No Performance Improvement after Upgrade

[lenhard]

I recently did an upgrade on my system but the performance at best stayed the same and even decreased about 20% in cuda-related computation (deep learning). Can it be my settings or was my choice of components simply that bad?
(I'm no expert when it comes to picking the right parts, but I thought I would get at least a bit more performance out of my new system)

Previous System:
Mainboard: Asus Sabertooth Mainboard Sockel 2011
CPU: Intel i7-3820 Prozessor
Graphics: Nvidia GTX 1060 6GB
RAM: DDR3 Kingston HyperX XMP 16GB (2133 MHz, 240-polig, CL11)
Disk: Samsung SSD 850 PRO 256GB
Power: Corsair AX850
Cooler: Fractal Design

Upgraded System:
Mainboard: ASUS PRIME Z390-A
CPU: Intel Core i7 9700K
Graphics: Nvidia GTX 1060 6GB
RAM: DDR4 G.Skill RGB F4 16GB
Disk: Samsung MZ-V7E500BW 970 EVO internal SSD
Power: Corsair AX850
Cooler: Noctua

Any suggestions on where and what to change? other RAM? For CUDA-computation probably a graphics-card upgrade(?) but which one? (another 800€ for a RTX 2080 would kinda hurt - is there a card that would provide me some significant increase in performance?)

Thanks in advance!
Lenny

 

[lthampton]

Well most games rely on the GPU and since you did not update it, then you're not going to see much of an improvement, however if you are looking for a large upgrade take a look at the gigabyte windforce 2070 cards ( msrp around 499), or perhaps even a 1080.

 

[sizzling]

Did you do a fresh install of Windows after the upgrade?

 

[lenhard]

Well most games rely on the GPU and since you did not update it, then you're not going to see much of an improvement, however if you are looking for a large upgrade take a look at the gigabyte windforce 2070 cards ( msrp around 499), or perhaps even a 1080.

The use case of this system is more machine learning and computational heavy big data operations rather than gaming. But your suggestion with the RTX 2070 seems very promising - although a little expensive... probably worth it though

 

[lenhard]

Did you do a fresh install of Windows after the upgrade?

Yes. With the upgrade of the motherboard I kind of had to

 

[jankerson]

I recently did an upgrade on my system but the performance at best stayed the same and even decreased about 20% in cuda-related computation (deep learning). Can it be my settings or was my choice of components simply that bad?
(I'm no expert when it comes to picking the right parts, but I thought I would get at least a bit more performance out of my new system)

Previous System:
Mainboard: Asus Sabertooth Mainboard Sockel 2011
CPU: Intel i7-3820 Prozessor
Graphics: Nvidia GTX 1060 6GB
RAM: DDR3 Kingston HyperX XMP 16GB (2133 MHz, 240-polig, CL11)
Disk: Samsung SSD 850 PRO 256GB
Power: Corsair AX850
Cooler: Fractal Design

Upgraded System:
Mainboard: ASUS PRIME Z390-A
CPU: Intel Core i7 9700K
Graphics: Nvidia GTX 1060 6GB
RAM: DDR4 G.Skill RGB F4 16GB
Disk: Samsung MZ-V7E500BW 970 EVO internal SSD
Power: Corsair AX850
Cooler: Noctua

Any suggestions on where and what to change? other RAM? For CUDA-computation probably a graphics-card upgrade(?) but which one? (another 800€ for a RTX 2080 would kinda hurt - is there a card that would provide me some significant increase in performance?)

Thanks in advance!
Lenny

Well you didn't upgrade the GPU so you are still stuck at mid range performance no matter what CPU you put in there.

As others have said an RTX 2070, GTX 1080 would be a nice performance boosts in games.

 

[ElectrO_90]

Sadly, you would have been better off upgrading the GPU before you upgraded the rest of the system. You sort of did it backwards.
It's better to have a powerful GPU over a Powerful CPU.