| Title |
BLAS3 optimization for the Godson-3B1500
|
|---|---|
| Published in |
SpringerPlus, November 2016
|
| DOI | 10.1186/s40064-016-3690-3 |
| Pubmed ID | |
| Authors |
Ming Zhang, Naijie Gu, Kaixin Ren |
| Abstract |
This paper proposes a performance model for general matrix multiplication (GEMM) on decoupled access/execute (DAE) architecture platforms, in order to guide improvements of the GEMM performance in the Godson-3B1500. This model focuses on the features of access processors (APs) and execute processors (EPs). To reduce the synchronization overhead between APs and EPs, a synchronization module selection mechanism (SMSM) is presented. Furthermore, two optimized algorithms of GEMM for DAE platforms based on the performance model are proposed for ideal performance. In the proposed algorithms, the kernel functions are optimized with single instruction multiple data (SIMD) vector instructions, and the overhead of AP is almost overlapped with EP by taking full advantage of the features of the architecture. Moreover, the synchronization overhead can be reduced according to the SMSM. In the end, the proposed algorithms are tested on the Godson-3B1500. The experimental results demonstrate that the computing performance of dGEMM reaches 91.9% of the theoretical peak performance and that zGEMM can reach 93% of the theoretical peak performance. |
X Demographics
The data shown below were collected from the profiles of 4 X users who shared this research output. Click here to find out more about how the information was compiled.
Geographical breakdown
| Country | Count | As % |
|---|---|---|
| Italy | 1 | 25% |
| Unknown | 3 | 75% |
Demographic breakdown
| Type | Count | As % |
|---|---|---|
| Members of the public | 4 | 100% |
Mendeley readers
The data shown below were compiled from readership statistics for 5 Mendeley readers of this research output. Click here to see the associated Mendeley record.
Geographical breakdown
| Country | Count | As % |
|---|---|---|
| Japan | 1 | 20% |
| Unknown | 4 | 80% |
Demographic breakdown
| Readers by professional status | Count | As % |
|---|---|---|
| Student > Bachelor | 1 | 20% |
| Professor | 1 | 20% |
| Student > Ph. D. Student | 1 | 20% |
| Researcher | 1 | 20% |
| Professor > Associate Professor | 1 | 20% |
| Other | 0 | 0% |
| Readers by discipline | Count | As % |
|---|---|---|
| Computer Science | 2 | 40% |
| Physics and Astronomy | 1 | 20% |
| Engineering | 1 | 20% |
| Unknown | 1 | 20% |
Attention Score in Context
This research output has an Altmetric Attention Score of 3. This is our high-level measure of the quality and quantity of online attention that it has received. This Attention Score, as well as the ranking and number of research outputs shown below, was calculated when the research output was last mentioned on 14 May 2017.
All research outputs
#13,396,068
of 23,340,595 outputs
Outputs from SpringerPlus
#657
of 1,856 outputs
Outputs of similar age
#202,665
of 418,382 outputs
Outputs of similar age from SpringerPlus
#42
of 85 outputs
Altmetric has tracked 23,340,595 research outputs across all sources so far. This one is in the 42nd percentile – i.e., 42% of other outputs scored the same or lower than it.
So far Altmetric has tracked 1,856 research outputs from this source. They typically receive a little more attention than average, with a mean Attention Score of 5.8. This one has gotten more attention than average, scoring higher than 64% of its peers.
Older research outputs will score higher simply because they've had more time to accumulate mentions. To account for age we can compare this Altmetric Attention Score to the 418,382 tracked outputs that were published within six weeks on either side of this one in any source. This one has gotten more attention than average, scoring higher than 51% of its contemporaries.
We're also able to compare this research output to 85 others from the same source and published within six weeks on either side of this one. This one has gotten more attention than average, scoring higher than 51% of its contemporaries.