BTC $64,833.39 +1.33%
ETH $1,814.68 +1.29%
BNB $593.04 +1.32%
XRP $0.9692 +2.20%
SOL $73.37 +3.79%
TRX $0.3055 +0.69%
DOGE $0.0593 +1.71%
ADA $0.1545 +2.42%
BCH $210.72 +3.40%
LINK $9.25 +1.98%
HYPE $69.23 +5.72%
AAVE $113.46 +4.83%
SUI $0.7357 +3.45%
XLM $0.1404 +2.53%
ZEC $948.29 +8.56%
AAPL $351.83 -1.69%
AMZN $259.30 -2.03%
GOOGL $283.89 -3.27%
MSFT $516.24 -1.49%
META $624.36 -1.89%
NVDA $158.35 +0.70%
TSLA $381.78 -0.90%
SNDK $830.30 +0.43%
INTC $55.69 -1.36%
SPCX $139.07 +0.03%
MU $737.59 +0.52%
AMD $518.77 +2.42%
BTC $64,833.39 +1.33%
ETH $1,814.68 +1.29%
BNB $593.04 +1.32%
XRP $0.9692 +2.20%
SOL $73.37 +3.79%
TRX $0.3055 +0.69%
DOGE $0.0593 +1.71%
ADA $0.1545 +2.42%
BCH $210.72 +3.40%
LINK $9.25 +1.98%
HYPE $69.23 +5.72%
AAVE $113.46 +4.83%
SUI $0.7357 +3.45%
XLM $0.1404 +2.53%
ZEC $948.29 +8.56%
AAPL $351.83 -1.69%
AMZN $259.30 -2.03%
GOOGL $283.89 -3.27%
MSFT $516.24 -1.49%
META $624.36 -1.89%
NVDA $158.35 +0.70%
TSLA $381.78 -0.90%
SNDK $830.30 +0.43%
INTC $55.69 -1.36%
SPCX $139.07 +0.03%
MU $737.59 +0.52%
AMD $518.77 +2.42%

cs-4

All
Article
Flash

hot_img Cerebras releases the fourth generation AI inference system CS-4: performance doubled, power consumption doubled, more flexible deployment

Cerebras released its fourth-generation AI inference system CS-4 this week, based on the same 5nm WSE-3 wafer, achieving double the performance by doubling the clock frequency and power consumption. A single CS-4 cabinet accommodates 3 wafers (CS-3 has 2), featuring a modular "backpack" design that simplifies manufacturing and deployment, with a TDP of approximately 125 to 135kW. The CS-4 can provide an inference speed of nearly 4000 tokens/second/user, about twice that of the CS-3, and supports decomposed inference with heterogeneous systems such as AMD and AWS Trainium.Cerebras claims that the CS-4 offers about 2000 times the on-chip memory bandwidth of NVIDIA's Rubin (43PB/s), but the 44GB SRAM capacity remains unchanged, and long-context inference still requires multi-wafer stacking. For example, with the DeepSeek V4 Pro (1.6T parameters), approximately 20 systems are needed for a 1M context window, and about 40 systems are required for 256 concurrent users, corresponding to a CAPEX exceeding 20 million USD. Cerebras is collaborating with clients such as OpenAI and plans to achieve approximately double performance improvements each year, aiming for a 20-fold throughput increase by 2027. The "backpack" cabinet design of the CS-4 will continue into the next-generation "Nexus" platform.
app_icon
ChainCatcher Building the Web3 world with innovations.