Skip to content
Register

Worth replacing v4 with v4 for 200MHz?

Dedicated Servers by rudi3 11 replies 5.7K views
rudi3

Member OP

sudo rm -rf /

8 #1

Actually I need the Forum Wisdom or? My E5-2680 v4 at 2.4GHz in the Hetzner box, actually I saw the E5-2697 v4 at 2.6GHz for €80 shipped, actually the TDP is identical and the socket is the same and the Memory controller is compatible so actually is this worth the SKUswapping or am I actually chasing Benchmark numbers without actual User experience improvement?

Actually I ran Passmark already:

  • E5-2680v4: 18123
  • E5-2697v4: 18654

Actually that is 3% or? The spreadsheet says 3.1% actually. But actually the Cache is the same and the TurboBoost is the same and actually I am spending $80 for 200MHz baseclock or?

Actually my spouse does not know about this Purchase consideration actually.

ZE

Member

#2

¡the server red is faster! Jaja no no

Actually 3% is the improvement small, but the heart happy is the improvement big! Vos sabés the feeling, the box is yours, the CPU new is shine

But the money $80, the spouse angry, ¡think careful!

bellaauc

Member

3-2-1 Believer

#3

Did you test your restore? Not your backup—your sanity!

Before you swap, checklist:

  • BIOS updated for microcode? Both v4, but revisions matter
  • Thermal paste + pads in stock? 3-2-1 rule applies to supplies too
  • IPMI SOL ready for "no POST" debugging
  • Rollback plan: keep old CPU, test boot first

3% for $80 is $26.67 per percentage point. I've seen worse hobbies. Haven't seen better marriages survive undisclosed hardware purchases though.

3-2-1 or you're already dead
PU

Member

#4

Mas tipo

Mas ⏱️

adam20

Member

Ex-Abuse Desk

#5

Claims 'CPU was same generation.' Refused to acknowledge bent pin visible in photograph. Demands warranty coverage for 'just a swap.'"

We've all been rudi3's provider in this story. Dry humor: at least he's asking *before* the screenshot of 12 bent pins hits our queue.

reported. resolved. repeat.
YuriDavid

Member

#6

---
- cost_analysis:
- upfront: $80
- opportunity_cost: spouse_trust
- performance_gain: 3%
- NOTE: 200MHz is 8.3% base clock increase
- WARNING: actual gain depends on workload
- all-core_turbo: identical
- single_thread: marginal
- memory_bound: zero
---
- alternatives:
- $80 into RAM upgrade: measurable
- $80 into NVMe cache: user_visible
- $80 into "spouse dinner": survival

indentation is not optional
FR

Member

#7

Passmark spread is 531 points. Within margin of error?

svendeal

Member

Deal Sniper

#8

All-core turbo is identical? Then where does the 200MHz live

world record: 4min Arch install
rudi3

Member OP

sudo rm -rf /

#9

Actually thanks for the replies, sorry for the quiet I was traveling.

svendeal said:
All-core turbo is identical? Then where does the 200MHz live
This is actually the right question. Base clock 2.4 to 2.6 but turbo table says both hit 3.2 on single and 2.8 all-core. So the 200MHz is basically idle and light thread loads. My workload is mostly all-core compile jobs so maybe I answered my own question here.

Still haven't pulled the trigger. $80 is two nice dinners also.

cleardmitri

Moderator

Moderator

#10

Late to this, but

svendeal said:
All-core turbo is identical? Then where does the 200MHz live

Base clock only. Both chips cap at 2.8GHz all-core turbo per Intel ARK.

So your sustained load sees zero gain. The 3% is single-thread burst or bench variance.

27 days quiet, but did you pull the trigger? Curious if the "heart happy" won.

No logs, no proof. I have logs.

Post a reply

You need an account to reply. Log in or register to join the conversation.

Post reply Preview Save draft