[ 01 // PRACTICING ENTITY STOREFRONT · GREQ-KERNELS V1.1.0 ] [ WEBGL GLSL: PURPLE PARTICLE TREE + PORTAL RING + GYROID ORB · 60FPS ]

Where contractive math ↪ becomes serving speed people remember

↪ We engineer and license production GPU serving components — contractive KV-cache admission, prefill/decode disaggregation, and spectral load balancing, delivered as Ed25519-signed Python wheels under 35 U.S.C. § 287(a).

Explore Kernel Licenses ⟶⟶ Self-Serve SKU Matrix

[ 01 / DISTRIBUTION ]v1.1.0 .whl

[ 02 / ENTRY WORKSTATION ]$499

[ 03 / COMMERCIAL SKUS ]05 Tiers

[ 04 / PATENT MARKING ]35 U.S.C. § 287a

[ 02 // COMMERCIAL KERNEL LICENSES ]

Technology ↪ as practicing-entity infrastructure

Every commercial seat, 8-GPU serving node, 64-GPU disaggregated cluster, and silicon fab license includes instant Stripe Card or US Bank Account (ACH) checkout and cryptographic delivery receipts.

[ SKU // 01 · WORKSTATION ] Single GPU Node

GR-EQ Kernel SDK — Developer Workstation

$499/ annual seat license

Commercial practicing-entity license for greq-kernels v1.1.0 (KV-cache admission, prefill/decode disaggregation, load-balancing router, and certification CLI).

Signed wheel & sdist tarball (greq_kernels-1.1.0)
Contractive KV-cache admission & page eviction
Ed25519 certificate-ledger CLI verifier

Optional Engineering Add-ons

Priority Review+$250 Artifact Closeout Packet+$500 Extra Implementation Pass+$750

Request Invoice

[ SKU // 02 · SERVING NODE ] vLLM / SGLang / TRT-LLM

GR-EQ Serving Node License (vLLM / SGLang / TRT-LLM)

$2,500/ per 8-GPU node / year

Production serving license for greq-kernels v1.1.0 with signed wheel delivery, KV-cache contraction gate, and 35 U.S.C. 287(a) patent marking compliance.

Up to 8 GPUs per production serving node
Drop-in KV-cache contraction gate for vLLM & SGLang
35 U.S.C. § 287(a) commercial practicing license

Request Invoice

[ SKU // 03 · DISAGGREGATED CLUSTER ] Up to 64 GPUs

GR-EQ Disaggregated Prefill/Decode Cluster Suite

$15,000/ per 64-GPU cluster / year

Cluster-scale license covering disaggregated prefill/decode transport, cross-node KV handoff, and spectral load balancing across up to 64 GPUs.

Prefill/decode disaggregation transport & KV handoff
Closed-loop spectral load balancing across 64 GPUs
Priority engineering SLA & telemetry receipts

Request Invoice

[ SKU // 04–05 · SILICON FAB & OEM ] Hardware & Cloud Site

Silicon Fab & Enterprise OEM Suite

$49,500–$125,000/ annual license

Hardware/ASIC spectral verification harness ($49,500) or unlimited-node Enterprise Cloud & OEM redistribution ($125,000).

RTL / ASIC spectral verification & certificate ledger
Unlimited-node cloud & OEM appliance redistribution
Dedicated principal architect & custom integration

Request Invoice

[ 03 // SELF-SERVE SKU MATRIX ]

Some of our ⟶⟶ commercial packages

Protected under U.S. Patent Application Nos. 19/846,444 (Track One), 19/846,449, 19/846,445 & PCT/US26/51731. Virtual Patent Marking: whoisjohngalt.ai/patents.

Commercial SKU	License Tier	Self-Serve Action
GR-EQ Kernel SDK — Developer Workstation	$499 · annual seat license
GR-EQ Serving Node License (vLLM / SGLang / TRT-LLM)	$2,500 · per 8-GPU node / year
GR-EQ Disaggregated Prefill/Decode Cluster Suite	$15,000 · per 64-GPU cluster / year
GR-EQ Silicon & Fab Certification Suite	$49,500 · annual fab license
GR-EQ Enterprise Cloud & OEM Appliance License	$125,000 · annual site license