Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.
C++
#pragma once
#include <cstdint>
namespace strata::kernels {
// Configure before session capture. Existing graphs retain their selected kernels.
// This experiment is off by default and does not modify the legacy router.
void native_router_set_enabled(bool enabled);
bool native_router_enabled();
// Pinned CUDA topk-moe contract for ONE token, 512 experts, top 10, softmax,
// no selection bias, lower normalization clamp 2^-14, and scale 1.
// Reads 512 finite F32 logits; writes 10 I32 IDs and 10 F32 weights. Equal
// computed probabilities select the lower expert index. All spans must be
// four-byte aligned and outputs disjoint from each other and the input.
// Requires a nonnull ordered CUDA stream. No allocation or synchronization.
void native_router_top10(const float* logits, int32_t* ids, float* weights, void* stream);
/// n_tok rows at once (logits [n,512], ids/weights [n,10]); each row exactly as the single call.
void native_router_top10_multi(const float* logits, int32_t* ids, float* weights, int n_tok, void* stream);
/// The same contract for a model family's (experts, top-k): 512 / 10 (Flash-Next: the two calls above, unchanged) and
/// 256 / 8 (Qwen3.6-35B-A3B). logits [n_tok, n_expert], ids / weights [n_tok, k]. CUDA / HIP.
bool native_router_topk_supported(int n_expert, int k);
void native_router_topk(const float* logits, int32_t* ids, float* weights, int n_expert, int k, int n_tok, void* stream);
/// STRATA_ROUTE_RESIDENT: swap non-resident tail-rank experts for resident ones within `margin` logits (see the .cu).
/// `stats` is nullptr or 4 device uint64 counters: tail entries seen, swaps, non-resident before, non-resident after.
void native_route_resident(const float* logits, int32_t* ids, float* weights, const int32_t* res_layer, int n_tok, float margin,
int rank_lo, int rank_hi, unsigned long long* stats, void* stream);
}