Quartz sync: Nov 4, 2024, 8:15 PM

content/5.Semester/PHPC/ÜB/ÜB1/ÜB - 1 PHPC.md

@@ -1,26 +1,36 @@
 ---
 date created: Monday, 4. November 2024, 19:10
-date modified: Monday, 4. November 2024, 19:20
+date modified: Monday, 4. November 2024, 20:12
 ---
 
-Here is the content in Markdown format with the full task statements included in the headers:
-
-# 1. Note
-
-Please join us here: https://matrix.lmu.de/#/room/#wa2425:matrix.lmu.de for questions and discussions.
-
-There is no bonus for the final exams from the lab exercises.
-
 # 2. The Top500 List
 
 ## (a) Explain the terms Rmax and Rpeak in the ranking
 
+- `Rmax`
+  - → Maximal LINPACK performance achieved
+  - LINPACK benchmark measures how fast it can solve a large System of linear equations
+- `Rpeak`
+  - → Theoretical peak performance
+  - calculated based on its hardware specifications, such as the number of processors and their clock speeds
+
 ## (b) Name 3 trends you can observe since the first release of the Top500 list in June 1996
 
+- **Exponential Performance Growth**: Supercomputers have experienced rapid increases in computational power, with the leading system's performance doubling approximately
+- **Rise of Parallel Processing**: There's been a significant increase in the number of processing cores, reflecting a shift towards parallel architectures to handle complex computations more efficiently.
+- **Enhanced Energy Efficiency**: Modern supercomputers are achieving higher performance while consuming less energy, emphasizing a focus on sustainable high-performance computing.
+
 ## (c) Choose any high ranked system from the Top500 list and discuss its specification compared to SuperMUC-NG
 
+| System      | Manufacturer | Cores     | Processor                                  | LINPACK Performance (Rmax) | Rpeak           | Power Consumption |
+| ----------- | ------------ | --------- | ------------------------------------------ | -------------------------- | --------------- | ----------------- |
+| Frontier    | HPE Cray     | 8,699,904 | AMD EPYC 64-core CPUs with AMD MI250X GPUs | 1.1 exaflops               | 2 exaflops      | ~21 MW            |
+| SuperMUC-NG | Lenovo       | 311,040   | Intel Xeon Platinum 8174 24-core CPUs      | 19.48 petaflops            | 26.87 petaflops | ~4 MW             |
+
 ## (d) Today, mobile devices, such as smartphones and tablets, are equipped with relatively powerful CPUs. Could a mobile device achieving a Linpack performance of 2 GFLOP/sec have been ranked in past Top500 lists? What is the best ranking it could have reached?
 
+A mobile device achieving 2 GFLOP/s could have ranked within the TOP500 lists up until June 1995. Beyond this date, the performance of the 500th-ranked system exceeded 2 GFLOP/s, making it ineligible for the list thereafter.
+
 ## (e) Assume a CPU with the following specification: 20 cores @ 2 GHz base frequency, 2x floating-point units per core with support for fused multiply-add (FMA) instructions, and a 256-bit vector length. Calculate the theoretical (double precision floating-point) peak performance of this CPU
 
 # 3. Moore’s Law
@@ -98,11 +108,11 @@ comments powered by Disqus.
 <div id="disqus_thread"></div>
 <script>
     /**
-    *  RECOMMENDED CONFIGURATION VARIABLES: EDIT AND UNCOMMENT THE SECTION BELOW TO INSERT DYNAMIC VALUES FROM YOUR PLATFORM OR CMS.
-    *  LEARN WHY DEFINING THESE VARIABLES IS IMPORTANT: https://disqus.com/admin/universalcode/#configuration-variables    */
+    * RECOMMENDED CONFIGURATION VARIABLES: EDIT AND UNCOMMENT THE SECTION BELOW TO INSERT DYNAMIC VALUES FROM YOUR PLATFORM OR CMS.
+    * LEARN WHY DEFINING THESE VARIABLES IS IMPORTANT: https://disqus.com/admin/universalcode/#configuration-variables */
     /*
     var disqus_config = function () {
-    this.page.url = PAGE_URL;  // Replace PAGE_URL with your page's canonical URL variable
+    this.page.url = PAGE_URL; // Replace PAGE_URL with your page's canonical URL variable
     this.page.identifier = PAGE_IDENTIFIER; // Replace PAGE_IDENTIFIER with your page's unique identifier variable
     };
     */