Old 6 core Xeons still good value in a world of used Ryzen's?

[emodan]

One of the advantages of going second-hand Ryzen is the upgrade path.

I recently built my friend a super budget Ryzen build with a Ryzen 3 1200 and an RX 480(which was free as hand me down from a mutual friend) as whilst this is all he could afford for now next year when say the Ryzen 5 2600 is £100 he can add this to his build himself without any fuss. For the same price could have made him a system with older better performance parts but that would leave him no easy upgrade path.

 

[MartinPrince]

Yeah I watched that video earlier and to be honest it comes across as purely trolling other channels that recommend old tech. He picked literally the worst combo possible and compared it to systems that were blatantly going to thrash it.....

I would tend to agree. If he had compared a Xeon 5650 and overclocked it to ~4.2Ghz, which most people should look to do (Mine was running stable at 4.6Ghz) then that would have made for a more interesting comparison.

 

[Poneros]

What drama-mongers like Hardware Unboxed are missing is that, for some people (and I've been in this position) the $50-80 difference between two systems is the difference between being able to afford the PC or not. That's why these Xeons are such great deals & why they very much have a place.

He doesn't care because he just gets free **** sent all the time, so saying just add X more $s is easy. Sadly he is an egomaniac and thus he can only think of himself - he can't put himself in other people's shoes. That's why he makes these videos taking shots at other Australian youtubers all the time, he can't stand not being the top dog in his region on youtube so he wants to take them down, even though in the end he's just making misleading click-bait videos.

 

[davida76]

I would tend to agree. If he had compared a Xeon 5650 and overclocked it to ~4.2Ghz, which most people should look to do (Mine was running stable at 4.6Ghz) then that would have made for a more interesting comparison.

Definitely mate, but then that wouldn't fit the agenda as we both know.

 

[Strife212]

The video in OP is reviewing them at stock, of course they're going to be bad. They're insane clockers and you can easily get 1.5Ghz+ overclocks on them.

Overclocked they're barely worse than Ryzens, it's a joke considering how old they are.

 

[humbug]

The video in OP is reviewing them at stock, of course they're going to be bad. They're insane clockers and you can easily get 1.5Ghz+ overclocks on them.

Overclocked they're barely worse than Ryzens, it's a joke considering how old they are.

Twice as fast and the 1400 overclocks a good way too.

 

[Steveocee]

New low end tech outperforms old higher end tech that is cheap because it's old.

/thread

In other news.
The sky is blue, the grass is green and bumhug continues to over justify Ryzen across the OCUK boards.

 

[stooeh]

My backup system running my old stuff has my GA-EX58-UD5 1366 board (will be 10 years old in September!) with a 4GHz ancient overclocked Xeon from ebay and a GTX 970. Still a decent low end system
Definitely wouldn't recommend buying a board now though...

 

[~>Dg<~]

New low end tech outperforms old higher end tech that is cheap because it's old.

/thread

In other news.
The sky is blue, the grass is green and bumhug continues to over justify Ryzen across the OCUK boards.

this can be the case but can sometimes also be misleading. people in the new zen thread are comparing a 3600 to a stock 5820k for eg and saying upto 30 percent difference on single core ipc. the 5820k is clocked at 3.3ghz lol. when many will do upto 4.5ghz+. the difference is night and day and nigh on as quick clocked as the 3600. or close enough not to warrant upgrading. sad thing is some will based on stupid benchmarks. upgrade and waste 700-1000 pound. think before posting stupid benchmarks which obviously arent done well. it could cost someone a lot of money for next to no upgrade . sometimes even worse performance. while many will have the spare funds . some wont and will of saved for years for that next upgrade. that stupid benchmark may swing their decisions.

 

[humbug]

this can be the case but can sometimes also be misleading. people in the new zen thread are comparing a 3600 to a stock 5820k for eg and saying upto 30 percent difference on single core ipc. the 5820k is clocked at 3.3ghz lol. when many will do upto 4.5ghz+. the difference is night and day and nigh on as quick clocked as the 3600. or close enough not to warrant upgrading. sad thing is some will based on stupid benchmarks. upgrade and waste 700-1000 pound. think before posting stupid benchmarks which obviously arent done well. it could cost someone a lot of money for next to no upgrade . sometimes even worse performance. while many will have the spare funds . some wont and will of saved for years for that next upgrade. that stupid benchmark may swing their decisions.

5820K @ 4.0Ghz

Well here's the AMD 3600 vs my 5820K in geekbench. Now my 5820K is only at 4Ghz

Ryzen 3000 is looking extremely promising for performance and features.

3600 +25% ST, +31% MT.

 

[Hades]

Anither interesting discussion here. After watching it I looked on ebay and you can buy a dual processor board and two 8 core CPU's for about £210 delivered. That's 16 cores and 32 threads for £210! Cracking value even if the single core efficiency isn't great.

 

[~>Dg<~]

5820K @ 4.0Ghz

just to compare and show real figures. 5820k everything running in background daily use. at 4.4 with 2400mhz ram. scores


this chip does 4.6 all day. most 5820ks will do 4.4-4.5 with right cooling. then remember that 3600 ram its using. 2400 on the 5820k . 3600 mhz on the zen. quite a bit of difference just on ram. so overclocked 5820k with same speed ram would be very close. you just chose the minimums to try and pro amd it. the 3600 is basically a 8700. non k chip. thats it.

 

[humbug]

just to compare and show real figures. 5820k everything running in background daily use. at 4.4 with 2400mhz ram. scores


this chip does 4.6 all day. most 5820ks will do 4.4-4.5 with right cooling. then remember that 3600 ram its using. 2400 on the 5820k . 3600 mhz on the zen. quite a bit of difference just on ram. so overclocked 5820k with same speed ram would be very close. you just chose the minimums to try and pro amd it. the 3600 is basically a 8700. non k chip. thats it.

Well that's not bad, 4.4Ghz vs 4.2Ghz, 700 points ST, near 20% IPC difference. You like that number, you've used it incorrectly a lot

 

[~>Dg<~]

not bad compare like for like. 3600 mhz vs 2400. thats a massive difference. i expect the 3600 to be quicker. it should be its new 5820k is 5 years old roughly. the point is with the same ram and the 5820k overclocked there isnt much in it. good though if you buying on budget now.

run your ryzen 1600 for comparision humbug.

just for comparision a 2700x scores 4800

https://browser.geekbench.com/processors/2080

 

[humbug]

not bad compare like for like. 3600 mhz vs 2400. thats a massive difference. i expect the 3600 to be quicker. it should be its new 5820k is 5 years old roughly. the point is with the same ram and the 5820k overclocked there isnt much in it. good though if you buying on budget now.

run your ryzen 1600 for comparision humbug.

Quick run at 3.8Ghz.

4420
21858

https://browser.geekbench.com/v4/cpu/13713977

 

[~>Dg<~]

so basically if the 3600 is as said single ipc is about the same as said 8700-8700k

8700 https://browser.geekbench.com/processors/2063 5327
8700k https://browser.geekbench.com/processors/2062 5891

giving some actual insight. so people can guage possible performance.

 

[humbug]

so basically if the 3600 is as said single ipc is about the same as said 8700-8700k

8700 https://browser.geekbench.com/processors/2063 5327
8700k https://browser.geekbench.com/processors/2062 5891

giving some actual insight. so people can guage possible performance.

Similar ST score but 4.7Ghz vs 4.2Ghz, the 8700K is 12% higher clocked. the 3950X will be clocked to 4.7Ghz.

Also, add the clock speed difference vs my 3.8Ghz ontop of your 5080 and i would score 5127.

 

[~>Dg<~]

the difference is mine will do as said ryzens wont.same as the 8700k. oc is valid if they can do it. thats why your 1600 is at 3.9. if it could do this that which it cant.pointless. the benchmark actually shows how slow ryzen chips are tbh. 2700x scoring below a 5820k. nevermind a 1600. the new cpus look promising though if this benchmark shows real relevance

 

[DragonQ]

I'm sorry but since when is Geekbench an indicator for any kind of performance? It's all very interesting looking into numbers but let's not kid ourselves here. All of these synthetic benchmarks are trash.

 

[~>Dg<~]

which is why i put the if this benchmark shows real relevance.