intel Xeon 6760P processors on a platform with socket FCLGA4710, with MRDIMM memory at 8,800 MT/s on 12 channels per socket. The order states a CPU subtotal of USD 2,184,750. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • MRDIMM memory on a processor that lists DDR5 only. The order specifies MRDIMM memory, but the Intel specifications page for the selected processor lists DDR5 at up to 6,400 MT/s only.
    • 
      
  • More memory channels than the processor has. The order populates 12 memory channels per socket, but the selected processor has 8 memory channels.
    • 
      
  • Stated CPU subtotal differs from the Intel-listed price. The order states a CPU subtotal of 2,184,750 USD, but 500 CPUs at the Intel-listed recommended customer price of 8,739 USD each come to 4,369,500 USD.
    • 
    

    
2. Memory and total: The agent resubmits after fixing the errors

    
The agent's request. Please quote 250 servers, each with two Intel Xeon 6760P processors on a platform with socket FCLGA4710, with DDR5 memory at 8,800 MT/s on 8 channels per socket. The order states a CPU subtotal of USD 4,369,500. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Memory speed above the listed maximum. The order asks for memory at 8,800 MT/s, but the listed maximum for the selected memory type is 6,400 MT/s.
    • 
    

    
3. Memory and total: The agent resubmits after fixing the errors

    
The agent's request. Please quote 250 servers, each with two Intel Xeon 6760P processors on a platform with socket FCLGA4710, with DDR5 memory at 6,400 MT/s on 8 channels per socket. The order states a CPU subtotal of USD 4,369,500. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 500 CPUs (250 servers, 2 per server), 32,000 cores in all, at an Intel-listed recommended customer price of 8,739 USD each, 4,369,500 USD in all.
    • 
    

    
4. Platform socket: A Xeon 6980P order on an FCLGA4710 platform

    
The agent's request. Please quote 400 servers, each with two Intel Xeon 6980P processors on a platform with socket FCLGA4710, with DDR5 memory at 6,400 MT/s on 12 channels per socket. The order states a CPU subtotal of USD 11,164,000. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Platform socket does not match the processor. The order pairs the selected processor (socket FCLGA7529) with a platform of the other socket type, FCLGA4710.
    • 
    

    
5. Platform socket: The agent resubmits after fixing the errors

    
The agent's request. Please quote 400 servers, each with two Intel Xeon 6980P processors on a platform with socket FCLGA7529, with DDR5 memory at 6,400 MT/s on 12 channels per socket. The order states a CPU subtotal of USD 11,164,000. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 800 CPUs (400 servers, 2 per server), 102,400 cores in all, at an Intel-listed recommended customer price of 13,955 USD each, 11,164,000 USD in all.
    • 
    

    
6. Sockets per server: A Xeon 6780E order with four processors per server

    
The agent's request. Please quote 200 servers, each with four Intel Xeon 6780E processors on a platform with socket FCLGA4710, with DDR5 memory at 6,400 MT/s on 8 channels per socket. The order states a CPU subtotal of USD 1,907,000. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • More sockets per server than the processor supports. The order puts four processors in each server, but Intel lists the selected Xeon 6 processor as 2S (up to two sockets).
    • 
      
  • Stated CPU subtotal differs from the Intel-listed price. The order states a CPU subtotal of 1,907,000 USD, but 800 CPUs at the Intel-listed recommended customer price of 9,535 USD each come to 7,628,000 USD.
    • 
    

    
7. Sockets per server: The agent resubmits after fixing the errors

    
The agent's request. Please quote 200 servers, each with two Intel Xeon 6780E processors on a platform with socket FCLGA4710, with DDR5 memory at 6,400 MT/s on 8 channels per socket. The order states a CPU subtotal of USD 3,814,000. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Order summary. 400 CPUs (200 servers, 2 per server), 57,600 cores in all, at an Intel-listed recommended customer price of 9,535 USD each, 3,814,000 USD in all.
    • 
    

  

  

    
Narration transcript
Click to expand
    
The text of the narrated walkthrough, step by step.

    
    
      
  1. Welcome to a demo of Predictika's Validator. An AI sales agent prepares an order for Intel Xeon 6 server processors with the CPU model, sockets per server. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the processor specifications Intel publishes and returns a verdict, every reason and the figures it computed.
    2. 
      
  2. I will start with the first request by the agent.
    3. 
      
  3. The agent says. Please quote 250 servers, each with two Intel Xeon 6760P processors on a platform with socket FCLGA4710, with MRDIMM memory at 8,800 MT/s on 12 channels per socket. The order states a CPU subtotal of USD 2,184,750. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with three findings, all in one answer. The order specifies MRDIMM memory, but the Intel specifications page for the selected processor lists DDR5 at up to 6,400 MT/s only. The order populates 12 memory channels per socket, but the selected processor has 8 memory channels. The order states a CPU subtotal of 2,184,750 USD, but 500 CPUs at the Intel-listed recommended customer price of 8,739 USD each come to 4,369,500 USD.
    6. 
      
  6. The validator returned three blocking findings in one answer. A check that reads one field at a time would pass most of this; it shows only when the choices in the quote are checked against each other and against the stated total.
    7. 
      
  7. What this shows: for memory type, speed and channels that this processor does not list, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the memory type to DDR5, the memory channels per socket to 8 and the stated CPU subtotal as to $4,369,500.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Still blocked, with one finding, all in one answer. The order asks for memory at 8,800 MT/s, but the listed maximum for the selected memory type is 6,400 MT/s.
    12. 
      
  12. Three findings cleared, but the change brings a different blocking finding: memory speed above the listed maximum.
    13. 
      
  13. The agent resubmits once more.
    14. 
      
  14. The revised request. It changes the memory speed to 6,400 MT/s.
    15. 
      
  15. It calls the validator once more.
    16. 
      
  16. This time the validator clears it.
    17. 
      
  17. The earlier finding is cleared, and the validator's total, USD 4,369,500, matches the quote.
    18. 
      
  18. Three problems caught in one call, fixed in two passes, and a clean $4,369,500 quote. That is the value of Predictika's validator: catching mistakes before they reach the customer.
    19.