-
### Current Behavior
In the absence of a `--cpus` option provided to `nextstrain build`, the AWS Batch runner uses the job definition's default CPU allocation (`vcpus`) which is not detectable by Sna…
-
Not "overloading NAT" as in the RFCs, but port overloading as used by Palo Alto Networks, Juniper, Cisco et al.
Basically what they mean is "using one port for more than one client", by doing desti…
-
If a task is currently running it has consumed VCU resources on the device.
Runahead scheduling allows a dependent to be launched while the task is still running. However, the dependent consumes i…
-
`max_workers` is just an alternative way of defining `cores_per_worker` (except on heterogenous queues, where you probably want `cores_per_worker` anyways). It is just another thing for users to have …
-
Dear mot developper,
when using the AI inference service, it seems that AI JSON format response from the server might be inconsistent over the time.
The error is not deterministically reproductib…
-
**改进内容**
资源看板中,不同的队列资源显示不相通,但是实际使用中,default中使用的就是a40的卡。应该在a40的队列中正确显示已在default中被使用的资源。
![image](https://github.com/t9k/user-support/assets/34251317/3d856bb2-de4e-4a89-9ec1-ba4b68a9ffcf)
**期望的效果…
-
### 🚀 The feature, motivation and pitch
In the typical single-node multi-gpu DDP + multi-threaded DataLoader + intra-op OpenMP parallelism, there might be CPU thread oversubscription leading to slo…
-
cc @laurae2
On benchmark on dual Xeon Gold 6154 vs MKL:
```
Warmup: 0.9943 s, result 224 (displayed to avoid compiler optimizing warmup away)
A matrix shape: (M: 2304, N: 2304)
B matrix sha…
-
Hi,
I'm curious what is going on with the memory management when using sycl::buffer inside of std::vector.
I have a very small example program:
```
#include
#include
using namespace sycl…
-
### Problem Description
The two programming models OpenMP and HIP provided by ROCm leverage the same HSA runtime. HIP holds its own pool HSA queues controlled by the `GPU_MAX_HW_QUEUES` environment…