5 ms·

> First trend: Moore’s Law, or rather, its lack thereof. Despite all the efforts of Apple to hide its flattening curve with ARM-based chips, it is vox populi th

> First trend: Moore’s Law, or rather, its lack thereof. Despite all the efforts of Apple to hide its flattening curve with ARM-based chips, it is vox populi that a computer from 2014 is just about as powerful as the one you bought as a Christmas gift in 2023, bar the amount of RAM or the speed and capacity of the SSDs therein. Is this really true? https://www.cpubenchmark.net/compare/2275vs5022/Intel-i7-4790K-vs-Intel-i9-13900K https://www.cpubenchmark.net/compare/2275vs5022/Intel-i7-479... Only on the CPU side, there's an 86.4% difference between the two top models. Even between the top Apple CPU at the time, there's an 80% difference. And in terms of GPU & energy consumption, the difference is exponential..

margalabargala 3y ago
Those passmark score increases are almost entirely due to increased thread count. Single thread scores have only doubles in that time. Compare to the decade before; picking a Pentium 4 from 2004, the single thread benchmark quintupled to 2014. Using a Pentium 4 computer was painful in 2014, I know, I remember doing it. Using a 2014 computer today isn't so bad at all. It's definitely the case that the rate of improvement has slowed.
- xwolfi 3y ago
  But my eyes see something different. Let's take a very good game in 2014, Dragon Age: Inquisition, and a good game this year, say Cyberpunk 2077. There is a progress between the two that sort of compensate for maybe single-thread benchmark lack of progress. Ofc the moore low is done now if you look at chips, but the pace of progress, if you look at the actual result where you need processing, seems to be satisfying. I do know what you mean though: since multi-core processors started, the experience on office-like work has been little changed and you can def run everything well with a 3Ghz 4-core CPU today. The SSD was really more of a game changer.
  - margalabargala 3y ago
    If the progress you're talking about in terms of processing efficiency of those games? Or something else? Of note, I would bet that if you took a modern GPU, SSD, 32gb of RAM, and stuck it on a 2014-era motherboard with a i7-4790K it would run Cyberpunk 2077 just fine. I don't think you could say the same about a Pentium 4 running Dragon Age: Inquisition, or even Skyrim. Agree on the SSD being a huge gamechanger.
- samus 3y ago
  Single-thread performance is only relevant for workloads that can't be parallelized, or for software that will never be adapted. It has been well known for more than a decade now that parallelization is the only way forward. Edit: application-specific accellerators are another promising approach that has been in vogue for a long time and which achieves speedups that are obviously not easily visible by looking at simple GHz metrics
  - pjc50 3y ago
    > only relevant for workloads that can't be parallelized This ends up dominating anyway. Partly because the easy stuff does get parallelized. So you're left with things like "browser layout engine" being single threaded. Almost all systems have single-threaded UIs, because trying to do anything else becomes unreasonably hard to reason about. The UI thread will delegate work to other threads, but there is nearly always a single "UI thread" bottleneck.
    - samus 3y ago
      For most desktop applications, this bottleneck is not that restricting though. Almost any noticeable UI delay is due to stuff that ought to be offloaded into a background thread, but isn't. For example, IntelliJ often freezes up when I enter a search string in the settings dialog, which is probably due to a search index being loaded or built on the UI thread. Layout engines for browsers do get parallelized though. And the point of modern 3D APIs like Vulkan APIs is precisely to allow calling it from more than one thread.

margalabargala 3y ago

Those passmark score increases are almost entirely due to increased thread count. Single thread scores have only doubles in that time. Compare to the decade before; picking a Pentium 4 from 2004, the single thread benchmark quintupled to 2014. Using a Pentium 4 computer was painful in 2014, I know, I remember doing it. Using a 2014 computer today isn't so bad at all. It's definitely the case that the rate of improvement has slowed.

xwolfi 3y ago
But my eyes see something different. Let's take a very good game in 2014, Dragon Age: Inquisition, and a good game this year, say Cyberpunk 2077. There is a progress between the two that sort of compensate for maybe single-thread benchmark lack of progress. Ofc the moore low is done now if you look at chips, but the pace of progress, if you look at the actual result where you need processing, seems to be satisfying. I do know what you mean though: since multi-core processors started, the experience on office-like work has been little changed and you can def run everything well with a 3Ghz 4-core CPU today. The SSD was really more of a game changer.
- margalabargala 3y ago
  If the progress you're talking about in terms of processing efficiency of those games? Or something else? Of note, I would bet that if you took a modern GPU, SSD, 32gb of RAM, and stuck it on a 2014-era motherboard with a i7-4790K it would run Cyberpunk 2077 just fine. I don't think you could say the same about a Pentium 4 running Dragon Age: Inquisition, or even Skyrim. Agree on the SSD being a huge gamechanger.
samus 3y ago
Single-thread performance is only relevant for workloads that can't be parallelized, or for software that will never be adapted. It has been well known for more than a decade now that parallelization is the only way forward. Edit: application-specific accellerators are another promising approach that has been in vogue for a long time and which achieves speedups that are obviously not easily visible by looking at simple GHz metrics
- pjc50 3y ago
  > only relevant for workloads that can't be parallelized This ends up dominating anyway. Partly because the easy stuff does get parallelized. So you're left with things like "browser layout engine" being single threaded. Almost all systems have single-threaded UIs, because trying to do anything else becomes unreasonably hard to reason about. The UI thread will delegate work to other threads, but there is nearly always a single "UI thread" bottleneck.
  - samus 3y ago
    For most desktop applications, this bottleneck is not that restricting though. Almost any noticeable UI delay is due to stuff that ought to be offloaded into a background thread, but isn't. For example, IntelliJ often freezes up when I enter a search string in the settings dialog, which is probably due to a search index being loaded or built on the UI thread. Layout engines for browsers do get parallelized though. And the point of modern 3D APIs like Vulkan APIs is precisely to allow calling it from more than one thread.