Ways of Managing NIH Resources by Office of Extramural Research - 未名空间MITBBS历史存档

国际科技财经博客移民网络热点娱乐民生时事公众号

Redian新闻

>未名空间

>Biology - 生物学

Ways of Managing NIH Resources by Office of Extramural Research

Ways of Managing NIH Resources by Office of Extramural Research# Biology - 生物学

c*k2011-10-20 07:10

1 楼

【以下文字转载自 Piebridge 讨论区】
发信人: beifang2008 (北方), 信区: Piebridge
标题: 关于搬运大妈
发信站: BBS 未名空间站 (Mon Mar 29 15:44:50 2010, 美东)
挖个严肃坑。
一个人过，生活太没有规律。想想还是有个人约束一下比较好。想起一个儿时好友，与
俺年龄相当，现已离异多时，有两个孩子，老大归她，已大学毕业成家，老二归男方。
几个问题：
1）虽然从小一起长大，彼此美国有隔阂。可是毕竟这么多年各自有不妥呢个的生活，
会不会性格和生活习惯有太多差异？
2）大妈英语不好（基本不会），如果搬过来，生活会不会不适应，如果因此她整天闷
闷不乐也不好。
3）她后来的生活中应该也有很多亲戚、朋友，俺并不了解。不知道自己能不能面对，
也许自己多虑了。
请鹊版的朋友们给说说。

m*52011-10-20 07:10

2 楼

垂直叠加这种粗暴的技术已经发了很多文章，但是我觉得最重要的问题根本没解决：
产热，散热
我估计等碳基芯片出来了，这个idea还在意淫
IBM已经与胶水制造商3M合作，联合打造用来“叠加”100层硅片，制造速度比现在快
1000倍的处理器
电脑制造商IBM已经与3M公司合作，利用胶水把一层层芯片粘在一起，研制“摩天楼”
电脑。希望通过这种方式，可以令手机和PC的速度提高1000倍。这种产品有望在2013年
上市。
3M公司还生产用于航空航天业的耐热胶水、黏合剂和胶带，但是它与IBM联合研制
的这种高科技胶水，事实上是计算机信息处理技术能否向前迈进一大步的关键。现在把
芯片垂直叠加在一起的技术面临产品过热等问题。新型胶水应该让叠加在一起的芯片具
备良好的导热性，确保线路不会因过热而烧毁。该研究打算制造由100层芯片组成的商
用微处理器。
IBM研究部副总裁伯尼-梅尔森在声明里说：“这种新型‘3D’方法能够大大提高平板电
脑等电子产品的速度，进而显著提高电脑的能力。”

y*82011-10-20 07:10

3 楼

N*******************[email protected]
October 2011
Options for Managing NIH Resources
A.Current Way of Managing
–Bottom out success rates (doing nothing but letting the system correct
itself)
B.Other Options
−Reducing or limiting size of awards
−Limiting number of awards held by an PI
−Limiting the amount of funds an PI can hold
−Limiting salaries of PIs
Current Way of Managing
•Generally project based
•Competitive peer review
•Average of $414K per year for 4.3 years
•Approximately 50 institutions get 70% of the funds
•Approximately 20% of the investigators get 50% of the funds
Using FY 2010 budget and actual numbers of applications, the effect of NIH
budget reductions on RPG success rate:
10% reduction in NIH RPG Budget = 12.3% Success Rate*
20% reduction in NIH RPG Budget = 4.1% Success Rate*
FY10 RPGs = 20.6% Success Rate
* Assumes no cuts to non-competing
Option: Reducing or Limiting Size of Award
Reducing Average Size of an Award
FY 2010 Competing RPG Awards: 9,287
Total Cost: $3,732,077,099
Average Cost: $401,860
Average Cost - $25k: $376,860
Total Savings: $232,175,000
Example of effect on number of RPG Awards in FY 2010:
FY10 Actual Estimated Effect
Competing RPG awards: 9,287 9,903
RPG success rate*: 20.5% 21.7%
Limiting the Number of Awards Per PI
FY 2010 – No Limits
–Projecting 9,455 competing NIH RPGs with 20.6% success rate
Limit to 5 RPGs per PI
–Estimate 9,470 competing (+15) NIH RPGs with 20.6% success rate
Limit to 3 RPGs per PI
–Estimate 9,719 competing (+264) NIH RPGs with 21.1% success rate
Limit to 2 RPGs per PI
–Estimate 10,411 competing (+956) NIH RPGs with 22.6% success rate
Limiting the Amount of Funds per PI
Limit PI’s Total RPG Support to $1 Million
•Affects 3,245 PIs with total current funding of $6.4 billion
•Funding for these PIs reduced to $3.245 billion
•Savings of $3.1 billion (estimated one-fourth available for competing
awards in any given year, or approximately $800 million)
•Allows for approximately 2,000 additional competing RPG awards at
average cost of $400,000
Limit PI’s Total RPG Support to $800,000
•Affects 4,629 PIs with total current funding of $7.6 billion
•Funding for these PIs reduced to $3.7 billion
•Savings of $3.9 billion (estimated one-fourth available for competing
awards in any given year, or approximately $975 million)
•Allows for approximately 2,400 additional competing RPG awards at
average cost of $400,000
Limit PI’s Total RPG Support to $400,000
•Affects 12,000 PIs with total current funding of $11.9 billion
•Funding for these PIs reduced to $4.8 billion
•Savings of $7.1 billion (estimated one-fourth available for competing
awards in any given year, or approximately $1.8 billion)
•Allows for approximately 4,400 additional competing RPG awards at
average cost of $400,000
Limiting Salaries of PIs

a*e2011-10-20 07:10

4 楼

叠个5~10层估计也就是几年的事
不过应该不是高性能应用

【在 m********5 的大作中提到】

: 垂直叠加这种粗暴的技术已经发了很多文章，但是我觉得最重要的问题根本没解决：
: 产热，散热
: 我估计等碳基芯片出来了，这个idea还在意淫
: IBM已经与胶水制造商3M合作，联合打造用来“叠加”100层硅片，制造速度比现在快
: 1000倍的处理器
: 电脑制造商IBM已经与3M公司合作，利用胶水把一层层芯片粘在一起，研制“摩天楼”
: 电脑。希望通过这种方式，可以令手机和PC的速度提高1000倍。这种产品有望在2013年
: 上市。
: 3M公司还生产用于航空航天业的耐热胶水、黏合剂和胶带，但是它与IBM联合研制
: 的这种高科技胶水，事实上是计算机信息处理技术能否向前迈进一大步的关键。现在把

y*82011-10-20 07:10

5 楼

FYI, NIH may get 10% cut for 2012 fiscal year.
I don't know how many people would support "Limiting Salaries of PIs".
I thought that "Limiting number of awards held by an PI" had been in effect
for a long while.

y*82011-10-20 07:10

6 楼

Simply making the NIH R01 non-renewable would be most efficient. My thought.

e*o2011-10-20 07:10

7 楼

555

effect

【在 y******8 的大作中提到】

: FYI, NIH may get 10% cut for 2012 fiscal year.
: I don't know how many people would support "Limiting Salaries of PIs".
: I thought that "Limiting number of awards held by an PI" had been in effect
: for a long while.