Predictika · 2U Intel Xeon 6 rack server configurations

An AI sales agent's 2U rack server build, checked before it goes out

An AI sales agent prepares a configuration for a 2U dual-socket rack server: chassis, processors, memory, drives, GPUs, PCIe cards and power supplies. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the configuration limits and returns a verdict, every reason and the figures it computed.

Scope: ASUS RS720-E12 series 2U rack servers (RS8G, RS12U and RS24U/G chassis): processors, memory, drives, GPUs, PCIe cards and power supplies. See everything this validator covers

This validator models configurations of the ASUS RS720-E12 series 2U Intel Xeon 6 rack servers, not just the scenarios shown below.

  • Chassis and drives — the RS720-E12-RS8G has 8 and the RS720-E12-RS24U/G 24 hot-swap 2.5-inch bays, the RS720-E12-RS12U has 12 hot-swap 3.5-inch bays; two M.2 connectors.
  • Processors and memory — dual-socket, processors up to 350 W TDP, 16 DIMM slots per processor, DDR5 RDIMMs of 64, 96 or 128 GB, MRDIMM at one DIMM per channel only; memory capacity and the maximum memory speed (6400 MT/s at one DIMM per channel, 5200 MT/s at two, 8000 MT/s for MRDIMM) are computed and compared with what the quote states.
  • GPUs, PCIe cards and power — up to 3 dual-slot GPUs; the full-height rear slots give 6 PCIe Gen5 x8 links or 3 x16 links, plus 2 half-height x16 slots and 2 OCP 3.0 slots; a 1+1 redundant pair of 3200 W or 2700 W Titanium power supplies; 10 to 35 degrees C operating temperature.

Any RS720-E12 series configuration that touches these rules is in scope — not only the scenarios pictured here. Not covered: prices (ASUS publishes no price list for these servers, so this validator produces no price build-up or line-item quote), the approved vendor list of specific CPUs, DIMMs, drives, GPUs and RAID or HBA cards, other ASUS server families, and regional power supply variants.

The AI sales agentBuilds the server configuration 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
Slot and bay limits that depend on the chassisOnly with care
A set of product rules per chassis. It works, but each chassis variant needs its own limits kept up to date.		The chassis choice and every count are in one request, so each limit is checked against the chassis selected.
GPUs and cards that share the same PCIe lanesOnly with workarounds
The full-height slots offer 6 x8 links or 3 x16 links, so a GPU uses two links. That needs a computed total of link units across several selections, which rule screens handle poorly.		Link units used are computed from every GPU and card on each call and compared with the 6 available.
Memory speed that depends on how the DIMMs are populatedOnly with workarounds
The speed drops from 6400 to 5200 MT/s above one DIMM per channel and MRDIMM works only at one per channel; this needs a calculation from the DIMM count and processor count.		Modules per processor and memory type determine the speed on every call, and a stated speed is compared with it.
Stated memory capacity that must match the modulesOnly with care
A stated figure is a free-entry field; a rule can compare it with a computed total, but someone has to write and maintain that comparison.		The stated figure travels with the configuration and is compared with the computed figure on every call.
Every problem returned in one answerOnly with workarounds
Rules tend to fire as the configuration 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 RS720-E12 series is a dual-socket server; the quote specifies 3 as the processor count.
2The RS720-E12 series supports processors up to 350 W TDP; the quote has a 400 W processor.
3The quote has 32 memory modules, but the processors on it provide only 16 DIMM slots (16 per processor).
4DDR5 RDIMMs for this server are 64 GB, 96 GB or 128 GB; the quote has 32 GB modules.
5MRDIMM memory is supported at one DIMM per channel only (up to 8 modules per processor); the quote has 32 modules on 2 processor(s).
6The quote has 12 drives, but the selected chassis has 8 hot-swap drive bays.
7The RS12U chassis takes 3.5-inch drives and the RS8G and RS24U/G chassis take 2.5-inch drives; the drives on the quote do not fit the selected chassis.
8The server has 2 M.2 connectors; the quote has 3 M.2 drives.
9The server supports at most 3 dual-slot GPUs; the quote has 4.
10The full-height rear slots provide 6 PCIe Gen5 x8 links or 3 x16 links; the GPUs and cards on the quote need 8 x8 links (a GPU or x16 card takes 2).
11The full-height rear slots provide 6 PCIe Gen5 x8 links or 3 x16 links; the GPUs and cards on the quote need 7 x8 links (a GPU or x16 card takes 2).
12The server has 2 half-height PCIe Gen5 x16 slots; the quote has 3 half-height cards.
13The server has 2 OCP 3.0 slots; the quote has 3 OCP 3.0 cards.
14The power supplies are 80 PLUS Titanium 3,200 W or 2,700 W units; the quote has 2,000 W units.
15The operating temperature range is 10 to 35 degrees C; the site is planned at 40 degrees C.
16The operating temperature range is 10 to 35 degrees C; the site is planned at 5 degrees C.
17The quote states 4,096 GB of memory, but the modules on it add up to 2,048 GB.
18The quote states memory speeds up to 6,400 MT/s, but this memory configuration supports up to 5,200 MT/s.
19The RS720-E12 series supports processors up to 350 W TDP; the quote has a 420 W processor.
20DDR5 RDIMMs for this server are 64 GB, 96 GB or 128 GB; the quote has 48 GB modules.
21The server uses a 1+1 redundant power supply pair; the power supply count on the quote is 1.
22The operating temperature range is 10 to 35 degrees C; the site is planned at 38 degrees C.
23The quote states 2,048 GB of memory, but the modules on it add up to 1,536 GB.

  

  

    
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. GPU server: A 24-bay GPU server with four dual-slot GPUs

    
The agent's request. Please quote an ASUS RS720-E12-RS24U/G (24 x 2.5-inch bays) with 2 Intel Xeon 6 processors of 350 W, 32 x 64 GB RDIMM memory, 24 x 2.5-inch drives, 4 dual-slot GPUs, 2 OCP 3.0 cards, 2 M.2 drives, 2 x 3200 W power supply units, for a site at 24 degrees C. The quote states 2,048 GB of memory running up to 5,200 MT/s. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Too many GPUs. The server supports at most 3 dual-slot GPUs; the quote has 4.
    • 
      
  • Not enough full-height PCIe slots. The full-height rear slots provide 6 PCIe Gen5 x8 links or 3 x16 links; the GPUs and cards on the quote need 8 x8 links (a GPU or x16 card takes 2).
    • 
    

    
2. GPU server: The agent resubmits after fixing the errors

    
The agent's request. Please quote an ASUS RS720-E12-RS24U/G (24 x 2.5-inch bays) with 2 Intel Xeon 6 processors of 350 W, 32 x 64 GB RDIMM memory, 24 x 2.5-inch drives, 3 dual-slot GPUs, 2 OCP 3.0 cards, 2 M.2 drives, 2 x 3200 W power supply units, for a site at 24 degrees C. The quote states 2,048 GB of memory running up to 5,200 MT/s. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 2,048 GB of memory (32 x 64 GB) running up to 5,200 MT/s; 6 of 6 PCIe x8 links used in the full-height slots.
    • 
    

    
3. Storage server: A 12-bay storage server with 2.5-inch drives and one power supply

    
The agent's request. Please quote an ASUS RS720-E12-RS12U (12 x 3.5-inch bays) with 2 Intel Xeon 6 processors of 270 W, 16 x 128 GB RDIMM memory, 12 x 2.5-inch drives, 4 x8 full-height cards, 1 OCP 3.0 card, 1 x 2700 W power supply, for a site at 22 degrees C. The quote states 2,048 GB of memory running up to 6,400 MT/s. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Drive size does not fit the chassis. The RS12U chassis takes 3.5-inch drives and the RS8G and RS24U/G chassis take 2.5-inch drives; the drives on the quote do not fit the selected chassis.
    • 
      
  • Power supplies must be 1+1 redundant. The server uses a 1+1 redundant power supply pair; the power supply count on the quote is 1.
    • 
    

    
4. Storage server: The agent resubmits after fixing the errors

    
The agent's request. Please quote an ASUS RS720-E12-RS12U (12 x 3.5-inch bays) with 2 Intel Xeon 6 processors of 270 W, 16 x 128 GB RDIMM memory, 12 x 3.5-inch drives, 4 x8 full-height cards, 1 OCP 3.0 card, 1 x 2700 W power supply, for a site at 22 degrees C. The quote states 2,048 GB of memory running up to 6,400 MT/s. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • Power supplies must be 1+1 redundant. The server uses a 1+1 redundant power supply pair; the power supply count on the quote is 1.
    • 
    

    
5. Storage server: The agent resubmits after fixing the errors

    
The agent's request. Please quote an ASUS RS720-E12-RS12U (12 x 3.5-inch bays) with 2 Intel Xeon 6 processors of 270 W, 16 x 128 GB RDIMM memory, 12 x 3.5-inch drives, 4 x8 full-height cards, 1 OCP 3.0 card, 2 x 2700 W power supply units, for a site at 22 degrees C. The quote states 2,048 GB of memory running up to 6,400 MT/s. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 2,048 GB of memory (16 x 128 GB) running up to 6,400 MT/s; 4 of 6 PCIe x8 links used in the full-height slots.
    • 
    

    
6. MRDIMM compute node: An 8-bay compute node with 32 MRDIMM modules

    
The agent's request. Please quote an ASUS RS720-E12-RS8G (8 x 2.5-inch bays) with 2 Intel Xeon 6 processors of 330 W, 32 x 64 GB MRDIMM memory, 8 x 2.5-inch drives, 2 dual-slot GPUs, 1 x16 full-height card, 1 OCP 3.0 card, 2 x 3200 W power supply units, for a site at 25 degrees C. The quote states 2,048 GB of memory running up to 5,200 MT/s. It is ready to send.

    
Verdict: Blocked.

    
    
      
  • MRDIMM with two modules per channel. MRDIMM memory is supported at one DIMM per channel only (up to 8 modules per processor); the quote has 32 modules on 2 processor(s).
    • 
    

    
7. MRDIMM compute node: The agent resubmits after fixing the errors

    
The agent's request. Please quote an ASUS RS720-E12-RS8G (8 x 2.5-inch bays) with 2 Intel Xeon 6 processors of 330 W, 16 x 64 GB MRDIMM memory, 8 x 2.5-inch drives, 2 dual-slot GPUs, 1 x16 full-height card, 1 OCP 3.0 card, 2 x 3200 W power supply units, for a site at 25 degrees C. The quote states 1,024 GB of memory running up to 8,000 MT/s. It is ready to send.

    
Verdict: Allowed.

    
    
      
  • Configuration summary. 1,024 GB of memory (16 x 64 GB) running up to 8,000 MT/s; 6 of 6 PCIe x8 links used in the full-height slots.
    • 
    

  

  

    
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 a configuration for a 2U dual-socket rack server: chassis, processors, memory, and more. Before it goes out, the agent sends it to Predictika, which checks it against a declarative model of the configuration limits 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 an ASUS RS720-E12-RS12U (12 x 3.5-inch bays) with 2 Intel Xeon 6 processors of 270 W, 16 x 128 GB RDIMM memory. The quote states 2,048 GB of memory running up to 6,400 MT/s. The rest of the request is on screen.
    4. 
      
  4. Before it goes out, the agent calls the validator.
    5. 
      
  5. Blocked, with two findings, all in one answer. The RS12U chassis takes 3.5-inch drives and the RS8G and RS24U/G chassis take 2.5-inch drives; the drives on the quote do not fit the selected chassis. The server uses a 1+1 redundant power supply pair; the power supply count on the quote is 1.
    6. 
      
  6. The validator returned two 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.
    7. 
      
  7. What this shows: for slot and bay limits that depend on the chassis, a rules-based CPQ is error-prone for this.
    8. 
      
  8. Now the agent resubmits.
    9. 
      
  9. The corrected request. It changes the drives to 12 x 3.5-inch.
    10. 
      
  10. It calls the validator again.
    11. 
      
  11. Still blocked, with one finding, all in one answer. The server uses a 1+1 redundant power supply pair; the power supply count on the quote is 1.
    12. 
      
  12. One finding cleared. Still blocking: power supplies must be 1+1 redundant.
    13. 
      
  13. The agent resubmits once more.
    14. 
      
  14. The revised request. It changes the power supplies to 2 x 2700 W.
    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 figure matches the quote.
    18. 
      
  18. Two problems caught in one call, fixed in two passes, and a clean quote. That is the value of Predictika's validator: catching mistakes before they reach the customer. Now you can try the other scenarios yourself.
    19.