F5F Stay Refreshed Hardware Desktop At 3nm it means advanced manufacturing, focusing on more processing power rather than simply adding more cores.

At 3nm it means advanced manufacturing, focusing on more processing power rather than simply adding more cores.

At 3nm it means advanced manufacturing, focusing on more processing power rather than simply adding more cores.

Pages (2): 1 2 Next
G
Greeklemon
Member
59
12-17-2019, 04:27 AM
#1
We're nearing it... around 2030, reaching 3nm. Continuing to add cores? What will the 3900x look like then compared to the latest AMD Ryzen 9 3900X? Let's see how it stacks up: https://cpu.userbenchmark.com/Compare/In...1980vs4044. Should we anticipate such a gap?
G
Greeklemon
12-17-2019, 04:27 AM #1

We're nearing it... around 2030, reaching 3nm. Continuing to add cores? What will the 3900x look like then compared to the latest AMD Ryzen 9 3900X? Let's see how it stacks up: https://cpu.userbenchmark.com/Compare/In...1980vs4044. Should we anticipate such a gap?

S
SnowBall_Girl
Junior Member
31
12-17-2019, 06:15 AM
#2
None of these will work with Windows 95.
S
SnowBall_Girl
12-17-2019, 06:15 AM #2

None of these will work with Windows 95.

T
Thuder2
Member
174
12-17-2019, 11:31 AM
#3
We've arrived at a stage where core details don't matter. They'll need to develop 128-bit solutions to maintain progress. X58 remains viable within certain limits.
T
Thuder2
12-17-2019, 11:31 AM #3

We've arrived at a stage where core details don't matter. They'll need to develop 128-bit solutions to maintain progress. X58 remains viable within certain limits.

M
MCtoto123
Member
56
01-04-2020, 03:36 PM
#4
M
MCtoto123
01-04-2020, 03:36 PM #4

A
Atyrio
Junior Member
49
01-06-2020, 09:19 AM
#5
We should anticipate such a gap if innovation returns in the CPU field. We’ll probably reach below 3nm, maybe before 2030. For consumer desktops we’re unlikely to see more than 32 cores, since gaming wouldn’t justify it without extreme optimization. Workstation and server CPUs, however, could scale higher thanks to virtualization enabling additional cores. P.S. Abandon Userbenchmark—it’s a dubious site with no real value. The performance gap between those CPUs is far greater than what the site claims because of its flawed scoring method.
A
Atyrio
01-06-2020, 09:19 AM #5

We should anticipate such a gap if innovation returns in the CPU field. We’ll probably reach below 3nm, maybe before 2030. For consumer desktops we’re unlikely to see more than 32 cores, since gaming wouldn’t justify it without extreme optimization. Workstation and server CPUs, however, could scale higher thanks to virtualization enabling additional cores. P.S. Abandon Userbenchmark—it’s a dubious site with no real value. The performance gap between those CPUs is far greater than what the site claims because of its flawed scoring method.

B
Baki13
Member
142
01-07-2020, 04:39 PM
#6
There could be a time when processors are built from materials other than silicon to achieve greater performance. For context, my father worked at CRAY for many years (a supercomputer firm). He observed how Seymour Cray attempted to create supercomputer chips using Gallium Arsenide instead of silicon because it offers certain performance and output benefits. The issue was that producing functional wafers—even a few working chips—yielded poor results, so he had to abandon the approach. You give a group of innovators like Seymour Cray (maybe rest his soul, he died in a car accident long before his era) sufficient time and resources, and they’ll discover alternatives to silicon for making CPUs that are more effective.
B
Baki13
01-07-2020, 04:39 PM #6

There could be a time when processors are built from materials other than silicon to achieve greater performance. For context, my father worked at CRAY for many years (a supercomputer firm). He observed how Seymour Cray attempted to create supercomputer chips using Gallium Arsenide instead of silicon because it offers certain performance and output benefits. The issue was that producing functional wafers—even a few working chips—yielded poor results, so he had to abandon the approach. You give a group of innovators like Seymour Cray (maybe rest his soul, he died in a car accident long before his era) sufficient time and resources, and they’ll discover alternatives to silicon for making CPUs that are more effective.

C
coolerijk
Member
125
01-09-2020, 04:23 AM
#7
organics
C
coolerijk
01-09-2020, 04:23 AM #7

organics

E
Ezmoth
Member
62
01-17-2020, 12:49 AM
#8
Also quantum computing will become more relevant in the coming years. I’m excited about completely different architectures beyond x86 and novel physical designs. Gone are the days of traditional wafer chips in sockets. Back in the day, HP suggested building CPUs as modular bricks with connections on top and bottom—upgrading meant simply adding another brick. Those concepts are exactly what I’m looking forward to.
E
Ezmoth
01-17-2020, 12:49 AM #8

Also quantum computing will become more relevant in the coming years. I’m excited about completely different architectures beyond x86 and novel physical designs. Gone are the days of traditional wafer chips in sockets. Back in the day, HP suggested building CPUs as modular bricks with connections on top and bottom—upgrading meant simply adding another brick. Those concepts are exactly what I’m looking forward to.

Z
Zesthzu66
Junior Member
29
01-17-2020, 10:48 PM
#9
Not always. The situation has reached a stage where electrons start interacting closely.
Z
Zesthzu66
01-17-2020, 10:48 PM #9

Not always. The situation has reached a stage where electrons start interacting closely.

F
Fred10244
Posting Freak
937
01-22-2020, 06:48 AM
#10
God was the original player in planned obsolescence.
F
Fred10244
01-22-2020, 06:48 AM #10

God was the original player in planned obsolescence.

Pages (2): 1 2 Next