[CORE] concurrent partial prefills (#2372)

# What this PR does / why we need it?

When processing a mix of large and small requests, the TTFT of responses
is significantly reduc\ed. Please refer to
https://github.com/vllm-project/vllm/pull/10235, which achieves the same
effect by simply limiting the number of prompt fills for long requests.
This solution can be applied to both AscendScheduler (V0) and vLLM
Scheduler (V1). Tests show that TTFT can be significantly improved when
handling such mixed requests. However, This capability is currently
missing when Ascend Scheduler is enabled.

This benchmark used the Qwen3-8B model, with a context length of 128K,
running on a single card.

Regarding dataset selection, the sharegpt_clean dataset is used, with
its content concatenated and cropped. Small requests with token=50 and
medium requests with token=10240 were constructed (there were also large
requests with token=102400, but these were ignored because when using
the Prefill First scheduling strategy, max_num_batched_tokens will not
be set to such a large value). When loading vLLM, set
max_num_batched_tokens=22000. This length can accommodate two
medium-sized requests and some short requests, reflecting an extreme
scenario where the budget is almost entirely occupied by longer
requests.

Next, we mix 990 small requests and 100 medium requests into one type of
load scenario (hereinafter referred to as 10%), and similarly generate
load scenarios with 5% medium requests and 1% load scenarios.

Performance tests were conducted separately for enabling vLLMScheduler,
AscendScheduler, and AscendScheduler (long prompt concurrency set to 1).

- vLLM version: v0.10.2
- vLLM main:
1dfea5f4a9

---------

Signed-off-by: Csrayz <jover@cmbchina.com>

docs/source/user_guide/configuration/additional_config.md

@@ -60,6 +60,8 @@ The details of each config option are as follows:
 | `enabled` | bool | `False` | Whether to enable ascend scheduler for V1 engine|
 | `enable_pd_transfer` | bool | `False` | Whether to enable pd transfer. When using it, decode is started only when prefill of all requests is done. This option only takes effects on offline inference. |
 | `decode_max_num_seqs` | int | `0` | Whether to change max_num_seqs of decode phase when enable pd transfer. This option only takes effects when enable_pd_transfer is True. |
+| `max_long_partial_prefills` | Union[int, float] | `float('inf')` | the maximum number of prompts longer than long_prefill_token_threshold that will be prefilled concurrently. |
+| `long_prefill_token_threshold` | Union[int, float] | `float('inf')` | a request is considered long if the prompt is longer than this number of tokens. |
 
 ascend_scheduler_config also support the options from [vllm scheduler config](https://docs.vllm.ai/en/stable/api/vllm/config.html#vllm.config.SchedulerConfig). For example, you can add `enable_chunked_prefill: True` to ascend_scheduler_config as well.
 
@@ -79,6 +81,8 @@ An example of additional configuration is as follows:
     "ascend_scheduler_config": {
         "enabled": True,
         "enable_chunked_prefill": True,
+        "max_long_partial_prefills": 1,
+        "long_prefill_token_threshold": 4096,
     },
     "multistream_overlap_shared_expert": True,
     "refresh": False,

tests/ut/core/test_schedule_config.py

@@ -50,6 +50,8 @@ class TestAscendSchedulerConfig(TestBase):
                 scheduler_cls="vllm_ascend.core.scheduler.AscendScheduler",
                 max_num_batched_tokens=2048,
                 max_model_len=2048,
+                max_long_partial_prefills=1,
+                long_prefill_token_threshold=512,
             ),
         )
         self.assertEqual(ascend_config.enable_chunked_prefill, False)
@@ -58,6 +60,8 @@ class TestAscendSchedulerConfig(TestBase):
                          "vllm_ascend.core.scheduler.AscendScheduler")
         self.assertEqual(ascend_config.max_num_batched_tokens, 2048)
         self.assertEqual(ascend_config.encoder_cache_size, 2048)
+        self.assertEqual(ascend_config.max_long_partial_prefills, 1)
+        self.assertEqual(ascend_config.long_prefill_token_threshold, 512)
 
     def test_not_implemented_policy(self):
         with self.assertRaises(NotImplementedError) as context:

tests/ut/core/test_scheduler.py

@@ -221,7 +221,7 @@ class TestAscendScheduler(TestBase):
                              len(requests) - i - 1)
 
     def test_schedule(self):
-        '''Test scheduling. 
+        '''Test scheduling.
         Two cases: default APC/no prompt logprobs; APC=True + prompt logprobs
         '''
         scheduler = self.create_scheduler()
@@ -279,6 +279,27 @@ class TestAscendScheduler(TestBase):
         for i, request in enumerate(requests):
             self.assertEqual(scheduler.running[i], request)
 
+    def test_concurrent_partial_prefills_schedule(self):
+        '''Test concurrent partial prefills scheduling.
+        total requests = 10, every request has 10 token.
+        while set long_prefill_token_threshold = 1, scheduler can
+        only schedule max_long_partial_prefills long request.
+        '''
+        scheduler = self.create_scheduler()
+        scheduler.scheduler_config.chunked_prefill_enabled = False
+        scheduler.scheduler_config.max_long_partial_prefills = 2
+        scheduler.scheduler_config.long_prefill_token_threshold = 1
+        requests = create_requests(num_requests=10, num_tokens=20)
+        for request in requests:
+            scheduler.add_request(request)
+
+        # Test initial scheduling
+        output = scheduler.schedule()
+        self.assertEqual(len(output.scheduled_new_reqs),
+                         scheduler.scheduler_config.max_long_partial_prefills)
+        self.assertEqual(output.scheduled_cached_reqs.num_reqs, 0)
+        self.assertEqual(len(output.finished_req_ids), 0)
+
     def test_schedule_enable_prefix_caching(self):
         '''Test scheduling.
         Two cases: default APC/no prompt logprobs; APC=True + prompt logprobs

vllm_ascend/core/schedule_config.py

@@ -20,10 +20,14 @@ from typing import Type, Union
 
 from vllm.config import SchedulerConfig
 
+MAX_INT = 2147483647
+
 
 @dataclass
 class AscendSchedulerConfig(SchedulerConfig):
     enable_chunked_prefill: bool = False
+    max_long_partial_prefills: int = MAX_INT
+    long_prefill_token_threshold: int = MAX_INT
     policy: str = "fcfs"
     scheduler_cls: Union[str, Type[object]] = (
         "vllm_ascend.core.scheduler.AscendScheduler")
@@ -42,6 +46,8 @@ class AscendSchedulerConfig(SchedulerConfig):
         }
         # Override default values into original SchedulerConfig
         scheduler_config["enable_chunked_prefill"] = False
+        scheduler_config["max_long_partial_prefills"] = None
+        scheduler_config["long_prefill_token_threshold"] = None
         scheduler_config["policy"] = "fcfs"
         scheduler_config["scheduler_cls"] = (
             "vllm_ascend.core.scheduler.AscendScheduler")
@@ -67,6 +73,28 @@ class AscendSchedulerConfig(SchedulerConfig):
                 "max_num_batched_tokens and makes vLLM reject longer "
                 "sequences. Please increase max_num_batched_tokens or "
                 "decrease max_model_len.")
+        # concurrent partial prefills. Default is inf
+        if self.max_long_partial_prefills is None:
+            self.max_long_partial_prefills = MAX_INT
+            self.long_prefill_token_threshold = MAX_INT
+
+        if self.long_prefill_token_threshold is None or \
+            self.long_prefill_token_threshold <= 0:
+            if self.max_model_len is None:
+                self.long_prefill_token_threshold = MAX_INT
+            else:
+                self.long_prefill_token_threshold = \
+                    max(1, int(self.max_model_len * 0.04))
+
+        if self.max_long_partial_prefills < 0:
+            raise ValueError(
+                f"max_long_partial_prefills must be non-negative, but got "
+                f"{self.max_long_partial_prefills}")
+        if self.long_prefill_token_threshold < 0:
+            raise ValueError(
+                f"long_prefill_token_threshold must be non-negative, but got "
+                f"{self.long_prefill_token_threshold}")
+
         if self.policy != "fcfs":
             raise NotImplementedError(
                 f"currently AscendScheduler only supports fcfs policy, got {self.policy}"

vllm_ascend/core/scheduler.py

@@ -102,6 +102,14 @@ class AscendScheduler(Scheduler):
             # all request prefilled, change phase to decode
             if not self.waiting and not self.running:
                 self.phase = "decode"
+        # Skip long prompt requests in prefill stage.
+        # long_prefill_budget is float('inf') if not use.
+        if self.vllm_config.scheduler_config.long_prefill_token_threshold == 0:
+            long_prefill_budget = float('inf')
+            long_prefill_token_threshold = float('inf')
+        else:
+            long_prefill_budget = self.vllm_config.scheduler_config.max_long_partial_prefills
+            long_prefill_token_threshold = self.vllm_config.scheduler_config.long_prefill_token_threshold
 
         # Schedule prefill requests first.
         while self.waiting and token_budget > 0:
@@ -217,6 +225,11 @@ class AscendScheduler(Scheduler):
                 skip_cur_request()
                 continue
 
+            if  num_new_tokens > long_prefill_token_threshold \
+                and long_prefill_budget <= 0:
+                skip_cur_request()
+                continue
+
             new_blocks = self.kv_cache_manager.allocate_slots(
                 request,
                 num_new_tokens + num_external_computed_tokens,
@@ -268,6 +281,8 @@ class AscendScheduler(Scheduler):
             # Update request info.
             num_scheduled_tokens[request.request_id] = num_new_tokens
             token_budget -= num_new_tokens
+            if num_new_tokens > long_prefill_token_threshold:
+                long_prefill_budget -= 1
             request.status = RequestStatus.RUNNING
             request.num_computed_tokens = num_computed_tokens
             # Count the number of prefix cached tokens.