Does Mhz Still Matter?
Key topics
The article 'Does MHz Still Matter?' explores the relevance of CPU clock speed in modern computing, sparking a discussion on the trade-offs between clock speed, core count, and other factors like PCIe lanes and memory bandwidth.
Snapshot generated from the HN discussion
Discussion Activity
Very active discussionFirst comment
23m
Peak period
32
0-6h
Avg / period
11.8
Based on 47 loaded comments
Key moments
- 01Story posted
Aug 22, 2025 at 10:47 AM EDT
5 months ago
Step 01 - 02First comment
Aug 22, 2025 at 11:09 AM EDT
23m after posting
Step 02 - 03Peak activity
32 comments in 0-6h
Hottest window of the conversation
Step 03 - 04Latest activity
Aug 24, 2025 at 4:01 PM EDT
5 months ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
That said, at such low core count the primary Epyc advantage is PCIe lanes no?
Also, EPYC's PCIe advantage doesn't hold for the Hetzner provided server setup unfortunately because the configurator allows the same number of devices to be attached to both servers.
But, but with fewer processes I can totally believe this works out to be the better option. Thank you for the write-up!
In our case, though, if we provide 1/48 of 10Gbit network, it really doesn't work for our end customers. So, we're trying to provide the VMs from smaller but more up-to-date lineup.
It's always the workload type. For the mixed environments (some with a heavy constant load while the other have only some occasional spikes) the increase of RAM per node is the most important part what allowed us to actually decrease the node count. The whole racks with multiple switches was replaced by a single rack with a modest amount of servers and a single stacked switch.
It is surprisingly hard to keep a modern CPU core properly saturated unless you are baking global illumination, searching primes or mining crypto currency. I/O and latency will almost always dominate at scale. Moving information is way more expensive than processing it.
And if there is a pointer chasey task that isn't utilizing all of the core and is waiting on memory and data dependencies, often that just makes threading a bigger win, 16 threads doing pointer chases instead of one if you aren't memory bandwidth bound.
Transputers just came 30+ years too early.
In most of the ways that matter, Transputers weren’t too early. And if you built one again now, you’d find little to no market for it (as per efforts by several startups and various related research efforts).
Sources: many conversations with the architect of Transputer, and various of the engineers that designed the hardware and Occam, and watching the Inmos 40th Anniversary lectures on YouTube (channel: @ednutting). Also being in the processor design startup space.
But I figure it is a broad field, so I'm curious what you're doing and if it is the best use of time and energy
I'm also assuming that the generative AI model wouldn't run on your machine well and need to be elsewhere
I think the above answer was just a hype parrot
I would love it if generative AI could get us even further, we are severely compute limited and also testing in the lab is 20k a pop ... I am strongly incentivised for generative AI to be the answer, but as someone who works deeply in the field, the hype is real.
Modern PS5 developments already shows SSD I/O is getting faster than CPU core can keep up. It is also not true when CPU is still the limiting factor on Web Server.
The larger server grade parts start to shine when the server is doing a lot of different things. The extra memory bandwidth helps keep the CPU fed and the higher core count reduces the need for context switching because your workloads aren’t competing as much.
The best part about the AMD consumer CPUs is that you can even use ECC RAM if you get the right motherboard.
https://www.asrock.com/mb/AMD/X870%20Taichi%20Creator/index....
Asus has options as well such as https://www.asus.com/motherboards-components/motherboards/pr...
I think it was more rare when AM5 first came out, there were a bunch of ECC supported consumer boards for AM4 and threadripper.
(I'm currently in the midst of refanning my CSE847-JBOD with bigger, quieter fans and swapping PSUs.)
I use a 5950X for running genetic programming and neuroevolution experiments and about once every 100 hours the machine will just not like the state/load it is experiencing and will restart. My approach is to checkpoint as often as possible. I restart the program the next morning and it deserializes the last snapshot from disk. Worst case, I lose 5 minutes of work.
This also helps with Windows updates, power outages, and EM/cosmic radiation.
If you really care, you can buy an Epyc branded AM4 or AM5 cpu which has remarkably similar specifications and MSRP to Ryzen offerings.
It's unfortunate that we can only have 16 core CPUs running at 5+ GHz. I would have loved to have a 32 or 64 core Ryzen 9. The software we use charge per core used, so 30% less performance is that much extra cost, which is easily an order of magnitude higher than a flagship server CPU. These licenses cost millions per year for couple 16 core seats.
So, at the end, CPU speed is determining how fast abd economically chips are developed.
About the technical problem: it's ODE solution for a sparse matrix of 10s of millions elements. Sparsity comes from the locality of interactions within a chip: not every transistor is connected to thd others. So the modern simulators make use of this sparsity to divide the circuit to independent chunks and spread over to multiple independent threads.
Scale of the compute time is not objective, but typically anywhere from couple hours to couple of months. Most top level chip integration verification jobs take weeks. Because of this we spend months for verification after the design is pretty much finish before the tapeout. This applies for every single reasonably complex chip.
The ROI on hiring a professional overclocker to build, tune and test a workstation is probably at least break even. As long as the right checksums are in place, extreme OC is just a business writeoff.
https://blackcoretech.com/
Given the urgency and the kind of money involved, I offered to set up a gaming PC for them using phase change cooling. Sadly they just made the staff work longer hours to catch up with the paperwork.
Time is money. Or the inverse of money. Ufff, my head hurts.
On that note I cant wait to see 256 Core Zen6c later this year. We will soon be able to buy a server with 512 Core, 1024 vCPU / Thread. 2TB of Memory, x TB of SSD all inside 1U.
Although said somewhat tongue in cheek, it has been a rough several years for tech hobbyist consumers. At least the end of Moore's law scaling and the bite of Dennard scaling combined to nerf generational improvements enough that getting by on existing hardware wasn't as nearly as bad as it would've been 20 yrs ago.
Now that maybe the AI bubble is just starting to burst, we've got tariffs to ensure tech consumers still won't see undistorted prices. The silver lining in all this is that it got me into retro gaming and computing which, frankly, is really great.